An issa-elml-based method and system for predicting potential for electric energy replacement

Abstract

The application discloses an electric energy substitution potential prediction method and system based on ISSA-ELM, relates to the technical field of electric energy substitution potential prediction, and comprises the following steps: analyzing influence factors of electric energy substitution potential and quantifying, inputting quantized values of the influence factors of electric energy substitution potential; configuring parameters of a sparrow search algorithm, performing global and local search optimization based on a chicken swarm optimization algorithm and a Cauchy-Gaussian mutation strategy; evaluating a fitness function to obtain an optimal adaptive individual, calculating optimal weights and thresholds of an extreme learning machine, establishing an optimal extreme learning machine prediction model, and outputting an electric energy substitution potential prediction result. According to the method, the parameters of the sparrow search algorithm are configured, the chicken swarm optimization algorithm and the Cauchy-Gaussian mutation strategy are combined, global and local search optimization of prediction model parameters is realized, the search efficiency of the algorithm is improved, the ability of the algorithm to jump out of a local optimal solution is enhanced, and therefore the generalization ability and the prediction precision of the prediction model are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of electric energy substitution potential prediction, in particular to an electric energy substitution potential prediction method and system based on ISSA-ELM. BACKGROUND

[0002] With the continuous growth of global energy demand and the increasing impact of energy consumption on the environment, the development and utilization of renewable energy have received widespread attention. Among renewable energy, electric energy, as a clean and efficient energy form, has attracted the attention of researchers in its ability to replace traditional energy. Electric energy substitution potential prediction refers to analyzing and quantifying various factors that affect electric energy substitution potential, establishing a prediction model, and predicting the ability of electric energy to replace traditional energy in the future. This is of great significance for formulating energy policies, optimizing energy structure, and promoting the use of renewable energy.

[0003] In recent years, researchers have proposed many electric energy substitution potential prediction methods based on machine learning and optimization algorithms. Among them, extreme learning machine (ELM) is a high-efficiency single-hidden layer feedforward neural network that has been widely used in electric energy substitution potential prediction due to its fast training speed and strong generalization ability. However, the traditional ELM algorithm has some shortcomings in the training process. First, the traditional ELM algorithm uses a random generation method to initialize the hidden layer node parameters, which may affect the stability and prediction accuracy of the algorithm. Second, the traditional ELM algorithm does not effectively utilize optimization algorithms to optimize network parameters during the training process, thereby limiting the further improvement of its prediction performance. SUMMARY

[0004] In view of the above problems, the present application is proposed.

[0005] Therefore, the technical problems solved by the present application are: the existing electric energy substitution potential prediction methods have low precision, low training speed, and low reliability, and how to solve the problems of parameter optimization and prediction accuracy in electric energy substitution potential prediction.

[0006] To solve the above technical problems, the present application provides the following technical scheme: an electric energy substitution potential prediction method based on ISSA-ELM, comprising analyzing the influencing factors of electric energy substitution potential and quantifying, inputting the quantized values of the electric energy substitution potential influencing factors; configuring the parameters of the sparrow search algorithm, performing global and local search optimization based on the chicken swarm optimization algorithm and the Cauchy-Gaussian mutation strategy; evaluating the fitness function to obtain the best adaptive individual, calculating the optimal weight and threshold of the extreme learning machine, establishing an optimal extreme learning machine prediction model and outputting the electric energy substitution potential prediction result.

[0007] As a preferred scheme of the ISSA-ELM-based electric energy substitution potential prediction method, the influencing factors of the electric energy substitution potential include regional renewable energy capacity and utilization, environmental quality index, and industrial and residential area electrification rate; the regional renewable energy capacity and utilization include installation capacity and actual energy production utilization rate of wind power, solar power, and hydroelectric power; the environmental quality index includes air quality index, total CO2 emission, and per capita emission; and the industrial and residential area electrification rate includes coverage rate of electric heating and electric cooking facilities.

