阅读说明：本技术 一种融合ais数据挖掘的港口水域规划方法及系统 (Port water area planning method and system integrating AIS data mining ) 是由 沈忱 齐越 王达川 房卓 方森松 杨靓 贾鹏鹏 杨琼 薛天寒 于 2021-06-23 设计创作，主要内容包括：本发明公开了一种融合AIS数据挖掘的港口水域规划方法及系统,包括以下步骤：步骤S1、构建用于港口水域规划量化仿真分析的港口水域样本矩阵；步骤S2、获得均衡港口水域规划量化仿真规划精度和规划效率且实现统一分析的港口水域规划集群；步骤S3、基于所述港口水域规划集群对所述港口水域的通行规划进行量化仿真分析。本发明将各个连通排布复杂的港口水域依照层次聚类算法将通行特征相似的港口水域划分为可进行统一规划的港口水域规划集群,实现对规划规划精度和规划效率的最佳均衡。(The invention discloses a port water area planning method and a system integrating AIS data mining, comprising the following steps of: s1, constructing a port water area sample matrix for quantitative simulation analysis of port water area planning; step S2, obtaining the quantified simulation planning precision and planning efficiency of the balanced port water area planning and realizing a port water area planning cluster of unified analysis; and step S3, carrying out quantitative simulation analysis on the passing plan of the port water area based on the port water area planning cluster. According to the invention, the port water areas with similar traffic characteristics are divided into port water area planning clusters capable of carrying out unified planning by all the port water areas with complicated communication arrangement according to a hierarchical clustering algorithm, so that the optimal balance of planning precision and planning efficiency is realized.)

1. A harbor water area planning method integrated with AIS data mining is characterized by comprising the following steps:

step S1, constructing a port water area sample matrix for quantitative simulation analysis of port water area planning based on all port water areas in the planning area and the passage logs of the port water areas;

step S2, cluster analysis is carried out on the port water area sample matrix by utilizing a multi-target cluster analysis algorithm so as to obtain the port water area planning cluster which is used for balancing the quantitative simulation planning precision and the planning efficiency of the port water area planning and realizing the unified analysis;

and step S3, carrying out quantitative simulation analysis on the passing plan of the port water area based on the port water area planning cluster so as to accurately and efficiently determine the passing plan of the port water area.

2. The method for planning the water area of the port integrated with the AIS data mining as claimed in claim 1, wherein the method comprises the following steps: in the step S1, the specific method for constructing the port water area sample matrix includes:

coding and quantizing each passing master station into matrix elements x of a port water area sample matrix_iAnd quantizing the traffic log of each traffic master station into element characteristic quantity [ y ] of matrix elements_j]；

Combining said matrix elements x_iAnd the element feature quantity [ y_j]Constructing and obtaining the port water area sample matrix as

Wherein i belongs to (1, N), j belongs to (1, M), N is the total number of harbor water areas, and M is the total number of element characteristics.

3. The method for planning the water area of the port integrated with the AIS data mining as claimed in claim 2, wherein the AIS data mining method comprises the following steps: in the step S1, the method further includes converting the port water area sample matrix a into a similarity adjacency matrix U, including:

calculating each element and the rest elements in the port water area sample matrix A in turnSimilarity and generating a similarity matrixThe similarity calculation formula is as follows:

similarity threshold based on presettingCarrying out binarization conversion on the similarity matrix to generate a similarity adjacent matrixThe binarization conversion formula is as follows:

wherein i₁、i₂Belongs to (1, N), j belongs to (1, M), N is the total number of harbor water areas, M is the total category of element characteristics,respectively is the ith of the port water area sample matrix A₁、i₂The j-th class of feature elements of an individual matrix element,is the ith of similarity adjacency matrix U₁Line i₂The matrix elements of the columns are arranged in a matrix,is the ith of the similarity matrix₁Line i₂Matrix elements of the columns.

