CN111523218B - Multi-target parameter optimization method based on dynamic multi-target evolution - Google Patents

Abstract

The embodiment of the application discloses a multi-objective parameter optimization method and system based on dynamic multi-objective evolution, an electronic terminal and a storage medium. According to the method, accurate prediction of the center points of the plurality of sub-areas of the Pareto solution set in the new environment is achieved through the center points of the plurality of sub-areas of the Pareto solution set at the current moment; and other individuals are randomly generated in a possible solution set range at the next moment by adopting a multi-region diversity maintaining strategy, so that population diversity is increased. The initial population at the next moment is obtained by adopting algorithm optimization, so that the initial population in a new environment is more similar to a real Pareto solution set, and the aim of accelerating the convergence speed of the algorithm is fulfilled.

Description

A Multi-objective Parameter Optimization Method Based on Dynamic Multi-objective Evolution

technical field

The invention relates to the field of computer technology, in particular to a multi-objective parameter optimization method based on dynamic multi-objective evolution, a multi-objective parameter optimization device based on dynamic multi-objective evolution, a storage medium and electronic equipment.

Background technique

In the engineering industry and other fields, such as metallurgy, aviation and other industries, when solving engineering problems or scientific research problems, the solution is expected to be optimal. However, when faced with a technical problem to be solved, the technical problem may consist of multiple objectives. For example, for the technical problem of dynamic path planning, when planning the path, it is generally necessary to consider the goals of the shortest distance, the shortest time, and good road conditions; while in the process of vehicle travel, traffic accidents, traffic control, etc. occur randomly. This requires the optimization algorithm to adjust the optimization results according to real-time information, and perform dynamic online optimization to optimize the driving route.

In view of the above situation, dynamic multi-objective optimization evolutionary algorithm can be used to solve it. For example, most of the existing dynamic multi-objective optimization evolutionary algorithms are designed for discrete time variables, or some static multi-objective optimization algorithms are directly used to solve DMOP (Dynamic Multi-objective Optimization Problem, dynamic multi-objective optimization problem). However, for DMOP, because it has multiple time-dependent (environment) conflicting goals, and its Pareto optimal solution will change with time, the optimization of DMOP is more difficult.

As for the dynamic optimization problem, it is mainly divided into the following three types: 1) Maintain the diversity of the population: such as the random migration evolution strategy proposed by Grefenstette, the hypervariation method proposed by Morrison, and the diversity maintenance strategy proposed by GanRuan et al. An effective method to improve population diversity; 2) Memory-based method: For dynamic evolutionary algorithms, add better solutions obtained in history, and restart these solutions when needed to use them for evolution, in the case of environmental changes This will greatly improve the efficiency and search ability of the algorithm for solving problems. Memory is usually divided into two types: implicit memory utilizing redundant representations and explicit memory stored by introducing additional memory sets. For example, the use of additional diploid implicit memory method proposed by Ryan, the gene hierarchical structure memory method proposed by Collins, etc. Although the above-mentioned various implicit memory methods can enable the evolutionary algorithm to indirectly store some effective information, it is not sure whether the algorithm can effectively use this information.

Contents of the invention

In order to solve the above technical problems, the embodiments of the present invention expect to provide a dynamic multi-objective evolution-based multi-objective parameter optimization method, a dynamic multi-objective evolution-based multi-objective parameter optimization device, storage media and electronic equipment.

The technical solution of the present invention is achieved as follows: according to the first aspect, a multi-objective parameter optimization method based on dynamic multi-objective evolution is provided, including:

Obtain the Pareto solution set of the object to be optimized at time t and the Pareto solution set at time t-1, so as to obtain the Pareto solution set at time t according to the Pareto solution set at time t and the center point position of each sub-region in the Pareto solution set at time t-1. Evolution step length; where Pareto solution set includes multiple sub-regions; t≥1;

According to the evolution step at the time t and the position of the center point of each sub-region in the Pareto solution set at the time t, obtain a predicted solution of the center point of each sub-region at time t+1; and

According to the maximum point and minimum point of each of the sub-regions at the time t, use a preset random function to obtain the random value of each of the sub-regions at time t+1; wherein, the number of random values of each of the sub-regions is related to the number of raw solutions in this subregion;

Generate the initial population at the time t+1 based on the predicted solution of the central point of each sub-region at the time t+1, and the random value of each sub-region at the time t+1, so as to utilize the time t+1 The initial population of is to obtain the multi-objective parameter optimization result at the time t+1.