[0008] As a preferred scheme of the ISSA-ELM-based electric energy substitution potential prediction method, the parameters of the sparrow search algorithm include setting the maximum number of iterations of the sparrow search algorithm, the population size, and the population proportion of explorers and followers; the position update function of the explorer is represented as:

[0009]

[0010] where t is the total number of iterations (t e 1, 2, iter max ), j is the dimension of the optimization problem j e 1, 2, d, represents the number and type of influencing factors in the electric energy substitution potential prediction, Q and a are random numbers generated randomly, Q follows a normal distribution, is used to simulate the randomness factor in the electric energy substitution potential prediction, L is a 1 x d all-1 matrix, is used to initialize or adjust specific parameters of the prediction model, R2 represents the warning value, is a safety measure of the current state of the population in the electric energy substitution potential prediction; when the warning value is less than the safety value, it indicates that the population is in a safe search area, and the foraging range is large, indicating that the prediction model is in a stable search area, and the prediction accuracy is high; when the warning value is greater than or equal to the safety value, it indicates that the population has approached a dangerous area, and the whole population will urgently forage in other areas, indicating that the prediction model is in an unstable or inaccurate search area; and the position update function of the follower is represented as:

[0011]

[0012] where X P represents the best position found by the current explorer, i.e., the best parameter configuration found by the prediction model in the current iteration, X worstrepresents the worst position of the current population foraging, that is, the worst parameter configuration currently explored by the prediction model, A is a 1xd matrix with values of 1 or -1, used to adjust the parameters or features in the prediction model, when i>n / 2, it represents that the i-th follower does not obtain food and is in an emergency state, and takes emergency measures to improve the performance of the prediction model, including adjusting parameters, reinitializing the model or changing the search strategy; when danger comes, the danger warning mechanism of the sparrow group is updated, and is represented as:

[0013]

[0014] wherein X best is the global best position, that is, the optimal parameter configuration found by the prediction model in all iterations, corresponding to the highest prediction accuracy, β and K are step control parameters, both are random numbers, β is subject to a standard normal distribution, K∈[-1, 1], the parameter adjustment affects the convergence speed of the search algorithm and the exploration degree of the global optimal solution, ε is a constant, used to control the convergence and stability of the algorithm, f i is the fitness value of the i-th sparrow, that is, the fitness score calculated in the prediction model according to the current parameter configuration, used to evaluate the advantages and disadvantages of the parameter configuration, f g and f ω are the global best fitness value and the worst fitness value, that is, the highest and lowest fitness scores in the current search process of the prediction model, used to guide the search direction and parameter update.

[0015] As a preferred scheme of the electric energy substitution potential prediction method based on ISSA-ELM, wherein: the global and local search optimization includes updating the follower position by the chicken swarm optimization algorithm and mutating the individual by the Cauchy-Gaussian mutation strategy, updating the positions of the explorers and followers under the warning mechanism, constantly iterating to generate new solutions and recording the current optimal solution and the global optimal solution of the explorer position; the chicken swarm optimization algorithm is introduced to update the follower position, and the chicken swarm optimization algorithm strategy is that hens approach roosters to seek local use and global optimization, and the hen position update function is represented as:

[0016]

[0017] S2=exp(f s -f i )

[0018] wherein r represents any r-th rooster in the hen partner, and is represented as a random selection mechanism of the simulation algorithm in the electric energy substitution potential prediction model, used to introduce the elements of global search, s represents any s-th rooster or hen in the chicken swarm, and r≠s.

[0019] As a preferred scheme of the ISSA-ELM-based electric energy substitution potential prediction method, the global and local search optimization further comprises: performing mutation on the individual with the highest fitness based on Cauchy-Gaussian mutation, and performing iteration on the positions before and after mutation, and the iteration is represented as:

[0020]

[0021] wherein, represents the position of the optimal individual after mutation, and is used to update the parameters of the electric energy substitution potential prediction model, and σ 2 represents the standard deviation of the Cauchy-Gaussian mutation, and is a parameter for controlling the degree of mutation, and the value of σ 2 adjusts the exploration range and convergence speed of the algorithm in the electric energy substitution potential prediction model, λ1саuсhy(0,σ 2 is a random variable satisfying the Cauchy distribution, and λ2Gauss(0,σ 2 is a random variable satisfying the Gaussian distribution, and λ1 and λ2 are parameters adjusted with the iteration number t, and the search ability of the global optimization algorithm in the electric energy substitution potential prediction model is dynamically adjusted.

[0022] As a preferred scheme of the ISSA-ELM-based electric energy substitution potential prediction method, the obtaining of the optimal individual comprises: evaluating the fitness function to obtain the optimal individual, optimizing the model parameters by using K-fold cross-validation, and selecting the mean square error (MSE) of the prediction model as the fitness function, and the function of the MSE is represented as:

[0023]

[0024] wherein, y i is the electric energy substitution potential prediction value, is the true value of the electric energy substitution potential, and n is the sample number of the electric energy substitution potential prediction.