4. The method for planning the water area of the port integrated with the AIS data mining as claimed in claim 3, wherein the AIS data mining method comprises the following steps: in step S2, a multi-objective division algorithm for performing element cluster division on the port water area sample matrix is constructed by using the port water area planning quantization simulation planning precision and the planning efficiency, and includes:

mathematically quantifying the quantified simulation planning precision of the port water area planning to generate a precision benefit target, wherein the precision benefit target formula is as follows:

mathematically quantifying the efficiency of the harbor water area planning quantification simulation planning to generate an efficiency benefit target, wherein the efficiency benefit target formula is as follows:

and constructing a multi-target function of the multi-target division algorithm by combining the precision benefit target and the efficiency benefit target according to a minimization solving principle, wherein the multi-target function is as follows:

wherein, C_kMatrix element sets of the k-th port water area planning cluster divided in the port water area sample matrix A are provided, respectively is the ith of the port water area sample matrix A₁、i₂And m is the total number of planning clusters of the port water area with the optimal quality after the port water area sample matrix A is divided.

5. The method for planning the water area of the port integrated with the AIS data mining as claimed in claim 4, wherein the AIS data mining method comprises the following steps: the specific method for cluster division of the port water area sample matrix A by utilizing the multi-target division algorithm comprises the following steps:

step S201: all matrix elements in the port water area sample matrix A are quantized into a single port water area planning cluster respectively, and iteration times are set;

step S202: calculating the fitness function of the two-to-two fusion of the port water area planning cluster and the rest port water area planning clusters in sequence, and reserving the two single-port water area planning clusters with the maximum fitness function for fusion and normalization to realize fusion iteration;

step S203: if the fusion iteration times are more than or equal to the set iteration times, selecting the port water area planning cluster in the step S202 as the optimal port water area planning cluster; if the fusion iteration number is less than the set iteration number, the process returns to step S202.

6. The method for planning the water area of the port integrated with the AIS data mining of claim 5, wherein the AIS data mining method comprises the following steps: the specific method constructed according to the fitness function comprises the following steps:

fitness function f is obtained based on linear calibration of target function₄＝-f₃+ δ, where δ is a perturbation constant;

sequentially calculating the fitness value of the fusion of the planning clusters of the two port water areas and selecting the fitness value f₄And merging the two highest port water area planning clusters.

7. The method for planning a harbor water area with integrated AIS data mining as claimed in claim 6, wherein the step S3 is performed by performing a quantitative simulation analysis on the passage plan of the harbor water area based on the harbor water area planning cluster, comprising:

s301, sequentially selecting the port water area with the narrowest drainage basin in each port water area planning cluster, and simulating dynamic flow of ships in the port water area with the narrowest drainage basin according to a real traffic log of the port water area with the narrowest drainage basin so as to count traffic flow of the port water area with the narrowest drainage basin in unit time and take the traffic flow as the maximum load-bearing traffic flow of all the port water areas in each port water area cluster;

s302, sequentially selecting the port water area with the widest drainage basin in each port water area planning cluster, and simulating dynamic flow of ships in the port water area with the widest drainage basin according to the port water area real traffic log with the widest drainage basin to count traffic flow of the port water area with the widest drainage basin in unit time to serve as the minimum just-needed traffic flow of all the port water areas in each port water area cluster;

and S303, setting planned traffic flow in unit time of all port water areas in each port water area cluster based on the minimum just-needed traffic flow and the maximum bearing traffic flow so as to ensure that traffic efficiency and traffic safety requirements are met.

8. The method for planning a water area of a port integrated with AIS data mining according to claim 7, wherein in step S303, the specific method for planning traffic flow setting includes:

if the minimum just-needed traffic flow is lower than the maximum bearing traffic flow, the planning traffic flow is set to be lower than the maximum bearing traffic flow and higher than the minimum just-needed traffic flow;

and if the minimum just-needed traffic flow is higher than the maximum bearing traffic flow, setting the planned traffic flow to be lower than the maximum bearing traffic flow.

