阅读说明：本技术 基于混合并行遗传和变邻域算法的生产调度方法和系统 (Production scheduling method and system based on hybrid parallel inheritance and variable neighborhood algorithm ) 是由 陆少军 马崇轺 刘心报 程浩 崔龙庆 胡朝明 江涛 赵婷 于 2021-07-19 设计创作，主要内容包括：本发明提供一种基于混合并行遗传和变邻域算法的生产调度方法、系统、存储介质和电子设备,涉及生产调度领域。本发明采用启发式算法获取初始化种群中每个个体的各个车间生产调度方案,将适应度值最高的个体作为全局最优解；在邻域结构中搜索新解；将更新后的全局最优解迁移到各个子群体；根据更新后的各个子群体中个体的适应度值,采用选择算子、交叉算子和变异算子获取下一代子群体；在当前群体选择适应度值最高的个体,更新全局最优解。通过混合粗粒度并行遗传和变邻域搜索优化算法的迭代找到近似最优解,避免遗传算法的早熟现象,也加快了算法的收敛程度；考虑机器加工恶化效应和资源投入带来的效率提升,兼顾生产调度决策和资源配置决策问题。(The invention provides a production scheduling method, a production scheduling system, a storage medium and electronic equipment based on a hybrid parallel inheritance and variable neighborhood algorithm, and relates to the field of production scheduling. The method adopts a heuristic algorithm to obtain each workshop production scheduling scheme of each individual in the initialized population, and takes the individual with the highest fitness value as a global optimal solution; searching a new solution in the neighborhood structure; migrating the updated global optimal solution to each sub-population; according to the updated fitness value of the individuals in each sub-population, adopting a selection operator, a crossover operator and a mutation operator to obtain a next sub-population; and selecting the individual with the highest fitness value in the current population, and updating the global optimal solution. An approximate optimal solution is found through iteration of a mixed coarse-grained parallel inheritance and variable neighborhood search optimization algorithm, the premature phenomenon of the genetic algorithm is avoided, and the convergence degree of the algorithm is accelerated; and considering the efficiency improvement brought by the machine processing deterioration effect and the resource investment, and considering the production scheduling decision and the resource allocation decision.)

1. A production scheduling method based on a hybrid parallel inheritance and variable neighborhood algorithm is characterized by comprising the following steps:

s1, initializing algorithm parameters, wherein the algorithm parameters at least comprise neighborhood searching times, neighborhood structure number, maximum genetic iteration times and maximum algorithm cycle iteration times;

s2, distributing and coding all workpieces to be processed, randomly generating an initialization group, and dividing the initialization group into a plurality of sub-groups; decoding by adopting a heuristic algorithm to obtain the processing sequence and the resource allocation scheme of each workshop workpiece corresponding to each individual in the initialized population, calculating the fitness value of each individual, and taking the individual with the highest fitness value as a global optimal solution;

s3, searching a new solution in the corresponding neighborhood structure according to the global optimal solution;

s4, judging whether the fitness value of the global optimal solution is smaller than that of the new solution, if so, updating the global optimal solution into the new solution, and turning to S3; otherwise, the global optimal solution is reserved, and the process goes to S5;

s5, judging whether a neighborhood search termination condition is met or not according to the neighborhood search times and the neighborhood structure number, if so, ending the variable neighborhood search, transferring the global optimal solution to each sub-population, and turning to S6; otherwise, go to S3;

s6, calculating the updated fitness value of the individuals in each sub-population;

s7, obtaining a next generation sub-population by adopting a preset selection operator, a preset crossover operator and a preset mutation operator according to the updated fitness value of the individuals in each sub-population; selecting an individual with the highest fitness value in the current group, and updating the global optimal solution according to the individual with the highest fitness value;

s8, judging whether the current genetic iteration number is smaller than the maximum genetic iteration number, if so, adding one to the current genetic iteration number, and switching to S6; otherwise, go to S9;

s9, judging whether the iteration number of the current algorithm loop is smaller than the maximum algorithm loop iteration number, if so, adding one to the iteration number of the current algorithm loop, and turning to S3; and if not, taking the processing sequence of the workpieces in each workshop corresponding to the global optimal solution and the resource allocation scheme as the optimal scheduling scheme.

2. The hybrid parallel inheritance and variable neighborhood algorithm-based production scheduling method according to claim 1, wherein the step S2 of decoding and acquiring the processing order and the resource allocation scheme of each workshop workpiece corresponding to each individual in the initialized population by using a heuristic algorithm comprises the steps of:

s10, arranging the workpieces in each workshop according to the respective non-decreasing sequence of initial processing time, and enabling fs to be 1; the initial processing time refers to the processing time a' before the workpiece is not allocated with resources;

s20, allocating alpha resources to the fs workpiece in the permutation;

s30, if fs α < R_jAnd R is_jIf- (fs +1) × α > 0, fs ═ fs +1, the process proceeds to S20; otherwise, go to S40; wherein R is_jThe number of resources of the jth workshop is shown, j is 1, …, M, and M shows the number of workshops;

s40, if fs α < R_jAnd R is_j- (fs + 1). alpha. < 0, then R_j-fs +1 th workpiece in the permutation is assigned fs + α resources;

and S50, reversing the arrangement sequence of the workpieces in the workshop, determining the processing sequence of the workpieces, and acquiring the processing sequence of the workpieces in the current workshop and the resource allocation scheme.

3. The hybrid parallel genetic and variable neighborhood algorithm-based production scheduling method of claim 1, wherein said number of neighborhood structures comprises four different neighborhood structures, defining a neighborhood structure N_kK is 1 or 2 or 3 or 4;

if k is 1, correspondingly, S3 includes:

s311, defining the global optimal solution as V^*＝(v₁，…，v_i，…v_N) Wherein v is₁，…，v_i，…v_NIs any integer from 1 to M; v. of_iIndicating the corresponding workshop to which the ith workpiece is distributed; n represents the number of all workpieces to be processed; m represents the number of workshops;

s312, making q equal to 1, counting the total actual processing time of each workshop corresponding to the global optimal solution,

wherein n is_jIndicating the number of the processed workpieces in the jth workshop; p is a radical of_iIndicating the initial deterioration time of the ith workpiece when no resource is allocated; r is_iIndicating the amount of the workpiece resource distributed at the ith; the actual deterioration time of the ith workpiece in the s-th machining in the corresponding workshop is (p)_i-θr_i)s； The actual processing time of the ith part for processing the workpieces in the ith workshop in the s-th sequence is shown, and a' represents the initial processing time of each workpiece; θ is the efficiency of each resource;

obtaining the maximum total actual machining time of a workshop

S313, randomly generating 1 random integer rand from 1 to M₁；

S314, if rand₁＝j_max1If yes, the operation goes to S313; otherwise, go to S315;

s315, if v_q＝rand₁Then v is_q＝j_max1(ii) a If v is_q＝j_max1Then v is_q＝rand₁(ii) a Otherwise, go to S316;

s316, if q is less than N, q is q +1, and go to S315; otherwise, switching to S4 after obtaining a new solution;

or if k is 2, correspondingly, S3 includes:

s321, defining the global optimal solution to be represented as V^*＝(v₁，…，v_i，…v_N)；

