ISA-PSO-RVM fault power failure prediction model based on simulated annealing-particle swarm optimization

Abstract

The invention relates to an ISA-PSO-RVM fault power failure prediction model based on a simulated annealing-particle swarm optimization algorithm, and the model building process comprises the steps: data collection: collecting the historical data of fault power failure, including fault time, fault type, fault position and influence factors; model construction: determining an RVM model structure; selecting proper kernel functions and parameters, and constructing a basic framework of the RVM model; sA-PSO algorithm optimization: initializing a particle swarm, evaluating fitness, updating particle positions and speeds, carrying out iterative search until a preset number of iterations is reached or a stop condition is met, and outputting an optimal solution: selecting a particle with an optimal fitness value as the optimal solution, so as to obtain optimized RVM model parameters; and training the fault power failure data by using the optimized RVM model, and establishing a mapping relation between the fault power failure and the influence factors.

Description

Technical Field

[0001] The present invention relates to the technical field of power outage due to fault in distribution network, and in particular to an ISA-PSO-RVM fault power outage prediction model based on simulated annealing-particle swarm algorithm. Background Art

[0002] The distribution network is the "last mile" for electricity to reach power customers. It is most closely connected with users and has the most direct impact on users. Reducing power outages caused by distribution network faults and improving power supply reliability are of great value in ensuring the quality of power consumption for users and safeguarding social and economic influence. With the deepening of the market-oriented reform of the power system, power outages caused by distribution system faults will inevitably cause economic losses to power users and have an impact on society. The power outage time can reflect the level of power supply reliability of a country. However, there is still a certain gap between the power supply reliability of the distribution network in my country and that in developed countries.

[0003] Most urban power distribution systems use radial wiring, which is sensitive to single faults and has a high fault rate. In addition, they are directly connected to users. If a fault occurs, it will directly lead to power outages for users and affect residents' lives. Therefore, accurate fault outage prediction provides a decision-making reference for fault repair and grid reinforcement, further reducing the probability and duration of fault outages and ensuring safe and stable operation of the grid.

[0004] In view of this, the present invention provides an ISA-PSO-RVM fault outage prediction model based on simulated annealing-particle swarm algorithm. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide an ISA-PSO-RVM fault outage prediction model based on simulated annealing-particle swarm algorithm.

[0006] To solve the above problems, the technical solution adopted by the present invention is:

[0007] An ISA-PSO-RVM fault outage prediction model based on simulated annealing-particle swarm algorithm, the model establishment process includes:

[0008] Step S1, data collection:

[0009] Collect historical data of power outages caused by faults, including fault time, fault type, fault location, and influencing factors;

[0010] Step S2: Model construction:

[0011] Step S201, determine the RVM model structure: select appropriate kernel functions and parameters to build the basic framework of the RVM model;

[0012] Step S202, SA-PSO algorithm optimization, which includes:

[0013] (1) Initialize the particle swarm: set the number, position, speed and other parameters of the particle swarm, and randomly generate the initial solution;

[0014] (2) Fitness evaluation: Calculate the fitness value of each particle based on the difference between the predicted result of the RVM model and the true value;

[0015] (3) Update particle position and velocity: Update the particle position and velocity according to the velocity and displacement formula of the particle swarm algorithm and the probability jump mechanism of the simulated annealing algorithm.

[0016] (4) Iterative search: Repeat the above steps until the preset number of iterations is reached or the stopping condition is met.

[0017] (5) Output the optimal solution: Select the particle with the best fitness value as the optimal solution, that is, obtain the optimized RVM model parameters.

[0018] Step S3: Use the optimized RVM model to train the fault power outage data and establish a mapping relationship between the fault power outage and the influencing factors.

[0019] As an implementation mode of the present invention, step S1 also includes: preprocessing the data, including data cleaning, data transformation and data normalization, etc., to improve the quality and availability of the data.

[0020] As an implementation of the present invention, step S201 includes:

[0021] (1) Select kernel function: Use Gaussian kernel function as kernel function;

[0022] (2) Determine model parameters, which include hyperparameters and kernel function parameters.

[0023] As an embodiment of the present invention, step (2) comprises:

[0024] The loss function is obtained using the following formula: MSE = 1 / n × ∑ (yi-yi') 2 ; Where MSE is the mean square error; n is the number of samples, yi is the true value, and yi' is the predicted value;

[0025] For each particle, the RVM model parameter combination it represents will be used to predict the sample data. After obtaining the prediction result, the loss function value corresponding to the particle can be calculated;

[0026] Directly use the opposite of the loss function value as the fitness value.