According to the second aspect, a multi-objective parameter optimization system based on dynamic multi-objective evolution is provided, including:

The evolution step calculation module is used to obtain the Pareto solution set of the object to be optimized at time t and the Pareto solution set of t-1 time, so as to obtain the Pareto solution set of the object at time t and the Pareto solution set of t-1 time for each sub-region The position of the center point of the center point obtains the evolution step at time t; where the Pareto solution set includes multiple sub-regions; t≥1;

The central point prediction solution acquisition module is used to obtain the position of the central point of each sub-region at t+1 time according to the evolution step size at the time t and the position of the central point of each sub-region in the Pareto solution set at the time t. predictive solutions; and

The random value calculation module is used to obtain the random value of each sub-area at time t+1 according to the maximum point and minimum point of each sub-area at time t, using a preset random function; wherein, each of the sub-areas The number of random values of is related to the number of original solutions in this sub-region;

An optimization result generating module, configured to generate the initial population at the time t+1 based on the predicted solution of the central point of each sub-region at the time t+1, and the random value of each sub-region at the time t+1, The multi-objective parameter optimization result at time t+1 is obtained by using the initial population at time t+1.

According to a third aspect, there is provided a storage medium on which a computer program is stored, and when the program is executed by a processor, the multi-objective parameter optimization method based on dynamic multi-objective evolution according to the above-mentioned embodiments is implemented.

According to a fourth aspect, an electronic terminal is provided, including:

processor; and

a memory for storing executable instructions of the processor;

Wherein, the processor is configured to execute the multi-objective parameter optimization method based on dynamic multi-objective evolution according to the above-mentioned embodiments by executing the executable instructions.

The embodiment of the present invention provides a multi-objective parameter optimization method based on dynamic multi-objective evolution. By using the center points of multiple sub-regions of the Pareto solution at the current moment, the accurate calculation of the center points of the multiple sub-regions of the Pareto solution in the new environment is realized. Prediction: Using a multi-regional diversity maintenance strategy to randomly generate other individuals in the range of possible solution sets at the next moment, increasing the diversity of the population. The algorithm optimization is used to obtain the initial population at the next moment, so that the initial population in the new environment is closer to the real Pareto solution set, and the purpose of speeding up the convergence of the algorithm is achieved.

Description of drawings

Figure 1 shows a schematic diagram of a multi-region multi-center point prediction strategy;

Figure 2 shows a schematic diagram of the multi-region multi-center point prediction strategy in different situations when there is no individual sub-region;

Fig. 3 shows a schematic diagram of a multi-regional diversity maintenance strategy;

Figure 4a and Figure 4b respectively illustrate the change trend of the reverse generation distance evaluation index (IGD) of the test functions DF2 and HE2;

Figure 5a and Figure 5b respectively show the comparison diagrams of the real frontier and the obtained frontier of DF2 and Fun9;

Fig. 6 shows a schematic flow chart of a multi-objective parameter optimization method based on dynamic multi-objective evolution;

Fig. 7 shows a schematic diagram of the composition of a multi-objective parameter optimization device based on dynamic multi-objective evolution;

Fig. 8 schematically shows a block diagram of an electronic device;

Fig. 9 schematically shows a schematic diagram of a program product.

Detailed ways

The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention.

At present, whether it is an engineering problem or a scientific research problem, the solution is hoped to be optimal, so as to achieve the purpose of improving efficiency and reducing energy consumption. However, most optimization problems are composed of multiple objectives, and the solutions of this series of objective functions are often contradictory and mutually reinforcing. In some engineering practices, it even includes various uncertain or dynamic factors, making the problem more complicated. For example, in dynamic route planning, in the planning scheme, the optimization goal generally considers the shortest distance, the shortest time, and good road conditions, etc., but in the process of vehicle travel, traffic accidents, traffic control and other situations occur randomly, which requires optimization. The algorithm adjusts the optimization results based on real-time information, and performs dynamic online optimization to optimize the driving route. This kind of dynamic multi-objective optimization problem exists in various fields, and finally this kind of problem can be abstracted into the following mathematical model (take minimization as an example):

Among them, F(x,t) is the objective function, t∈[t ₀ ,t _s ] is the time variable, x=(x ₁ ,x ₂ ,···,x _n )∈R ⁿ is the n-dimensional decision vector, a _i b _i is the upper and lower limits of the i-th dimension of the decision variable x, i=1,2,...,n.