[0025] As a preferred scheme of the ISSA-ELM-based electric energy substitution potential prediction method, the output of the electric energy substitution potential prediction result comprises: constantly iterating to obtain the optimal weight and threshold of the extreme learning machine, and the iteration stopping conditions comprise: the iteration number reaches a set standard, and the model error reaches an expected level; when the iteration satisfies the stopping conditions, the optimal initial weight and threshold of the extreme learning machine are outputted, the best fitness function value and the best parameters, i.e., the best input weight and threshold of the extreme learning machine, are obtained, the optimal extreme learning machine prediction model is established based on the improved optimal input weight and threshold, and the regional electric energy substitution potential prediction result is outputted.

[0026] Another object of the present application is to provide an ISSA-ELM-based electric energy substitution potential prediction system, which can solve the problem of low training speed in current electric energy substitution potential prediction by configuring parameters of the sparrow search algorithm, performing global and local search optimization based on the chicken swarm optimization algorithm and the Cauchy-Gaussian mutation strategy.

[0027] As a preferred scheme of the ISSA-ELM-based electric energy substitution potential prediction system, the system comprises a factor quantification module, an optimization algorithm module and a prediction module; the factor quantification module is used for analyzing and quantifying influencing factors of electric energy substitution potential and inputting quantized values of the influencing factors; the optimization algorithm module is used for configuring parameters of the sparrow search algorithm, performing global and local search optimization based on the chicken swarm optimization algorithm and the Cauchy-Gaussian mutation strategy; and the prediction module is used for evaluating a fitness function to obtain an optimal adaptive individual, calculating optimal weights and thresholds of the extreme learning machine, establishing an optimal extreme learning machine prediction model and outputting an electric energy substitution potential prediction result.

[0028] A computer device comprises a memory and a processor, the memory stores a computer program, and the processor executes the computer program to implement steps of the ISSA-ELM-based electric energy substitution potential prediction method.

[0029] A computer-readable storage medium stores a computer program, and the computer program is executed by a processor to implement steps of the ISSA-ELM-based electric energy substitution potential prediction method.

[0030] The present application has the following advantages: the ISSA-ELM-based electric energy substitution potential prediction method provided by the present application can analyze and quantify influencing factors of electric energy substitution potential, input quantized values of the influencing factors, ensure that input data of a prediction model have high relevance and accuracy, and thus improve reliability of a prediction result; by configuring parameters of the sparrow search algorithm and combining the chicken swarm optimization algorithm and the Cauchy-Gaussian mutation strategy, the present application can realize global and local search optimization of parameters of the prediction model, improve search efficiency of the algorithm, enhance the ability of the algorithm to jump out of a local optimal solution, and thus improve generalization ability and prediction precision of the prediction model; in addition, by introducing the mutation strategy, the present application can increase exploration ability of the algorithm, help to find better solutions in a complex search space, evaluate a fitness function to obtain an optimal adaptive individual, calculate optimal weights and thresholds of the extreme learning machine, which indicates that a prediction model training process is completed and ensures that the model has the highest accuracy and stability when predicting electric energy substitution potential; by establishing an optimal extreme learning machine prediction model, the present application can output reliable electric energy substitution potential prediction results, and the present application can achieve better results in terms of accuracy, efficiency and reliability. BRIEF DESCRIPTION OF DRAWINGS

[0031] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative effort on the basis of these drawings.

[0032] Figure 1 The overall flowchart of the ISSA-ELM-based electric energy substitution potential prediction method provided for the first embodiment of the present application.

[0033] Figure 2 The overall flowchart of the ISSA-ELM-based electric energy substitution potential prediction system provided for the third embodiment of the present application. DETAILED DESCRIPTION

[0034] In order to make the above-mentioned objects, features and advantages of the present application more apparent and comprehensible, the specific embodiments of the present application will be described in detail below with reference to the drawings in the specification. Obviously, the described embodiments are only some of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present application.

[0035] Embodiment 1

[0036] Reference Figure 1 For an embodiment of the present application, an ISSA-ELM-based electric energy substitution potential prediction method is provided, which comprises the following steps.

[0037] S1: analyze the influencing factors of electric energy substitution potential and quantify, and input the quantized values of the influencing factors of electric energy substitution potential.

[0038] Further, the influencing factors of electric energy substitution potential include regional renewable energy capacity and utilization rate, environmental quality index, and industrial and residential area electrification rate; the regional renewable energy capacity and utilization rate includes installation capacity and actual energy utilization rate of wind energy, solar energy and hydroelectric power; the environmental quality index includes air quality index, total CO2 emission and per capita emission; and the industrial and residential area electrification rate includes coverage rate of electric heating and electric cooking facilities.