S322, counting the total actual processing time P of each workshop corresponding to the global optimal solution_jObtaining the maximum total actual machining time of the workshopAnd obtaining the j_max2Initial deterioration time p of processed workpiece in each workshop_imax2Largest workpiece, p_imax2＝max{p_i|v_i＝j_max2}；

S323, randomly generating 1 random integer rand from 1 to M₂；

S324, if rand₂＝j_max2If yes, go to S323; otherwise, let v_imax＝rand₂And then the new solution is transferred to S4;

or if k is 3, correspondingly, S3 includes:

s331, defining the global optimal solution to be V^*＝(v₁，…，v_i，…v_N)；

S332, counting the total actual processing time P of each workshop corresponding to the global optimal solution_jObtaining the maximum total actual machining time of the workshopAnd obtaining the j_max3Workpiece p with largest initial deterioration time of processed workpiece in each workshop_imax3＝max{p_i|v_i＝j_max3}；

S333, randomly generating 1 random integer rand from 1 to N₃；

S334, ifThen the process proceeds to S333; otherwise, proceed to S335

S335, setting a new variable c, and orderv_imax3C, obtaining a new solution and then transferring to S4;

or if k is 4, correspondingly, S3 includes:

s341, defining the global optimal solution as V^*＝(v₁，…，v_i，…v_N)；

S342, randomly generating random integers a and b in a range from 1 to N, and setting variables x, y, z', and d, wherein x is 0;

s343, if a is equal to b, the process returns to S342; otherwise, go to S344;

s344, if a > b, let z be b, b be a, and a be z; let d be b-a +1, otherwise, go to S345;

s345, let z ═ v_a+x，v_a+x＝v_b-x，v_b-x＝z′，x＝x+1；

S346, if x is less than y, turning to S345; otherwise, the process proceeds to S4 after a new solution is obtained.

4. The method for production scheduling based on hybrid parallel genetic and variable neighborhood algorithm according to claim 3,

and S2, distributing and coding all the workpieces to be processed by adopting an integer coding mode, and randomly generating an initialization group O, wherein the number of individuals in the group O is G x H, and the initialization group O is averagely divided into H sub-groups O_hThen the number of individuals in each sub-population is G;

the function of the fitness value is:

wherein the content of the first and second substances,representing the g individual in the h sub-population,and C represents the maximum value of the actual completion time in the current population.

5. The hybrid parallel inheritance and variable neighborhood algorithm-based production scheduling method according to claim 4, wherein the preset selection operator in S7 is a roulette mode, and the updated total number of individuals in each sub-population is G ', wherein G' is G + 1;

the updated selection process of each sub-population comprises the following steps:

s10a, setting a variable G ', and calculating G' as 1,2 Value of (A), initial P (X)₀) 0, initial n is 1; wherein the content of the first and second substances,representing the probability of each individual being selected;representing the sum of the fitness of each sub-population;

s20a, randomly generating a random number sx1 from 0 to 1;

s30a, finding that P (X) is satisfied_g′-1)＜sx1≤P(X_g′) G' selecting individualsAdding the next generation parent O'_h，n＝n+1；

S40a, if n is less than or equal to G, the process is switched to S20a, otherwise, the selection process is ended.

6. The method for production scheduling based on hybrid parallel inheritance and variable neighborhood algorithm as claimed in claim 5, wherein the intersection process of each sub-population updated in S7 comprises the following steps:

s10b, mixing the parent substance O'_hThe G individuals in the group are randomly divided into L pairing groups, andtwo sub-individuals in the first matched group are respectivelyAnd wherein the content of the first and second substances,the distribution workshop of the ith workpiece of the kth individual of the ith distribution group is shown, and k is 1 or 2; setting variables m, z₁、z₂Let m equal to 1;

s20b, randomly generating 2 random integers rand from 1 to N₄、rand₅；

S30b, if rand₄＝rand₅Then go to S20 b; otherwise, go to S40 b;

s40 b: order toOrder to

S50b, if m is less than L, then m is m +1, and the process proceeds to S20 b; otherwise, obtaining the next generation group and ending.

7. The method for production scheduling based on hybrid parallel genetic and variable neighborhood algorithm of claim 6,

defining the next generation population obtained by the crossover process is denoted O_h′， Represents the g 'th individual in the h' th sub-population in the next generation population, wherein,the value of (a) indicates the cell to which the ith workpiece is assigned;

the mutation process of each sub-population updated in S7 includes:

s10c, initializing g 'to 1, and setting a mutation probability p';

s20c, randomly generating a random number sx2 from 0 to 1, and switching to S30c if sx2 is less than p'; otherwise, go to S60 c;

s30c, randomly generating a random integer rand from 1 to N₆；

S40c, randomly generating a random integer rand from 1 to M₇；

S50c, ifThen proceed to S40 c; otherwise, it ordersProceeding to S60 c;

s60c, if G' < G, G ═ G +1, proceed to S20 c; otherwise, ending after obtaining the next generation of sub-population.

8. A production scheduling system based on a hybrid parallel inheritance and variable neighborhood algorithm is characterized by comprising:

the initialization module is used for executing S1 and initializing algorithm parameters, wherein the algorithm parameters at least comprise neighborhood searching times, neighborhood structure number, maximum genetic iteration times and maximum algorithm cycle iteration times;

the solving module is used for executing S2, distributing all the workpieces to be processed for coding, randomly generating an initialization group and dividing the initialization group into a plurality of sub-groups; decoding by adopting a heuristic algorithm to obtain the processing sequence and the resource allocation scheme of each workshop workpiece corresponding to each individual in the initialized population, calculating the fitness value of each individual, and taking the individual with the highest fitness value as a global optimal solution;

the searching module is used for executing S3 and searching a new solution in the corresponding neighborhood structure according to the global optimal solution;

the first judging module is used for executing S4 and judging whether the fitness value of the global optimal solution is smaller than that of the new solution, if so, the global optimal solution is updated to the new solution, and the searching module is switched to execute S3; otherwise, the global optimal solution is reserved, and the step is switched to a second judgment module to execute S5;