In the dynamic multi-objective optimization algorithm, the key lies in how the algorithm responds to changes in the objective function and the optimal Pareto solution set brought about by environmental changes when the environment changes. In order to solve the problems existing in the prior art, a multi-objective parameter optimization method based on dynamic multi-objective evolution that can realize multi-region co-evolution is provided. Specifically, it will be described with reference to the following examples.

The basic framework of the dynamic multi-objective evolutionary algorithm corresponding to the method of the present invention may include:

1 Set the time parameter t=0;

2 Initialize the population P ₀ , calculate the objective function value F ₀ , define a set of weight vectors λ(t)=λ ¹ (t), λ ² (t),...,λ ^N(t) (t);

3 while termination instruction == 0do

4 ifChange()then

5 Set t=t+1;

6 Use the multi-region central point strategy to generate _PMRP ;

7 Utilize multi-regional diversity maintenance strategies to generate _PRDM ;

8 P _archive = P _MRP ∪ _{P RDM} ;

9 else

10 Use the newly generated population P _archive as the initial population, and use the MOEA/D-DE algorithm to optimize the multi-objective problem at time t;

11 end

12 end

Under this framework, according to the center points of multiple sub-regions of the Pareto solution set at the current moment, the accurate prediction of the center points of the multiple sub-regions of the Pareto solution set in the new environment is realized; the possible solution set at the next moment is adopted using the multi-region diversity maintenance strategy The scope randomly generates other individuals, increasing population diversity. The MOEA/D algorithm is used to optimize the initial population at the next moment. The MOEA/D algorithm mentioned in the framework is an existing algorithm and will not be described here.

The main aspects of this method are the multi-region multi-center point prediction strategy and the multi-region diversity maintenance strategy.

For the multi-region multi-center point prediction strategy, it may specifically include the following contents.

Step S11, obtain the Pareto solution set of the object to be optimized at time t and the Pareto solution set at time t-1, and obtain the center point position of each sub-region according to the Pareto solution set at time t and the Pareto solution set at time t-1 The evolution step at time t; where the Pareto solution set includes multiple sub-regions; t≥1.

In this embodiment, as shown in the method flow shown in FIG. 6 , the evolution step size may be determined first. Specifically, it can include:

The evolution step size is determined by the position of the center point of each subregion in the previous two moments, assuming that the solution of each subregion has a similar motion to the center point order /> and /> Respectively represent the center point at time t-1 time, evolution step size /> The expression is as follows:

where k represents the number of sub-regions.

Step S12, according to the evolution step at time t and the position of the center point of each sub-region in the Pareto solution set at time t, a prediction solution of the center point of each sub-region at time t+1 is obtained.

In this example, a predicted solution is generated:

After determining the evolution step size of each region solution at time t, the solution at time t+1 can be predicted, and the specific expression is as follows:

Among them, i=1,2,...,k, is the solution obtained after the environment changes at time t, /> is the predicted solution after the environment changes at time t+1;

The multi-region multi-center point prediction strategy is shown in Figure 1.

In multi-area central point prediction, there may be a phenomenon of unrelated individuals in some areas, specifically:

The evolution step of the individual-free sub-area can be determined by using at least one nearest individual of the individual-free sub-area; wherein, the individual-free sub-area includes: an individual-free boundary sub-area and/or an individual-free central sub-area.

When there is an individual in the no-individual sub-region at time t-1 and no individual at time t, the no-individual sub-region is obtained according to the position of the center point of the no-individual area at time t-1 and the evolution step at time t. The center point of the sub-region at time t+1 predicts the solution.

Or, when there is no individual in the no-individual sub-region at time t-1 and time t; for the no-individual boundary sub-region, according to the central position of the first sequence sub-region in the Pareto solution set at time t+1, the prediction solution and t The evolution step at time, and the predicted solution of the central position of the final sequence sub-region at time t-1 and the evolution step at time t obtain the predicted solution of the said no-individual sub-region at time t+1; or

For the center sub-area without individuals, according to the center position of the first adjacent sub-area at time t, and the position of the first nearest individual at time t and time t-1, the center point prediction solution at time t+1 is obtained; according to the first The center position of the two adjacent sub-regions at time t, and the position of the second nearest neighbor individual at time t and time t-1 obtain the center point prediction solution at time t+1.