[0039] It should be noted that the quantification process includes data collection, by consulting official statistics, research reports, energy consumption records, etc., to collect relevant data, data standardization, standardization processing of the collected data to facilitate comparison of data at different times and places, data conversion, conversion of raw data into values that can be used for model calculation, such as scaling data to a specific range through normalization processing, data verification, through cross-validation or other methods to ensure the accuracy and reliability of the quantification data.

[0040] S2: Configure the parameters of the sparrow search algorithm, and perform global and local search optimization based on the chicken swarm optimization algorithm and the Cauchy-Gaussian mutation strategy.

[0041] Further, the related parameters of the sparrow search algorithm are set, including the maximum number of iterations, the population size, the population proportion of explorers and followers, the sparrow search algorithm has good optimization ability, solves the problem of low prediction accuracy and accuracy in electric energy substitution potential prediction, the explorers in the population are responsible for finding food and determining the position and foraging direction, while the followers find food by obtaining the information shared by the explorers, the foraging behavior of sparrows can be further divided into the following exploration strategies, each sparrow in the sparrow population can learn the strategies and behaviors adopted by other sparrows, the aggressor in the population will compete for food resources with the same species with high intake to improve the predation rate, in addition, sparrows have a warning mechanism that allows them to move to a safe area when danger is imminent, the explorers provide food for the entire sparrow population and determine the foraging direction of the followers, therefore, the position update range of the explorers is greater than that of the followers, the explorer foraging position update function is expressed as:

[0042]

[0043] Where, t is the total number of iterations (t e 1, 2, iter max), represents the total number of iterations of the sparrow search algorithm in the prediction method of the potential of electricity substitution; j is the dimension of the optimization problem j ∈ 1, 2, d, representing the number and variety of various influencing factors that need to be considered in the prediction of the potential of electricity substitution; Q and a are random numbers generated randomly, Q follows a normal distribution, and is used to simulate the randomness in the prediction of the potential of electricity substitution; L is a 1 × d all-1 matrix, which is used to initialize or adjust the specific parameters of the prediction model; R2 represents the warning value, which is a safety measure of the current state of the population in the prediction of the potential of electricity substitution, and a value lower than R2 indicates that the prediction model may be in a relatively stable search area, and the prediction accuracy may be higher; ST represents the safety value, which indicates that the population is in a safe search area when the warning value is less than the safety value, and the foraging range is larger; when the warning value is greater than or equal to the safety value, it indicates that the population has approached the dangerous area, and the whole population will urgently go to other areas to forage, which represents the dangerous threshold that the population should avoid in the prediction of the potential of electricity substitution, and when R2 is greater than or equal to ST, it indicates that the prediction model may be in an unstable or inaccurate search area, and the model or algorithm parameters may need to be adjusted to improve the prediction accuracy. The position update function of the follower is represented as:

[0044]

[0045] where X P represents the best position found by the current explorer, i.e. the best parameter configuration found by the prediction model in the current iteration, which is used to improve the prediction accuracy; X worst represents the worst position of the current population foraging, i.e. the worst parameter configuration currently explored by the prediction model, which may lead to a decrease in prediction accuracy or instability; A is a 1 × d matrix with values of 1 or -1, which can be used to adjust certain parameters or features in the prediction model to optimize the prediction effect of the model; i > n / 2 indicates that the i-th follower with a lower fitness value has not obtained food and is in an emergency state, and emergency measures should be taken to improve the performance of the prediction model, which may include adjusting parameters, reinitializing the model or changing the search strategy, etc. When danger comes, the danger warning mechanism of the sparrow group is updated according to the following equation:

[0046]

[0047] where X best is the global best position, i.e. the optimal parameter configuration found by the prediction model in all iterations, corresponding to the highest prediction accuracy; β and K are step control parameters, both of which are random numbers, β follows a standard normal distribution, and K ∈ [-1, 1], these parameter adjustments will affect the convergence speed of the search algorithm and the exploration degree of the global optimal solution; ε is a constant small enough to control the convergence and stability of the algorithm; f iThe fitness value of the i-th sparrow is the fitness score calculated in the prediction model according to the current parameter configuration, and is used to evaluate the advantages and disadvantages of the parameter configuration. g And w The global best fitness value and the worst fitness value are the highest and the lowest fitness scores of the prediction model in the current search process, and are used to guide the search direction and parameter update.

