Novel intelligent operation and maintenance method for power system transformer substation based on improved SSA-PSO algorithm

Abstract

The invention discloses a novel electric power system transformer substation intelligent operation maintenance method based on an improved SSA-PSO algorithm, and the method comprises the steps: analyzing the state data of a transformer substation, and providing a comprehensive index calculation method for the intelligent operation maintenance of the transformer substation; a correlation coefficient and a deep regression analysis method are adopted to deeply analyze the relevance between the comprehensive index and the operation fault risk of the transformer substation; constructing an intelligent operation maintenance model of the transformer substation with the aims of minimizing the maintenance cost and reducing the fault operation risk value to be maximized; and introducing a local factor and an SSA algorithm to improve a PSO algorithm, solving the intelligent operation maintenance model of the transformer substation, and outputting an intelligent operation maintenance strategy of the transformer substation. According to the method, the aims of minimizing the maintenance cost and maximizing the fault risk value can be achieved, the operation maintenance efficiency and reliability of the transformer substation are effectively improved, the maintenance cost is reduced, faults are prevented, and safe and stable operation of a power system is ensured.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent operation and maintenance of substations, and in particular relates to a novel intelligent operation and maintenance method for substations in a power system based on an improved SSA-PSO algorithm. Background Art

[0002] As a vital component of the new power system, substations play an important role as a link between users and power grids, ensuring smooth operation and interconnection between the two. After long-term practice and technological progress, substations have been widely popularized and have achieved remarkable results in improving the operating performance of traditional substations. However, with the popularization and development of substations, the need to ensure the long-term safe and stable operation of their equipment has become increasingly prominent. To this end, the implementation of reasonable operation, care and maintenance strategies has become a vital task. Therefore, it is particularly urgent to study strategies for the stable operation of substations to ensure the reliability and continuous power supply capacity of the power system.

[0003] In terms of improving the maintenance efficiency of substations, different strategies have been applied, such as corrective maintenance strategies and maintenance strategies based on status data information. However, corrective maintenance strategies are largely affected by the environment and their own technology, and cannot cover all operating conditions, which can easily lead to wrong decisions and operations. The maintenance strategy based on status data information has the problem of less sample information for decision-making, and some key indicators and influencing factors are often ignored. At the same time, the risk indicators of substation operation are not fully considered, making it difficult to judge whether the substation is close to the edge of failure.

[0004] In summary, the existing maintenance strategies have a series of problems such as inefficient resource utilization and inaccurate solution algorithms. Therefore, how to overcome the impact of substation operation risks on the power system and improve the accuracy of maintenance models to ensure the safe, stable and economical operation of the power grid has become a very challenging task. Summary of the invention

[0005] In view of the problems in the prior art of substation operation and maintenance, such as large limitations of corrective maintenance strategies, insufficient sample information of maintenance strategies based on status data information and neglect of key indicators, low resource utilization efficiency and inaccurate solution algorithm, the present invention proposes a new intelligent operation and maintenance method for substations in a power system based on an improved SSA-PSO algorithm, which analyzes the status data of the substation and proposes a comprehensive indicator calculation method; uses correlation coefficients and deep regression analysis methods to explore the relationship between comprehensive indicators and substation operation failure risks; constructs an intelligent operation and maintenance model with the goals of minimizing maintenance costs and maximizing the reduction of failure operation risk values; introduces local factors and SSA algorithm to improve the PSO algorithm and solve the substation operation and maintenance model; and solves the deficiencies of the prior art by improving the algorithm and combining local factors.

[0006] In order to achieve the above object, the present invention is implemented by the following technical solutions:

[0007] A novel intelligent operation and maintenance method for power system substation based on improved SSA-PSO algorithm includes the following steps:

[0008] Step S1: Analyze the status data of the substation and propose a comprehensive indicator calculation method for intelligent operation and maintenance of the substation;

[0009] Step S2: using correlation coefficient and deep regression analysis method to deeply analyze the correlation between comprehensive indicators and substation operation failure risks;

[0010] Step S3: constructing a substation intelligent operation and maintenance model with the goal of minimizing maintenance costs and maximizing the risk value of reducing fault operation;

[0011] Step S4: Introduce local factors and SSA algorithm to improve the PSO algorithm and solve the substation operation and maintenance model.

[0012] As a preferred solution of the present invention, in step S1, the comprehensive index calculation method for substation intelligent operation and maintenance includes: reliability index and substation operation status index;

[0013] The reliability indicators specifically include:

[0014] System reliability R1 is:

[0015]

[0016] Where: T fail is the time when the fault occurs; T total is the total operation time of the substation;

[0017] The mean time to repair a fault, R2, is:

[0018]

[0019] Where: T mttr is the sum of the total repair time required for all failures; N mttr is the total number of all failures that occurred during the same period;

[0020] The substation operation status indicators specifically include:

[0021] The power factor qualification rate R3 is:

[0022]

[0023] Where: θ d is the number of qualified data of power factor of substation; θ cosθ The total number of power factor data of the substation;

[0024] The voltage deviation qualification rate R4 is:

[0025]

[0026] Where: N V,d N is the number of qualified data collected for substation voltage deviation; V,S To collect the total number of qualified data of substation voltage deviation.