In this embodiment, first, determine the evolution step size of the individual-free region:

When the phenomenon of unrelated individuals occurs, it can be divided into two cases: no individual in the boundary area and no individual in the middle area. Since the individual evolutionary step size difference in adjacent areas is small, the evolutionary step size of the unindividual area can be passed through the adjacent Individuals in the area are obtained. For sub-areas without individuals, the corresponding area boundaries can also be determined. The multi-area central point prediction strategy first calculates the Euclidean distance between other individuals and the boundary L, and finds the individual x ^{near closest} to the boundary L, If the no-individual area is located in the boundary area, only one x ^near is used to determine the evolution step size. If the no-area individual is located in the middle area, there will be two adjacent boundaries, so two nearest individuals x ^near1 and x ^near2 will be found. For no The individual area is located in the boundary area and the evolution step determination expression is as follows:

For the no-individual area located in the middle area, the evolution step determination expression is as follows:

Among them, F _p is the step size adjustment parameter. According to the comparative analysis of the experiment, the effect is better when F _p = 0.4

Individual-free regions yield solutions:

The case of no-individual regions may occur in the current generation or in the previous generation, which can be divided into three cases: a. There are individuals in this generation, but there are no individuals in the previous generation. In this case, the solution is generated by formula 8; b There is no individual in this generation, and there are individuals in the previous generation. In this case, the way to generate a solution is as follows:

C has no individuals in this generation and the previous generation. In this case, for the boundary area, the solution is generated as follows:

For intermediate individuals, the solution is generated as follows:

For the above embodiments, reference may be made to what is shown in FIG. 2 .

For example, the implementation process based on multi-region and multi-center point forecasting strategy may include the following:

1 fori=1,2,...,kdo

2

3 Calculate the center point according to formula 6

4 ifi=1ori=kthen

5 Calculate the evolution step size according to formula 7;

6 else

7 Calculate the evolution step size according to formula 10;

8 end

9 Generate a new solution according to formula 8;

10 end

11 end

Step S13, according to the maximum point and minimum point of each sub-region at the time t, use a preset random function to obtain the random value of each sub-region at time t+1; wherein, the random value of each sub-region The number of values is related to the number of raw solutions for that subregion.

Step S14, based on the prediction solution of the central point of each sub-region at the time t+1, and the random value of each sub-region at the time t+1, generate the initial population at the time t+1, so as to use the t+1 time The initial population at time +1 obtains the multi-objective parameter optimization result at time t+1.

In this embodiment, other individual P _RDMs can be randomly generated based on the multi-region diversity maintenance strategy, specifically:

The multi-region diversity maintenance strategy generates a random solution in each sub-region to maintain population diversity. The random solution is generated through the minimum point and maximum point of each sub-region. For the minimum point lowi _t and maximum point highi of each sub-region at time t _t , i=1,2,...,k, the specific expression is as follows:

Among them, i=1,2,...,k; j=1,2,...,n; k is the number of sub-regions, n is the dimension of the decision space, for the i-th element of the minimum point and maximum point and /> The specific definition is as follows:

Where P is the number of solutions in the i-th sub-region, and the expressions of the moving directions lowiD _t and highiD _t of the minimum point and maximum point at time t are as follows:

Among them, i=1,2,...,k; j=1,2,..,n; k is the number of sub-regions, n is the dimension of the decision space, the i-th direction of the minimum point and maximum point movement direction element and /> The expression is as follows:

Where i=1, 2,...,k, then the multi-region diversity maintenance strategy is used to generate the solutions for each region as follows:

x _t+1 = random(low _t+1 ，high _t+1 ) (19)

Among them, random(a,b) is a function that generates random numbers, which can generate a random number of a and b. The number of random solutions generated by each sub-area is equal to the number of original solutions of the sub-area. For some sub-areas, In the case of no individual, the multi-region diversity maintenance strategy solution is generated as follows:

Among them, Δ _i represents the i-th sub-region, i=1,2,...,k,k is the number of sub-regions, uppe and lower represent the upper and lower boundary values of the decision space.

The multi-regional diversity maintenance strategy is shown in Figure 3.

