Genetic-random forest algorithm-based power grid disaster assessment system and method

Abstract

The invention discloses a power grid disaster assessment system and method based on a genetic-random forest algorithm, and relates to the technical field of smart power grids, and the method comprises the steps: obtaining an assessment index through the calculation of an analysis and assessment module for the damage of an extreme event to a high-proportion new energy power grid; inputting the evaluation indexes into a high-proportion new energy power grid data acquisition and processing module based on regional grid partition for acquisition and preprocessing to obtain a normalized data set; inputting the normalized data set into a high-proportion new energy power grid disaster condition solving module based on a genetic-random forest algorithm to obtain an optimal parameter of a prediction model for high-proportion new energy power grid damage under an extreme event; and evaluating the power grid disaster through the prediction model for the high-proportion new energy power grid damage under the extreme event. Aiming at the extreme event problem possibly occurring in the future, an evaluation model for power grid damage under the condition of the extreme event is creatively fused, and a brand-new pre-disaster prevention work scheme is provided for the power grid.

Description

Technical Field

[0001] The present invention relates to the technical field of smart grids, and in particular to a grid disaster assessment system and method based on a genetic-random forest algorithm. Background Art

[0002] With the increasing complexity of the power system, especially the advancement of high-proportion renewable energy (such as wind power, solar energy, etc.) grid connection, the security and stability of the power grid are facing unprecedented challenges. In recent years, with the frequent occurrence of extreme climate events (such as typhoons, earthquakes, snowstorms, etc.), the disaster resistance and post-disaster recovery capabilities of the power system have received more and more attention. New energy power grids are particularly vulnerable to extreme weather conditions because of their dependence on climate conditions and renewable energy characteristics. These extreme events may cause large-scale power grid shutdowns, equipment damage, and long-term power outages, which in turn may cause large-scale economic losses, social chaos, and even safety accidents. Therefore, how to effectively assess the impact of extreme events on the power grid, predict the power grid disaster situation in advance, and propose targeted emergency recovery measures has become a technical problem that needs to be solved urgently in the power industry.

[0003] At present, the field of power grid disaster assessment at home and abroad mainly relies on risk assessment methods based on traditional statistical methods or simplified models. These traditional methods usually lack comprehensive consideration of the variable factors in complex power grid systems, and have low accuracy and timeliness in extreme event prediction and post-disaster recovery assessment. For example, most of the existing disaster assessment systems use simple assessment models based on experience or regional historical data. These models are often insufficient for damage assessment of high-proportion new energy power grids, and it is difficult to fully capture the complex dynamic characteristics of power grids and the interaction of disaster factors. In addition, existing technologies usually do not make full use of big data and artificial intelligence algorithms to process multi-dimensional data of power grid disasters under extreme events, nor do they take into account the impact of micro-topography characteristics on power grid disasters, and lack effective prediction and emergency response mechanisms. Therefore, the existing assessment system fails to meet the safety requirements of modern power systems under extreme climates, especially for pre-disaster pre-assessment, risk prediction and post-disaster recovery planning of high-proportion new energy power grids.

[0004] The present invention proposes an innovative solution to the deficiencies of the prior art by introducing a disaster assessment method based on a genetic-random forest algorithm. First, the potential threat of extreme events to the power grid is comprehensively assessed through four dimensions: the disaster-inducing impact coefficient, the disaster-producing impact coefficient, the disaster-affected impact coefficient, and the disaster resistance coefficient. Combined with micro-topography data and historical disaster data, the defects of insufficient consideration of environmental factors and power grid complexity in the existing methods are made up. Secondly, by combining genetic algorithms and random forest algorithms, the disaster prediction model of the power grid can be automatically learned and optimized from a large amount of nonlinear data, which not only improves the accuracy of the assessment, but also enhances the early warning capability of extreme events. Through this method, the disaster risk of the power grid in extreme events can be predicted more scientifically and accurately, providing the power sector with a more effective pre-disaster pre-assessment plan and post-disaster recovery strategy, thereby reducing the losses caused by disasters and improving the disaster resistance and recovery capabilities of the power grid, with obvious technical advantages and application prospects. Summary of the invention

[0005] In view of the above-mentioned problems, the present invention is proposed.