[0027] As an embodiment of the present invention, step (3) comprises:

[0028] Calculate the new speed based on the current position, the individual optimal position and the global optimal position, and update the new position based on the new speed;

[0029] For each particle, a prediction is made using the RVM model, and the fitness value is calculated based on the difference between the predicted result and the true value.

[0030] Individual optimal position update: If the fitness value of the current particle is better than the fitness value of its individual optimal position, the individual optimal position is updated to the current position of the particle; if the fitness value of the current particle is better than the fitness value of the global optimal position, the global optimal position is updated to the current position of the particle;

[0031] Repeat the above steps until the preset number of iterations is reached or the stopping condition is met.

[0032] As an implementation of the present invention, the method further includes: step S4, evaluating the performance of the model by cross-validation and other methods to ensure the stability and reliability of the model. .

[0033] The beneficial effects of adopting the above technical solution are:

[0034] The ISA-PSO-RVM fault power outage prediction model based on simulated annealing-particle swarm algorithm provided by the present invention takes into account that the PSO algorithm and the FA algorithm are both swarm intelligent bionic algorithms. Compared with other bionic algorithms, the PSO algorithm requires fewer parameters to be set, the algorithm concept is concise, easy to implement, and has a fast convergence speed. It is an efficient optimization algorithm. However, the PSO algorithm is prone to fall into local extreme values ​​during the optimization process, causing the "premature" phenomenon. The SA algorithm has the ability of probabilistic sudden jumps, which can make the algorithm jump out of the local extreme value. Therefore, the SA algorithm and the PSO algorithm are used for algorithm fusion, and the SA-PSO algorithm is proposed. The SA-PSO algorithm is used to optimize the RVM kernel function, and the SA-PSO-RVM fault power outage prediction model is built to predict the distribution network fault power outage index, thereby increasing the distribution network fault power outage index prediction method. DETAILED DESCRIPTION

[0035] In order to make the objectives, technical solutions and advantages of the present invention more clear, the present invention is clearly and completely described below in conjunction with specific embodiments.

[0036] The embodiment of the present invention provides an ISA-PSO-RVM fault outage prediction model based on simulated annealing-particle swarm algorithm, and the model establishment process includes:

[0037] Step S1, data collection:

[0038] Collect historical data of power outages caused by faults, including fault time, fault type, fault location, and influencing factors; it also includes:

[0039] Preprocess the data, including data cleaning, data transformation and data normalization, to improve the quality and availability of the data.

[0040] Step S2: Model construction:

[0041] Step S201, determine the RVM model structure: select appropriate kernel functions and parameters to build the basic framework of the RVM model; it includes:

[0042] (1) Select kernel function: Use Gaussian kernel function as kernel function;

[0043] (2) Determine model parameters, which include hyperparameters and kernel function parameters.

[0044] The loss function is obtained using the following formula: MSE = 1 / n × ∑ (yi-yi') 2 ; Where MSE is the mean square error; n is the number of samples, yi is the true value, and yi' is the predicted value;

[0045] For each particle, the RVM model parameter combination it represents will be used to predict the sample data; after obtaining the prediction result, the loss function value corresponding to the particle can be calculated;

[0046] Directly use the opposite of the loss function value as the fitness value.

[0047] Step S202, SA-PSO algorithm optimization, which includes:

[0048] (1) Initialize the particle swarm: set the number, position, speed and other parameters of the particle swarm, and randomly generate the initial solution;

[0049] (2) Fitness evaluation: Calculate the fitness value of each particle based on the difference between the predicted result of the RVM model and the true value;

[0050] (3) Update particle position and velocity: Update the particle position and velocity according to the velocity and displacement formula of the particle swarm algorithm and the probability jump mechanism of the simulated annealing algorithm;

[0051] Specifically include:

[0052] Calculate the new speed based on the current position, the individual optimal position and the global optimal position, and update the new position based on the new speed;

[0053] For each particle, a prediction is made using the RVM model, and the fitness value is calculated based on the difference between the predicted result and the true value.

[0054] If the fitness value of the current particle is better than the fitness value of its individual optimal position, the individual optimal position is updated to the current position of the particle; if the fitness value of the current particle is better than the fitness value of the global optimal position, the global optimal position is updated to the current position of the particle;

[0055] Repeat the above steps until the preset number of iterations is reached or the stopping condition is met.

[0056] (4) Iterative search: Repeat the above steps until the preset number of iterations is reached or the stopping condition is met.

[0057] (5) Output the optimal solution: Select the particle with the best fitness value as the optimal solution, that is, obtain the optimized RVM model parameters.