[0027] As a preferred solution of the present invention, in step S2, the correlation coefficient is to use the Pearson correlation coefficient method to screen out representative feature data, and the steps are as follows:

[0028] S21: collecting data sets including reliability indicators and power grid operation status indicators, and storing the reliability indicators and power grid operation status indicators in different data sets respectively, and performing data integration;

[0029] S22: For each pair of reliability indicators and power grid operation status indicators, the Pearson correlation coefficient formula is used to calculate the correlation between them;

[0030] Among them, the Pearson correlation coefficient formula is as follows:

[0031]

[0032] Where: r is the Pearson correlation coefficient; x i ,y i are two sets of characteristic values ​​of the i-th pair of reliability index and substation operation status index; are the means of the two sets of features respectively; the value range of the Pearson correlation coefficient is between -1 and 1, close to 1 indicates a strong positive correlation, close to -1 indicates a strong negative correlation, and close to 0 indicates no linear correlation;

[0033] S23: The correlation coefficient thresholds between the reliability index and the grid operation status index are set, with positive and negative correlations of 0.4 and -0.4 respectively;

[0034] S24: According to the set threshold, the grid operation status indicators whose correlation with the reliability indicators reaches the threshold are screened out.

[0035] As a preferred solution of the present invention, in step S2, a correlation model between comprehensive indicators and substation operation failure risks is constructed using a deep regression analysis method, and the steps are as follows:

[0036] Construct a regression model containing different feature data items. The specific representation of the model is as follows:

[0037]

[0038] Where: Y is the dependent variable, indicating the risk of operational failure; is the independent variable, representing different indicator data sets; β1, β2....β6 are the coefficients of the model; ε is the error term;

[0039] The Lasso regression method is introduced and the calculation is as follows:

[0040]

[0041] Where: Y z The reliability index and the substation operation status index are the indicators that affect the operation failure risk value; is the mean value of all indicators affecting the operating failure risk of reliability index and substation operating status index; β is the model parameter; λ is the regularization parameter; Z is the total number of reliability index and substation operating status index;

[0042] Find the model coefficient β that minimizes the Y value j , model coefficient β j Specifically include: Two target regression coefficients, the specific formula is:

[0043]

[0044]

[0045] Where: is the target regression coefficient; is the mean of the indicator series data; is the mean value of operational failure risk;

[0046] Through the regression coefficient, the regression equation is obtained, which is expressed as follows:

[0047]

[0048] Where: is the index x i Run the corresponding failure risk value.

[0049] As a preferred solution of the present invention, in step S3, the substation intelligent operation and maintenance model F is:

[0050]

[0051] Where: A F Substation maintenance costs.

[0052] As a preferred solution of the present invention, the step S4 improves the PSO algorithm by introducing local factors and the SSA algorithm, and applies it to the solution process of the substation operation and maintenance model. The specific steps include:

[0053] S41: Initialize the relevant parameters of the PSO algorithm and set the maximum number of iterations;

[0054] S42: Initialize the particle swarm using Tent chaotic mapping;

[0055] S43: Update the local search factor parameters with the help of SSA algorithm to optimize the search strategy;

[0056] S44: Introduce the updated local search factor parameter to iteratively update the state of the particle swarm;

[0057] S45: Determine whether the current number of iterations reaches a preset maximum value. If the condition is met, output the optimal substation operation and maintenance method; if not, return to step S43 to continue iterative optimization.

[0058] As a preferred solution of the present invention, in step S42, the process of initializing the particle swarm using Tent chaotic mapping is as follows:

[0059]

[0060] x k+1 =E k+1 ub+lb(1-E k+1 )

[0061] Where: k is the number of iterations; E k is the chaotic sequence value at the kth iteration; ub and lb are the upper and lower boundary values ​​of the particle position respectively; xk+1 For the final next step iterative population; E k+1 is the chaotic sequence value at k+1 iterations;

[0062] As a preferred solution of the present invention, in step S43, when the SSA algorithm updates the local search factor parameters, the following steps are specifically performed:

[0063] S43.1: Set a m A population of sparrows and the initial position of each sparrow are determined;

[0064] S43.2: According to the algorithm rules, update the current position parameters of the discoverer, joiner and sparrow for optimization;

[0065] S43.3: Parameter Y for initial amplitude step_0 Optimize

[0066] S43.4: Check whether the preset maximum number of iterations has been reached. If the condition is met, record the parameters of the current local optimal search disturbance amplitude; if not, return to step S43.2 and continue the iterative process until the condition is met.

[0067] As a preferred solution of the present invention, in step S43.2, updating the current positions of the discoverer, the joiner and the sparrow specifically includes the following steps:

[0068] Update discoverer:

[0069]

[0070] Where: is the position of the pth sparrow after k iterations in the q dimension; θ is the distribution factor; k max is the maximum number of iterations; R is a random number that obeys a normal distribution; L is a matrix with a distribution; Q is the sparrow escape factor; ST k is the k-th iteration dynamic safety threshold; is the position of the pth sparrow after k+1 iterations in the q dimension; θ×k max is the product of the distribution factor and the maximum number of iterations.