9. An analysis system for a harbor water area planning method fused with AIS data mining according to any one of claims 1-8, comprising:

the system comprises a sample processing unit (1) and a control unit, wherein the sample processing unit is used for constructing a port water area sample matrix for planning, quantifying and simulating and analyzing a port water area based on all port water areas in a planning area and the passage logs of the port water areas;

the cluster division unit (2) is used for carrying out cluster analysis on the port water area sample matrix by utilizing a multi-target cluster analysis algorithm so as to obtain a port water area planning cluster which is used for balancing the quantitative simulation planning precision and the planning efficiency of the port water area planning and realizing uniform analysis;

and the passage planning unit (3) is used for carrying out quantitative simulation analysis on the passage planning of the port water area based on the port water area planning cluster so as to accurately and efficiently determine the passage planning of the port water area.

10. The analysis system of claim 9, wherein the traffic planning unit comprises a ship flow simulation component for fusing AIS data mining and simulating dynamic flow of ships over the port water area according to a real traffic log, a ship flow statistics component for counting the number of ships dynamically flowing over the port water area, and a traffic planning component for determining a traffic plan for the port water area according to the number of ships.

Technical Field

The invention relates to the technical field of port planning, in particular to a port water area planning method and system integrating AIS data mining.

Background

The port system composed of ships, channels, anchor sites, berths and the like has randomness, complexity and uncertainty, dynamic change and real-time interaction of multi-factor information. The port water area planning is to plan and design many factors such as anchor land, channel, convolute water area, and whether the design is scientific and reasonable for the port with large traffic and complex arrival ships will directly affect the safety of ships entering and leaving the port and the port operation efficiency.

At present, aiming at the research of the quantitative simulation analysis of harbor waters, a specific or individual harbor area is mostly taken as a prototype, the simulation strategy based on discrete events is mainly adopted to carry out the quantitative analysis of harbor water planning, and the input parameters of the model are generally determined by the general design specifications of harbors or empirical data. The port water area planning quantitative simulation analysis method generally regards a port ship navigation operation system as a discrete event service system, provides service for permanent elements in the system, namely an anchor ground, a channel, a berth and a loading and unloading system, and provides service objects for ships. The anchor ground provides a waiting place for the ship to sail and park; the channel provides a passage for the ship to enter and exit the port; the berth provides a site for loading and unloading operations for the ship. The research method needs a large amount of prior knowledge, and port water area planning analysis is carried out on the data only according to manual experiences such as threshold values, so that planning precision and safety cannot be guaranteed.

Disclosure of Invention

The invention aims to provide a port water area planning method and system integrating AIS data mining, and aims to solve the technical problems that a large amount of prior knowledge is needed in the prior art, port water area planning analysis is carried out on data only according to manual experiences such as a threshold value and the like, and planning accuracy and safety cannot be guaranteed.

In order to solve the technical problems, the invention specifically provides the following technical scheme:

a port water area planning method integrating AIS data mining comprises the following steps:

step S1, constructing a port water area sample matrix for quantitative simulation analysis of port water area planning based on all port water areas in the planning area and the passage logs of the port water areas;

step S2, cluster analysis is carried out on the port water area sample matrix by utilizing a multi-target cluster analysis algorithm so as to obtain the port water area planning cluster which is used for balancing the quantitative simulation planning precision and the planning efficiency of the port water area planning and realizing the unified analysis;

and step S3, carrying out quantitative simulation analysis on the passing plan of the port water area based on the port water area planning cluster so as to accurately and efficiently determine the passing plan of the port water area.

As a preferred embodiment of the present invention, in the step S1, the specific method for constructing the port water area sample matrix includes:

coding and quantizing each passing master station into matrix elements x of a port water area sample matrix_iAnd quantizing the traffic log of each traffic master station into element characteristic quantity [ y ] of matrix elements_j]；

Combining said matrix elements x_iAnd the element feature quantity [ y_j]Constructing and obtaining the port water area sample matrix as

Wherein i belongs to (1, N), j belongs to (1, M), N is the total number of harbor water areas, and M is the total number of element characteristics.