[0006] Therefore, the technical problem solved by the present invention is: the deficiencies of the existing power system post-disaster recovery assessment system in dealing with extreme events (such as typhoons, earthquakes, snowstorms, etc.), especially the lack of effective pre-disaster pre-assessment and post-disaster emergency response capabilities before extreme events occur, and the inability to predict in advance the impact of disasters on high-proportion new energy power grids, and the inability to scientifically plan emergency repair plans to reduce the losses caused by disasters. Specifically, the present invention constructs an assessment system through four aspects: disaster-induced impact coefficient, disaster-produced impact coefficient, disaster-affected impact coefficient, and disaster resistance coefficient. Combined with historical data and micro-topography features, the genetic-random forest algorithm is used to predict power grid disasters, thereby achieving pre-disaster scientific assessment and post-disaster rapid recovery planning, improving the disaster response capability of the power system under extreme events, shortening recovery time, reducing economic losses, and ensuring social stability.

[0007] In order to solve the above technical problems, the present invention provides the following technical solution: a power grid disaster assessment system based on a genetic-random forest algorithm, which comprises the following steps:

[0008] An analysis and assessment module for the damage caused by extreme events to power grids with a high proportion of renewable energy, a data collection and processing module for power grids with a high proportion of renewable energy based on regional grid partitioning, and a disaster situation solution module for power grids with a high proportion of renewable energy based on a genetic-random forest algorithm.

[0009] The analysis and evaluation module of the damage of extreme events to the high-proportion new energy power grid is used to calculate the evaluation index and input it into the high-proportion new energy power grid data acquisition and processing module based on regional grid partitioning;

[0010] The high-proportion new energy power grid data acquisition and processing module based on regional grid partition processes the evaluation index to obtain a normalized data set, and inputs it into the high-proportion new energy power grid disaster situation solving module based on genetic-random forest algorithm;

[0011] The high-proportion new energy power grid disaster situation solving module based on the genetic-random forest algorithm completes the output of optimal parameters based on the normalized data set and constructs a prediction model for damage to the high-proportion new energy power grid under extreme events.

[0012] As a preferred solution of the power grid disaster assessment system based on the genetic-random forest algorithm described in the present invention, the analysis and assessment module of the damage of extreme events to a high-proportion new energy power grid includes a disaster-induced impact coefficient assessment submodule, a disaster-produced impact coefficient assessment submodule, a disaster-stricken impact coefficient assessment submodule, a disaster resistance coefficient assessment submodule and an assessment index statistics submodule.

[0013] The high-proportion new energy power grid data acquisition and processing module based on regional grid partitioning includes a data acquisition module and a data processing module.

[0014] The high-proportion new energy power grid disaster situation solving module based on the genetic-random forest algorithm includes a training set and a test set division and parameter individual encoding submodule, a random forest establishment submodule, a data set testing submodule, an individual parameter selection, crossover, and mutation submodule, and an optimal parameter detection output module.

[0015] As a preferred solution of the power grid disaster assessment system based on the genetic-random forest algorithm described in the present invention, the disaster impact coefficient assessment submodule constructs a disaster impact coefficient index through the local precipitation frequency evaluation index and the annual precipitation.

[0016] The disaster impact coefficient assessment submodule constructs a disaster impact coefficient index through the elevation, slope aspect and forest coverage of the region.

[0017] The disaster impact coefficient assessment submodule constructs a disaster impact coefficient index through the gross domestic product and population of the region.

[0018] The disaster resistance coefficient assessment submodule constructs a disaster resistance coefficient index through the region's per capita GDP and economic growth rate.

[0019] The disaster-induced impact coefficient index, the disaster-inducing impact coefficient index, the disaster-affected impact coefficient index, and the disaster resistance coefficient index are input into the evaluation index statistical submodule to obtain the statistically evaluated evaluation index.

[0020] As a preferred solution of the power grid disaster assessment system based on the genetic-random forest algorithm described in the present invention, the data acquisition module collects the assessment indicators to obtain a data set.

[0021] The data processing module is responsible for normalizing the data set.

[0022] As a preferred solution of the power grid disaster assessment system based on the genetic-random forest algorithm described in the present invention, the training set and test set division and parameter individual encoding submodule divides the data set according to the set ratio and encodes the individual parameters to obtain input data.

