阅读说明：本技术 一种基于改进混沌引力搜索算法的活性污泥过程辨识方法 (Activated sludge process identification method based on improved chaotic gravity search algorithm ) 是由 李俊红 宋伟成 袁银龙 蒋一哲 李政 芮佳丽 蒋泽宇 于 2021-07-19 设计创作，主要内容包括：本发明提供了一种基于改进混沌引力搜索算法的活性污泥过程辨识方法,属于污水处理系统辨识技术领域。其技术方案为：一种基于改进混沌引力搜索算法的活性污泥过程辨识方法,所述具体包括以下步骤：步骤1)建立活性污泥过程的多输入单输出模型；步骤2)构建改进混沌引力搜索算法的辨识流程。本发明的有益效果为：本发明提出的改进混沌引力搜索算法是一种群智能优化算法,它有较快的收敛速度和较高的收敛精度,能较好的适用于对活性污泥过程的建模和参数辨识。(The invention provides an activated sludge process identification method based on an improved chaotic gravity search algorithm, and belongs to the technical field of sewage treatment system identification. The technical scheme is as follows: an activated sludge process identification method based on an improved chaotic gravity search algorithm specifically comprises the following steps: step 1) establishing a multi-input single-output model of an activated sludge process; and 2) constructing an identification process for improving the chaotic gravity search algorithm. The invention has the beneficial effects that: the improved chaotic gravitation search algorithm provided by the invention is a group intelligent optimization algorithm, has higher convergence speed and higher convergence precision, and can be better suitable for modeling and parameter identification of an activated sludge process.)

1. An activated sludge process identification method based on an improved chaotic gravity search algorithm is characterized by comprising the following steps:

step 1) establishing a multi-input single-output model of an activated sludge process;

and 2) constructing an identification process for improving the chaotic gravity search algorithm.

2. The method for identifying the activated sludge process based on the improved chaotic gravity search algorithm according to claim 1, wherein the modeling step of the step 1) is as follows:

step 1-1) constructing a multi-input single-output model of the activated sludge process: in equation (1), y (t) is the output of the system, w (t) is colored noise, x (t) is the non-interfering output of the system,

y(t)＝x(t)+w(t), (1)

wherein x (t) and w (t) are respectively represented as:

w(t)＝D(z)v(t),

A(z)，B_i(z) and D (z) are with respect to the postshift operator z^-1The polynomial of (c):

step 1-2) obtaining output y (t) and input u according to formulas (2) and (3)_i(t), intermediate variablesA relation between the non-interfering output x (t) and the error v (t), wherein,

for non-linear partIs defined as follows:

whereinThe parameter vector of the non-linear part, equation (2) can be re-expressed as:

3. the method for identifying the activated sludge process based on the improved chaotic gravity search algorithm according to claim 1, wherein the model in the step 1) is a model of a multi-input single-output system.

4. The method for identifying the activated sludge process based on the improved chaotic gravity search algorithm according to claim 1, wherein the specific steps of the step 2) constructing the identification process of the improved chaotic gravity search algorithm are as follows:

step 2-1) initializing a population to generate a population with N particles;

step 2-2) adding Chemical Oxygen Demand (COD) and ammonia Nitrogen (NH)₃-N) and sludge reflux ratio as input data for an activated sludge process model, water produced by the secondary sedimentation tank as output data;

step 2-3) defining a fitness function fit (theta) as:

wherein the content of the first and second substances,is the estimated value of the output, and y (t) is the actual value of the output;

step 2-4) recording the maximum fitness value as f_w(t), the minimum fitness value is denoted as f_b(t)；

Step 2-5) calculating the intermediate mass m according to the formulas (8) and (9)_i(t) mass of particle M_i(t)；

Step 2-6) calculating the attraction between two particles according to the formulas (10) and (11)Sum of external forces F to which the particles are subjected_i ^d(t)；

Wherein M is_i(t) and M_j(t) masses of particles i and j, respectively, R_ijIs the Euclidean distance between the two particles, ξ is a very small constant, k is a constant that makes no sense in order to avoid the situation where the Euclidean distance is 0_bRepresenting a set of K individuals before the maximum inertial mass is reached or before the optimal fitness function value is obtained, g (t) is the universal gravitation constant:

k_b＝(N-δ)×(L-h)/L+δ×z_h, (13)

z_h+1＝λ×z_h×(1-z_h), (14)

in the formula (12), H represents the maximum number of iterations, β represents the number of iterations at that time, and G₀Represents the initial value of the universal gravitation coefficient, wherein epsilon is a constant; in formulae (13) and (14), z_hIs a random number between 0 and 1, δ being the percentage of force applied by one particle to the other, λ being a normal number;

steps 2-7) according to formula (15)) Calculating acceleration of particles

Step 2-8) updating the particle speed according to the formulas (16) and (17)And position