[0071] Updated joiners are:

[0072]

[0073] Where: is the global worst position at the tth iteration; is the best position of the sparrow in its historical position at the kth iteration; ψ is an array of randomly assigned values ​​1 or -1;

[0074] Update Sparrow position:

[0075]

[0076] Where: Y is the best position of the sparrow among all the historical positions of the sparrows in the entire sparrow group at the kth iteration; w is the parameter of the worst disturbance amplitude; Y is the parameter of the initial disturbance amplitude; the sparrow population continuously updates its position to find the optimal step_0

[0077] Parameter of the local search perturbation amplitude; Y step_size k+1 is the parameter of the k+1th disturbance amplitude; B is a random parameter between [0,1], where the parameter Y that optimizes the initial disturbance amplitude step_0 It is expressed as follows:

[0078]

[0079] In the formula, is the parameter of the k+1th disturbance amplitude; is the position of the pth sparrow after k+1 iterations in the q dimension;

[0080] As a preferred solution of the present invention, in step S4, a local search factor parameter is introduced to update the particle state, which specifically includes the following steps:

[0081] The position vector L of each particle l l =(L l1 ,L l2 ,...,L ld ) and the velocity vector V l =(V l1 ,V l2 ,...,V ld ), L l1 , L l2 is the position of particle l in the first and second dimensions; L ld is the position of particle l in the dth dimension; V l1 、V l2 is the velocity of particle l in the first and second dimensions; V ld is the velocity of particle l in the dth dimension;

[0082] In each iteration, the particle's velocity and position are updated according to the following formula:

[0083]

[0084] Where: V l tis the velocity of particle l at the tth iteration; L t l is the position of particle l at the tth iteration; V l t+1 is the velocity of particle l at the t+1th iteration; The position of particle l at iteration t+1; is the best position of particle l in the historical position at the tth iteration; is the best position of particle l among all the historical positions of particles in the entire particle swarm at the tth iteration; r1 and r2 are two randomly distributed random numbers in the range [0,1]; b1 and b2 are acceleration factors, representing the learning factors of individuals and society respectively; c1 is a random number generation function, which is used to generate random numbers within a certain range and randomly perturb the position during local search; Y step_size Parameter that controls the amplitude of the local search disturbance.

[0085] Compared with the prior art, the beneficial effects of the present invention are as follows: the present invention can help identify and understand the overall operation status of the substation by analyzing the status data of the substation and proposing a comprehensive index calculation method. These comprehensive indicators include system reliability, mean fault repair time, power factor qualification rate and voltage deviation qualification rate, which can effectively evaluate the operation risk level of the substation. Using correlation coefficients and deep regression analysis to explore the relationship between comprehensive indicators and substation operation failure risks helps to gain a deep understanding of the interaction between different factors and determine which factors have the greatest impact on failure risks. Based on this, operation and maintenance personnel can better formulate maintenance plans, adjust operation strategies and resource allocation to achieve intelligent substation operation and maintenance management. Constructing an intelligent operation and maintenance model with the goal of minimizing maintenance costs and maximizing the reduction of fault operation risk values ​​can provide decision support for substation operations, thereby improving operation efficiency and reducing operation costs. In particular, the PSO algorithm is improved by introducing local factors and the SSA algorithm, and the substation operation and maintenance model is solved, which significantly improves the solution efficiency and accuracy. The method of the present invention not only shows a relatively superior response repair time, but also greatly improves the accuracy of fault repair, reduces the risk of fault operation, and further reduces maintenance costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0086] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative labor. Among them:

[0087] Figure 1is a flow chart of a method according to an embodiment of the present invention;

[0088] Figure 2 A flow chart of updating the local search factor of the SSA algorithm in an embodiment of the present invention;

[0089] Figure 3 This is a flowchart of improving the SSA-PSO algorithm solution in an embodiment of the present invention;

[0090] Figure 4 This is the main electrical wiring diagram of the substation in the embodiment of the present invention;

[0091] Figure 5 is the average qualified rate of the substation operation status indicators in the embodiment of the present invention;

[0092] Figure 6 It is a line chart comparing the optimization spaces of different algorithms in the embodiments of the present invention. DETAILED DESCRIPTION

[0093] In order to make the purpose, technical solution and advantages of the embodiment of the present invention clearer, the technical solution of the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings of the embodiment of the present invention. Obviously, the described embodiment is a part of the embodiment of the present invention, not all of the embodiments. Based on the described embodiment of the present invention, all other embodiments obtained by ordinary technicians in this field belong to the scope of protection of the present invention.

[0094] Example 1

[0095] like Figure 1 As shown in FIG. 1 , an embodiment of the present invention is provided, which provides a novel intelligent operation and maintenance method for a power system substation based on an improved SSA-PSO algorithm, including:

[0096] Step S1: Analyze the status data of the substation and propose a comprehensive indicator calculation method for intelligent operation and maintenance of the substation;

[0097] Based on the calculation results of comprehensive indicators, intelligent decision-making support can be provided for substation operation and maintenance personnel to help them better formulate maintenance plans, adjust operation strategies and resource allocation, so as to achieve more intelligent substation operation and maintenance management.

