Operation and maintenance method, system and equipment based on electric vehicle charging and battery swap facility and medium

Abstract

The invention relates to the technical field of electric vehicle charging, and discloses an operation and maintenance method, system and equipment based on an electric vehicle charging facility and a medium, and the method comprises the steps: obtaining and preprocessing the equipment and environment data of the electric vehicle charging facility, guaranteeing the data to be accurate and reliable, and laying a foundation for the assessment of the state of the facility. A first state evaluation model is established to accurately evaluate the facility state, and potential problems are found in time. Based on this, a second fault probability model is constructed to predict the fault possibility, and operation and maintenance are scientifically arranged. And establishing a third cost-benefit evaluation model to optimize the operation and maintenance cost in combination with the fault probability and the operation and maintenance cost parameters. And constructing a fourth improved Markov decision model to formulate a reasonable operation and maintenance strategy. And calling the model through an upper-layer strategy function to generate an optimal strategy and implementing the optimal strategy through a lower-layer execution function. Intelligent and precise operation and maintenance are realized, the failure rate is reduced, the reliability, stability and operation benefits are improved, and a guarantee is provided for healthy development of the industry.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of electric vehicle charging and battery swapping, and particularly relates to an operation and maintenance method, system, device and medium based on electric vehicle charging and battery swapping facilities. BACKGROUND

[0002] With the increasing popularity of electric vehicles in modern society, the number of electric vehicle charging and battery swapping facilities is also increasing. The stable operation of these facilities is crucial to ensure the normal use and popularization of electric vehicles. However, the current operation and maintenance of electric vehicle charging and battery swapping facilities faces many complex and severe challenges. On the one hand, due to the high frequency of use of the equipment and the complex and variable operating environment, equipment failures occur frequently, and the types of failures are diverse, which not only greatly increases the difficulty of operation and maintenance, but also significantly increases the operation and maintenance cost. On the other hand, traditional operation and maintenance methods are often limited to the treatment of single equipment or local problems, and lack comprehensive consideration of the overall state of the equipment and the surrounding environmental factors, making it difficult to achieve efficient and accurate operation and maintenance management.

[0003] In actual application scenarios, electric vehicle charging and battery swapping facilities in different regions may face completely different environmental conditions. For example, in high-temperature and high-humidity areas, equipment is prone to accelerated aging due to overheating or moisture; in areas with more dust, equipment may be damaged due to dust intrusion. These adverse environmental factors will undoubtedly further accelerate the aging and damage of the equipment. At the same time, different brands and models of charging and battery swapping equipment on the market have significant differences in performance and reliability, which undoubtedly pose more stringent requirements on the operation and maintenance method. Therefore, how to develop scientific and effective operation and maintenance strategies for different environmental conditions and equipment characteristics has become an important issue to be solved in the field of operation and maintenance of electric vehicle charging and battery swapping facilities. SUMMARY

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

[0005] Therefore, the present application provides an operation and maintenance method, system, device and medium based on electric vehicle charging and battery swapping facilities, which can solve many problems faced by the current operation and maintenance of electric vehicle charging and battery swapping facilities, such as frequent equipment failures, high operation and maintenance costs, and lack of comprehensive consideration in traditional operation and maintenance methods.

[0006] To solve the above technical problems, the present application provides the following technical solutions:

[0007] In a first aspect, the present application provides an operation and maintenance method based on electric vehicle charging and battery swapping facilities, comprising:

[0008] Obtaining equipment data and environment data of a target electric vehicle charging and swapping facility, and preprocessing the equipment data and the environment data;

[0009] Establishing a first state evaluation model according to the preprocessed equipment data and environment data;

[0010] Establishing a second fault probability model according to the first state evaluation model;

[0011] Establishing a third cost-benefit evaluation model according to the second fault probability model and an operation and maintenance cost parameter;

[0012] Constructing a fourth improved Markov decision model based on the first state evaluation model, the second fault probability model and the third cost-benefit evaluation model, wherein the state space includes equipment health states and environment factors, the action space includes several operation and maintenance behaviors, and a reward function introduces a fault risk weighting mechanism;

[0013] Generating an optimal operation and maintenance strategy by calling the fourth improved Markov decision model through a preset upper strategy function, and triggering a corresponding lower execution function to implement operation and maintenance operations.

[0014] As a preferred scheme of the electric vehicle charging and swapping facility operation and maintenance method, the method further comprises:

[0015] Obtaining operation and maintenance operation execution result data;

[0016] Feeding back the operation and maintenance operation execution result data to the first state evaluation model, the second fault probability model, the third cost-benefit evaluation model and the fourth improved Markov decision model for parameter iterative updating;

[0017] Until the iteration condition is met.

[0018] The preferred scheme can ensure that the model always matches the actual operation and maintenance situation, and improves the accuracy and reliability of the model. As the operation and maintenance operations are continuously carried out, the actual operation status of the equipment and the environment factors will change. If the model parameters are not updated, the evaluation results of the model may deviate from the actual situation. By feeding back the operation and maintenance operation execution result data to each model for parameter iterative updating, the model can adapt to these changes in a timely manner.