[0048] It should be noted that the positions of the explorers and followers under the early warning mechanism are updated by the chicken swarm optimization algorithm (CSO) and the individuals are mutated by the Cauchy-Gaussian mutation strategy (CG), and at the same time, new solutions are continuously generated and the current optimal solution and the global optimal solution of the explorer position are recorded, further solving the problem of low prediction accuracy and accuracy in electric energy substitution potential prediction. The traditional sparrow search algorithm for updating the position of the follower is prone to local optimization, and in order to avoid this shortcoming, the chicken swarm optimization algorithm is introduced to update the position of the follower. The chicken swarm optimization algorithm has excellent global optimization ability, and the algorithm strategy is that the hens approach the cocks with a certain probability, which can consider local utilization and global optimal seeking at the same time. The hen position updating function is represented as:

[0049]

[0050] S2=exp(f s -f i )

[0051] Wherein, r represents any rth cock in the hen partner, which can be understood as a random selection mechanism of the simulation algorithm in the electric energy substitution potential prediction model, and is used to introduce the element of global search to ensure that the search process will not fall into a local optimal solution; f represents the fitness value, i.e. the fitness score calculated in the electric energy substitution potential prediction model according to the current parameter configuration, which is used to evaluate the advantages and disadvantages of the parameter configuration and guide the parameter update and search direction selection of the algorithm; s represents any sth cock or hen in the chicken swarm, and r≠s, so as to increase the diversity and global search ability of the algorithm in the electric energy substitution potential prediction model.

[0052] It should also be noted that in the traditional sparrow search algorithm, when the sparrow individuals are iterated to the later stage, the population is easily assimilated into a local optimum, and in order to overcome this shortcoming, the Cauchy-Gaussian mutation method is introduced to mutate the individual with the highest current fitness, and the positions before and after mutation are compared to select a better position for iteration, which is represented as:

[0053]

[0054] Wherein, represents the position of the optimal individual after mutation, used to update the parameters of the electricity substitution potential prediction model to improve the prediction accuracy and reliability; σ 2 represents the standard deviation of the Cauchy-Gaussian variation, which is a parameter to control the degree of variation, adjusting the value of σ 2 can affect the amplitude and direction of the mutation operation, and thus affect the exploration range and convergence speed of the algorithm in the electricity substitution potential prediction model; λ1Gauss(0, σ 2 ) is a random variable satisfying the Cauchy distribution, λ2Gauss(0, σ 2 ) is a random variable satisfying the Gaussian distribution, λ1and λ2are parameters adjusted with the iteration number t, aiming to improve the search ability of the global optimization algorithm in the electricity substitution potential prediction model.

[0055] S3: Evaluate the fitness function to obtain the best adapted individual, calculate the optimal weight and threshold of the extreme learning machine, establish the optimal extreme learning machine prediction model and output the electricity substitution potential prediction result.

[0056] Further, the fitness function is evaluated to obtain the best adapted individual, and SSA should take appropriate measures to ensure the best parameter settings to improve the prediction accuracy of the model. The K-fold cross-validation (K-CV) method is used to optimize the model parameters. K-fold cross-validation can effectively solve the problem of model overfitting and poor generalization ability. The mean square error (MSE) of the prediction model is selected as the fitness function, and the function of MSE is represented as:

[0057]

[0058] where y i is the electricity substitution potential prediction value, is the true value of the electricity substitution potential, and n is the number of samples of the electricity substitution potential prediction.

[0059] It should be noted that the iteration is continuously performed to obtain the optimal weight and threshold of the extreme learning machine. The iteration stopping criteria are two: one is that the iteration number reaches the set standard, and the other is that the model error reaches the expected level. When the iteration meets the stopping condition, the optimal initial weight and threshold of the extreme learning machine are obtained. Otherwise, repeat the third and fourth steps until the iteration number reaches the preset value. Through the loop iteration, the best fitness function value is obtained, and thus the best parameters, i.e. the best input weight and threshold of the extreme learning machine, are obtained. The optimal extreme learning machine prediction model is established and the electricity substitution potential prediction result is output. According to the optimal input weight and threshold of the ISSA improved extreme learning machine prediction model, the optimal extreme learning machine prediction model is established and the regional electricity substitution potential prediction result is output.

[0060] Example 2

[0061] An embodiment of the present application provides an ISSA-ELM-based electric energy substitution potential prediction method. In order to verify the beneficial effects of the present application, economic benefit calculation and simulation experiments are used for scientific demonstration.

