Power distribution network reliability evaluation method, device and system, and storage medium

Abstract

The invention discloses a power distribution network reliability evaluation method, device and system, and a storage medium, and the method comprises the steps: S1, carrying out the feature extraction of multi-modal data of a power distribution network, and obtaining feature data; s2, the feature data are encrypted through a Paillier algorithm and then are stored; s3, training a PSO-ELM (Particle Swarm Optimization-Extreme Learning Machine) model of an extreme learning machine optimized by a particle swarm optimization algorithm according to the decrypted, extracted and stored feature data and environmental data; and S4, inputting real-time power distribution service user data and environment data into the trained PSO-ELM model, and outputting to obtain a power distribution network reliability evaluation result. By adopting the technical scheme of the invention, the possibility that the power distribution network data is tampered or leaked is reduced, and the accuracy and speed of evaluation are improved.

Description

Technical Field

[0001] The present invention belongs to the field of information security technology, and in particular relates to a distribution network reliability assessment method and device, a system, and a storage medium. Background Art

[0002] As various industries are increasingly demanding that power companies provide reliable, efficient and clean electricity, power companies are constantly promoting the construction of smart grids. Among them, the distribution network, as the final part of the link, focuses on distribution network monitoring and dispatching services, intelligent services, and distribution network information collection. It has the characteristics of many nodes, wide business coverage, widely distributed equipment, rich information, and diversified business and communication methods.

[0003] In the continuous development of smart distribution networks and cyber-physical fusion systems, data transmission in distribution networks is vulnerable to exploitation and malicious use due to loopholes in management systems and identification programs, and there is a risk of malicious tampering by attackers. At present, the construction of power communication networks is advancing in an orderly manner. At the same time, the number of access devices on the distribution side has further increased, and the access methods have also become more diversified. The business of the power system has derived a rich category, and these business data are transmitted in the smart distribution network. Due to the significant increase in the number of devices, access methods and business types, it is impossible to accurately evaluate the reliability of the smart distribution network. Summary of the invention

[0004] The technical problem to be solved by the present invention is to provide a distribution network reliability assessment method and device, system, and storage medium, which extracts features from the multi-mode business data of the distribution network, encrypts the data through the Paillier algorithm, and stores the data, thereby reducing the possibility of tampering or leakage of the distribution network data. In the reliability assessment of the distribution network, after decrypting and extracting the stored data, the extreme learning machine model is enhanced using the particle swarm optimization algorithm to improve the accuracy and speed of the assessment.

[0005] To achieve the above object, the present invention adopts the following technical solution:

[0006] A distribution network reliability assessment method, comprising:

[0007] Step S1, extracting features from multimodal data of the distribution network to obtain feature data;

[0008] Step S2, storing the characteristic data after encryption by Paillier algorithm;

[0009] Step S3, training a PSO-ELM model of an extreme learning machine optimized by a particle swarm optimization algorithm according to the decrypted extracted stored feature data and environmental data;

[0010] Step S4: input the real-time distribution service user data and environmental data into the trained PSO-ELM model, and output the distribution network reliability assessment result.

[0011] Preferably, in step S1, natural language processing technology is used to judge the importance of entries in the power distribution service user data according to the frequency of occurrence, so as to extract data features.

[0012] Preferably, step S3 comprises:

[0013] According to the reliability influencing factor system of distribution network, select reliability evaluation indicators;

[0014] Extract the stored characteristic data and environmental data according to the decryption to obtain data related to the reliability influencing factors;

[0015] Based on the reliability evaluation index and the data related to the reliability influencing factors, an evaluation decision table is formed;

[0016] The PSO-ELM model is trained according to the evaluation decision table; wherein, the data related to the reliability influencing factors are used as the input of the PSO-ELM model, and the reliability evaluation index of the distribution network is used as the output to train the PSO-ELM model.

[0017] The present invention also provides a distribution network reliability assessment device, comprising:

[0018] A feature extraction module is used to extract features from multi-modal data of the distribution network to obtain feature data;

[0019] An encryption module, used to store feature data after encryption through Paillier algorithm;

[0020] A training module, used to train a PSO-ELM model of an extreme learning machine optimized by a particle swarm optimization algorithm based on the decrypted extracted stored feature data and environmental data;

[0021] The evaluation module is used to input real-time distribution business user data and environmental data into the trained PSO-ELM model and output the distribution network reliability evaluation results.