[0023] The random forest establishment submodule collects input data according to the Bagging method, and assigns the collected results to the root node of the decision tree as the parent node.

[0024] By calculating the Gini coefficient between the parent node and the possible child nodes, the nodes with a Gini coefficient less than the limit value are selected for division, that is, the internal nodes of the tree are obtained. By repeatedly calculating the Gini coefficient between the parent and child nodes, selection is performed, and finally the leaf nodes of the tree are obtained, and the random forest data is output.

[0025] The data set testing submodule uses the comparison between real data and predicted data as an indicator for evaluating the accuracy of the random forest algorithm to test the output random forest data and output a test data set.

[0026] The individual parameter selection, crossover, and mutation submodules optimize the test data set, use a genetic algorithm to optimize the number N of decision trees and the number M of feature selection of the random forest, and finally output the optimization parameters.

[0027] The optimized parameters are fed back to the random forest establishment submodule for iteration. When the number of iterations reaches the requirement, the optimized parameters are input into the optimal parameter detection output module to complete the output of the optimal parameters.

[0028] Another object of the present invention is to provide a method for evaluating power grid disasters based on a genetic-random forest algorithm, which can scientifically evaluate and predict the impact of extreme events on the power grid through four dimensions: disaster-induced impact coefficient, disaster-produced impact coefficient, disaster-affected impact coefficient, and disaster resistance coefficient, combined with historical data and micro-topographic features, using a genetic-random forest algorithm, thereby effectively predicting the disaster situation of a high-proportion new energy power grid in extreme events. This method solves the problem that the existing power grid disaster assessment system lacks a pre-disaster warning mechanism based on big data and complex algorithms, and cannot identify disaster risks and plan emergency response measures in advance.

[0029] In order to solve the above technical problems, the present invention provides the following technical solutions: a power grid disaster assessment method based on a genetic-random forest algorithm, comprising:

[0030] The evaluation index is calculated through the analysis and evaluation module of the damage caused by extreme events to the high proportion of new energy power grid.

[0031] The evaluation indicators are input into the high-proportion new energy power grid data acquisition and processing module based on regional grid partitioning.

[0032] The evaluation values ​​are collected and preprocessed through the high-proportion new energy power grid data acquisition and processing module based on regional grid partitioning to obtain a normalized data set.

[0033] The normalized data set is input into the disaster situation solution module of the high-proportion renewable energy power grid based on the genetic-random forest algorithm to obtain the optimal parameters of the prediction model for damage to the high-proportion renewable energy power grid under extreme events.

[0034] The damage to the power grid is assessed through a prediction model for damage to a power grid with a high proportion of renewable energy under extreme events.

[0035] As a preferred scheme of the power grid disaster assessment method based on genetic-random forest algorithm described in the present invention, the disaster impact coefficient assessment submodule in the analysis and assessment module of the damage of extreme events to high-proportion new energy power grid constructs a disaster impact coefficient index through the local precipitation frequency evaluation index and annual precipitation.

[0036] The precipitation frequency evaluation index is expressed as:

[0037]

[0038] Among them, A represents the precipitation frequency evaluation index, b n Indicates the frequency of precipitation level n.

[0039] The disaster impact coefficient assessment submodule in the analysis and assessment module of the damage of extreme events to high-proportion new energy power grids constructs a disaster impact coefficient index through the elevation, slope aspect and forest coverage of the area.

[0040] Divide the grid into 5km×5km, and use the forest coverage area in the grid and the total area of ​​the grid to get the forest coverage rate of the grid, expressed as,

[0041]

[0042] Among them, m n The forest coverage area of ​​the nth grid, M represents the total area of ​​the grid, R n Represents the forest land coverage of the grid.

[0043] As a preferred solution of the power grid disaster assessment method based on the genetic-random forest algorithm described in the present invention, wherein: the acquisition module in the high-proportion new energy power grid data acquisition and processing module based on regional grid partitioning collects the assessment indicators, and the sample data of the dth grid under extreme event conditions is as follows,

[0044] c dq =[c d1 ,c d2 ,c d3 ,…c dn ],q=1…n

[0045] Among them, c dq It represents the value of the qth input variable in the dth network sample, and the data set is obtained after collection.