Wherein rand_iIs [0,1 ]]A random number in between, alpha is [0,1 ]]Constant between;

step 2-9) judging whether the maximum iteration times are reached, if not, skipping the program to step 2-3), and if so, entering step 2-10);

and 2-10) outputting a result to finish identification.

Technical Field

The invention relates to the technical field of sewage treatment system identification, in particular to an activated sludge process identification method based on an improved chaotic gravity search algorithm.

Background

With the development of society, the problem of environmental pollution is more and more serious, so that the quality of water is also seriously influenced. Water is an indispensable resource for people's daily life. In order to better analyze and predict the activated sludge process, a corresponding system model needs to be established for the activated sludge process, and parameters of the established model are identified. For this reason, many researchers have proposed different identification methods, such as: random gradient algorithm, least square algorithm, particle swarm algorithm and the like.

The random gradient algorithm has low identification precision and low convergence speed, so that the identification effect in actual production is poor; the least square algorithm has the problem of data saturation caused by the fact that the data quantity is increased in the process of tracking the time-varying parameters; although the particle swarm algorithm serving as the swarm intelligence algorithm can be well applied to different working conditions, the problems of local optimization and large calculation amount are also caused.

How to solve the above technical problems is the subject of the present invention.

Disclosure of Invention

The invention aims to provide an activated sludge process identification method based on an improved chaotic gravitation search algorithm, and the improved chaotic gravitation search algorithm provided by the invention is a group intelligent optimization algorithm, has higher convergence speed and higher convergence precision, and can be better suitable for modeling and parameter identification of an activated sludge process.

The invention is realized by the following measures: the method specifically comprises the following steps:

step 1) establishing a multi-input single-output model of the activated sludge process.

And 2) constructing an identification process for improving the chaotic gravity search algorithm.

As a further optimization scheme of the activated sludge process identification method based on the improved chaotic gravity search algorithm, the specific modeling step of the step 1) is as follows:

step 1-1) constructing a multi-input single-output model of the activated sludge process: in equation (1), y (t) is the output of the system, w (t) is colored noise, x (t) is the non-interfering output of the system,

y(t)＝x(t)+w(t), (1)

wherein x (t) and w (t) are respectively represented as:

w(t)＝D(z)v(t),

A(z)，B_i(z) and D (z) are with respect to the postshift operator z^-1The polynomial of (c):

step 1-2) obtaining output y (t) and input u according to formulas (2) and (3)_i(t), intermediate variablesA relation between the non-interfering output x (t) and the error v (t), wherein,

for non-linear partIs defined as follows:

whereinThe parameter vector of the non-linear part, equation (2) can be re-expressed as:

as a further optimization scheme of the activated sludge process identification method based on the improved chaotic gravity search algorithm, the model in the step 1) is a model of a general multi-input single-output system.

As a further optimization scheme of the activated sludge process identification method based on the improved chaotic gravitation search algorithm, the step 2) of establishing the identification process of the improved chaotic gravitation search algorithm comprises the following specific steps:

step 2-1) initializing a population to generate a population with N particles;

step 2-2) adding Chemical Oxygen Demand (COD) and ammonia Nitrogen (NH)₃-N) and sludge reflux ratio as input data for an activated sludge process model, water produced by the secondary sedimentation tank as output data;

step 2-3) defining a fitness function fit (theta) as:

wherein the content of the first and second substances,is the estimated value of the output, and y (t) is the actual value of the output.