For example, the implementation process based on multi-regional diversity maintenance strategies may include:

1 fori=1,2,...,kdo

2 if there is no individual area then

3 Calculate the minimum point and maximum point at time t according to formula 15;

4 else

5 Calculate the minimum point and maximum point at time t according to formula 20;

6 end

7 Calculate the direction of movement of the minimum and maximum points according to formula 17;

8 Generate the minimum point and maximum point at time t+1 according to formula 18;

9 end

10 Generate individual P _RDM according to Equation 19

Compared with the prior art, the embodiment of the present invention adopts a multi-region center point prediction strategy, and according to the center points of multiple sub-regions of the Pareto solution set at the current moment, the accurate prediction of the center points of the multiple sub-regions of the Pareto solution set under the new environment is realized; The multi-region diversity maintenance strategy is used to randomly generate other individuals in the range of possible solution sets at the next moment to increase the diversity of the population, thereby improving the convergence speed and convergence accuracy of the algorithm in the new environment. The method of the invention can fully capture the change of the solution set, realize accurate prediction of multi-area center points in the new environment, randomly generate other individuals in areas where solutions may be generated, and increase population diversity.

Since the present invention focuses on the general dynamic multi-objective optimization problem, the superiority of the algorithm is illustrated by an internationally used test function. The specific expression of the test function is shown in Table 1.

Table 1 Test function

The closer the predicted population is to the real Pareto front under environment t, the faster the algorithm can converge to the real Pareto front.

In order to quantitatively analyze the two important indicators of convergence and distribution of the algorithm, the dynamic reverse generation distance is used as the evaluation criterion, and the expression is as follows:

in, is the real Pareto front at time t, P ^t is the approximate Pareto front obtained by the algorithm at time t, d(v, P ^t ) represents the minimum Euclidean distance between point v on the real Pareto front and the approximate Pareto front.

In 20 independent runs, n _t = 10, τ _t = 10, any experimental result was selected at random, and the change trend of the inverse generational distance evaluation index (IGD) of the test functions DF2 and HE2 is shown in Fig. 4a and Fig. 4b. It can be seen that, with the continuous change of environmental factors, IGD fluctuates in a small range of values and tends to be stable.

Figure 5a and Figure 5b show the comparison diagrams of the real Pareto frontier and the approximate frontier obtained by the algorithm of DF2 in the environment t=53 and Fun9 at t=10, 17, 40 and below, the dark part in the figure is the real Pareto frontier, The light colored part is the approximate Pareto front. It can be clearly seen from the figure that the approximate frontier almost coincides with the real frontier, indicating that the patented algorithm can achieve high convergence accuracy.

In order to quantitatively illustrate the superiority of the method of the present invention, Table 2 provides the mean (MIGD) value and mean square error of the reverse generation distance evaluation index of different objective functions of the algorithm. The algorithm goes through 100 environment changes. It can be seen directly from the table that the average value of MIGD obtained by the patented algorithm is very small, and there is little difference in each stage, indicating that the patented algorithm has high convergence accuracy and distribution breadth, and has high stability.

Table 2 The mean value and mean square error of different objective functions MIGD

For example, as one of the most important raw material industries, the most fundamental task of the iron and steel industry is to provide society with sufficient quantities and High-quality high-performance steel products meet the needs of social development, national security, and people's lives. Cold rolling is an important link in steel production, and there are many process quality parameters. The high-precision setting and multi-objective precise control of these process parameters are the key to ensure product precision and quality. The cold-rolling unsteady-state variable-speed process is in the stage of dramatic changes in the rolling conditions of the plate, and its process quality parameters will also change accordingly. Therefore, the optimization of the process quality parameters is a typical dynamic multi-objective optimization problem. According to different working conditions and quality requirements, different process quality parameters are specified, and the process quality parameters are dynamically optimized through the method of the present invention, so that the rolling unsteady-state speed change process is always in the optimal state, and finally this dynamic can be improved. Process control accuracy, improve product quality.

In addition, the application of the present invention can not only be applied to the above-mentioned industrial engineering, but also has a wide range of applications in other fields. For example, in aviation operations, affected by bad weather, aircraft failures, random accidents of customers, etc., aviation scheduling needs to consider passenger retention time and operating costs In the production scheduling, the sequence of orders changes, the sudden failure of production equipment and other factors cause the original plan to change, and the scheduling system must dynamically adjust the production strategy to optimize the production strategy. way to complete production tasks. These dynamic multi-objective optimization problems that commonly exist in the real world urgently need suitable algorithms to improve the ability to solve the problems.

It should be noted that the above-mentioned figures are only schematic illustrations of the processing included in the method according to the exemplary embodiment of the present invention, and are not intended to be limiting. It is easy to understand that the processes shown in the above figures do not imply or limit the chronological order of these processes. In addition, it is also easy to understand that these processes may be executed synchronously or asynchronously in multiple modules, for example.