a second judging module, configured to execute S5, judge whether a neighborhood search termination condition is satisfied according to the number of neighborhood search times and the number of neighborhood structures, end the variable neighborhood search if satisfied, migrate the global optimal solution to each sub-population, and shift to the calculating module to execute S6; otherwise, the step is shifted to a searching module to execute S3;

the calculating module is used for executing S6 and calculating the updated fitness value of the individual in each sub-population;

the updating module is used for executing S7 and acquiring the next generation sub-population by adopting a preset selection operator, a preset crossover operator and a preset mutation operator according to the updated fitness value of the individual in each sub-population; selecting an individual with the highest fitness value in the current group, and updating the global optimal solution according to the individual with the highest fitness value;

a third judging module, configured to execute S8, judge whether the current genetic iteration number is smaller than the maximum genetic iteration number, if so, add one to the current genetic iteration number, and shift to the calculating module to execute S6; otherwise, the step is shifted to a fourth judging module to execute S9;

a fourth judging module, configured to execute S9, judge whether the current algorithm loop iteration number is smaller than the maximum algorithm loop iteration number, if yes, add one to the current algorithm loop iteration number, and shift to the searching module to execute S3; and if not, taking the processing sequence of the workpieces in each workshop corresponding to the global optimal solution and the resource allocation scheme as the optimal scheduling scheme.

9. A storage medium storing a computer program for hybrid parallel genetic and variable neighborhood algorithm based production scheduling, wherein the computer program causes a computer to execute the hybrid parallel genetic and variable neighborhood algorithm based production scheduling method according to any one of claims 1 to 7.

10. An electronic device, comprising:

one or more processors;

a memory; and

one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the programs comprising instructions for performing the hybrid parallel genetic and variable neighborhood algorithm-based production scheduling method of any one of claims 1-7.

Technical Field

The invention relates to the technical field of production scheduling, in particular to a production scheduling method, a production scheduling system, a storage medium and electronic equipment based on a hybrid parallel inheritance and variable neighborhood algorithm.

Background

With the continuous development of new generation information technology, the manufacturing process of high-end equipment such as mobile phone chips, airplanes, large-scale machines and the like is continuously refined, the manufacturing of one high-end equipment generally comprises a plurality of different types of parts (workpieces), and the high-end equipment has a relatively complex manufacturing process. In a high-end equipment manufacturing system, a factory may produce various components and parts, and there are many plants, each of which has different resources. Furthermore, there is a deteriorating effect on the operation of the machine, i.e. in the actual production process, the machining time of the workpiece is influenced by the starting machining time, which is the later the longer the machining time is. Therefore, the enterprise cost can be reduced, the production efficiency can be improved, the competitiveness of the enterprise can be improved, and the resource utilization and the customer satisfaction can be maximized as far as possible.

At present, scheduling decision of various parts and configuration decision problem of resources are solved. Conventional scheduling algorithms include: the Variable neighborhood Search algorithm (VNS) is an algorithm that diversifies Search directions by changing a neighborhood structure, thereby enhancing Search capability and optimizing computational performance. Genetic Algorithm (GA) is a global search Algorithm that simulates biological processes in nature, and usually abstracts a problem space into a population, and generates a new generation of population by performing Genetic operator operations such as selection, crossing, and mutation on individual population. The genetic algorithm has the advantages of high efficiency, simplicity, robustness and the like on global parallel search, and is applied to various fields. However, in the face of a large population scale, the genetic algorithm needs to calculate and evaluate fitness of a large number of individuals, and the algorithm may run slowly; premature phenomena may also occur with common genetic problems.

In addition, the traditional scheduling algorithm is usually based on a specific production situation, so that a complex actual production process can deviate from a theory, a production scheduling decision and a resource allocation decision cannot be taken into consideration, a reasonable scheduling decision cannot be made for each main enterprise within a limited time, or decision support can be provided for decision makers of the enterprises, the overall capacity of a manufacturing system is improved, and the production cost is reduced.

Disclosure of Invention

Technical problem to be solved

Aiming at the defects of the prior art, the invention provides a production scheduling method, a system, a storage medium and electronic equipment based on a hybrid parallel inheritance and variable neighborhood algorithm, and solves the technical problem that a production scheduling decision and a resource allocation decision cannot be taken into consideration.

(II) technical scheme

In order to achieve the purpose, the invention is realized by the following technical scheme:

a production scheduling method based on a hybrid parallel inheritance and variable neighborhood algorithm comprises the following steps:

s1, initializing algorithm parameters, wherein the algorithm parameters at least comprise neighborhood searching times, neighborhood structure number, maximum genetic iteration times and maximum algorithm cycle iteration times;

s2, distributing and coding all workpieces to be processed, randomly generating an initialization group, and dividing the initialization group into a plurality of sub-groups; decoding by adopting a heuristic algorithm to obtain the processing sequence and the resource allocation scheme of each workshop workpiece corresponding to each individual in the initialized population, calculating the fitness value of each individual, and taking the individual with the highest fitness value as a global optimal solution;

s3, searching a new solution in the corresponding neighborhood structure according to the global optimal solution;

s4, judging whether the fitness value of the global optimal solution is smaller than that of the new solution, if so, updating the global optimal solution into the new solution, and turning to S3; otherwise, the global optimal solution is reserved, and the process goes to S5;

s5, judging whether a neighborhood search termination condition is met or not according to the neighborhood search times and the neighborhood structure number, if so, ending the variable neighborhood search, transferring the global optimal solution to each sub-population, and turning to S6; otherwise, go to S3;

s6, calculating the updated fitness value of the individuals in each sub-population;

s7, obtaining a next generation sub-population by adopting a preset selection operator, a preset crossover operator and a preset mutation operator according to the updated fitness value of the individuals in each sub-population; selecting an individual with the highest fitness value in the current group, and updating the global optimal solution according to the individual with the highest fitness value;

s8, judging whether the current genetic iteration number is smaller than the maximum genetic iteration number, if so, adding one to the current genetic iteration number, and switching to S6; otherwise, go to S9;

s9, judging whether the iteration number of the current algorithm loop is smaller than the maximum algorithm loop iteration number, if so, adding one to the iteration number of the current algorithm loop, and turning to S3; and if not, taking the processing sequence of the workpieces in each workshop corresponding to the global optimal solution and the resource allocation scheme as the optimal scheduling scheme.