Step 2-4) recording the maximum fitness value as f_w(t), the minimum fitness value is denoted as f_b(t)；

Step 2-5) calculating the intermediate mass m according to the formulas (8) and (9)_i(t) mass of particle M_i(t)；

Step 2-6) calculating the attraction between two particles according to the formulas (10) and (11)Sum of external forces to which the particles are subjected

Wherein M is_i(t) and M_j(t) masses of particles i and j, respectively, R_ijIs the Euclidean distance between the two particles, ξ is a very small constant, k is a constant that makes no sense in order to avoid the situation where the Euclidean distance is 0_bExpressed is a group containing the maximum inertial mass before reaching the maximum inertial mass or before obtaining the optimal fitness function valueA set of K individuals, g (t) is the universal gravitation constant:

k_b＝(N-ζ)×(L-h)/L+ζ×z_h， (13)

z_h+1＝λ×z_h×(1-z_h)， (14)

in the formula (12), H represents the maximum number of iterations, β represents the number of iterations at that time, and G₀Represents the initial value of the universal gravitation coefficient, and eta is a constant. In formulae (13) and (14), z_hIs a random number between 0 and 1, where ζ is the percentage of force applied by one particle to the other and λ is a normal number.

Step 2-7) calculating the acceleration of the particle according to the formula (15)

Step 2-8) updating the particle speed according to the formulas (16) and (17)And position

Wherein rand_iIs [0,1 ]]A random number therebetween, alpha is [ [ alpha ] ]0，1]Constant between.

Step 2-9) judging whether the maximum iteration times are reached, if not, skipping the program to step 2-3), and if so, entering step 2-10);

and 2-10) outputting a result to finish identification.

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

(1) the invention establishes a model for identifying the process parameters of the activated sludge, and uses Chemical Oxygen Demand (COD) and ammonia Nitrogen (NH)₃-N) and the sludge reflux ratio are used as input data, the model parameters are identified by using an improved chaotic gravity search algorithm, and as can be seen from figure 4, the model parameters can be well identified by the algorithm.

(2) Compared with a chaotic gravitation search algorithm and a gravitation search algorithm, the chaotic gravitation search algorithm is improved, the position of the particles is updated, the search range is increased, the convergence speed is improved, the improved chaotic gravitation search algorithm can better identify a nonlinear system, the identification precision is higher, and the obtained estimation error is smaller; meanwhile, the identification method has better applicability to the activated sludge process.

Drawings

The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the principles of the invention and not to limit the invention.

Fig. 1 is an overall flow chart of the activated sludge process identification method based on the improved chaotic gravity search algorithm provided by the invention.

Fig. 2 is a schematic view of an activated sludge process of the activated sludge process identification method based on the improved chaotic gravity search algorithm provided by the invention.

Fig. 3 is a general model schematic diagram of a multi-input single-output system of the activated sludge process identification method for improving the chaotic gravity search algorithm provided by the invention.

FIG. 4 is a schematic diagram of the error between the identification parameter and the true value according to the present invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. Of course, the specific embodiments described herein are merely illustrative of the invention and are not intended to be limiting.

Example 1

Referring to fig. 1 to 4, the technical scheme provided by the invention is that an activated sludge process identification method based on an improved chaotic gravity search algorithm comprises the following specific steps:

step 1) establishing a multi-input single-output model of an activated sludge process;

and 2) constructing an identification process for improving the chaotic gravity search algorithm.

Further preferably, the specific modeling steps of step 1) are as follows:

step 1-1) constructing a multi-input single-output model of the activated sludge process: in equation (1), y (t) is the output of the system, w (t) is colored noise, x (t) is the non-interfering output of the system,

y(t)＝x(t)+w(t)， (1)

wherein x (t) and w (t) are respectively represented as:

w(t)＝D(z)v(t)，

A(z)，B_i(z) and D (z) are with respect to the postshift operator z^-1The polynomial of (c):

step 1-2) obtaining output y (t) and input u according to formulas (2) and (3)_i(t), intermediate variablesA relation between the non-interfering output x (t) and the error v (t), wherein,

for non-linear partIs defined as follows:

whereinThe parameter vector of the non-linear part, equation (2) can be re-expressed as:

further preferably, the model of step 1) is a model of a general multi-input single-output system.

Further preferentially, the specific steps of constructing the identification process of the improved chaotic gravity search algorithm in the step 2) are as follows:

step 2-1) initializing a population to generate a population with N particles;

step 2-2) adding Chemical Oxygen Demand (COD) and ammonia Nitrogen (NH)₃-N) and the sludge reflux ratio are used as input data of an activated sludge process model, water generated by the secondary sedimentation tank is used as output data, and the data are recorded;

step 2-3) defining a fitness function fit (theta) as:

wherein the content of the first and second substances,is the estimated value of the output, and y (t) is the actual value of the output.