[0022] Preferably, the feature extraction module uses natural language processing technology to judge the importance of entries in the power distribution service user data according to the frequency of occurrence, so as to extract data features.

[0023] Preferably, the training module includes:

[0024] A selection unit, used for selecting a reliability evaluation index according to a distribution network reliability influencing factor system;

[0025] A screening unit, used to extract the stored characteristic data and environmental data according to the decryption, and obtain data related to reliability influencing factors;

[0026] A building unit for forming an evaluation decision table based on reliability evaluation indicators and data related to reliability influencing factors;

[0027] The training unit is used to train the PSO-ELM model according to the evaluation decision table; wherein the data related to the reliability influencing factors are used as the input of the PSO-ELM model, and the reliability evaluation index of the distribution network is used as the output to train the PSO-ELM model.

[0028] An embodiment of the present invention further provides a distribution network reliability assessment system, comprising: a memory and a processor, wherein the memory stores a computer program executed by the processor, and the computer program executes a distribution network reliability assessment method when executed by the processor.

[0029] An embodiment of the present invention further provides a storage medium, on which a computer program is stored, and the computer program executes the distribution network reliability assessment method when running.

[0030] The present invention utilizes data feature extraction and transmission encryption and decryption to perform reliability assessment on distribution network data, extracts features from data of different services of the distribution network, encrypts and stores them in a classified manner, and combines artificial intelligence with a particle swarm optimization algorithm to optimize the extreme learning machine model on the premise of protecting data security, so as to more quickly and effectively perform reliability assessment on the distribution network. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0032] Figure 1 This is a flow chart of a method for evaluating the reliability of a distribution network according to an embodiment of the present invention;

[0033] Figure 2 This is a schematic diagram of the communication network architecture of the distribution network;

[0034] Figure 3 Flowchart for training the PSO-ELM model. DETAILED DESCRIPTION

[0035] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0036] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0037] Embodiment 1:

[0038] like Figure 1 As shown, an embodiment of the present invention provides a distribution network reliability assessment method, comprising:

[0039] Step S1, extracting features from multimodal data of the distribution network to obtain feature data;

[0040] Step S2, storing the characteristic data after encryption by Paillier algorithm;

[0041] Step S3, training a PSO-ELM model of an extreme learning machine optimized by a particle swarm optimization algorithm according to the decrypted extracted stored feature data and environmental data;

[0042] Step S4: input the real-time distribution service user data and environmental data into the trained PSO-ELM model, and output the distribution network reliability assessment result.

[0043] As an implementation method of an embodiment of the present invention, in step S1, the construction of the energy Internet and the application of the Internet of Things technology have continuously advanced the construction of the smart distribution network, promoting the generation of differentiated service needs of users and the process of electricity market reform. The traditional "vertically closed" distribution network architecture is difficult to meet the safety, reliability, and economic requirements of the current distribution business. The smart distribution network focuses on the production capacity, transportation capacity, and energy storage on the distribution side, and continues to extend to the user side. In this scenario, the distribution network communication network architecture formed by the combination of the distribution network and the Internet of Things technology is similar to the structure of the power Internet of Things. It is composed of a communication network, a distribution Internet of Things platform, an edge computing terminal, and smart terminal devices. Figure 1 shown.

[0044] from Figure 2It can be seen that the intelligent terminal equipment is directly connected to the distribution area equipment, user meters, sensors, etc., collects and uploads distribution business user data and environmental data, and executes relevant control instructions issued by the distribution Internet of Things platform and edge computing terminals. With the rapid development of Internet of Things technology and sensor technology, intelligent terminals are evolving in the direction of miniaturization and modularization, and can support multiple communication methods and adapt well to different work requirements and network transmission environments. These collected data constitute multimodal data, and different types of data have different processing and transmission requirements. Feature extraction technology is used to process these data.