As a preferred embodiment of the present invention, the step S1 further includes converting the port water area sample matrix a into a similarity adjacency matrix U, including:

sequentially calculating the similarity of each element and the rest elements in the port water area sample matrix A, and generating a similarity matrixThe similarity calculation formula is as follows:

similarity threshold based on presettingCarrying out binarization conversion on the similarity matrix to generate a similarity adjacent matrixThe binarization conversion formula is as follows:

wherein i₁、i₂Belongs to (1, N), j belongs to (1, M), N is the total number of harbor water areas, M is the total category of element characteristics,respectively is the ith of the port water area sample matrix A₁、i₂The j-th class of feature elements of an individual matrix element,is the ith of similarity adjacency matrix U₁Line i₂The matrix elements of the columns are arranged in a matrix,is degree of similarityIth of matrix₁Line i₂Matrix elements of the columns.

As a preferred embodiment of the present invention, in step S2, the method for constructing a multi-objective division algorithm for performing element cluster division on a port water area sample matrix by using the quantified simulation planning precision and planning efficiency of the port water area planning includes:

mathematically quantifying the quantified simulation planning precision of the port water area planning to generate a precision benefit target, wherein the precision benefit target formula is as follows:

mathematically quantifying the efficiency of the harbor water area planning quantification simulation planning to generate an efficiency benefit target, wherein the efficiency benefit target formula is as follows:

and constructing a multi-target function of the multi-target division algorithm by combining the precision benefit target and the efficiency benefit target according to a minimization solving principle, wherein the multi-target function is as follows:

wherein, C_kMatrix element sets of the k-th port water area planning cluster divided in the port water area sample matrix A are provided,i₁、i₂∈(1,N)，respectively is the ith of the port water area sample matrix A₁、i₂And m is the total number of planning clusters of the port water area with the optimal quality after the port water area sample matrix A is divided.

As a preferred scheme of the present invention, the specific method for cluster division of the port water area sample matrix a by using a multi-objective division algorithm comprises:

step S201: all matrix elements in the port water area sample matrix A are quantized into a single port water area planning cluster respectively, and iteration times are set;

step S202: calculating the fitness function of the two-to-two fusion of the port water area planning cluster and the rest port water area planning clusters in sequence, and reserving the two single-port water area planning clusters with the maximum fitness function for fusion and normalization to realize fusion iteration;

step S203: if the fusion iteration times are more than or equal to the set iteration times, selecting the port water area planning cluster in the step S202 as the optimal port water area planning cluster; if the fusion iteration number is less than the set iteration number, the process returns to step S202.

As a preferred scheme of the present invention, the specific method constructed according to the fitness function includes:

fitness function f is obtained based on linear calibration of target function₄＝-f₃+ δ, where δ is a perturbation constant;

sequentially calculating the fitness value of the fusion of the planning clusters of the two port water areas and selecting the fitness value f₄And merging the two highest port water area planning clusters.

As a preferred aspect of the present invention, in the step S3, the specific method for performing quantitative simulation analysis on the traffic plan of the port water area based on the port water area planning cluster includes:

s301, sequentially selecting the port water area with the narrowest drainage basin in each port water area planning cluster, and simulating dynamic flow of ships in the port water area with the narrowest drainage basin according to a real traffic log of the port water area with the narrowest drainage basin so as to count traffic flow of the port water area with the narrowest drainage basin in unit time and take the traffic flow as the maximum load-bearing traffic flow of all the port water areas in each port water area cluster;

s302, sequentially selecting the port water area with the widest drainage basin in each port water area planning cluster, and simulating dynamic flow of ships in the port water area with the widest drainage basin according to the port water area real traffic log with the widest drainage basin to count traffic flow of the port water area with the widest drainage basin in unit time to serve as the minimum just-needed traffic flow of all the port water areas in each port water area cluster;

and S303, setting planned traffic flow in unit time of all port water areas in each port water area cluster based on the minimum just-needed traffic flow and the maximum bearing traffic flow so as to ensure that traffic efficiency and traffic safety requirements are met.