[0046] The data set is input into the data processing module for normalization, which is expressed as:

[0047]

[0048] Among them, c represents sample metadata, S represents normalized data, and c M Indicates the maximum value in the sample data, c m It represents the minimum value in the sample data, and the normalized data set is obtained by calculation and sorting.

[0049] The random forest establishment submodule in the high-proportion new energy power grid disaster situation solving module based on the genetic-random forest algorithm outputs random forest data.

[0050] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the steps of the power grid disaster assessment method based on the genetic-random forest algorithm are implemented.

[0051] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the power grid disaster assessment method based on the genetic-random forest algorithm as described above.

[0052] Beneficial effects of the present invention: The present invention innovatively integrates an assessment model for power grid damage under extreme event conditions to address the problem of extreme events that may occur in the future, providing the power grid with a new pre-disaster prevention solution. The assessment model for power grid damage under extreme event conditions collects and processes data on a 5km×5km grid, which not only ensures the high accuracy of input data, but also significantly reduces the difficulty of data collection, speeds up the collection speed, and uses a genetic random forest algorithm for prediction, which has a high accuracy rate and the predicted data is more in line with the actual data, providing strong support for pre-disaster prevention measures and post-disaster rapid repair work. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0054] Figure 1 The overall framework diagram of the power grid disaster assessment system based on the genetic-random forest algorithm provided in the first embodiment of the present invention.

[0055] Figure 2 A structural diagram of an analysis and evaluation module for damages caused by extreme events to a high proportion of new energy power grids in a power grid disaster assessment system based on a genetic-random forest algorithm provided in the first embodiment of the present invention.

[0056] Figure 3 This is a structural diagram of a high-proportion new energy power grid data collection and processing module based on regional grid partitioning in a power grid disaster assessment system based on a genetic-random forest algorithm provided in the first embodiment of the present invention.

[0057] Figure 4 A structural diagram of a high-proportion new energy power grid disaster solving module based on a genetic-random forest algorithm in a power grid disaster assessment system based on a genetic-random forest algorithm provided in the first embodiment of the present invention.

[0058] Figure 5 An overall flow chart of a power grid disaster assessment method based on a genetic-random forest algorithm provided for a second embodiment of the present invention.

[0059] Figure 6 An accuracy analysis diagram of the genetic-random forest algorithm in the power grid disaster assessment method based on the genetic-random forest algorithm provided in the third embodiment of the present invention.

[0060] Figure 7A prediction, assessment and analysis diagram of power grid load loss in a certain area under extreme events in a power grid disaster assessment method based on a genetic-random forest algorithm provided in the third embodiment of the present invention.

[0061] Figure 8 This is a diagram of the actual results of power grid load loss in a certain area under extreme events in the power grid disaster assessment method based on the genetic-random forest algorithm provided in the third embodiment of the present invention.

[0062] Fig. 9 A distribution map of disaster-affected points in a certain area under extreme events in a power grid disaster assessment method based on a genetic-random forest algorithm provided in the third embodiment of the present invention. DETAILED DESCRIPTION

[0063] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0064] Example 1, reference Figure 1 to Figure 4 , is an embodiment of the present invention, and provides a power grid disaster assessment system based on a genetic-random forest algorithm, which is characterized by:

[0065] An analysis and evaluation module 100 for damages caused by extreme events to a high-proportion renewable energy power grid, a data collection and processing module 200 for a high-proportion renewable energy power grid based on regional grid partitioning, and a disaster situation solving module 300 for a high-proportion renewable energy power grid based on a genetic-random forest algorithm.

[0066] The analysis and evaluation module 100 of damages of extreme events to a high-proportion new energy power grid is used to calculate the evaluation index and input it into the high-proportion new energy power grid data acquisition and processing module 200 based on regional grid partitioning;

[0067] The high-proportion new energy power grid data collection and processing module 200 based on regional grid partitioning processes the evaluation indicators to obtain a normalized data set, and inputs it into the high-proportion new energy power grid disaster situation solving module 300 based on the genetic-random forest algorithm.