Preferably, in S2, the obtaining, by decoding using a heuristic algorithm, the processing order and the resource allocation plan of each workshop workpiece corresponding to each individual in the initialized population includes:

s10, arranging the workpieces in each workshop according to the respective non-decreasing sequence of initial processing time, and enabling fs to be 1; the initial processing time refers to the processing time a' before the workpiece is not allocated with resources;

s20, allocating alpha resources to the fs workpiece in the permutation;

s30, if fs ^ alpha<R_jAnd R is_j-(fs+1)*α>If fs is 0, fs +1, the process proceeds to S20; otherwise, go to S40; wherein R is_jThe number of resources of the jth workshop is shown, j is 1, …, M, and M shows the number of workshops;

s40, if fs ^ alpha<R_jAnd R is_j-(fs+1)*α<0, then R is_j-fs +1 th workpiece in the permutation is assigned fs + α resources;

and S50, reversing the arrangement sequence of the workpieces in the workshop, determining the processing sequence of the workpieces, and acquiring the processing sequence of the workpieces in the current workshop and the resource allocation scheme.

Preferably, the number of the neighborhood structures comprises four different neighborhood structures, and the neighborhood structure N is defined_kK is 1 or 2 or 3 or 4;

if k is 1, correspondingly, S3 includes:

s311, defining the global optimal solution as V^*＝(v₁，…，v_i，…v_N) Wherein v is₁,…,v_i,…v_NIs any integer from 1 to M; v. of_iIndicating the corresponding workshop to which the ith workpiece is distributed; n represents the number of all workpieces to be processed; m represents the number of workshops;

s312, making q equal to 1, counting the total actual processing time of each workshop corresponding to the global optimal solution,

wherein n is_jIndicating the number of the processed workpieces in the jth workshop; p is a radical of_iIndicating the initial deterioration time of the ith workpiece when no resource is allocated; r is_iIndicating the amount of the workpiece resource distributed at the ith; the actual deterioration time of the ith workpiece in the s-th machining in the corresponding workshop is (p)_i-θr_i)s； The actual processing time of the ith part for processing the workpieces in the ith workshop in the s-th sequence is shown, and a' represents the initial processing time of each workpiece; θ is the efficiency of each resource;

obtaining the maximum total actual machining time of a workshop

S313, randomly generating 1 random integer rand from 1 to M₁；

S314, if rand₁＝j_max1If yes, the operation goes to S313; otherwise, go to S315;

s315, if v_q＝rand₁Then v is_q＝j_max1(ii) a If v is_q＝j_max1Then v is_q＝rand₁(ii) a Otherwise, go to S316;

s316, if q < N, q ═ q +1, and go to S315; otherwise, switching to S4 after obtaining a new solution;

or if k is 2, correspondingly, S3 includes:

s321, defining the global optimal solution to be represented as V^*＝(v₁,…,v_i,…v_N)；

S322, counting the total actual processing time P of each workshop corresponding to the global optimal solution_jObtaining the maximum total actual machining time of the workshopAnd obtaining the j_max2Initial deterioration time p of processed workpiece in each workshop_imax2Largest workpiece, p_imax2＝max{p_i|v_i＝j_max2}；

S323, randomly generating 1 random integer rand from 1 to M₂；

S324, if rand₂＝j_max2If yes, go to S323; otherwise, let v_imax＝rand₂And then the new solution is transferred to S4;

or if k is 3, correspondingly, S3 includes:

s331, defining the global optimal solution to be V^*＝(v₁，…，v_i,…v_N)；

S332, counting the total actual processing time P of each workshop corresponding to the global optimal solution_jObtaining the maximum total actual machining time of the workshopAnd obtaining the j_max3Workpiece p with largest initial deterioration time of processed workpiece in each workshop_imax3＝max{p_i|v_i＝j_max3}；

S333, randomly generating 1 random integer rand from 1 to N₃；

S334, ifThen the process proceeds to S333; otherwise, proceed to S335

S335, setting a new variable c, and orderv_imax3C, obtaining a new solution and then transferring to S4;

or if k is 4, correspondingly, S3 includes:

s341, defining the global optimal solution as V^*＝(v₁,…,v_i,…v_N)；

S342, randomly generating random integers a and b in a range from 1 to N, and setting variables x, y, z', and d, wherein x is 0;

s343, if a is equal to b, the process returns to S342; otherwise, go to S344;

s344, if a>b, let z be b, b be a, a be z; let d be b-a +1 and y beOtherwise, go to S345;

s345, let z ═ v_a+x，v_a+x＝v_b-x，v_b-x＝z′，x＝x+1；

S346, if x is less than y, turning to S345; otherwise, the process proceeds to S4 after a new solution is obtained.

Preferably, the S2 encodes all the workpieces to be processed by using an integer encoding method, and randomly generates an initialization population O, where the number of the individual workpieces in O is G × H, and the initialization population O is divided into H sub-populations O on average_hThen the number of individuals in each sub-population is G;

the function of the fitness value is:

wherein the content of the first and second substances,representing the g individual in the h sub-population,and C represents the maximum value of the actual completion time in the current population.

Preferably, the preset selection operator in S7 is a roulette mode, and the updated total number of individuals in each sub-population is G', where G ═ G + 1;

the updated selection process of each sub-population comprises the following steps:

s10a, setting a variable G ', and accumulating the probability when G ' is 1,2, … and G ' is calculated Value of (A), initial P (X)₀) 0, initial n is 1; wherein the content of the first and second substances,representing the probability of each individual being selected;representing the sum of the fitness of each sub-population;

s20a, randomly generating a random number sx1 from 0 to 1;

s30a, finding that P (X) is satisfied_g′-1)<sx1≤P(X_g′) G' selecting individualsAdding the next generation parent O'_h，n＝n+1；

S40a, if n is less than or equal to G, the process is switched to S20a, otherwise, the selection process is ended.

Preferably, the crossing process of each sub-population updated in S7 includes:

s10b, mixing the parent substance O'_hThe G individuals in the group are randomly divided into L pairing groups, andtwo sub-individuals in the first matched group are respectivelyAnd wherein the content of the first and second substances,the distribution workshop of the ith workpiece of the kth individual of the ith distribution group is shown, and k is 1 or 2; setting variables m, z₁、z₂Let m equal to 1;

s20b, randomly generating 2 random integers rand from 1 to N₄、rand₅；

S30b, if rand₄＝rand₅Then go to S20 b; otherwise, go to S40 b;

s40 b: order toOrder to

S50b, if m < L, then m is m +1, and the process proceeds to S20 b; otherwise, obtaining the next generation group and ending.