Further, as shown in FIG. 7 , the implementation of this example also provides a multi-objective parameter optimization system 20 based on dynamic multi-objective evolution, including: an evolution step calculation module 201, a central point prediction solution acquisition module 202, A random value calculation module 203 and an optimization result generation module 204 . in:

The evolution step calculation module 201 can be used to obtain the Pareto solution set of the object to be optimized at time t and the Pareto solution set at time t-1, so as to obtain the Pareto solution set at time t and the Pareto solution set at time t-1 The position of the center point of each sub-region obtains the evolution step at time t; the Pareto solution set includes multiple sub-regions; t≥1.

The central point prediction solution acquisition module 202 can be used to obtain the position of the central point of each sub-region in the Pareto solution set at the time t according to the evolution step at the time t and the position of the center point of each sub-region in the Pareto solution set at the time t+1. The predicted solution for the center point.

The random value calculation module 203 can be used to obtain the random value of each sub-region at time t+1 according to the maximum point and minimum point of each sub-region at time t, using a preset random function; wherein, each The number of random values in the sub-area is related to the number of original solutions in the sub-area.

The optimization result generation module 204 may be configured to generate the predicted solution at the time t+1 based on the predicted solution of the center point of each sub-region at the time t+1, and the random value of each sub-region at the time t+1. The initial population is used to obtain the multi-objective parameter optimization result at the time t+1 by using the initial population at the time t+1.

The specific details of each module in the above-mentioned dynamic multi-objective evolution-based multi-objective parameter optimization system 20 have been described in detail in the corresponding dynamic multi-objective evolution-based multi-objective parameter optimization method, so details will not be repeated here.

It should be noted that although several modules or units of the device for action execution are mentioned in the above detailed description, this division is not mandatory. Actually, according to the embodiment of the present disclosure, the features and functions of two or more modules or units described above may be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided to be embodied by a plurality of modules or units.

In an exemplary embodiment of the present disclosure, an electronic device capable of implementing the above method is also provided.

Those skilled in the art can understand that various aspects of the present invention can be implemented as systems, methods or program products. Therefore, various aspects of the present invention can be embodied in the following forms, that is: a complete hardware implementation, a complete software implementation (including firmware, microcode, etc.), or a combination of hardware and software implementations, which can be collectively referred to herein as "circuit", "module" or "system".

An electronic device 600 according to this embodiment of the present invention is described below with reference to FIG. 8 . The electronic device 600 shown in FIG. 8 is only an example, and should not limit the functions and scope of use of this embodiment of the present invention.

As shown in FIG. 8, electronic device 600 takes the form of a general-purpose computing device. The components of the electronic device 600 may include, but are not limited to: at least one processing unit 610, at least one storage unit 620, a bus 630 connecting different system components (including the storage unit 620 and the processing unit 610), and a display unit 640.

Wherein, the storage unit stores program codes, and the program codes can be executed by the processing unit 610, so that the processing unit 610 executes various exemplary methods according to the present invention described in the "Exemplary Methods" section of this specification. Implementation steps.

The storage unit 620 may include a readable medium in the form of a volatile storage unit, such as a random access storage unit (RAM) 6201 and/or a cache storage unit 6202 , and may further include a read-only storage unit (ROM) 6203 .

Storage unit 620 may also include a program/utility 6204 having a set (at least one) of program modules 6205, such program modules 6205 including but not limited to: an operating system, one or more application programs, other program modules, and program data, Implementations of networked environments may be included in each or some combination of these examples.

Bus 630 may represent one or more of several types of bus structures, including a memory cell bus or memory cell controller, a peripheral bus, an accelerated graphics port, a processing unit, or a local area using any of a variety of bus structures. bus.

The electronic device 600 can also communicate with one or more external devices 700 (such as keyboards, pointing devices, Bluetooth devices, etc.), and can also communicate with one or more devices that enable the user to interact with the electronic device 600, and/or communicate with Any device (eg, router, modem, etc.) that enables the electronic device 600 to communicate with one or more other computing devices. Such communication may occur through input/output (I/O) interface 650 . Moreover, the electronic device 600 can also communicate with one or more networks (such as a local area network (LAN), a wide area network (WAN) and/or a public network such as the Internet) through the network adapter 660 . As shown, the network adapter 660 communicates with other modules of the electronic device 600 through the bus 630 . It should be appreciated that although not shown, other hardware and/or software modules may be used in conjunction with electronic device 600, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives And data backup storage system, etc.