[0068] The high-proportion new energy power grid disaster situation solving module 300 based on the genetic-random forest algorithm completes the output of optimal parameters based on the normalized data set and constructs a prediction model for damage to the high-proportion new energy power grid under extreme events.

[0069] Reference Figure 2The analysis and evaluation module 100 of extreme event damage to a high-proportion new energy power grid includes a disaster-induced impact coefficient evaluation submodule 101, a disaster-produced impact coefficient evaluation submodule 102, a disaster-affected impact coefficient evaluation submodule 103, a disaster resistance coefficient evaluation submodule 104, and an evaluation index statistics submodule 105.

[0070] Reference Figure 3 The high-proportion new energy power grid data acquisition and processing module 200 based on regional grid partitioning includes a data acquisition module 201 and a data processing module 202 .

[0071] Reference Figure 4 The high-proportion new energy power grid disaster situation solving module 300 based on the genetic-random forest algorithm includes a training set and a test set division and parameter individual encoding submodule 301, a random forest establishment submodule 302, a data set testing submodule 303, an individual parameter selection, crossover, and mutation submodule 304, and an optimal parameter detection and output module 305.

[0072] The disaster impact coefficient evaluation submodule 101 constructs a disaster impact coefficient index through the local precipitation frequency evaluation index and the annual precipitation.

[0073] The disaster impact coefficient assessment submodule 102 constructs a disaster impact coefficient index through the elevation, slope aspect and forest coverage of the region.

[0074] The disaster impact coefficient assessment submodule 103 constructs a disaster impact coefficient index through the gross domestic product and population of the region.

[0075] The disaster resistance coefficient evaluation submodule 104 constructs a disaster resistance coefficient index through the region's per capita GDP and economic growth rate.

[0076] The disaster-induced impact coefficient index, the disaster-produced impact coefficient index, the disaster-affected impact coefficient index, and the disaster-resistance coefficient index are input into the evaluation index statistical submodule 105 to obtain the statistically evaluated evaluation index.

[0077] The data collection module 201 collects evaluation indicators to obtain a data set.

[0078] The data processing module 202 is responsible for normalizing the data set.

[0079] The training set and test set division and parameter individual coding submodule 301 divides the data set according to the set ratio and encodes the individual parameters to obtain input data.

[0080] The random forest establishment submodule 302 collects input data according to the Bagging method, and distributes the collected results to the root node of the decision tree as the parent node.

[0081] By calculating the Gini coefficient between the parent node and the possible child nodes, the nodes with a Gini coefficient less than the limit value are selected for division, that is, the internal nodes of the tree are obtained. By repeatedly calculating the Gini coefficient between the parent and child nodes, selection is performed, and finally the leaf nodes of the tree are obtained, and the random forest data is output.

[0082] The data set testing submodule 303 uses the comparison between the real data and the predicted data as an indicator for evaluating the accuracy of the random forest algorithm to test the output random forest data and output a test data set.

[0083] The individual parameter selection, crossover, and mutation submodule 304 optimizes the test data set, optimizes the number N of decision trees and the number M of feature selection of the random forest using a genetic algorithm, and finally outputs the optimized parameters;

[0084] The optimized parameters are fed back to the random forest establishment submodule 302 for iteration. When the number of iterations reaches the requirement, the optimized parameters are input to the optimal parameter detection output module 305 to complete the output of the optimal parameters.

[0085] Example 2, reference Figure 5 , is an embodiment of the present invention, and provides a method for a power grid disaster assessment system based on a genetic-random forest algorithm, characterized in that:

[0086] S1: The evaluation index is calculated through the analysis and evaluation module of the damage caused by extreme events to the high proportion of new energy power grid.

[0087] The disaster impact coefficient assessment submodule in the analysis and assessment module of extreme events' damage to high-proportion new energy power grids constructs a disaster impact coefficient index through the local precipitation frequency evaluation index and annual precipitation.

[0088] The precipitation frequency evaluation index is expressed as:

[0089]

[0090] Among them, A represents the precipitation frequency evaluation index, b n It indicates the frequency of precipitation of the nth level;

[0091] The disaster impact coefficient assessment submodule in the analysis and assessment module of extreme events' damage to high-proportion new energy power grids constructs a disaster impact coefficient index through the region's elevation, slope aspect and forest coverage.