[0019] As a preferred scheme of the electric vehicle charging and swapping facility operation and maintenance method, the method further comprises:

[0020] Predefining a health degree index set, wherein the health degree index set includes several health degree indexes for evaluating the target electric vehicle charging and swapping facility;

[0021] Select a health index based on the pre-processed equipment data and environment data;

[0022] Characterize the selected health index by a health score formula, and calculate the health score of the target electric vehicle charging and replacing facility according to the score formula;

[0023] According to the health score, the health state of the equipment is graded to form a first state evaluation model.

[0024] As a preferred scheme of the electric vehicle charging and replacing facility operation and maintenance method, the second fault probability model is established according to the first state evaluation model, which includes:

[0025] A set of preset fault types is provided, which includes a plurality of different fault types and different fault characteristics corresponding to the plurality of different fault types;

[0026] Feature extraction is performed on the output of the first state evaluation model, and feature information related to the fault characteristics in the set of preset fault types is extracted;

[0027] The extracted features are matched and compared with different fault characteristics in the set of preset fault types, and the matching degree corresponding to each fault type is calculated;

[0028] According to the matching degree and historical fault data, the probability of each fault type occurring under the current equipment state is determined, and a second fault probability model is constructed.

[0029] As a preferred scheme of the electric vehicle charging and replacing facility operation and maintenance method, the third cost-benefit evaluation model is established according to the second fault probability model and the operation and maintenance cost parameter, which includes:

[0030] The operation and maintenance cost parameter of the target electric vehicle charging and replacing facility is determined;

[0031] A third cost-benefit evaluation model is established, which represents the goal of minimizing operation and maintenance cost and maximizing operation and maintenance benefit.

[0032] As a preferred scheme of the electric vehicle charging and replacing facility operation and maintenance method, the fourth improved Markov decision model is called by a preset upper strategy function, which includes:

[0033] Different strategy rules under the combination of different equipment health states and environmental factors are preset;

[0034] The upper strategy function selects the corresponding rule from the preset strategy rules according to the current obtained equipment health state and environmental factor;

[0035] The selected rule is taken as an input condition to the fourth improved Markov decision model;

[0036] Different operation and maintenance behaviors are comprehensively evaluated to find an operation and maintenance strategy that can achieve the optimal cost-benefit ratio in the current state.

[0037] The preferred embodiment can effectively combine the actual health status of the equipment and the environmental factors, and accurately formulate the most suitable operation and maintenance strategy. Through the preset strategy rules, the upper strategy function can quickly and accurately screen the rules that meet the current situation, and provide targeted input conditions for the fourth improved Markov decision model. In this way, when comprehensively evaluating different operation and maintenance behaviors, the specific conditions of the equipment and the environmental impact can be fully considered, so as to find the operation and maintenance strategy that can truly achieve the optimal cost-benefit ratio. On the one hand, it avoids the waste of resources caused by blind operation and maintenance, and reduces unnecessary operation and maintenance costs; on the other hand, it improves the efficiency and effectiveness of operation and maintenance, ensures that the electric vehicle charging and swapping facilities can maintain good operating conditions in various situations, and improves the overall operation and maintenance benefits. Moreover, this method has strong adaptability and flexibility, and can adjust the operation and maintenance strategy in time according to different combinations of equipment health status and environmental factors, better meeting the needs of actual operation and maintenance work.

[0038] As a preferred embodiment of the operation and maintenance method based on electric vehicle charging and swapping facilities according to the application, the operation and maintenance operation triggered by the corresponding lower execution function includes:

[0039] When the optimal operation and maintenance strategy is determined, the upper strategy function triggers the corresponding lower execution function;

[0040] The lower execution function generates specific operation and maintenance operation instructions according to the optimal operation and maintenance strategy;

[0041] The instructions include but are not limited to equipment maintenance, replacement of parts, and adjustment of operating parameters;

[0042] The instructions are sent to the corresponding execution equipment or personnel to implement specific operation and maintenance operations.

[0043] In a second aspect, the application provides an operation and maintenance system based on electric vehicle charging and swapping facilities, comprising:

[0044] A data acquisition and processing module is configured to acquire equipment data and environmental data of a target electric vehicle charging and swapping facility, and to preprocess the equipment data and environmental data;

[0045] A first model establishment module is configured to establish a first state evaluation model according to the preprocessed equipment data and environmental data;

[0046] A second model building module, configured to build a second fault probability model based on the first state assessment model;

[0047] A third model building module, configured to build a third cost-benefit evaluation model based on the second failure probability model and the operation and maintenance cost parameter;

[0048] a fourth model building module, configured to construct a fourth improved Markov decision model based on the first state assessment model, the second failure probability model, and the third cost-benefit assessment model, wherein the state space includes the equipment health state and environmental factors, the action space includes several operation and maintenance behaviors, and the reward function introduces a failure risk weighting mechanism;

[0049] The operation and maintenance module is used to call the fourth improved Markov decision model through a preset upper-layer strategy function to generate an optimal operation and maintenance strategy, and trigger the corresponding lower-layer execution function to implement the operation and maintenance operation.