As a preferable scheme of the present invention, in step S303, a specific method for planning traffic flow setting includes:

if the minimum just-needed traffic flow is lower than the maximum bearing traffic flow, the planning traffic flow is set to be lower than the maximum bearing traffic flow and higher than the minimum just-needed traffic flow;

and if the minimum just-needed traffic flow is higher than the maximum bearing traffic flow, setting the planned traffic flow to be lower than the maximum bearing traffic flow.

As a preferred aspect of the present invention, the present invention provides an analysis system for a harbor water area planning method based on the integrated AIS data mining, comprising:

the system comprises a sample processing unit, a port water area planning and simulation analysis unit and a port water area planning and simulation analysis unit, wherein the sample processing unit is used for constructing a port water area sample matrix for the quantitative simulation analysis of the port water area planning based on all port water areas in a planning area and the passage logs of the port water areas;

the cluster division unit is used for carrying out cluster analysis on the port water area sample matrix by utilizing a multi-target cluster analysis algorithm so as to obtain a port water area planning cluster which is used for balancing the quantitative simulation planning precision and the planning efficiency of the port water area planning and realizing uniform analysis;

and the passage planning unit is used for carrying out quantitative simulation analysis on the passage planning of the port water area based on the port water area planning cluster so as to accurately and efficiently determine the passage planning of the port water area.

As a preferred aspect of the present invention, the transit planning unit includes a ship flow simulation component, a ship flow statistics component and a transit planning component, the ship flow simulation component is configured to merge AIS data mining and simulate dynamic movement of a ship on the port water area according to a real transit log, the ship flow statistics component is configured to count the number of ships dynamically flowing on the port water area, and the transit planning component is configured to determine a transit plan of the port water area according to the number of ships.

Compared with the prior art, the invention has the following beneficial effects:

according to the invention, the port water areas with similar traffic characteristics are divided into the port water area planning clusters capable of carrying out unified planning by all the port water areas with complicated communication and arrangement according to the hierarchical clustering algorithm, so that the optimal balance of planning precision and planning efficiency is realized, the port water area planning integrated with AIS data mining is finally ensured to be separated from artificial experience or threshold value comparison, the planning result is more reliable, and the prior knowledge is avoided being relied on in the process of dividing the port water area clusters, so that the calculation complexity is reduced.

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 should be apparent that the drawings in the following description are merely exemplary, and that other embodiments can be derived from the drawings provided by those of ordinary skill in the art without inventive effort.

FIG. 1 is a flow chart of a method for planning a water area of a port according to an embodiment of the present invention;

FIG. 2 is a schematic view of a water area structure of a port according to an embodiment of the present invention;

fig. 3 is a schematic diagram of a water area planning cluster structure of a port according to an embodiment of the present invention;

fig. 4 is a block diagram of an analysis system according to an embodiment of the present invention.

The reference numerals in the drawings denote the following, respectively:

1-a sample processing unit; 2-cluster dividing unit; 3-pass planning unit.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and 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.

As shown in fig. 1-4, the invention provides a port water area planning method integrating AIS data mining, comprising the following steps:

step S1, constructing a port water area sample matrix for quantitative simulation analysis of port water area planning based on all port water areas in the planning area and the passage logs of the port water areas;

in step S1, the specific method for constructing the port water area sample matrix includes:

coding and quantizing each passing master station into matrix elements x of a port water area sample matrix_iAnd quantizing the traffic log of each traffic master station into element characteristic quantity [ y ] of matrix elements_j]；

Combining matrix elements x_iAnd element characteristic quantity [ y_j]Constructing and obtaining a port water area sample matrix as

Wherein i belongs to (1, N), j belongs to (1, M), N is the total number of harbor water areas, and M is the total number of element characteristics.

Elemental feature quantity [ y_j]＝[y₁,y₂,…,y_M]Wherein [ y_j]Is a tabulated summary of various characteristics of port waters, such as: the coordinates, anchorage, channel, berth, loading and unloading system, etc. of the port water areas are determined according to the actual conditions during the specific implementation, and the characteristic quantities are used for dividing the similarity of all scattered port water areas into clusters for uniform planning.