[0045] The acquisition end collects distribution network-related business data, uses natural language processing technology to extract data features, and judges the importance of terms based on the frequency of their appearance. A sentence consists of multiple words, and different sentences have different functions in power grid data. Similarly, their importance is also different. Therefore, the determination of term weights also plays an important role in distribution network data transmission. The formula is as follows:

[0046]

[0047] TF-IDF=TF×IDF (3)

[0048] Among them, f represents the number of times the specified term t appears in document d, |d| represents the total number of terms in d, D| represents the total number of expected documents, and |{d∈D:t∈d}| represents the result of the number of documents containing the specified term t. This is also a representative information retrieval and data mining weighting technology, known as the Term Frequency-Inverse Document Frequency (TF-IDF) algorithm.

[0049] As an implementation method of an embodiment of the present invention, in step S2, there is a close connection between the smart distribution network and the Internet of Things, and the security of their business transmission plays an important role in the connection between the two. In the security protection mechanism of the Internet of Things, an important option is to encrypt data information. The safe and accurate transmission of data at the distribution network acquisition end is an important part of the reliability assessment of the distribution network. In the communication between the smart distribution network and the Internet of Things devices, there are three main ways to implement information encryption processing, namely, encryption of the Internet of Things front-end hardware equipment and retrograde encryption, encryption using software, and encryption during the acquisition communication process.

[0050] After feature extraction, the Paillier homomorphic encryption algorithm is used to encrypt the distribution network business data, transmit the encrypted data and store it. When the data needs to be read and used by the user or manager, it is decrypted. In this process, the encryption algorithm can prevent data leakage or tampering.

[0051] The specific process of the Paillier encryption algorithm is as follows:

[0052] (1) Generate public key G and private key S:

[0053] Select two large prime numbers p and q, denote the product of p and q as N, and denote the greatest common divisor as λ.

[0054] N=p*q (4)

[0055] λ=lcm(p-1,q-1) (5)

[0056] Choose a random number g, And satisfy:

[0057] gcd(L(g λ modN 2 ),N)=1 (6)

[0058] in,

[0059]

[0060] Then there is a corresponding plaintext d:

[0061] d=(L(g λ modN 2 )) -1 (8)

[0062] In summary, we get the public key G(N,g) and private key S(λ,d).

[0063] (2) Encrypt the processed data:

[0064] Take any random number r, and r,d <N, The ciphertext c is calculated using the following formula:

[0065] c=g d r N modN 2 (9)

[0066] (3) After the user or administrator reads the data, the obtained data is decrypted:

[0067]

[0068] As an implementation method of an embodiment of the present invention, in step S3, the reliability assessment of the distribution network has certain reliability indicators as measurement standards, which are mainly related to its own equipment, related grid structure and the natural environment in which it is located. Therefore, in the process of reliability assessment, it is necessary not only to consider the collection of related equipment and user information by the distribution network acquisition end, but also to consider the influence of external factors such as weather, screen and process the massive data inside and outside the distribution network, analyze and evaluate the data, and further train the extreme learning machine (ELM) model optimized by the particle swarm optimization algorithm (PSO, Particle Swarm Optimization), hereinafter referred to as the PSO-ELM model, so as to evaluate the reliability level of the distribution network, so as to better provide guidance for the overall planning and operation management of the distribution network.

[0069] According to the relevant standards "Technical Guidelines and Preparation Instructions for Distribution Network Planning and Design" and "Distribution Network Comprehensive Evaluation Index System Table" issued by power grid companies, the factors affecting the reliability index of long-term distribution networks can be mainly divided into five categories, namely weather conditions and grid structure level, equipment quality, operation and maintenance level, and equipment level. Since the five categories contain multiple indicators, a total of 20 typical and representative indicators were selected to form a reliability influencing factor system. The specific names and calculation methods are shown in Table 1.

[0070] Table 1

[0071]

[0072]

[0073] The reliability evaluation index selected is the reliability on service index (RS), which corresponds to the average service availability index (ASAI) in the reliability index of the distribution network system, that is, the ratio of the possible power supply hours of the distribution system to the power supply hours required by users in a year.