Preferably, the next generation population defining the acquisition of the crossover process is denoted O_h′， Represents the g 'th individual in the h' th sub-population in the next generation population, wherein,the value of (a) indicates the cell to which the ith workpiece is assigned;

the mutation process of each sub-population updated in S7 includes:

s10c, initializing g 'to 1, and setting a mutation probability p';

s20c, randomly generating a random number sx2 from 0 to 1, and switching to S30c if sx2< p'; otherwise, go to S60 c;

s30c, randomly generating a random integer rand from 1 to N₆；

S40c, randomly generating a random integer rand from 1 to M₇；

S50c, ifThen proceed to S40 c; otherwise, it ordersProceeding to S60 c;

s60c, if G' < G, G ═ G +1, proceed to S20 c; otherwise, ending after obtaining the next generation of sub-population.

A production scheduling system based on a hybrid parallel inheritance and variable neighborhood algorithm comprises:

the initialization module is used for executing S1 and initializing algorithm parameters, wherein the algorithm parameters at least comprise neighborhood searching times, neighborhood structure number, maximum genetic iteration times and maximum algorithm cycle iteration times;

the solving module is used for executing S2, distributing all the workpieces to be processed for coding, randomly generating an initialization group and dividing the initialization group into a plurality of sub-groups; decoding by adopting a heuristic algorithm to obtain the processing sequence and the resource allocation scheme of each workshop workpiece corresponding to each individual in the initialized population, calculating the fitness value of each individual, and taking the individual with the highest fitness value as a global optimal solution;

the searching module is used for executing S3 and searching a new solution in the corresponding neighborhood structure according to the global optimal solution;

the first judging module is used for executing S4 and judging whether the fitness value of the global optimal solution is smaller than that of the new solution, if so, the global optimal solution is updated to the new solution, and the searching module is switched to execute S3; otherwise, the global optimal solution is reserved, and the step is switched to a second judgment module to execute S5;

a second judging module, configured to execute S5, judge whether a neighborhood search termination condition is satisfied according to the number of neighborhood search times and the number of neighborhood structures, end the variable neighborhood search if satisfied, migrate the global optimal solution to each sub-population, and shift to the calculating module to execute S6; otherwise, the step is shifted to a searching module to execute S3;

the calculating module is used for executing S6 and calculating the updated fitness value of the individual in each sub-population;

the updating module is used for executing S7 and acquiring the next generation sub-population by adopting a preset selection operator, a preset crossover operator and a preset mutation operator according to the updated fitness value of the individual in each sub-population; selecting an individual with the highest fitness value in the current group, and updating the global optimal solution according to the individual with the highest fitness value;

a third judging module, configured to execute S8, judge whether the current genetic iteration number is smaller than the maximum genetic iteration number, if so, add one to the current genetic iteration number, and shift to the calculating module to execute S6; otherwise, the step is shifted to a fourth judging module to execute S9;

a fourth judging module, configured to execute S9, judge whether the current algorithm loop iteration number is smaller than the maximum algorithm loop iteration number, if yes, add one to the current algorithm loop iteration number, and shift to the searching module to execute S3; and if not, taking the processing sequence of the workpieces in each workshop corresponding to the global optimal solution and the resource allocation scheme as the optimal scheduling scheme.

A storage medium storing a computer program for hybrid parallel genetic and variable neighborhood algorithm based production scheduling, wherein the computer program causes a computer to execute the hybrid parallel genetic and variable neighborhood algorithm based production scheduling method as described above.

An electronic device, comprising:

one or more processors;

a memory; and

one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the programs comprising instructions for performing a hybrid parallel genetic and variable neighborhood algorithm based production scheduling method as described above.

(III) advantageous effects

The invention provides a production scheduling method, a production scheduling system, a production scheduling storage medium and electronic equipment based on a hybrid parallel inheritance and variable neighborhood algorithm. Compared with the prior art, the method has the following beneficial effects:

in the invention, a heuristic algorithm is adopted to decode and obtain each workshop workpiece production scheduling scheme corresponding to each individual in an initialized population, and the individual with the highest fitness value is taken as a global optimal solution; searching a new solution in the corresponding neighborhood structure; when the neighborhood search termination condition is met, migrating the updated global optimal solution to each sub-population; according to the updated fitness value of the individuals in each sub-population, adopting a preset selection operator, a preset crossover operator and a preset mutation operator to obtain a next sub-population; and selecting the individual with the highest fitness value in the current population, and updating the global optimal solution. An approximate optimal solution is found through iteration of a mixed coarse-grained parallel inheritance and variable neighborhood search optimization algorithm, the premature phenomenon of a single genetic algorithm is avoided, the diversity of population characteristics is kept, and meanwhile, the convergence degree of the algorithm is accelerated; the heuristic algorithm considers the efficiency improvement brought by the machine processing deterioration effect and the resource investment and considers the problems of production scheduling decision and resource allocation decision.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.

Fig. 1 is a schematic flowchart of a production scheduling method based on a hybrid parallel inheritance and variable neighborhood algorithm according to an embodiment of the present invention;

fig. 2 is another schematic flow chart of a production scheduling method based on a hybrid parallel inheritance and variable neighborhood algorithm according to an embodiment of the present invention;

fig. 3 is a block diagram of a production scheduling system based on a hybrid parallel inheritance and variable neighborhood algorithm according to an embodiment of the present invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention are clearly and completely described, and it is obvious that the described embodiments are a part of the embodiments of the present invention, but not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

The embodiment of the application solves the technical problem that a production scheduling decision and a resource allocation decision cannot be considered at the same time by providing a production scheduling method, a production scheduling system, a storage medium and electronic equipment based on a hybrid parallel inheritance and variable neighborhood algorithm.

In order to solve the technical problems, the general idea of the embodiment of the application is as follows:

in the embodiment of the invention, a heuristic algorithm is adopted to decode and obtain each workshop workpiece production scheduling scheme corresponding to each individual in an initialized population, and the individual with the highest fitness value is taken as a global optimal solution; searching a new solution in the corresponding neighborhood structure; when the neighborhood search termination condition is met, migrating the updated global optimal solution to each sub-population; according to the updated fitness value of the individuals in each sub-population, adopting a preset selection operator, a preset crossover operator and a preset mutation operator to obtain a next sub-population; and selecting the individual with the highest fitness value in the current population, and updating the global optimal solution. An approximate optimal solution is found through iteration of a mixed coarse-grained parallel inheritance and variable neighborhood search optimization algorithm, the premature phenomenon of a single genetic algorithm is avoided, the diversity of population characteristics is kept, and meanwhile, the convergence degree of the algorithm is accelerated; the heuristic algorithm considers the efficiency improvement brought by the machine processing deterioration effect and the resource investment and considers the problems of production scheduling decision and resource allocation decision.