[0098] Furthermore, the reliability indicators specifically include:

[0099] By analyzing system reliability and mean fault repair time indicators, we can optimize maintenance plans and strategies, reasonably arrange preventive maintenance and corrective maintenance, reduce maintenance costs and improve maintenance efficiency.

[0100] System reliability R1 is:

[0101] It indicates the frequency of failures per unit time. The lower the frequency, the more stable and reliable the equipment operation.

[0102]

[0103] Where: T fail is the time when the fault occurs; T total is the total operation time of the substation;

[0104] The mean time to repair a fault, R2, is:

[0105] Indicates the average time required to repair each fault, which can help evaluate maintenance efficiency and the speed at which equipment can be restored to operation.

[0106]

[0107] Where: T mttr is the sum of the total repair time required for all failures; N mttr is the total number of all failures that occurred during the same period;

[0108] Furthermore, the substation operation status indicators specifically include:

[0109] When the power factor of a substation decreases or the voltage deviation increases, it may lead to increased energy loss, unbalanced grid load, reduced equipment life, decreased system stability, and damage to load equipment. Therefore, maintaining proper power factor and stable voltage is the key to maintaining stable and efficient operation of the power system.

[0110] The power factor qualification rate R3 is:

[0111] A decrease in power factor will lead to an increase in reactive power in the system, a decrease in the energy conversion efficiency in the system, and an increase in energy loss.

[0112]

[0113] Where: θ d is the number of qualified data of power factor of substation; θ cosθ The total number of power factor data of substation;

[0114] The voltage deviation qualification rate R4 is:

[0115] Increased voltage deviation will lead to reduced stability of the power system and increase the risk of substation system failure or power outage.

[0116]

[0117] Where: N V,d N is the number of qualified data collected for substation voltage deviation; V,STo collect the total number of qualified data of substation voltage deviation.

[0118] In one of the embodiments, step S2: using correlation coefficient and deep regression analysis method to deeply analyze the correlation between the comprehensive index and the substation operation failure risk;

[0119] If the dimension of the feature data is too high, the computational complexity of the training model will increase, thereby slowing down the calculation speed. On the contrary, if the dimension of the feature data is too low, the recognition accuracy will decrease. In order to solve this problem, the present invention adopts the Pearson correlation coefficient method to screen out the most representative feature data, and then uses the deep regression analysis method to establish the correlation model between the comprehensive index and the risk of operational failure.

[0120] Furthermore, in step S2, the correlation coefficient is to use the Pearson correlation coefficient method to screen out representative feature data, and the steps are as follows:

[0121] S21: collecting data sets including reliability indicators and power grid operation status indicators, and storing the reliability indicators and power grid operation status indicators in different data sets respectively, and performing data integration;

[0122] S22: For each pair of reliability indicators and power grid operation status indicators, the Pearson correlation coefficient formula is used to calculate the correlation between them;

[0123] Among them, the Pearson correlation coefficient formula is as follows:

[0124]

[0125] Where: r is the Pearson correlation coefficient; x i ,y i are two sets of characteristic values ​​of the i-th pair of reliability index and substation operation status index; are the means of the two sets of features respectively; the correlation coefficient ranges from -1 to 1, close to 1 indicates a strong positive correlation, close to -1 indicates a strong negative correlation, and close to 0 indicates no linear correlation;

[0126] S23: The correlation coefficient thresholds between the reliability index and the grid operation status index are set, with positive and negative correlations of 0.4 and -0.4 respectively;

[0127] S24: According to the set threshold, the grid operation status indicators whose correlation with the reliability indicators reaches the threshold are screened out.

[0128] Furthermore, in step S2, a deep regression analysis method is used to construct a correlation model between the comprehensive index and the substation operation failure risk, and the steps are as follows:

[0129] A regression model is established, with reliability index and grid operation status index as independent variables and operation failure risk as dependent variable, to explore the functional relationship between them. The change of operation failure risk under different comprehensive index levels is predicted to help grid operators formulate more effective preventive operation and maintenance strategies.

[0130] When dealing with complex data, you can consider using a multiple linear regression model, which may involve the relationship between multiple independent variables and a dependent variable, and add interaction terms to increase the direct connection between reliability indicators and grid operation status indicators;

[0131] Construct a regression model containing different feature data items. The specific representation of the model is as follows:

[0132]

[0133] Where: Y is the dependent variable, indicating the risk of operational failure; is the independent variable, representing different indicator data sets; β1, β2....β6 are the coefficients of the model; ε is the error term;

[0134] In order to estimate the model parameters, the Lasso (Least Absolute Shrinkage and Selection Operator) regression method is introduced to fit the data by minimizing the loss function, and a regularization term is added to the loss function to penalize the complexity of the model and prevent overfitting. The calculation is as follows:

[0135]

[0136] Where: Y z The reliability index and the substation operation status index are the indicators that affect the operation failure risk value; is the mean value of the operating failure risk affected by all the reliability indicators and substation operating status indicators; β is the model parameter; λ is the regularization parameter; Z is the total number of reliability indicators and substation operating status indicators.