Through the description of the above implementations, those skilled in the art can easily understand that the example implementations described here can be implemented by software, or by combining software with necessary hardware. Therefore, the technical solutions according to the embodiments of the present disclosure can be embodied in the form of software products, and the software products can be stored in a non-volatile storage medium (which can be CD-ROM, U disk, mobile hard disk, etc.) or on the network , including several instructions to make a computing device (which may be a personal computer, a server, a terminal device, or a network device, etc.) execute the method according to the embodiments of the present disclosure.

In an exemplary embodiment of the present disclosure, there is also provided a computer-readable storage medium on which a program product capable of implementing the above-mentioned method in this specification is stored. In some possible implementations, various aspects of the present invention can also be implemented in the form of a program product, which includes program code, and when the program product is run on a terminal device, the program code is used to make the The terminal device executes the steps according to various exemplary embodiments of the present invention described in the "Exemplary Method" section above in this specification.

As shown in FIG. 9 , a program product 800 for implementing the above method according to an embodiment of the present invention is described, which can adopt a portable compact disk read-only memory (CD-ROM) and include program codes, and can be used in terminal equipment, For example running on a personal computer. However, the program product of the present invention is not limited thereto. In this document, a readable storage medium may be any tangible medium containing or storing a program, and the program may be used by or in combination with an instruction execution system, apparatus or device.

The program product may reside on any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. The readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or device, or any combination thereof. More specific examples (non-exhaustive list) of readable storage media include: electrical connection with one or more conductors, portable disk, hard disk, random access memory (RAM), read only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

A computer readable signal medium may include a data signal carrying readable program code in baseband or as part of a carrier wave. Such propagated data signals may take many forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the foregoing. A readable signal medium may also be any readable medium other than a readable storage medium that can transmit, propagate, or transport a program for use by or in conjunction with an instruction execution system, apparatus, or device.

Program code embodied on a readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

Program code for carrying out the operations of the present invention may be written in any combination of one or more programming languages, including object-oriented programming languages—such as Java, C++, etc., as well as conventional procedural programming languages. Programming language - such as "C" or a similar programming language. The program code may execute entirely on the user's computing device, partly on the user's device, as a stand-alone software package, partly on the user's computing device and partly on a remote computing device, or entirely on the remote computing device or server to execute. In cases involving a remote computing device, the remote computing device may be connected to the user computing device through any kind of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computing device (for example, using an Internet service provider). business to connect via the Internet).

Other embodiments of the present disclosure will be readily apparent to those skilled in the art from consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any modification, use or adaptation of the present disclosure, and these modifications, uses or adaptations follow the general principles of the present disclosure and include common knowledge or conventional technical means in the technical field not disclosed in the present disclosure . The specification and examples are to be considered exemplary only, with the true scope and spirit of the disclosure indicated by the appended claims.

The above is only a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Anyone skilled in the art can easily think of changes or substitutions within the technical scope disclosed in the present invention. Should be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention should be determined by the protection scope of the claims.

Claims (9)