Step 2-4) recording the maximum fitness value as f_w(t), the minimum fitness value is denoted as f_b(t)；

Step 2-5) calculating the intermediate mass m according to the formulas (8) and (9)_i(t) mass of particle M_i(t)；

Step 2-6) calculating the attraction between two particles according to the formulas (10) and (11)Sum of external forces to which the particles are subjected

Wherein M is_i(t) and M_j(t) masses of particles i and j, respectively, R_ijIs the Euclidean distance between the two particles, ξ is a very small constant, k is a constant that makes no sense in order to avoid the situation where the Euclidean distance is 0_bRepresenting a set of K individuals before the maximum inertial mass is reached or before the optimal fitness function value is obtained, g (t) is the universal gravitation constant:

k_b＝(N-ζ)×(L-h)/L+ζ×z_h， (13)

z_h+1＝λ×z_h×(1-z_h)， (14)

in the formula (12), H represents the maximum number of iterations, β represents the number of iterations at that time, and G₀Represents the initial value of the universal gravitation coefficient, and eta is a constant. In formulae (13) and (14), z_hIs a random number between 0 and 1, where ζ is the percentage of force applied by one particle to the other and λ is a normal number.

Step 2-7) calculating the acceleration of the particle according to the formula (15)

Step 2-8) updating the particle speed according to the formulas (16) and (17)And position

Wherein rand_iIs [0,1 ]]A random number in between, alpha is [0,1 ]]Constant between.

Step 2-9) judging whether the maximum iteration times are reached, if not, skipping the program to step 2-3), and if so, entering step 2-10);

and 2-10) outputting a result to finish identification.

The schematic diagram of the activated sludge process used in this example is shown in FIG. 2. Wherein u is₁(t)、u₂(t) and u₃(t) respectively being sewage, activated sludge, air, y₁(t) is treated water.

With the above mentioned general multiple input single output model, the following model can be established for this embodiment:

y(t)＝x(t)+w(t)

comparing the above model with step 1), it is possible to obtain

a₁＝0.73，a₂＝-0.22，b₁₁＝0.4，b₁₂＝0.27，b₂₁＝0.52，b₂₂＝0.26，b₃₁＝0.78，b₃₂＝0.30，

d₁＝0.73，d₂＝0.42，γ₁₁＝0.37，γ₁₂＝0.95，γ₂₁＝0.80，γ₂₂＝0.64，γ₃₁＝0.70，γ₃₂＝0.52.

A fitness function, fit, is determined for the above model for use in the modified chaotic gravity search algorithm, the fitness function being defined as follows:

in the formula (I), the compound is shown in the specification,is the estimated value of the output, and y (t) is the actual value of the output.

In order to improve the chaos gravitation search algorithm on behalf of the parameters to be identified, the parameters to be identified are formed into a parameter vector theta, and the parameters to be identified are as follows:

θ＝[a₁，a₂，b₁₁，b₁₂，b₂₁，b₂₂，b₃₁，b₃₂，d₁，d₂，γ₁₁，γ₁₂，γ₂₁，γ₂₂，γ₃₁，γ₃₂]

initializing a population according to step 2-1);

obtaining input and output data of the activated sludge process model according to the step 2-2);

obtaining a fitness function value fit (theta) according to the step 2-3);

recording the maximum fitness value as f according to the step 2-4)_w(t), will be the mostSmall fitness value is noted as f_b(t)；

Calculating the intermediate mass m according to step 2-5)_i(t) mass of particle M_i(t)；

Calculating the attractive force between two particles according to the step 2-6)Sum of external forces to which the particles are subjected

Calculating the acceleration of the particles according to step 2-7)

Updating the velocity of the particles according to step 2-8)And positionNamely updating the estimated value of the parameter vector theta;

and completing circulation according to the steps 2-9) and the steps 2-10) and outputting a result.

Wherein, an initial value G of the gravitational constant is set₀And the number of populations N need to be considered for several reasons: the initial value of the gravity constant is too small, so that the position change of the population individual is small after each iteration, and finally the convergence speed is slow. The initial value of the gravity constant is too large, so that the position change of the population individual is large after each iteration, an optimal value cannot be found, and finally the identification result cannot be converged; the population number is too small, which causes the problem of poor population optimization effect and low identification precision. The problem of large calculation amount is caused by too large population number.

The parameter identification result of the activated sludge process identification method based on the improved chaotic gravitation search algorithm is shown in fig. 4, and it can be seen that the identification precision of the method is high, the estimated value of the parameter to be identified is very close to the true value, and meanwhile, the identification method has good applicability to an activated sludge process model.

The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.