[0137] Find the model coefficient β that minimizes the Y value j , model coefficient β j Specifically include: Two target regression coefficients, the specific formula is:

[0138]

[0139]

[0140] Where: is the target regression coefficient; is the mean of the indicator series data ; is the mean value of operational failure risk;

[0141] Through the regression coefficient, the regression equation is obtained, which is expressed as follows:

[0142]

[0143] Where: is the index x i Run the corresponding failure risk value.

[0144] In a specific embodiment, step S3: constructing a substation intelligent operation and maintenance model with the goal of minimizing maintenance costs and maximizing the reduction of fault operation risk value;

[0145] After an unexpected failure occurs, maintenance of the faulty component must not only consider the maintenance cost, but also minimize the risk value of the fault operation. Therefore, the maintenance cost is minimized and the risk value of the fault operation is minimized.

[0146] In step S3, the substation intelligent operation and maintenance model F is:

[0147]

[0148] Where: A F Substation maintenance costs.

[0149] In a specific embodiment, step S4: introducing local factors and SSA algorithm to improve the PSO algorithm and solve the substation operation and maintenance model.

[0150] The PSO algorithm integrates the experience of individuals and groups, but it is easy to fall into the local optimal solution when dealing with high-dimensional space functions and multi-constraint problems. This situation is particularly prominent in the field of intelligent operation and maintenance of substations, because these problems often have complex nonlinear characteristics and changeable constraints. In high-dimensional space, the search space of particles grows exponentially, resulting in a decrease in the search efficiency of the PSO algorithm, making it difficult to effectively explore the global optimal solution. At the same time, when faced with multiple constraints, the PSO algorithm may get stuck in the case of constraint conflicts and it is difficult to find a solution that satisfies all constraints. In order to overcome these problems, the local search mechanism is introduced to improve the robustness and adaptability of the algorithm, and the SSA algorithm is combined to optimize the parameters. In the field of intelligent operation and maintenance of substations, these improvements and combined strategies can better solve complex problems and improve system performance and operational stability.

[0151] like Figure 2 , Figure 3As shown, step S4 improves the PSO algorithm by introducing local factors and SSA algorithm, and applies it to the solution process of the substation operation and maintenance model. The specific steps include:

[0152] S41: Initialize the relevant parameters of the PSO algorithm and set the maximum number of iterations;

[0153] S42: Initialize the particle swarm using Tent chaotic mapping;

[0154] By using the Tent chaos map method to initialize parameters, the algorithm's search efficiency and global optimization capabilities can be improved. Since the initial solutions initialized by the Tent chaos map are more evenly distributed in the search space, the algorithm can converge faster toward the optimal solution, reducing the number of unnecessary iterations.

[0155] In step S42, the process of initializing the particle swarm using Tent chaotic mapping is as follows:

[0156]

[0157] x k+1 =E k+1 ub+lb(1-E k+1 )

[0158] Where: k is the number of iterations; E k is the chaotic sequence value at the kth iteration; ub and lb are the upper and lower boundary values ​​of the particle position respectively; x k+1 For the final next step iterative population; E k+1 is the chaotic sequence value at k+1 iterations

[0159] S43: Update the local search factor parameters with the help of SSA algorithm to optimize the search strategy;

[0160] In step S43, when the SSA algorithm updates the local search factor parameters, it specifically performs the following steps:

[0161] S43.1: Set a m A population of sparrows and the initial position of each sparrow are determined;

[0162] S43.2: According to the algorithm rules, update the current position parameters of the discoverer, joiner and sparrow for optimization;

[0163] In step S43.2, the current positions of the discoverer, joiner and sparrow are updated, which specifically includes the following steps:

[0164] Update discoverer:

[0165]

[0166] Where: is the position of the pth sparrow after k iterations in the q dimension; θ is the distribution factor; k max is the maximum number of iterations; R is a random number that obeys a normal distribution; L is a matrix with a distribution; Q is the sparrow escape factor; ST k is the k-th iteration dynamic safety threshold; is the position of the pth sparrow after k+1 iterations in the q dimension; θ×k max is the product of the distribution factor and the maximum number of iterations.

[0167] Updated joiners are:

[0168]

[0169] Where: is the global worst position at the tth iteration; is the best position of the sparrow in its historical position at the kth iteration; ψ is an array of randomly assigned values ​​1 or -1;

[0170] Update Sparrow position:

[0171]

[0172] Where: Y is the best position of the sparrow among all the historical positions of the sparrows in the entire sparrow group at the kth iteration; w is the parameter of the worst disturbance amplitude; Y is the parameter of the initial disturbance amplitude; the sparrow population continuously updates its position to find the optimal step_0

[0173] Parameter of the local search perturbation amplitude; Y step_size k+1 is the parameter of the k+1th disturbance amplitude; B is a random parameter between [0,1], where the parameter Y that optimizes the initial disturbance amplitude step_0 It is expressed as follows:

[0174]

[0175] In the formula, is the parameter of the k+1th disturbance amplitude; is the position of the pth sparrow after k+1 iterations in the q dimension;

[0176] S43.3: Parameter Y for initial amplitude step_0 Optimize

[0177] S43.4: Check whether the preset maximum number of iterations has been reached. If the condition is met, record the parameters of the current local optimal search disturbance amplitude; if not, return to step S43.2 and continue the iterative process until the condition is met.