1. A multi-objective parameter optimization method based on dynamic multi-objective evolution, characterized in that, comprising: Obtain the Pareto solution set of the object to be optimized at time t and the Pareto solution set at time t-1, so as to obtain the Pareto solution set at time t according to the Pareto solution set at time t and the center point position of each sub-region in the Pareto solution set at time t-1. Evolution step; where Pareto solution set includes multiple sub-regions; t≥1, the object to be optimized includes multiple process quality parameters of cold rolling process in steel production, or includes aviation scheduling scheme in aviation operation, or includes production scheduling production strategy in According to the evolution step at the time t and the position of the center point of each sub-region in the Pareto solution set at the time t, obtain a predicted solution of the center point of each sub-region at time t+1; and According to the maximum point and minimum point of each of the sub-regions at the time t, use a preset random function to obtain the random value of each of the sub-regions at time t+1; wherein, the number of random values of each of the sub-regions is related to the number of raw solutions in this subregion; Generate the initial population at the time t+1 based on the predicted solution of the central point of each sub-region at the time t+1, and the random value of each sub-region at the time t+1, so as to utilize the time t+1 The initial population of the multi-objective parameter optimization result at the time t+1 is obtained, wherein, when the object to be optimized includes multiple process quality parameters of the cold rolling process in steel production, the multi-objective parameter optimization result is the lowest resource and energy consumption, the lowest environmental and ecological load, and the highest efficiency and labor productivity are obtained by target optimization; when the object to be optimized includes the aviation scheduling scheme in aviation operations, the multi-objective parameter optimization result is as follows The multi-objective parameter optimization result is obtained by optimizing the objective of completing the production task in an optimal way when the object to be optimized includes the production strategy in production scheduling. 2. method according to claim 1, is characterized in that, described Pareto solution set comprises some quantity without individual subregion, and described method also comprises: determining an evolutionary step size of the individual-free sub-region using at least one nearest neighbor individual of the individual-free sub-region; Wherein, the individual-free sub-area includes: an individual-free boundary sub-area and/or an individual-free central sub-area. 3. The method according to claim 2, further comprising: when there are individuals in the no-individual sub-area at time t and no individual at time t-1, then according to the no-individual sub-area at time t The position of the center point and the evolution step at time t obtain the prediction solution of the center point of the individual-free sub-region at time t+1. 4. The method according to claim 2, characterized in that the method further comprises: when there is an individual in the no-individual sub-region at time t-1 and there is no individual at time t, then according to the no-individual sub-region at time t-1 The center point position of the individual region and the evolution step at time t obtain the center point prediction solution of the individual-free sub-region at time t+1. 5. method according to claim 2, it is characterized in that, described method also comprises: described no individual subregion when t-1 moment, t moment does not have individual; For no individual boundary subregion, according to Pareto solution The center position prediction solution of the first sequence sub-region at time t+1 and the evolution step size at time t time, and the center position prediction solution of the last sequence sub-region at time t-1 time and the evolution step size at time t time are obtained. The predicted solution of the no-individual sub-region at time t+1; or For the center sub-area without individuals, according to the center position of the first adjacent sub-area at time t, and the position of the first nearest individual at time t and time t-1, the center point prediction solution at time t+1 is obtained; according to the first The center position of the two adjacent sub-regions at time t, and the position of the second nearest neighbor individual at time t and time t-1 obtain the center point prediction solution at time t+1. 6. The method according to claim 2, characterized in that, the method further comprises: for the individual-free sub-region, according to the maximum point and minimum point of the adjacent sub-region at the time t, the top of the decision space The random value is determined in the boundary value and the lower boundary value. 7. A multi-objective parameter optimization system based on dynamic multi-objective evolution, characterized in that it comprises: The evolution step calculation module is used to obtain the Pareto solution set of the object to be optimized at time t and the Pareto solution set of t-1 time, so as to obtain the Pareto solution set of the object at time t and the Pareto solution set of t-1 time for each sub-region The position of the central point of the center point obtains the evolution step at time t; where the Pareto solution set includes multiple sub-regions; t≥1, the object to be optimized includes multiple process quality parameters of the cold rolling process in steel production, or includes Aviation scheduling schemes, or including production strategies in production scheduling; The central point prediction solution acquisition module is used to obtain the position of the central point of each sub-region at t+1 time according to the evolution step size at the time t and the position of the central point of each sub-region in the Pareto solution set at the time t. predictive solutions; and The random value calculation module is used to obtain the random value of each sub-area at time t+1 according to the maximum point and minimum point of each sub-area at time t, using a preset random function; wherein, each of the sub-areas The number of random values of is related to the number of original solutions in this sub-region; An optimization result generating module, configured to generate the initial population at the time t+1 based on the predicted solution of the central point of each sub-region at the time t+1, and the random value of each sub-region at the time t+1, The multi-objective parameter optimization result at time t+1 is obtained by using the initial population at time t+1, wherein when the object to be optimized includes multiple process quality parameters of the cold rolling process in steel production, the multiple The target parameter optimization result is obtained by optimizing the target with the lowest resource and energy consumption, the lowest environmental and ecological load, and the highest efficiency and labor productivity; when the object to be optimized includes the aviation scheduling scheme in aviation operations, the described The multi-objective parameter optimization result is obtained by considering the passenger residence time and operating cost as the objective optimization; when the object to be optimized includes the production strategy in production scheduling, the multi-objective parameter optimization result is to complete the production in an optimal way The task is obtained by optimizing the target. 8. A storage medium, on which a computer program is stored, and when the program is executed by a processor, the multi-objective parameter optimization method based on dynamic multi-objective evolution according to any one of claims 1 to 6 is implemented. 9. An electronic terminal, characterized in that it comprises: processor; and a memory for storing executable instructions of the processor; Wherein, the processor is configured to execute the multi-objective parameter optimization method based on dynamic multi-objective evolution according to any one of claims 1 to 6 by executing the executable instructions.