Example (b):

in a first aspect, as shown in fig. 1, an embodiment of the present invention provides a production scheduling method based on a hybrid parallel inheritance and variable neighborhood algorithm, including:

s1, initializing algorithm parameters, wherein the algorithm parameters at least comprise neighborhood searching times, neighborhood structure number, maximum genetic iteration times and maximum algorithm cycle iteration times;

s2, distributing and coding all workpieces to be processed, randomly generating an initialization group, and dividing the initialization group into a plurality of sub-groups; decoding by adopting a heuristic algorithm to obtain the processing sequence and the resource allocation scheme of each workshop workpiece corresponding to each individual in the initialized population, calculating the fitness value of each individual, and taking the individual with the highest fitness value as a global optimal solution;

s3, searching a new solution in the corresponding neighborhood structure according to the global optimal solution;

s4, judging whether the fitness value of the global optimal solution is smaller than that of the new solution, if so, updating the global optimal solution into the new solution, and turning to S3; otherwise, the global optimal solution is reserved, and the process goes to S5;

s5, judging whether a neighborhood search termination condition is met or not according to the neighborhood search times and the neighborhood structure number, if so, ending the variable neighborhood search, transferring the global optimal solution to each sub-population, and turning to S6; otherwise, go to S3;

s6, calculating the updated fitness value of the individuals in each sub-population;

s7, obtaining a next generation sub-population by adopting a preset selection operator, a preset crossover operator and a preset mutation operator according to the updated fitness value of the individuals in each sub-population; selecting an individual with the highest fitness value in the current group, and updating the global optimal solution according to the individual with the highest fitness value;

s8, judging whether the current genetic iteration number is smaller than the maximum genetic iteration number, if so, adding one to the current genetic iteration number, and switching to S6; otherwise, go to S9;

s9, judging whether the iteration number of the current algorithm loop is smaller than the maximum algorithm loop iteration number, if so, adding one to the iteration number of the current algorithm loop, and turning to S3; and if not, taking the processing sequence of the workpieces in each workshop corresponding to the global optimal solution and the resource allocation scheme as the optimal scheduling scheme.

According to the embodiment of the invention, a corresponding heuristic algorithm, a neighborhood structure and a coarse-grained parallel genetic algorithm are designed according to the characteristics of a production mode of required part processing, a reasonable scheduling decision is made for each main enterprise within a limited time, or decision support is provided for decision makers of the enterprises, the overall capacity of a manufacturing system is improved, and the production cost is reduced. Specifically, an approximate optimal solution is found through iteration of a mixed coarse-grained parallel inheritance and variable neighborhood search optimization algorithm, the premature phenomenon of a single genetic algorithm is avoided, the diversity of population characteristics is kept, and meanwhile, the convergence degree of the algorithm is accelerated; the heuristic algorithm considers the efficiency improvement brought by the machine processing deterioration effect and the resource investment and considers the problems of production scheduling decision and resource allocation decision.

The following will describe each step of the above scheme in detail with reference to the specific content:

firstly, it should be clear that in the manufacture of high-end products according to the embodiments of the present invention, it is assumed that the number of all the workpieces to be processed is N, and each workpiece is produced in only one workshop; a total of M workshops, without considering the resource holding amount and resource allocation, each workshop has the same processing efficiency on the same workpiece in a specific processing sequence.

S1, initializing algorithm parameters, wherein the algorithm parameters at least comprise neighborhood searching times, neighborhood structure number, maximum genetic iteration times and maximum algorithm cycle iteration times.

As shown in fig. 2, the number of neighborhood search times T and the number of neighborhood structures are initializedMaximum number of genetic iterationsMaximum algorithm loop iteration number t_max。

S2, distributing and coding all workpieces to be processed, randomly generating an initialization group, and dividing the initialization group into a plurality of sub-groups; and decoding by adopting a heuristic algorithm to obtain the processing sequence and the resource allocation scheme of each workshop workpiece corresponding to each individual in the initialized population, calculating the fitness value of each individual, and taking the individual with the highest fitness value as a global optimal solution.

All workpieces to be processed are distributed and coded in an integer coding mode, and a solution vector represents V-V₁,v₂,v₃,...，v_N) Randomly generating an initialization population O, wherein the number of individuals in the initialization population O is G x H, and averagely dividing the initialization population O into H sub-populations O_hThen the number of individuals per sub-population is G.

In S2, a heuristic algorithm is used to decode and obtain the processing order and the resource allocation plan of each workshop workpiece corresponding to each individual in the initialized population, including:

s10, arranging the workpieces in each workshop according to the respective non-decreasing sequence of initial processing time, and enabling fs to be 1; the initial processing time refers to the processing time alpha' before the workpiece is not allocated with resources;

s20, allocating alpha resources to the fs workpiece in the permutation;

s30, if fs ^ alpha<R_jAnd R is_j-(fs+1)*α>If fs is 0, fs +1, the process proceeds to S20; otherwise, go to S40; wherein R is_jThe number of resources of the jth workshop is shown, j is 1, …, M, and M shows the number of workshops;

s40, if fs ^ alpha<R_jAnd R is_j-(fs+1)*α<0, then R is_j-fs +1 th workpiece in the permutation is assigned fs + α resources;

and S50, reversing the arrangement sequence of the workpieces in the workshop, determining the processing sequence of the workpieces, and acquiring the processing sequence of the workpieces in the current workshop and the resource allocation scheme.

The decoding process can be divided into two stages, in the first stage, each workpiece is distributed in each processing workshop, and the part processing sequence of each workshop is determined according to the initial processing time; in the second phase, the resources owned by each plant are allocated to the components. In the process, the deterioration effect of machining is mainly considered in the first stage, and the machining time of the workpiece machined later is prolonged; while in the second phase, resource allocation is primarily considered, the goal is to reduce production time by efficiently batching and sorting and resource allocation.

In the previous research, the processing time of the workpiece in the actual production process rarely considers the situation of resource allocation, so that some researched algorithms can deviate from the theory in the actual production process and cannot obtain better effect. Conventional scheduling algorithms are based on a specific production situation, and do not consider the deterioration effect of machining and the efficiency improvement caused by resource inclination, and the resources owned by each workshop of a factory are different.

The embodiment of the invention respectively considers the deterioration effect of machine processing and the efficiency improvement brought by resource investment based on the actual production condition, provides the heuristic algorithm to solve the corresponding problems, fits the actual production condition and considers the production scheduling decision and the resource allocation decision.

The function of the fitness value is:

wherein the content of the first and second substances,representing the g individual in the h sub-population,and C represents the maximum value of the actual completion time in the current population.