[0062] In this embodiment, for the prediction of electric energy substitution potential, an extreme learning machine (ELM) model based on sparrow search algorithm (SSA), chicken swarm optimization algorithm (CSO) and Cauchy-Gaussian (CG) mutation strategy is constructed. The experiment mainly includes several key steps: analysis and quantification of influencing factors, algorithm configuration, fitness evaluation and model establishment, and on this basis, the prediction results are output. First, the data collection and preprocessing are carried out. The experiment selects three representative regions: region A (developing market, high renewable energy potential), region B (developed market, moderate renewable energy potential) and region C (low development market, low renewable energy potential). For each region, the installed capacity and actual energy utilization rate data of wind energy, solar energy and hydroelectric power are collected; air quality index, total CO2 emission and per capita emission; and industrial and residential electricity rate, such as coverage rate of electric heating and electric cooking facilities. In the algorithm configuration stage, the parameters of sparrow search algorithm are set, including the maximum number of iterations (100 iterations), the population size (50 population members), the ratio of explorers and followers (20% explorers, 80% followers). The chicken swarm optimization algorithm is used for global search optimization, and the Cauchy-Gaussian mutation strategy is used for local search optimization, which ensures that the optimal solution can be found in global and local levels. The mean square error (MSE) is used as the main fitness index for the evaluation of the fitness function. The K-fold cross-validation (K=5) method is used to optimize the model parameters, so that the model has good generalization ability. Finally, the configured parameters and algorithms are used to run the model for learning and iteration until the iteration stopping condition is met. Through the iteration process, the model parameters are gradually optimized and adjusted until the optimal model weight and threshold are obtained. Then the optimal extreme learning machine model is established, and the electric energy substitution potential prediction results of each region are output. The experimental data are recorded and analyzed with reference to Table 1.

[0063] Table 1 Experimental data record table

[0064]

[0065]

[0066] Analyzing the above table data, some key trends and results can be observed. Region A has a higher renewable energy utilization rate and a lower air quality index, indicating a cleaner environment and a high degree of dependence on renewable energy, thus the predicted value of electric energy substitution potential is higher. In contrast, region C has lower indicators, indicating insufficient electric energy substitution potential. The application of sparrow search algorithm and chicken optimization strategy can effectively find the optimal solution from both global and local levels, which is often difficult to achieve in traditional algorithms. For example, the Cauchy-Gaussian mutation strategy introduces random variation in the local search process, effectively avoiding the problems of premature convergence and local optimization. The introduction of this strategy improves the accuracy and robustness of the prediction model.

[0067] Embodiment 3

[0068] Reference Figure 2 For an embodiment of the present application, an ISSA-ELM-based electric energy substitution potential prediction system is provided, including a factor quantification module, an optimization algorithm module, and a prediction module.

[0069] The factor quantification module is used to analyze and quantify the influencing factors of electric energy substitution potential, and input the quantified values of electric energy substitution potential influencing factors. The optimization algorithm module is used to configure the parameters of the sparrow search algorithm, and perform global and local search optimization based on the chicken optimization algorithm and the Cauchy-Gaussian mutation strategy. The prediction module is used to evaluate the fitness function to obtain the best adaptive individual, calculate the optimal weight and threshold of the extreme learning machine, establish the optimal extreme learning machine prediction model, and output the electric energy substitution potential prediction result.

[0070] If the functions are implemented in the form of software function units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the present application or the parts that essentially contribute to the prior art or parts of the technical solutions can be embodied in the form of a software product, which is stored in a storage medium and includes instructions to make a computer device (which can be a personal computer, server, or network device, etc.) execute all or part of the steps of the embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disk or optical disk, and various program code storage media.

[0071] The logic and / or steps represented in the flow diagrams or otherwise described herein, for example, can be considered as a sequence of executable instructions, and can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, processor-containing system, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions. In the context of this specification, a "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device. The computer-readable medium can be, for example but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, device, or propagation medium.

[0072] More specific examples (a non-exhaustive list) of the computer-readable medium include the following: an electrical connection (electronic) having one or more wires, a portable computer diskette (magnetic), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium can also be paper or another suitable medium upon which the program is printed, as the program can be electronically captured, for example via an optical scanner, then compiled, interpreted, or otherwise processed, and stored in a computer memory in a form that can be later executed by a computer. In some embodiments, the machine-readable medium can be a computer- readable storage medium.

[0073] It should be understood that aspects of the application can be implemented in hardware, software, firmware or combinations thereof. In the above embodiments, various steps or methods can be implemented in software or firmware that is stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, and in another embodiment, any of the following can be used: discrete logic circuitry having logic gates for implementing logic functions upon data signals, application specific integrated circuits having appropriate combinational logic gates, programmable gate arrays (PGA), field programmable gate arrays (FPGA), and so forth. It should be noted that the foregoing embodiments are merely examples of implementations of the technical solutions of the present application and are not limiting. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present application, and all such modifications or replacements should be included in the scope of the claims of the present application.