[0178] S44: Introduce the updated local search factor parameter to iteratively update the state of the particle swarm;

[0179] S45: Determine whether the current number of iterations reaches a preset maximum value. If the condition is met, output the optimal substation operation and maintenance method; if not, return to step S43 to continue iterative optimization.

[0180] In one embodiment, in step S4, introducing a local search factor parameter to update the particle state specifically includes the following steps:

[0181] The position vector L of each particle l l =(L l1 ,L l2 ,...,L ld ) and the velocity vector V l =(V l1 ,V l2 ,...,V ld ), L l1 , L l2 is the position of particle l in the first and second dimensions; L ld is the position of particle l in the dth dimension. V l1 、V l2 is the velocity of particle l in the first and second dimensions; V ld is the velocity of particle l in the dth dimension;

[0182] In each iteration, the particle's velocity and position are updated according to the following formula:

[0183]

[0184] Where: V l t is the velocity of particle l at the tth iteration; is the position of particle l at the tth iteration; V l t+1 is the velocity of particle l at the t+1th iteration; The position of particle l at iteration t+1; is the best position of particle l in the historical position at the tth iteration; is the best position of particle l among all the historical positions of particles in the entire particle swarm at the tth iteration; r1 and r2 are two randomly distributed random numbers in the range [0,1]; b1 and b2 are acceleration factors, representing the learning factors of individuals and society respectively; c1 is a random number generation function, which is used to generate random numbers within a certain range and randomly perturb the position during local search; Y step_size Parameter that controls the amplitude of the local search disturbance.

[0185] Example 2

[0186] like Figure 4 As shown, another embodiment of the present invention provides an electrical main wiring diagram of a 10kV substation, which includes four parallel operation and maintenance units. Each operation and maintenance unit is composed of a transformer and two lines of different voltage levels, and power is transmitted from the 10kV side to the 220V kilovolt side.

[0187] Affected by extreme weather, transformer T1 fails and loses power, affecting the normal operation of the load under substation T1, and the substation is not operating normally. Comparative experiment group 1 is set up: substation status maintenance strategy optimization method based on multi-source data analysis; comparative experiment group 2 is set up: substation fault diagnosis system research based on optimized decision tree algorithm. Compared with the optimization method proposed in this invention, the operation of the three groups of schemes is shown in Table 1:

[0188] Table 1: Results of execution strategies of three methods

[0189]

[0190] Program No. Repair time / h Repair cost / thousand yuan Accuracy / % Failure risk value reduction rate / % References

[12] 1.9 998.54 94.16 2.37 References

[13] 2.5 2095.82 97.25 5.95 Methods 1.5 1080.05 98.95 10.6

[0191] As shown in Table 1, the substation operation and maintenance strategy proposed by comparative experimental group 1 focuses on cost factors. Although this strategy can significantly reduce costs and minimize costs, it ignores the risk factors of substation operation, resulting in the accuracy being affected and the fault risk value not being effectively reduced; the substation operation and maintenance strategy proposed by comparative experimental group 2 only focuses on fault maintenance and ignores the substation operation and maintenance costs, resulting in the highest cost consumption, but effectively improving the accuracy and fault risk value. The operation and maintenance strategy proposed in the present invention not only takes into account the operation and maintenance costs, but also takes into account the fault operation risk value. This method not only shows a relatively superior response and repair time, but also greatly improves the rate reduction rate and reduces the risk of fault operation, thereby reducing maintenance costs. The average qualified rate of substation operation status indicators changes, such as Figure 5 shown.

[0192] Depend on Figure 5It can be seen that when a fault occurs, the method of the present invention can react in about 0.5s, the comparative experiment 1 reacts in about 0.7s, and the comparative experiment 2 reacts in about 0.9s. All the methods of the present invention can react quickly when facing a fault; at the same time, after 1.65s, the voltage qualification rate of the method of the present invention is stabilized at about 97%, while after 2.35s of the comparative experiment 1, the voltage qualification rate is stabilized at about 93%, and after 2.25s of the comparative experiment 2, the voltage qualification rate is stabilized at about 95%. Therefore, the method of the present invention can more quickly and effectively prevent the grid operation status qualification rate from decreasing, and quickly restore the stable operation state. It further illustrates that the method proposed by the present invention has certain advantages in the task of substation operation and maintenance.

[0193] The average qualified rates of the grid operation status indicators corresponding to the three methods all dropped sharply. Since the method proposed in this paper takes into account the fault repair time indicator, it can quickly and effectively prevent the grid operation status qualified rate from decreasing and quickly restore the stable operation state. This further illustrates that the method proposed in this invention has certain advantages in the task of substation operation and maintenance.

[0194] In order to further illustrate the superiority of the intelligent algorithm of the present invention, the method proposed in the present invention is compared with the traditional particle swarm optimization algorithm (PSO), the traditional sparrow optimization algorithm (SSA) and the comparative experiment group 2 in the optimization space. Figure 6 Assuming that the same maximum number of iterations is set for different algorithms, the method of the present invention reaches stability after 370 iterations, which is completely superior to any other method.