[0092] Divide the grid into 5km×5km, and use the forest coverage area in the grid and the total area of ​​the grid to get the forest coverage rate of the grid, expressed as,

[0093]

[0094] Among them, m nThe forest coverage area of ​​the nth grid, M represents the total area of ​​the grid, R n Represents the forest land coverage of the grid.

[0095] S2: Input the evaluation index into the high-proportion new energy power grid data acquisition and processing module based on regional grid partitioning.

[0096] S3: The evaluation values ​​are collected and preprocessed through the high-proportion new energy power grid data collection and processing module based on regional grid partitioning to obtain a normalized data set.

[0097] The collection module in the data collection and processing module of the high-proportion new energy power grid based on regional grid partitioning collects the evaluation indicators. The sample data of the dth grid under extreme event conditions is as follows:

[0098] c dq =[c d1 ,c d2 ,c d3 ,…c dn ],q=1…n

[0099] Among them, c dq It represents the value of the qth input variable in the dth network sample, and the data set is obtained after collection;

[0100] The data set is input into the data processing module for normalization, which is expressed as:

[0101]

[0102] Among them, c represents sample metadata, S represents normalized data, and c M Indicates the maximum value in the sample data, c m It represents the minimum value in the sample data, and the normalized data set is obtained by calculation and sorting;

[0103] S4: The normalized data set is input into the high-proportion renewable energy power grid disaster situation solution module based on the genetic-random forest algorithm to obtain the optimal parameters of the prediction model for damage to the high-proportion renewable energy power grid under extreme events.

[0104] The random forest establishment submodule in the high-proportion new energy power grid disaster situation solution module based on the genetic-random forest algorithm outputs random forest data.

[0105] Among them, the Gini coefficient expression of probability distribution is as follows:

[0106]

[0107] Among them, the probability that S takes the value of a is g a .

[0108] Applied to the decision tree, for sample Y, the sample size is Q, and the Gini coefficient expression of sample Y is as follows:

[0109]

[0110] Among them, P a Indicates the number of the a-th category in the sample.

[0111] The data set testing submodule 33 uses the comparison between the real data and the predicted data as an indicator for evaluating the accuracy of the random forest algorithm to test the output random forest data and output a test data set.

[0112] Considering that the classification effect (error rate) of random forest is related to the number of decision trees N in the forest and the number of feature selections M of the decision trees. The test data set is optimized through individual parameter selection, crossover, and mutation submodules, and the number of decision trees N and the number of feature selections M of the random forest are optimized using genetic algorithms to finally output the optimization parameters. The optimized parameters will return to the random forest establishment submodule for iteration. When the number of iterations reaches the requirement, the optimized parameters will be input into the optimal parameter detection output module to complete the output of the optimal parameters.

[0113] S5: Assess the damage to the power grid through a prediction model for damage to a high proportion of renewable energy power grids under extreme events.

[0114] Specifically, the prediction model established based on the genetic-random forest algorithm first combines evaluation indicators such as the disaster-induced impact coefficient, the disaster-produced impact coefficient, the disaster-affected impact coefficient, and the disaster resistance coefficient, and uses regional grid partition data to model the possible impact on the power grid when an extreme event occurs. Through training on historical extreme event data, the model can accurately predict the degree of damage and regional distribution that the power grid may encounter before a new extreme event occurs. Through the prediction results of the model, the weak areas of the power grid can be identified, the risks that the power grid may face can be assessed, and targeted disaster prevention and mitigation measures can be taken in advance, such as adjusting the power grid load, optimizing the allocation of emergency repair resources, and preparing necessary equipment and personnel to improve the recovery speed of the power grid in disasters and minimize economic losses. In addition, the prediction model can also provide decision-making support for the post-disaster repair and recovery planning of the power grid, ensure the stability and continuity of power supply, and ensure that the social and economic activities in the region are not affected.

[0115] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program code.