[0074] It should be noted that the above examples are only used to illustrate the technical solutions of the present application but not limit the present application. Although the present application is described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or equivalently replaced, without departing from the spirit and scope of the technical solutions of the present application, which should be covered in the scope of the claims of the present application.

Claims

1. A method for predicting electric energy substitution potential based on ISSA-ELM, characterized in that: include: Analyze and quantify the factors affecting the potential of electric energy substitution, and input the quantitative values ​​of the factors affecting the potential of electric energy substitution; Configure the parameters of the sparrow search algorithm and perform global and local search optimization based on the swarm optimization algorithm and the Cauchy-Gaussian mutation strategy. The sparrow search algorithm is responsible for the overall search process. The swarm optimization algorithm improves the follower's position update strategy to prevent followers from easily falling into local optimal solutions. The Cauchy-Gaussian mutation strategy mutates the individuals with the current highest fitness, increasing the diversity of the population and preventing the algorithm from converging to a local optimal solution too early. Evaluate the fitness function to obtain the best adapted individual, calculate the optimal weights and thresholds of the extreme learning machine, establish the optimal extreme learning machine prediction model and output the electric energy substitution potential prediction results; The global and local search optimization also includes mutating the individual with the highest current fitness based on Cauchy-Gauss mutation, and iterating by comparing the positions before and after the mutation, which is expressed as: Among them, U t best It represents the position of the optimal individual after mutation and is used to update the parameters of the electric energy substitution potential prediction model. σ² represents the standard deviation of the Cauchy-Gaussian mutation, which is a parameter that controls the degree of mutation. Adjusting the value of σ² affects the exploration range and convergence speed of the algorithm in the electric energy substitution potential prediction model. λ1cauchy(0, σ 2 ) is a random variable that satisfies the Cauchy distribution, λ2Gauss(0,σ 2 ) is a random variable that satisfies the Gaussian distribution, λ1 and λ2 are parameters adjusted with the number of iterations t, which dynamically adjust the search capability of the global optimization algorithm in the electric energy substitution potential prediction model, X represents the location, and X i is the position of forager i, X t best is the optimal position at t iterations.

2. The method for predicting electric energy substitution potential based on ISSA-ELM according to claim 1, characterized in that: The factors affecting the potential of electric energy substitution include regional renewable energy capacity and utilization rate, environmental quality index, and electrification rate of industrial and residential areas; Regional renewable energy capacity and utilization rate includes installed capacity and actual capacity utilization rate of wind, solar, and hydropower; Environmental quality index includes air quality index, total CO2 emissions and per capita emissions; The electrification rate of industrial and residential areas includes the coverage of electric heating and electric cooking facilities.

3. The method for predicting electric energy substitution potential based on ISSA-ELM according to claim 2, characterized in that: The parameters of the sparrow search algorithm include setting the maximum number of iterations of the sparrow search algorithm, the population size, and the population ratio of explorers to followers; The explorer's foraging position update function is expressed as: Where t is the total number of iterations, t∈(1, 2, ..., iter max ), represents the total number of iterations of the sparrow search algorithm in the prediction of electric energy substitution potential, j is the dimension of the optimization problem, j∈(1, 2, …, d), represents the number and type of influencing factors in the prediction of electric energy substitution potential, X t+1 i，j In the example, X represents the position, and the whole represents the position of forager i in the j-dimensional optimization problem at the t+1th iteration. i，j represents the position of forager i in the j-dimensional optimization problem. Q and a are randomly generated numbers. Q follows a normal distribution and is used to simulate the random factors in the prediction of electric energy substitution potential. L is a 1×d all-one matrix used to initialize or adjust specific parameters of the prediction model. R2 represents the warning value and is a safety measure of the current state of the population in the prediction of electric energy substitution potential. When the alert value is less than the safety value, it indicates that the population is in a safe search area and has a large foraging range, which means that the prediction model is in a stable search area and has high prediction accuracy; When the alert value is greater than or equal to the safety value, it indicates that the population is approaching a dangerous area and the entire population will urgently move to other areas to forage, indicating that the prediction model is in an unstable or inaccurate search area; The follower's position update function is expressed as: Among them, X P represents the best position found by the current explorer, that is, the best parameter configuration found by the prediction model in the current iteration, X worst represents the worst position of the current population foraging, that is, the worst parameter configuration currently explored by the prediction model, X t worst represents the worst position of the population foraging at the tth iteration, X t i，j represents the position of forager i in the j-dimensional optimization problem at the tth iteration, X p t+1 represents the best position found by the explorer at the t+1th iteration. A is a 1×d matrix with a value of 1 or -1, which is used to adjust the parameters or features in the prediction model. When i>n / 2, it means that the i-th follower has not obtained food and is in an emergency state. Emergency measures should be taken to improve the performance of the prediction model, including adjusting parameters, reinitializing the model, or changing the search strategy. When danger comes, the sparrow flock's danger warning mechanism is updated, which is expressed as: Among them, X best is the global best position, that is, the optimal parameter configuration found by the prediction model in all iterations, corresponding to the highest prediction accuracy. β and K are step size control parameters, both of which are random numbers. β obeys the standard normal distribution, K∈[-1,1]. Parameter adjustment affects the convergence speed of the search algorithm and the degree of exploration of the global optimal solution. ε is a constant used to control the convergence and stability of the algorithm. f i is the fitness value of the i-th sparrow, that is, the fitness score calculated according to the current parameter configuration in the prediction model, which is used to evaluate the quality of the parameter configuration, f g and f ω are the global best fitness value and the worst fitness value, that is, the highest and lowest fitness scores of the prediction model in the current search process, which are used to guide the search direction and parameter update.