And S3, searching a new solution in the corresponding neighborhood structure according to the global optimal solution.

The number of the neighborhood structures in the implementation of the invention comprises four different neighborhood structures, and the neighborhood structure N is defined_kAnd k is 1 or 2 or 3 or 4.

If k is 1, the workpiece in the workshop and the workpiece in the other workshop are all exchanged until the condition is met. Correspondingly, the S3 includes:

s311, defining the global optimal solution as V^*＝(v₁,…,v_i,…v_N) Wherein v is₁,…,v_i,…v_NIs any integer from 1 to M; v. of_iIndicating the corresponding workshop to which the ith workpiece is distributed; n represents the number of all workpieces to be processed; m represents the number of workshops;

s312, making q equal to 1, counting the total actual processing time of each workshop corresponding to the global optimal solution,

wherein n is_jIndicates the number of the jth workshop processed workpieces；p_iIndicating the initial deterioration time of the ith workpiece when no resource is allocated; r is_iIndicating the amount of the workpiece resource distributed at the ith; the actual deterioration time of the ith workpiece in the s-th machining in the corresponding workshop is (p)_i-θr_i)s； The actual processing time of the ith part for processing the workpieces in the ith workshop in the s-th sequence is shown, and a' represents the initial processing time of each workpiece; θ is the efficiency of each resource;

obtaining the maximum total actual machining time of a workshop

S313, randomly generating 1 random integer rand from 1 to M₁；

S314, if rand₁＝j_max1If yes, the operation goes to S313; otherwise, go to S315;

s315, if v_q＝rand₁Then v is_q＝j_max1(ii) a If v is_q＝j_max1Then v is_q＝rand₁(ii) a Otherwise, go to S316;

s316, if q < N, q ═ q +1, and go to S315; otherwise, the process proceeds to S4 after a new solution is obtained.

Or if k is 2, specifically, the workpiece with the largest initial deterioration time in the workshop with the longest total actual machining time is transferred to other workshops. Correspondingly, the S3 includes:

s321, defining the global optimal solution to be represented as V^*＝(v₁,…,v_i,…v_N)；

S322, counting the total actual processing time P of each workshop corresponding to the global optimal solution_jObtaining the maximum total actual machining time of the workshopAnd obtaining the j_max2Initial deterioration time p of processed workpiece in each workshop_imax2Largest workpiece, p_imax2＝max{p_i|v_i＝j_max2}；

S323, randomly generating 1 random integer rand from 1 to M₂；

S324, if rand₂＝j_max2If yes, go to S323; otherwise, let v_imax＝rand₂And the new solution is obtained and then the process proceeds to S4.

Or if k is 3, specifically, the workpiece with the largest initial deterioration time in the longest total actual machining time workshop is exchanged with the workpieces in other workshops. Correspondingly, the S3 includes:

s331, defining the global optimal solution to be V^*＝(v₁,…,v_i,…v_N)；

S332, counting the total actual processing time P of each workshop corresponding to the global optimal solution_jObtaining the maximum total actual machining time of the workshopAnd obtaining the j_max3Workpiece p with largest initial deterioration time of processed workpiece in each workshop_imax3＝max{p_i|v_i＝j_max3}；

S333, randomly generating 1 random integer rand from 1 to N₃；

S334, ifThen the process proceeds to S333; otherwise, proceed to S335

S335, setting a new variable c, and orderv_imax3After a new solution is obtained, S4 is entered.

Or if k is 4, specifically code that reverses a portion of the globally optimal solution. Correspondingly, the S3 includes:

s341, defining the global optimal solution tableShown as V^*＝(v₁,…,v_i,…v_N)；

S342, randomly generating random integers a and b in a range from 1 to N, and setting variables x, y, z', and d, wherein x is 0;

s343, if a is equal to b, the process returns to S342; otherwise, go to S344;

s344, if a>b, let z be b, b be a, a be z; let d be b-a +1 and y beOtherwise, go to S345;

s345, let z ═ v_a+x，v_a+x＝v_b-x，v_b-x＝z′，x＝x+1；

S346, if x is less than y, turning to S345; otherwise, the process proceeds to S4 after a new solution is obtained.

S4, as shown in fig. 2, determining whether the fitness value of the global optimal solution is smaller than the fitness value of the new solution, if so, updating the global optimal solution to the new solution, and going to S3; otherwise, the global optimal solution is retained, and the process proceeds to S5.

S5, judging whether a neighborhood search termination condition is met or not according to the neighborhood search times and the neighborhood structure number, if so, ending the variable neighborhood search, transferring the global optimal solution to each sub-population, and turning to S6; otherwise, the process proceeds to S3.

As shown in fig. 2, the neighborhood search termination condition includes: judging whether the current neighborhood searching frequency it is smaller than the neighborhood searching frequency T, and if so, switching to the S3; if not, judging the number of the current neighborhood structuresWhether or not less than the number of neighborhood structuresIf yes, the process proceeds to step S3, and if not, the process proceeds to step S6.

And S6, calculating the fitness value of the individuals in each updated sub-population.

S7, obtaining a next generation sub-population by adopting a preset selection operator, a preset crossover operator and a preset mutation operator according to the updated fitness value of the individuals in each sub-population; and selecting the individual with the highest fitness value in the current population, and updating the global optimal solution according to the individual with the highest fitness value.

The coarse-grained parallel genetic algorithm adopted by the embodiment of the invention can keep the characteristics of each sub-population, ensures the characteristic diversity of the total population and can effectively inhibit the premature phenomenon. . By combining with the variable neighborhood algorithm, the global optimal solution can search for new solutions in a plurality of neighborhoods, the solved new solutions have higher fitness, and the convergence of the algorithm can be effectively improved.

The preset selection operator in S7 is a roulette mode, and the total number of individuals in each updated sub-population is G', where G ═ G + 1.

The updated selection process of each sub-population comprises the following steps:

s10a, setting a variable G ', and accumulating the probability when G ' is 1,2, … and G ' is calculated Value of (A), initial P (X)₀) 0, initial n is 1; wherein the content of the first and second substances,representing the probability of each individual being selected;representing the sum of the fitness of each sub-population;

s20a, randomly generating a random number sx1 from 0 to 1;

s30a, finding that P (X) is satisfied_g′-1)<sx1≤P(X_g′) G' selecting individualsAdding the next generation parent O'_h，n＝n+1；

S40a, if n is less than or equal to G, the process is switched to S20a, otherwise, the selection process is ended.