[0074] Further, step S3 includes:

[0075] According to the reliability influencing factor system of distribution network, select reliability evaluation indicators;

[0076] Extract the stored characteristic data and environmental data according to the decryption to obtain data related to the reliability influencing factors;

[0077] Based on the reliability evaluation index and the data related to the reliability influencing factors, an evaluation decision table is formed;

[0078] The PSO-ELM model is trained according to the evaluation decision table; wherein, the data related to the reliability influencing factors are used as the input of the PSO-ELM model, and the reliability evaluation index of the distribution network is used as the output to train the PSO-ELM model.

[0079] Further, if Figure 3 As shown, the training PSO-ELM model is as follows:

[0080] The extreme learning machine ELM is a fast single hidden layer feedforward neural network. Compared with the traditional artificial neural network, it has a faster learning speed. There is no need for iteration from the hidden layer to the output layer. The weight matrix of the input layer and the threshold matrix of the hidden layer are randomly generated and do not need to be adjusted in the subsequent processing and calculation process. At the same time, the output layer weights are obtained by minimizing the least squares solution of the square loss function, and finally converted into the solution of the Moore-Penrose generalized inverse of a matrix. Under such steps, the integrity of the network is enhanced, and there is no need for iterative steps like traditional artificial neural networks, which is of great significance to reducing the time for adjusting network parameters.

[0081] Assuming that the trained ELM model has Y sets of training data and p hidden layer nodes, the model can be represented by the following formula:

[0082]

[0083] Among them, w j represents the input weight of the jth hidden layer node; β j represents the output weight of the jth hidden layer node; b j represents the threshold of the jth hidden layer node; o i represents the output of the hidden layer.

[0084] Output weight β in ELM j It should be obtained by minimizing the training error. In this process, w j 、b i It should be randomly generated and remain unchanged during the training process. The ELM model is required to minimize the error close to the training sample, that is, there is β j 、w j 、b j Can satisfy the equation:

[0085]

[0086] If H is used to mark the output matrix of the hidden layer, the above formula can be expressed as:

[0087] Hβ=T (13)

[0088]

[0089] For β j 、w j 、b j Solve, where i = 1, 2, ..., Y; j = 1, 2, ..., p.

[0090]

[0091] It is equivalent to minimizing the loss function E.

[0092]

[0093] When H is uniquely determined, the weight w j and threshold b j The output weight β is randomly generated and does not change during the training process. The least squares solution of the linear equation Hβ=T is used to obtain the output weight β j .

[0094] β=H'T (17)

[0095] Among them, H' is the Moore-Penrose generalized inverse of the hidden layer matrix H, which can minimize the final training error.

[0096] Since the random determination of the number of hidden layer nodes in the ELM algorithm will cause differences in the accuracy and performance of the training results, it may cause redundancy of hidden layer nodes. When the random value is 0, some nodes will fail directly, which will greatly affect the prediction accuracy; when the gradient of the random value explodes, it will make it difficult to update deep learning, or even stagnate. At the same time, although such random assignment can increase the learning speed of the network, the instability of the network is greatly increased. Under the condition that other conditions remain unchanged, the results of two experiments may be very different. In order to solve the problem of hidden layer weights w in the ELM algorithm j and threshold b j In order to overcome the shortcomings of random assignment, we choose to introduce the particle swarm optimization algorithm PSO to optimize these two parameters, and use its fast global optimization capability to enhance the optimization of the ELM model parameters.

[0097] The PSO algorithm is relatively simple and can quickly find the local optimum, which meets the optimization requirements in ELM. The speed and position evolution of each particle in the algorithm is shown as follows:

[0098] v i (t+1)=Wv i (t)+c 1 r 1 (P i (t)-X i (t))+c 2 r2 (P g (t)-X i (t)) (18)

[0099] X i (t+1)=X i (t)+v i (t+1) (19)

[0100] Among them, the velocity of particle i after t iterations is recorded as v i (t); weight inertia is recorded as W; uniformly distributed random number function is recorded as r 1 (·), r 2 (·); the learning factor is denoted by c 1 、c 2 ; The best historical position of particle i is recorded as P i (t); the global optimal position of the cluster is denoted as P g (t); the position of particle i after iteration t is recorded as X i (t).