In step S1, the method further includes converting the port water area sample matrix a into a similarity adjacency matrix U, including:

sequentially calculating the similarity of each element and the rest elements in the port water area sample matrix A and generatingSimilarity matrixThe similarity calculation formula is as follows:

similarity threshold based on presettingThe similarity matrix is subjected to binarization conversion to generate a similarity adjacent matrixThe binarization conversion formula is as follows:

wherein i₁、i₂Belongs to (1, N), j belongs to (1, M), N is the total number of harbor water areas, M is the total category of element characteristics,respectively is the ith of the port water area sample matrix A₁、i₂The j-th class of feature elements of an individual matrix element,is the ith of similarity adjacency matrix U₁Line i₂The matrix elements of the columns are arranged in a matrix,is the ith of the similarity matrix₁Line i₂Matrix elements of the columns.

The port water areas divided into the same port water area planning cluster have the same type of passing functional modules, the same type of planning analysis can be adopted, all the port water areas belonging to the same cluster have linkage attributes, for example, once the passing state of a certain port water area is in an abnormal state, the other port water areas affected by the abnormal state can be obtained through association tracing in the cluster to which the port water area belongs, the influence on the other port water areas can be avoided by performing association tracing on the passing plan of any port water area, the comprehensive consideration on all the port water areas can be realized by performing unified passing planning on the same port water area planning cluster, the discrete independent planning on the port water areas is avoided, and the planning efficiency and the planning precision can be quickly improved.

The element characteristic quantity is the characteristic quantitative extraction of a passage log, the passage log comprises passage data of port water areas, the element characteristic quantity is the extraction of the passage data and is used for calculating the similarity of the passage data among the port water areas, the correlation characteristics among the port water areas can be known, and finally a similarity adjacent matrix among the port water areas is constructed, for example: u. of₁₂1, then, the matrix element x characterizing the port water area is denoted₁And x₂There is a correlation between, i.e. the matrix element x characterizing the port water area₁Matrix element x tracing to characteristic port water area in association with abnormal occurrence₂，u₁₂0 denotes the matrix element x characterizing the port water area₁And x₂There is no correlation between them, i.e. the matrix element x characterizing the port water area₁Matrix element x for abnormity and characterization of harbor water area₂There is no effect.

It can be understood that the pass planning of the port water area based on the cluster analysis of the embodiment makes full use of the passing relevance of the port water areas of the port water area, and the cutting analysis is not performed on the port water area singly, so that the complexity, uncertainty and ambiguity of the port water area are ignored, and the analysis accuracy and rationality are finally improved.

Step S2, carrying out cluster analysis on the port water area sample matrix by using a multi-target cluster analysis algorithm to obtain a port water area planning cluster which is used for balancing the quantitative simulation planning precision and the planning efficiency of the port water area planning and realizing uniform analysis;

in step S2, a multi-objective division algorithm for performing element cluster division on the port water area sample matrix is constructed using the port water area planning quantization simulation planning precision and the planning efficiency, including:

the precision of the mathematical quantitative port water area planning quantitative simulation planning generates a precision benefit target, and the precision benefit target formula is as follows:

the efficiency of the mathematical quantitative port water area planning quantitative simulation planning generates an efficiency benefit target, and the efficiency benefit target formula is as follows:

and constructing a multi-target function of the multi-target division algorithm by combining the precision benefit target and the efficiency benefit target by utilizing a minimization solving principle, wherein the multi-target function is as follows:

wherein, C_kMatrix element sets of the k-th port water area planning cluster divided in the port water area sample matrix A are provided,i₁、i₂∈(1,N)，respectively is the ith of the port water area sample matrix A₁、i₂And m is the total number of planning clusters of the port water area with the optimal quality after the port water area sample matrix A is divided.