The intersection process of each updated sub-population in S7 includes:

s10b, mixing the parent substance O'_hThe G individuals in the group are randomly divided into L pairing groups, andtwo sub-individuals in the first matched group are respectivelyAnd wherein the content of the first and second substances,the distribution workshop of the ith workpiece of the kth individual of the ith distribution group is shown, and k is 1 or 2; setting variables m, z₁、z₂Let m equal to 1;

s20b, randomly generating 2 random integers rand from 1 to N₄、rand₅；

S30b, if rand₄＝rand₅Then go to S20 b; otherwise, go to S40 b;

s40 b: order toOrder to

S50b, if m < L, then m is m +1, and the process proceeds to S20 b; otherwise, obtaining the next generation group and ending.

Defining the next generation population obtained by the crossover process is denoted O_h′， Represents the g 'th individual in the h' th sub-population in the next generation population, wherein,the value of (a) indicates the cell to which the ith workpiece was assigned from the g' th individual.

The mutation process of each sub-population updated in S7 includes:

s10c, initializing g 'to 1, and setting a mutation probability p';

s20c, randomly generating a random number sx2 from 0 to 1, and switching to S30c if sx2< p'; otherwise, go to S60 c;

s30c, randomly generating a random integer rand from 1 to N₆；

S40c, randomly generating a random integer rand from 1 to M₇；

S50c, ifThen proceed to S40 c; otherwise, it ordersProceeding to S60 c;

s60c, if G' < G, G ═ G +1, proceed to S20 c; otherwise, ending after obtaining the next generation of sub-population.

The embodiment of the invention designs an optimization algorithm of a coarse grain parallel genetic algorithm and a variable neighborhood search algorithm aiming at the scheduling problem of production and resource allocation of workpieces, firstly determines the allocation of the processed workpieces through coding, then determines the processing sequence and the resource amount of each workshop workpiece according to the scheduling algorithm, and finally finds out an approximate optimal solution through iteration of the coarse grain parallel genetic algorithm and the variable neighborhood search algorithm. The premature phenomenon of the genetic algorithm is avoided, the diversity of population characteristics is kept, and meanwhile, the convergence degree of the algorithm is accelerated.

In a second aspect, as shown in fig. 3, an embodiment of the present invention provides a production scheduling system based on a hybrid parallel inheritance and variable neighborhood algorithm, including:

the initialization module is used for executing S1 and initializing algorithm parameters, wherein the algorithm parameters at least comprise neighborhood searching times, neighborhood structure number, maximum genetic iteration times and maximum algorithm cycle iteration times;

the solving module is used for executing S2, distributing all the workpieces to be processed for coding, randomly generating an initialization group and dividing the initialization group into a plurality of sub-groups; decoding by adopting a heuristic algorithm to obtain the processing sequence and the resource allocation scheme of each workshop workpiece corresponding to each individual in the initialized population, calculating the fitness value of each individual, and taking the individual with the highest fitness value as a global optimal solution;

the searching module is used for executing S3 and searching a new solution in the corresponding neighborhood structure according to the global optimal solution;

the first judging module is used for executing S4 and judging whether the fitness value of the global optimal solution is smaller than that of the new solution, if so, the global optimal solution is updated to the new solution, and the searching module is switched to execute S3; otherwise, the global optimal solution is reserved, and the step is switched to a second judgment module to execute S5;

a second judging module, configured to execute S5, judge whether a neighborhood search termination condition is satisfied according to the number of neighborhood search times and the number of neighborhood structures, end the variable neighborhood search if satisfied, migrate the global optimal solution to each sub-population, and shift to the calculating module to execute S6; otherwise, the step is shifted to a searching module to execute S3;

the calculating module is used for executing S6 and calculating the updated fitness value of the individual in each sub-population;

the updating module is used for executing S7 and acquiring the next generation sub-population by adopting a preset selection operator, a preset crossover operator and a preset mutation operator according to the updated fitness value of the individual in each sub-population; selecting an individual with the highest fitness value in the current group, and updating the global optimal solution according to the individual with the highest fitness value;

a third judging module, configured to execute S8, judge whether the current genetic iteration number is smaller than the maximum genetic iteration number, if so, add one to the current genetic iteration number, and shift to the calculating module to execute S6; otherwise, the step is shifted to a fourth judging module to execute S9;

a fourth judging module, configured to execute S9, judge whether the current algorithm loop iteration number is smaller than the maximum algorithm loop iteration number, if yes, add one to the current algorithm loop iteration number, and shift to the searching module to execute S3; and if not, taking the processing sequence of the workpieces in each workshop corresponding to the global optimal solution and the resource allocation scheme as the optimal scheduling scheme.

In a third aspect, an embodiment of the present invention provides a storage medium storing a computer program for hybrid parallel genetic and variable neighborhood algorithm-based production scheduling, wherein the computer program causes a computer to execute the hybrid parallel genetic and variable neighborhood algorithm-based production scheduling method as described above.

In a fourth aspect, an embodiment of the present invention provides an electronic device, including:

one or more processors;

a memory; and

one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the programs comprising instructions for performing a hybrid parallel genetic and variable neighborhood algorithm based production scheduling method as described above.

It should be understood that the production scheduling system, the storage medium, and the electronic device based on the hybrid parallel inheritance and variable neighborhood algorithm provided in the embodiments of the present invention correspond to the production scheduling method based on the hybrid parallel inheritance and variable neighborhood algorithm, and the explanation, examples, and beneficial effects of the relevant contents thereof may refer to the corresponding contents in the production scheduling method based on the hybrid parallel inheritance and variable neighborhood algorithm, and are not described herein again.

In summary, compared with the prior art, the method has the following beneficial effects:

in the embodiment of the invention, a heuristic algorithm is adopted to decode and obtain each workshop workpiece production scheduling scheme corresponding to each individual in an initialized population, and the individual with the highest fitness value is taken as a global optimal solution; searching a new solution in the corresponding neighborhood structure; when the neighborhood search termination condition is met, migrating the updated global optimal solution to each sub-population; according to the updated fitness value of the individuals in each sub-population, adopting a preset selection operator, a preset crossover operator and a preset mutation operator to obtain a next sub-population; and selecting the individual with the highest fitness value in the current population, and updating the global optimal solution. An approximate optimal solution is found through iteration of a mixed coarse-grained parallel inheritance and variable neighborhood search optimization algorithm, the premature phenomenon of a single genetic algorithm is avoided, the diversity of population characteristics is kept, and meanwhile, the convergence degree of the algorithm is accelerated; the heuristic algorithm considers the efficiency improvement brought by the machine processing deterioration effect and the resource investment and considers the problems of production scheduling decision and resource allocation decision.

It is noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.

The above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.