4. The method for predicting electric energy substitution potential based on ISSA-ELM according to claim 3, characterized in that: The global and local search optimization includes updating the follower position through the chicken swarm optimization algorithm and mutating the individual using the Cauchy-Gauss mutation strategy, updating the positions of the explorer and the follower under the early warning mechanism, continuously iterating to generate new solutions and recording the current optimal solution and the global optimal solution of the explorer position; The chicken flock optimization algorithm is introduced to update the follower position. The chicken flock optimization algorithm strategy is that the hen approaches the rooster, performs local utilization and seeks the global optimality. The hen position update function is expressed as: Where r represents any r-th rooster in a hen's companion, and represents a random selection mechanism of the simulation algorithm in the electric energy substitution potential prediction model, which is used to introduce the element of global search. S represents any S-th rooster or hen in the chicken flock, r is not equal to S, and f i represents the random selection mechanism of forager i in the swarm optimization algorithm, f r Represents the random selection mechanism of the rth rooster.

5. The method for predicting electric energy substitution potential based on ISSA-ELM according to claim 4, characterized in that: The method of obtaining the best-fit individual includes evaluating the fitness function to obtain the best-fit individual, optimizing the model parameters by using K-fold cross validation, and selecting the mean square error (MSE) of the prediction model as the fitness function. The MSE function is expressed as: Among them, y i is the predicted value of electric energy substitution potential, is the true value of the electric energy substitution potential, and n is the number of samples for the electric energy substitution potential prediction.

6. The method for predicting electric energy substitution potential based on ISSA-ELM according to claim 5, characterized in that: The output electric energy substitution potential prediction result includes continuously iterating to obtain the optimal weights and thresholds of the extreme learning machine, and the iterative stopping conditions include the number of iterations reaching a set standard and the model error reaching an expected level; When the iteration meets the stopping condition, the optimal initial weights and thresholds of the extreme learning machine are output to obtain the optimal fitness function value and the optimal parameters, that is, the optimal input weights and thresholds of the extreme learning machine. Based on the improved optimal input weights and thresholds, the optimal extreme learning machine prediction model is established, and the regional electricity substitution potential prediction results are output.

7. A system using the ISSA-ELM-based electric energy substitution potential prediction method according to any one of claims 1 to 6, characterized in that: Including factor quantification module, optimization algorithm module, and prediction module; The factor quantification module is used to analyze and quantify the factors affecting the potential of electric energy substitution, and input the quantified values ​​of the factors affecting the potential of electric energy substitution; The optimization algorithm module is used to configure the parameters of the sparrow search algorithm and perform global and local search optimization based on the chicken swarm optimization algorithm and the Cauchy-Gauss mutation strategy; The prediction module is used to evaluate the fitness function to obtain the best adaptation individual, calculate the optimal weight and threshold of the extreme learning machine, establish the optimal extreme learning machine prediction model and output the electric energy substitution potential prediction result.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the electric energy substitution potential prediction method based on ISSA-ELM according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the electric energy substitution potential prediction method based on ISSA-ELM according to any one of claims 1 to 6 are implemented.

Images