[0058] Step S3: Use the optimized RVM model to train the fault power outage data and establish a mapping relationship between the fault power outage and the influencing factors.

[0059] Step S4: Evaluate the performance of the model through cross-validation and other methods to ensure the stability and reliability of the model.

[0060] The ISA-PSO-RVM fault power outage prediction model based on simulated annealing-particle swarm algorithm provided by the present invention takes into account that the PSO algorithm and the FA algorithm are both swarm intelligent bionic algorithms. Compared with other bionic algorithms, the PSO algorithm requires fewer parameters to be set, the algorithm concept is concise, easy to implement, and has a fast convergence speed. It is an efficient optimization algorithm. However, the PSO algorithm is prone to fall into local extreme values ​​during the optimization process, causing the "premature" phenomenon. The SA algorithm has the ability of probabilistic sudden jumps, which can make the algorithm jump out of the local extreme value. Therefore, the SA algorithm and the PSO algorithm are used for algorithm fusion, and the SA-PSO algorithm is proposed. The SA-PSO algorithm is used to optimize the RVM kernel function, and the SA-PSO-RVM fault power outage prediction model is built to predict the distribution network fault power outage index, thereby increasing the distribution network fault power outage index prediction method.

[0061] Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art may still modify the technical solutions described in the aforementioned embodiments, or make equivalent substitutions for some of the technical features therein; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An ISA-PSO-RVM fault outage prediction model based on simulated annealing-particle swarm algorithm, characterized in that: The model building process includes: Step S1, data collection: Collect historical data of power outages caused by faults, including fault time, fault type, fault location, and influencing factors; Step S2: Model construction: Step S201, determine the RVM model structure: select appropriate kernel functions and parameters to build the basic framework of the RVM model; Step S202, SA-PSO algorithm optimization, which includes: (1) Initialize the particle swarm: set the number, position, speed and other parameters of the particle swarm, and randomly generate the initial solution; (2) Fitness evaluation: Calculate the fitness value of each particle based on the difference between the predicted result of the RVM model and the true value; (3) Update particle position and velocity: Update the particle position and velocity according to the velocity and displacement formula of the particle swarm algorithm and the probability jump mechanism of the simulated annealing algorithm; (4) Iterative search: Repeat the above steps until the preset number of iterations is reached or the stop condition is met; (5) Output the optimal solution: Select the particle with the best fitness value as the optimal solution, that is, obtain the optimized RVM model parameters; Step S3: Use the optimized RVM model to train the fault power outage data and establish a mapping relationship between the fault power outage and the influencing factors.

2. The ISA-PSO-RVM fault outage prediction model based on simulated annealing-particle swarm algorithm according to claim 1 is characterized in that: Step S1 also includes: preprocessing the data, including data cleaning, data transformation and data normalization, etc., to improve the quality and availability of the data.

3. The ISA-PSO-RVM fault outage prediction model based on simulated annealing-particle swarm algorithm according to claim 1 is characterized in that: Step S201 includes: (1) Select kernel function: Use Gaussian kernel function as kernel function; (2) Determine model parameters, which include hyperparameters and kernel function parameters.

4. The ISA-PSO-RVM fault outage prediction model based on simulated annealing-particle swarm algorithm according to claim 1 is characterized in that: Step (2) comprises: The loss function is obtained using the following formula: MSE = 1 / n × ∑ (yi-yi') 2 ; Where MSE is the mean square error; n is the number of samples, yi is the true value, and yi' is the predicted value; For each particle, the RVM model parameter combination it represents will be used to predict the sample data; after obtaining the prediction result, the loss function value corresponding to the particle can be calculated; Directly use the opposite of the loss function value as the fitness value.

5. The ISA-PSO-RVM fault outage prediction model based on simulated annealing-particle swarm algorithm according to claim 1 is characterized in that: Step (3) includes: Calculate the new speed based on the current position, the individual optimal position and the global optimal position, and update the new position based on the new speed; For each particle, a prediction is made using the RVM model, and the fitness value is calculated based on the difference between the predicted result and the true value. If the fitness value of the current particle is better than the fitness value of its individual optimal position, the individual optimal position is updated to the current position of the particle; if the fitness value of the current particle is better than the fitness value of the global optimal position, the global optimal position is updated to the current position of the particle; Repeat the above steps until the preset number of iterations is reached or the stopping condition is met.

6. The ISA-PSO-RVM fault outage prediction model based on simulated annealing-particle swarm algorithm according to claim 1 is characterized in that: The method also includes: step S4, evaluating the performance of the model by cross-validation and other methods to ensure the stability and reliability of the model.