[0195] In summary, the present invention selects an indicator with greater influence from reliability indicators and substation operation status indicators; then, establishes the relationship between the indicator and the substation fault operation risk value, and establishes a substation operation and maintenance model with the goal of maximizing the reduction of the fault operation risk value and maximizing the profit; finally, introduces local factors and SSA algorithm to improve the PSO algorithm, and uses the improved PSO algorithm to solve the substation operation and maintenance model, which greatly improves the algorithm's solution efficiency and reduces the fault operation risk.

[0196] The present invention proposes four indicators, namely system availability, mean fault repair time, power factor qualification rate and voltage deviation qualification rate, to construct the correlation between substation fault operation risks and better evaluate the risk level of substation operation. At the same time, an improved SSA-PSO algorithm is proposed to solve the substation intelligent operation and maintenance model, which can have better performance and convergence speed when finding the global optimal solution and reduce the difficulty of setting algorithm parameters.

[0197] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine different embodiments or examples described in this specification and the features of different embodiments or examples, unless they are contradictory.

[0198] Any process or method description in the flow chart or otherwise described herein can be understood to represent a module, fragment or portion of a code including one or more executable instructions for implementing the steps of a specific logical function or process. And the scope of the preferred embodiment of the present application includes other implementations, in which the functions may not be performed in the order shown or discussed, including in a substantially simultaneous manner or in a reverse order according to the functions involved.

[0199] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of various changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A new intelligent operation and maintenance method for power system substation based on improved SSA-PSO algorithm, characterized in that: The steps include: Step S1: Analyze the status data of the substation and propose a comprehensive index calculation method for intelligent operation and maintenance of the substation; Step S2: using correlation coefficient and deep regression analysis method to deeply analyze the correlation between comprehensive indicators and substation operation failure risks; Step S3: constructing a substation intelligent operation and maintenance model with the goal of minimizing maintenance costs and maximizing the risk value of reducing fault operation; Step S4: introducing local factors and SSA algorithm to improve the PSO algorithm, and solving the substation intelligent operation and maintenance model.

2. According to claim 1, a new type of power system substation intelligent operation and maintenance method based on improved SSA-PSO algorithm is characterized in that: In the step S1, the comprehensive index calculation method for substation intelligent operation and maintenance includes: reliability index and substation operation status index; The reliability indicators specifically include: System reliability R1 is: Where: T fail is the time when the fault occurs; T total is the total operation time of the substation; The mean time to repair a fault, R2, is: Where: T mttr is the sum of the total repair time required for all faults; N mttr is the total number of all faults occurring in the same period of time; the substation operation status indicators specifically include: The power factor qualification rate R3 is: Where: θ d is the number of qualified data of power factor of substation; θ cosθ The total number of power factor data of substation; The voltage deviation qualification rate R4 is: Where: N V,d N is the number of qualified data collected for substation voltage deviation; V,S To collect the total number of qualified data of substation voltage deviation.

3. According to a novel power system substation intelligent operation and maintenance method based on improved SSA-PSO algorithm according to claim 1, it is characterized in that: In step S2, the correlation coefficient is to use the Pearson correlation coefficient method to screen out representative feature data, and the steps are as follows: S21: collecting data sets including reliability indicators and substation operation status indicators, and storing the reliability indicators and substation operation status indicators in different data sets respectively, and performing data integration; S22: For each pair of reliability index and substation operation status index, the Pearson correlation coefficient formula is used to calculate the correlation between them; Among them, the Pearson correlation coefficient formula is as follows: Where: r is the Pearson correlation coefficient; x i ,y i are two sets of characteristic values ​​of the i-th pair of reliability index and substation operation status index; are the means of the two sets of features respectively; the value range of the Pearson correlation coefficient r is between -1 and 1, close to 1 indicates a strong positive correlation, close to -1 indicates a strong negative correlation, and close to 0 indicates no linear correlation; S23: setting the correlation coefficient threshold between the reliability index and the substation operation status index, the positive and negative correlation thresholds are both 0.4 and -0.4; S24: According to the set correlation coefficient threshold, the substation operation status indicators whose correlation with the reliability indicator reaches the correlation coefficient threshold are screened out.

4. A novel power system substation intelligent operation and maintenance method based on improved SSA-PSO algorithm according to claim 1, characterized in that: In step S2, a deep regression analysis method is used to construct a correlation model between the comprehensive index and the substation operation failure risk, and the steps are as follows: Construct a multivariate linear regression model containing different feature data items. The specific representation of the model is as follows: Where: Y is the dependent variable, indicating the risk of operational failure; are independent variables, representing the indicator data sets of system reliability R1, mean fault repair time R2, power factor qualified rate R3, and voltage deviation qualified rate R4; β1, β2...β6 are model coefficients; ε is the error term; Introduce Lasso regression method to calculate the estimated model coefficient β j , j∈[1,6], is calculated as follows: Where: Y z The reliability index and the substation operation status index are the indicators that affect the operation failure risk value; The reliability index and substation operation status index are all indicators that affect the mean value of operation failure risk; β is the model parameter; λ is the regularization parameter; Z is the total number of reliability indicators and substation operation status indicators; find the model coefficient β that minimizes the Y value j , model coefficient β j Specifically include: Two target regression coefficients, the specific formula is: Where: is the target regression coefficient; is the mean of the indicator series data; is the mean value of operational failure risk; Through the regression coefficient, the regression equation is obtained, which is expressed as follows: Where: is the index x i Run the corresponding failure risk value.