It will be appreciated that f is utilized₁And f₂As an objective function of cluster division, f₁The larger the cluster size is, the larger the average degree of the cluster is, namely the closer the connection relation of the nodes in the cluster is, the larger the number of the clusters is, and the smaller the cluster size is generally; f. of₂The smaller the cluster size, the smaller the average externality of the cluster, that is, the more sparse the node connection relationship among the clusters, and the smaller the number of clusters, the larger the cluster size is generally. The two complementary items embody two aspects of the high-quality cluster division, and need to be in addition to the twoTaking a trade-off point.

To f₁Performing a minimization correction, which can be converted into a multi-objective optimization, -f₁The reduction in (f) will cause the network to tend to be divided into many small clusters of very high self-compactness, and f₂The reduction in (b) will tend to divide the network into a few large clusters that are sparsely connected to other parts of the network. Thus, the two complementary terms reflect two fundamental aspects of a good cluster partitioning, and the modularity density is an essential compromise between the two. f. of₂And-f₁All tend to balance the tendency of each other to want to increase or decrease the number of clusters, and all are related to modularity density, which can overcome the problem of resolution limitation, so f is set₂And-f₁As an objective function of cluster division, the hierarchical cohesion property of the network is utilized to avoid relying on prior knowledge, so that the real-time acquisition of various prior knowledge is reduced, and the pressure of data transmission, operation and storage is finally reduced.

Using f₂And-f₁The cluster can be controlled to be in the optimal scale as a target function of cluster division, the highest precision of the cluster division is guaranteed, the optimal scale guarantees the highest efficiency of the port water area traffic planning, the phenomenon that the time of the port water area traffic planning is too long due to the fact that the super-large scale occurs, the final analysis efficiency is affected is avoided, the highest precision of the cluster division guarantees the highest precision of the port water area traffic planning, and the phenomenon that the analysis error of the port water area traffic planning is affected due to the fact that the cluster is divided incorrectly is avoided.

The specific method for cluster division of the port water area sample matrix A by utilizing the multi-target division algorithm comprises the following steps:

step S201: all matrix elements in the port water area sample matrix A are quantized into a single port water area planning cluster respectively, and iteration times are set;

step S202: calculating the fitness function of the two-to-two fusion of the port water area planning cluster and the rest port water area planning clusters in sequence, and reserving the two single-port water area planning clusters with the maximum fitness function for fusion and normalization to realize fusion iteration;

step S203: if the fusion iteration times are more than or equal to the set iteration times, selecting the port water area planning cluster in the step S202 as the optimal port water area planning cluster; if the fusion iteration number is less than the set iteration number, the process returns to step S202.

The concrete method constructed according to the fitness function comprises the following steps:

fitness function f is obtained based on linear calibration of target function₄＝-f₃+ δ, where δ is a perturbation constant;

sequentially calculating the fitness value of the fusion of the planning clusters of the two port water areas and selecting the fitness value f₄And merging the two highest port water area planning clusters.

It can be understood that the accuracy benefit and the efficiency benefit can be ensured by selecting the analysis clusters with high fitness values for fusion, so that a more accurate and more comprehensive cluster structure can be obtained.

And step S3, carrying out quantitative simulation analysis on the passage plan of the port water area based on the port water area planning cluster so as to accurately and efficiently determine the passage plan of the port water area.

In step S3, the specific method for performing quantitative simulation analysis on the passage plan of the port water area based on the port water area planning cluster includes:

s301, sequentially selecting the port water area with the narrowest drainage basin in each port water area planning cluster, and simulating dynamic flow of ships in the port water area with the narrowest drainage basin according to a real traffic log of the port water area with the narrowest drainage basin so as to count traffic flow in unit time of the port water area with the narrowest drainage basin as the maximum load-bearing traffic flow of all the port water areas in each port water area cluster;

s302, sequentially selecting the port water area with the widest drainage basin in each port water area planning cluster, and simulating dynamic flow of ships in the port water area with the widest drainage basin according to the port water area real traffic log with the widest drainage basin so as to count traffic flow in unit time of the port water area with the widest drainage basin and take the traffic flow as the minimum just-needed traffic flow of all the port water areas in each port water area cluster;

and S303, setting planned traffic flow in unit time of all port water areas in each port water area cluster based on the minimum just-needed traffic flow and the maximum bearing traffic flow so as to ensure that traffic efficiency and traffic safety requirements are met.