[0101] In the PSO algorithm, the particle fitness value f i It is also an important parameter. The fitness function of the i-th particle is the inverse of the mean square error. If the particle fitness value f i Greater than individual extreme value P i (t), the current fitness value is higher and the individual extreme value is updated; otherwise, the individual extreme value remains unchanged. i and the global extreme value P g (t) By comparing the size, the global extreme value can be updated.

[0102] The specific optimization steps are as follows:

[0103] 1. Determine the initial parameters of PSO, input the training data of ELM and perform preprocessing;

[0104] 2. Input the PSO related parameters into the ELM model and initialize the number of nodes p in the ELM hidden layer;

[0105] 3. The output weight β of the ELM model is obtained by formula (17): j ;

[0106] 4. Input the test data, get the test accuracy, and input the test accuracy as the PSO fitness value;

[0107] 5. Solve the individual extreme value and global extreme value of PSO;

[0108] 6. Use equations (18) and (19) to update the particle velocity and position respectively and calculate the particle fitness value;

[0109] 7. Update individual extreme values ​​and global extreme values ​​according to fitness values;

[0110] 8. Determine the termination condition. If the maximum iteration value has been reached, proceed to the next step; otherwise, jump back to step 6;

[0111] 9. Get the weight w of the ELM hidden layer j and threshold b j Substitute into ELM as input;

[0112] 10. Output reliability evaluation value.

[0113] As the construction of smart distribution networks is increasingly improved and the types and complexity of distribution network services are constantly increasing, the embodiment of the present invention uses natural language processing technology to process the data features of the extracted massive data of the distribution network, and the data text is divided according to the frequency of occurrence to achieve classified storage of distribution network data. For the collected data, the embodiment of the present invention uses the Paillier encryption algorithm to encrypt it to prevent data leakage or tampering during transmission and storage. When it is needed, the relevant data is decrypted to ensure the privacy of distribution network data and the security of access and use; the embodiment of the present invention uses the particle swarm optimization algorithm to optimize and enhance the extreme learning machine model. In view of the problem of random value of ELM parameters, the particle swarm optimization algorithm is used for iterative calculation to obtain the optimal parameters, thereby forming a distribution network reliability assessment model based on the PSO-ELM model to improve the speed and accuracy of reliability assessment.

[0114] The embodiments of the present invention have the following advantages:

[0115] 1. Considering the diversity of distribution network services, natural language processing technology is used to extract multi-mode data features for the massive data of the distribution network. The weight is judged based on the number of occurrences of the terms. The feature extraction of the multi-mode business data of the distribution network can be completed in a short time. At the same time, the importance of the terms in a single business (TF) and in all businesses (IDF) is considered, and the representativeness of the terms for specific documents can be measured;

[0116] 2. Information security is increasingly becoming an important requirement for power system operation. The data after feature extraction is encrypted using the Paillier homomorphic encryption algorithm to prevent data leakage or tampering during transmission and storage. Only when the user or manager needs to read the data, it is decrypted using the Paillier algorithm to ensure the information security of the distribution network.

[0117] 3. Combining the currently widely used deep learning and adopting the extreme learning machine model, in order to prevent the customized ELM random parameters from causing gradient disappearance or gradient explosion during the iteration process, the particle swarm optimization algorithm is used to obtain the optimal ELM random parameters to form a PSO-ELM evaluation model, which can realize efficient and accurate distribution network reliability evaluation.

[0118] Embodiment 2:

[0119] The embodiment of the present invention further provides a distribution network reliability assessment device, comprising:

[0120] A feature extraction module is used to extract features from multi-modal data of the distribution network to obtain feature data;

[0121] An encryption module, used to store feature data after encryption through Paillier algorithm;

[0122] A training module, used to train a PSO-ELM model of an extreme learning machine optimized by a particle swarm optimization algorithm based on the decrypted extracted stored feature data and environmental data;

[0123] The evaluation module is used to input real-time distribution business user data and environmental data into the trained PSO-ELM model and output the distribution network reliability evaluation results.

[0124] As an implementation method of the embodiment of the present invention, the feature extraction module uses natural language processing technology to judge the importance of entries in the power distribution service user data according to the frequency of occurrence, so as to extract data features.