5. According to claim 1, a novel power system substation intelligent operation and maintenance method based on improved SSA-PSO algorithm is characterized in that: In step S3, the substation intelligent operation and maintenance model F is expressed as: Where: A F Substation maintenance costs.

6. A novel power system substation intelligent operation and maintenance method based on improved SSA-PSO algorithm according to claim 1, characterized in that: The step S4 improves the PSO algorithm by introducing local factors and the SSA algorithm, and applies it to the solution process of the substation intelligent operation and maintenance model. The specific steps include: S41: Initialize the relevant parameters of the PSO algorithm and set the maximum number of iterations; S42: Initialize the particle swarm using Tent chaotic mapping; S43: Update local search factor parameters using SSA algorithm; S44: Introduce the updated local search factor parameter to iteratively update the state of the particle swarm; S45: Determine whether the current number of iterations reaches the set maximum number of iterations. If the condition is met, output the optimal substation operation and maintenance model; if not, return to step S43 to continue iterative optimization.

7. A novel power system substation intelligent operation and maintenance method based on improved SSA-PSO algorithm according to claim 6, characterized in that: In step S42, the particle swarm is initialized using Tent chaotic mapping, and the process is expressed as follows: x k+1 =E k+1 ub+lb(1-E k+1 ) Where: k is the number of iterations; E k is the chaotic sequence value at the kth iteration; ub and lb are the upper and lower boundary values ​​of the particle position respectively; x k+1 Iterate the population for the final next step; E k+1 is the chaotic sequence value at k+1 iterations.

8. A novel power system substation intelligent operation and maintenance method based on improved SSA-PSO algorithm according to claim 6, characterized in that: In step S43, when the SSA algorithm updates the local search factor parameters, it specifically performs the following steps: S43.1: Set a m A population of sparrows and the initial position of each sparrow are determined; S43.2: Update the current position of the discoverer, joiner and sparrow according to the algorithm rules; S43.3: Parameter Y for initial amplitude step_0 Optimize S43.4: Check whether the preset maximum number of iterations has been reached. If the condition is met, record the parameters of the current local optimal search disturbance amplitude; if not, return to step S43.2 and continue the iterative process until the condition is met.

9. A novel power system substation intelligent operation and maintenance method based on improved SSA-PSO algorithm according to claim 8, characterized in that: In step S43.2, the current positions of the discoverer, joiner and sparrow are updated, which specifically includes the following steps: Update discoverer: Where: is the position of the pth sparrow after k iterations in the q dimension; θ is the distribution factor; k max is the maximum number of iterations; R is a random number that obeys a normal distribution; L is a matrix with a distribution; Q is the sparrow escape factor; ST k is the k-th iteration dynamic safety threshold; is the position of the pth sparrow after k+1 iterations in the q dimension; θ×k max is the product of the distribution factor and the maximum number of iterations. Updated joiners are: Where: is the global worst position at the tth iteration; is the best position of the sparrow in the history at the kth iteration; ψ is an array of randomly assigned values ​​1 or -1; n p is the step length of the pth sparrow; n m is the maximum stride length of the sparrow; Update the sparrow position to: Where: Y is the best position of the sparrow among all the historical positions of the sparrows in the entire sparrow group at the kth iteration; w is the parameter of the worst disturbance amplitude; Y is the parameter of the initial disturbance amplitude; the sparrow population continuously updates its position to find the optimal step_0 Parameters for the local search perturbation amplitude; is the parameter of the k+1th disturbance amplitude; B is a random parameter between [0,1], where the parameter Y that optimizes the initial disturbance amplitude step_0 It is expressed as follows: In the formula, is the parameter of the k+1th disturbance amplitude; is the position of the pth sparrow after k+1 iterations in the q dimension.

10. A novel power system substation intelligent operation and maintenance method based on improved SSA-PSO algorithm according to claim 1, characterized in that: In step S4, a local search factor parameter is introduced to update the particle state, which specifically includes the following steps: The position vector L of each particle l l =(L l1 ,L l2 ,...,L ld ) and the velocity vector V l =(V l1 ,V l2 ,...,V ld ), L l1 , L l2 is the position of particle l in the first and second dimensions; L ld is the position of particle l in the dth dimension; V l1 、V l2 is the velocity of particle l in the first and second dimensions; V ld is the velocity of particle l in the dth dimension; In each iteration, the particle's velocity and position are updated according to the following formula: Where: V l t is the velocity of particle l at the tth iteration; is the position of particle l at the tth iteration; V l t+1 is the velocity of particle l at the t+1th iteration; The position of particle l at iteration t+1; is the best position of particle l in the historical position at the tth iteration; is the best position of particle l among all the historical positions of particles in the entire particle swarm at the tth iteration; r1 and r2 are two randomly distributed random numbers in the range [0,1]; b1 and b2 are acceleration factors, representing the learning factors of individuals and society respectively; c1 is a random number generation function, which is used to generate random numbers within a certain range and randomly perturb the position during local search; Y step_size Parameter that controls the amplitude of the local search disturbance.

Images