[0116] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0117] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0118] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0119] Example 3, reference Figure 6-Figure 9 In this embodiment, in order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation experiments. Figure 6 This is the accuracy analysis chart of the genetic-random forest algorithm. The genetic population size is set to 100, the crossover rate of the algorithm is 0.8, the mutation rate is 0.2, and the maximum number of iterations is 200 generations to ensure that the algorithm can converge quickly. The accuracy of the optimized random forest algorithm is as follows: Figure 6 As shown, the optimal parameter individuals are 232 decision trees, 8 feature selections, and the highest accuracy is 92.31% (a total of 208, 192 correct).

[0120] Figure 7-Figure 8 This is a comparative analysis chart of the predicted evaluation and actual results of power grid load loss in a certain area under extreme events. Figure 7 For prediction, Figure 8 To be realistic, a parameter combination with 232 decision trees and 8 feature selections was selected to train the random forest algorithm, and the number of power outage users in the target area under flood disasters was predicted and compared with the actual number of power outage users. The results showed that 954 out of a total of 1040 grids were predicted correctly, with an accuracy rate of 91.73%.

[0121] Fig. 9 It is a location distribution map of disaster-stricken points in a certain area under extreme events. In order to better arrange power repair tasks and realize power supply to users in the area as soon as possible, grids that are more severely affected by the disaster are selected for repair, and the center point of each grid in the area and the location of a local power supply company are calculated to facilitate route planning for the power repair team.

[0122] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention 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 invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A power grid disaster assessment system based on genetic-random forest algorithm, characterized in that: include: An analysis and evaluation module (100) for damages caused by extreme events to a high-proportion new energy power grid, a data collection and processing module (200) for a high-proportion new energy power grid based on regional grid partitioning, and a disaster situation solving module (300) for a high-proportion new energy power grid based on a genetic-random forest algorithm; The analysis and evaluation module (100) of damage caused by extreme events to a high-proportion new energy power grid is used to calculate evaluation indicators and input them into a high-proportion new energy power grid data acquisition and processing module (200) based on regional grid partitions; The high-proportion new energy power grid data collection and processing module (200) based on regional grid partition processes the evaluation index to obtain a normalized data set, and inputs the data set into the high-proportion new energy power grid disaster situation solving module (300) based on the genetic-random forest algorithm; The high-proportion new energy power grid disaster situation solving module (300) based on the genetic-random forest algorithm completes the output of optimal parameters based on the normalized data set and constructs a prediction model for damage to the high-proportion new energy power grid under extreme events.

2. The power grid disaster assessment system based on genetic-random forest algorithm as claimed in claim 1, characterized in that: The analysis and evaluation module (100) for damage caused by extreme events to a high-proportion new energy power grid comprises a disaster-inducing impact coefficient evaluation submodule (101), a disaster-producing impact coefficient evaluation submodule (102), a disaster-affected impact coefficient evaluation submodule (103), a disaster resistance coefficient evaluation submodule (104) and an evaluation index statistics submodule (105); The high-proportion new energy power grid data acquisition and processing module (200) based on regional grid partitioning comprises a data acquisition module (201) and a data processing module (202); The high-proportion new energy power grid disaster situation solving module (300) based on the genetic-random forest algorithm includes a training set and a test set division and parameter individual coding submodule (301), a random forest establishment submodule (302), a data set testing submodule (303), an individual parameter selection, crossover, and mutation submodule (304), and an optimal parameter detection and output module (305).

3. The power grid disaster assessment system based on genetic-random forest algorithm as claimed in claim 2, characterized in that: The disaster impact coefficient evaluation submodule (101) constructs a disaster impact coefficient index through the local precipitation frequency evaluation index and annual precipitation; The disaster impact coefficient assessment submodule (102) constructs a disaster impact coefficient index through the elevation, slope and forest coverage of the region; The disaster impact coefficient assessment submodule (103) constructs a disaster impact coefficient index through the gross domestic product and population of the region; The disaster resistance coefficient evaluation submodule (104) constructs a disaster resistance coefficient index through the region's per capita GDP and economic growth rate; The disaster-inducing impact coefficient index, the disaster-producing impact coefficient index, the disaster-affected impact coefficient index, and the disaster-resistance coefficient index are input into an evaluation index statistical submodule (105) to obtain statistically evaluated evaluation indicators.

4. The power grid disaster assessment system based on genetic-random forest algorithm as claimed in claim 3, characterized in that: The data collection module (201) collects the evaluation indicators to obtain a data set; The data processing module (202) is responsible for normalizing the data set.