In step S303, the specific method for planning traffic flow setting includes:

if the minimum just-needed traffic flow is lower than the maximum bearing traffic flow, the planning traffic flow is set to be lower than the maximum bearing traffic flow and higher than the minimum just-needed traffic flow;

and if the minimum just-needed traffic flow is higher than the maximum bearing traffic flow, setting the planned traffic flow to be lower than the maximum bearing traffic flow.

The steps realize a principle method for unified planning in the port water area planning cluster so as to ensure that the circulation requirements of the ships in all the port water areas in the whole port water area planning cluster are met and the traffic jam of the narrowest port water area is avoided, wherein if the circulation requirements of the ships are greater than the bearing traffic requirements of the narrowest port water area, the safety is taken as priority, and the planned traffic flow is still kept lower than the maximum bearing traffic flow.

As shown in fig. 2-3, for example, the port water area planning clusters of the port water area sample matrix are 1, 2 and 3 clusters, the port water area planning cluster 1 includes port water areas 1-6, the port water area planning cluster 2 includes port water areas 7-10, and the port water area planning cluster 3 includes port water areas 12-14, wherein the port water areas in the port water area planning clusters 1, 2 and 3 are subjected to traffic planning by a uniform planning and analyzing method, and response analysis is performed in a cluster form, so as to rapidly improve analysis efficiency.

And counting the narrowest port water area and the widest port water area in the port water area planning clusters 1, 2 and 3, and performing ship traffic simulation to obtain the minimum just-needed traffic flow and the maximum bearing traffic flow of the port water area planning clusters 1, 2 and 3 as the determination reference data of the planned traffic flow.

As shown in fig. 4, the present invention provides an analysis system for a port water area planning method based on the integrated AIS data mining, comprising:

the system comprises a sample processing unit 1, a port water area planning and simulation analysis unit and a port water area planning and simulation analysis unit, wherein the sample processing unit is used for constructing a port water area sample matrix for the quantitative simulation analysis of the port water area planning based on all port water areas in a planning area and traffic logs of the port water areas;

the cluster division unit 2 is used for carrying out cluster analysis on the port water area sample matrix by utilizing a multi-target cluster analysis algorithm so as to obtain a port water area planning cluster which is used for balancing the quantitative simulation planning precision and the planning efficiency of the port water area planning and realizing uniform analysis;

and the passage planning unit 3 is used for carrying out quantitative simulation analysis on the passage planning of the port water area based on the port water area planning cluster so as to accurately and efficiently determine the passage planning of the port water area.

The traffic planning unit comprises a ship flow simulation component, a ship flow counting component and a traffic planning component, wherein the ship flow simulation component is used for fusing AIS data mining and simulating dynamic flow of ships on a port water area according to a real traffic log, the ship flow counting component is used for counting the number of the ships which dynamically flow on the port water area, and the traffic planning component is used for determining the traffic planning of the port water area according to the number of the ships.

According to the invention, the port water areas with similar traffic characteristics are divided into the port water area planning clusters capable of carrying out unified planning by all the port water areas with complicated communication and arrangement according to the hierarchical clustering algorithm, so that the optimal balance of planning precision and planning efficiency is realized, the port water area planning integrated with AIS data mining is finally ensured to be separated from artificial experience or threshold value comparison, the planning result is more reliable, and the prior knowledge is avoided being relied on in the process of dividing the port water area clusters, so that the calculation complexity is reduced.

The above embodiments are only exemplary embodiments of the present application, and are not intended to limit the present application, and the protection scope of the present application is defined by the claims. Various modifications and equivalents may be made by those skilled in the art within the spirit and scope of the present application and such modifications and equivalents should also be considered to be within the scope of the present application.