[0125] As an implementation of an embodiment of the present invention, the training module includes:

[0126] A selection unit, used for selecting a reliability evaluation index according to a distribution network reliability influencing factor system;

[0127] A screening unit, used to extract the stored characteristic data and environmental data according to the decryption, and obtain data related to reliability influencing factors;

[0128] A building unit for forming an evaluation decision table based on reliability evaluation indicators and data related to reliability influencing factors;

[0129] The training unit is used to train the PSO-ELM model according to the evaluation decision table; wherein the data related to the reliability influencing factors are used as the input of the PSO-ELM model, and the reliability evaluation index of the distribution network is used as the output to train the PSO-ELM model.

[0130] Embodiment 3:

[0131] An embodiment of the present invention further provides a distribution network reliability assessment system, comprising: a memory and a processor, wherein the memory stores a computer program executed by the processor, and the computer program executes a distribution network reliability assessment method when executed by the processor.

[0132] Embodiment 4:

[0133] An embodiment of the present invention further provides a storage medium, on which a computer program is stored, and the computer program executes the distribution network reliability assessment method when running.

[0134] The embodiments described above are only descriptions of the preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should all fall within the protection scope determined by the claims of the present invention.

Claims

1. A distribution network reliability assessment method, characterized in that: include: Step S1, extracting features from multimodal data of the distribution network to obtain feature data; Step S2, storing the characteristic data after encryption by Paillier algorithm; Step S3, training a PSO-ELM model of an extreme learning machine optimized by a particle swarm optimization algorithm according to the decrypted extracted stored feature data and environmental data; Step S4: input the real-time distribution service user data and environmental data into the trained PSO-ELM model, and output the distribution network reliability assessment result.

2. The distribution network reliability assessment method according to claim 1, characterized in that: In step S1, natural language processing technology is used to judge the importance of entries in the power distribution service user data according to their frequency of occurrence, so as to extract data features.

3. The distribution network reliability assessment method according to claim 2, characterized in that: Step S3 includes: According to the reliability influencing factor system of distribution network, select reliability evaluation indicators; Extract the stored characteristic data and environmental data according to the decryption to obtain data related to the reliability influencing factors; Based on the reliability evaluation index and the data related to the reliability influencing factors, an evaluation decision table is formed; The PSO-ELM model is trained according to the evaluation decision table; wherein, the data related to the reliability influencing factors are used as the input of the PSO-ELM model, and the reliability evaluation index of the distribution network is used as the output to train the PSO-ELM model.

4. A distribution network reliability assessment device, characterized in that: include: A feature extraction module is used to extract features from multi-modal data of the distribution network to obtain feature data; An encryption module, used to store feature data after encryption through Paillier algorithm; A training module, used to train a PSO-ELM model of an extreme learning machine optimized by a particle swarm optimization algorithm based on the decrypted extracted stored feature data and environmental data; The evaluation module is used to input real-time distribution business user data and environmental data into the trained PSO-ELM model and output the distribution network reliability evaluation results.

5. The distribution network reliability assessment device according to claim 4, characterized in that: The feature extraction module uses natural language processing technology to judge the importance of entries in the power distribution business user data according to their frequency of occurrence in order to extract data features.

6. The distribution network reliability assessment device according to claim 5, characterized in that: The training modules include: A selection unit, used for selecting a reliability evaluation index according to a distribution network reliability influencing factor system; A screening unit, used to extract the stored characteristic data and environmental data according to the decryption, and obtain data related to reliability influencing factors; A building unit for forming an evaluation decision table based on reliability evaluation indicators and data related to reliability influencing factors; The training unit is used to train the PSO-ELM model according to the evaluation decision table; wherein the data related to the reliability influencing factors are used as the input of the PSO-ELM model, and the reliability evaluation index of the distribution network is used as the output to train the PSO-ELM model.

7. A distribution network reliability assessment system, characterized in that: include: A memory and a processor, wherein the memory stores a computer program executed by the processor, and when the computer program is executed by the processor, the distribution network reliability assessment method according to any one of claims 1 to 3 is executed.

8. A storage medium, characterized in that: The storage medium stores a computer program, and the computer program executes the distribution network reliability assessment method according to any one of claims 1 to 3 when running.

Images