5. The power grid disaster assessment system based on genetic-random forest algorithm as claimed in claim 4, characterized in that: The training set and test set division and parameter individual coding submodule (301) divides the data set according to the set ratio and codes the individual parameters to obtain input data; The random forest establishment submodule (302) collects input data according to the Bagging method, and distributes the collected results to the root node of the decision tree as the parent node; By calculating the Gini coefficient between the parent node and the possible child nodes, the nodes with a Gini coefficient less than the limit value are selected for division, that is, the internal nodes of the tree are obtained. By repeatedly calculating the Gini coefficient between the parent and child nodes, the selection is performed, and finally the leaf nodes of the tree are obtained, and the random forest data is output; The data set testing submodule (303) uses the comparison between the real data and the predicted data as an indicator for evaluating the accuracy of the random forest algorithm to test the output random forest data and output a test data set; The individual parameter selection, crossover, and mutation submodule (304) optimizes the test data set, optimizes the number N of decision trees and the number M of feature selection of the random forest using a genetic algorithm, and finally outputs the optimized parameters; The optimized parameters are fed back to the random forest establishment submodule (302) for iteration. When the number of iterations reaches the requirement, the optimized parameters are input to the optimal parameter detection output module (305) to complete the output of the optimal parameters.

6. A method for a power grid disaster assessment system based on a genetic-random forest algorithm as claimed in any one of claims 1 to 5, characterized in that: include: The evaluation index is calculated through the analysis and evaluation module of the damage of extreme events to the high-proportion new energy power grid; Input the evaluation index into the high-proportion new energy power grid data acquisition and processing module based on regional grid partitioning; The evaluation values ​​are collected and preprocessed by a high-proportion new energy grid data collection and processing module based on regional grid partitioning to obtain a normalized data set; The normalized data set is input into the high-proportion renewable energy power grid disaster situation solution module based on the genetic-random forest algorithm to obtain the optimal parameters of the prediction model for damage to the high-proportion renewable energy power grid under extreme events; The damage to the power grid is assessed through a prediction model for damage to a power grid with a high proportion of renewable energy under extreme events.

7. The power grid disaster assessment method based on genetic-random forest algorithm according to claim 6, characterized in that: The disaster impact coefficient assessment submodule in the analysis and assessment module of the damage of extreme events to high-proportion new energy power grids constructs a disaster impact coefficient index through the local precipitation frequency evaluation index and annual precipitation; The precipitation frequency evaluation index is expressed as: Among them, A represents the precipitation frequency evaluation index, b n It indicates the frequency of precipitation of the nth level; The disaster impact coefficient assessment submodule in the analysis and assessment module of the extreme event damage to the high-proportion new energy power grid constructs a disaster impact coefficient index through the elevation, slope and forest coverage of the region; Divide the grid into 5km×5km, and use the forest coverage area in the grid and the total area of ​​the grid to get the forest coverage rate of the grid, expressed as, Among them, m n The forest coverage area of ​​the nth grid, M represents the total area of ​​the grid, R n Represents the forest land coverage of the grid.

8. The method for power grid disaster assessment based on genetic-random forest algorithm according to claim 7, characterized in that: The acquisition module in the high-proportion new energy grid data acquisition and processing module based on regional grid partitioning acquires the evaluation indicators. The sample data of the dth grid under extreme event conditions is as follows: c dq =[c d1 ,c d2 ,c d3 ,…c dn ],q=1…n Among them, c dq It represents the value of the qth input variable in the dth network sample, and the data set is obtained after collection; The data set is input into the data processing module for normalization, which is expressed as: Among them, c represents sample metadata, S represents normalized data, and c M Indicates the maximum value in the sample data, c m It represents the minimum value in the sample data, and the normalized data set is obtained by calculation and sorting; The random forest establishment submodule in the high-proportion new energy power grid disaster situation solving module based on the genetic-random forest algorithm outputs random forest data.

9. 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 power grid disaster assessment system based on the genetic-random forest algorithm as described in any one of claims 6 to 8 are implemented.

10. 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 power grid disaster assessment system based on the genetic-random forest algorithm as described in any one of claims 6 to 8 are implemented.

Images