[0050] In a third aspect, the present invention provides an electronic device comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the method described above when executing the computer program.

[0051] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, which implements the steps of the method described above when the computer program is executed by a processor.

[0052] Compared with existing technologies, the present invention offers the following advantages: It proposes an operation and maintenance method for electric vehicle charging and swapping facilities. This method obtains and preprocesses the equipment and environmental data of target electric vehicle charging and swapping facilities, ensuring the accuracy and reliability of the data used in subsequent model building. The equipment data reflects the operational status of the charging and swapping facility itself, while the environmental data accounts for the impact of external factors on the facility. The combination of the two lays the foundation for a comprehensive assessment of the facility's status.

[0053] The first state assessment model is built based on the preprocessed data, which can accurately assess the current state of the charging and swapping facilities. This helps to promptly identify potential problems with the facilities and prepare for them in advance.

[0054] Building a second failure probability model based on the first condition assessment model can predict the likelihood of facility failure. By analyzing and modeling large amounts of data, we can more scientifically determine the probability of failure and thus arrange operations and maintenance work in a targeted manner.

[0055] The third cost-benefit evaluation model is established by combining the second failure probability model and the operation and maintenance cost parameter, so that the operation and maintenance cost can be optimized under the premise of ensuring normal operation of the facility, resource waste caused by excessive operation and maintenance is avoided, and additional cost caused by frequent failure of the facility due to insufficient operation and maintenance is also prevented.

[0056] The fourth improved Markov decision model is constructed, the device health state and the environmental factor are introduced into the state space, various operation and maintenance behaviors are taken as the action space, and the reward function of the failure risk weighting mechanism is introduced, so that the model is more in line with the actual situation, and a more reasonable operation and maintenance strategy can be formulated.

[0057] The optimal operation and maintenance strategy is generated by calling the fourth improved Markov decision model through the preset upper layer strategy function, and the lower layer execution function is triggered to implement the operation and maintenance operation, so that seamless connection from strategy formulation to actual execution is realized, and the efficiency and quality of operation and maintenance work are improved.

[0058] The intelligent and precise operation and maintenance of the electric vehicle charging and battery swapping facility are realized, the failure rate of the facility can be effectively reduced, the reliability and stability of the facility are improved, the operation and maintenance cost is optimized, the operation benefit is improved, meanwhile, the environmental factor and various operation and maintenance behaviors are considered, so that the operation and maintenance strategy is more comprehensive and flexible, and can better adapt to different actual scenes, and a powerful guarantee is provided for the healthy development of the electric vehicle charging and battery swapping industry. BRIEF DESCRIPTION OF DRAWINGS

[0059] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced, and obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0060] Figure 1 A method flowchart of an operation and maintenance method based on an electric vehicle charging and battery swapping facility is provided for an embodiment of the present application.

[0061] Figure 2 An internal structure diagram of an electronic device of an operation and maintenance method based on an electric vehicle charging and battery swapping facility is provided for an embodiment of the present application. DETAILED DESCRIPTION

[0062] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the specific embodiments of the present application will be described in detail below with reference to the drawings in the specification, and obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor should belong to the protection scope of the present application.

[0063] Embodiment 1, Reference Figure 1 As a first embodiment of the present application, the embodiment provides a method for operating and maintaining electric vehicle charging and battery swapping facilities, comprising:

[0064] In the prior art, there are some problems, such as lack of comprehensive utilization of equipment data and environmental data, which leads to inaccurate assessment of equipment status and failure probability. Traditional operation and maintenance methods often only focus on the running parameters of the equipment itself, ignoring the influence of external environmental factors on the equipment, making it difficult to accurately adapt the operation and maintenance strategy to different environmental conditions.

[0065] The present application provides a method that can effectively solve the above-mentioned problems. Next, how to implement the method for operating and maintaining electric vehicle charging and battery swapping facilities will be described in detail in combination with multiple embodiments;

[0066] Figure 1 A method flowchart of a method for operating and maintaining electric vehicle charging and battery swapping facilities is shown, comprising:

[0067] S101, obtaining equipment data and environmental data of a target electric vehicle charging and battery swapping facility, and preprocessing the equipment data and environmental data, wherein:

[0068] It should be noted that in order to implement the operation and maintenance of electric vehicle charging and battery swapping facilities, the equipment data and environmental data of the electric vehicle charging and battery swapping facilities must be accurately obtained to determine the extent of operation and maintenance and whether to perform operation and maintenance operation;

[0069] In some specific embodiments, the equipment data covers various running parameters of the facility, such as charging power, battery temperature, voltage and current, etc., which directly reflect the working status of the equipment. Environmental data includes information such as the geographical location of the facility, climate conditions, and surrounding electromagnetic interference, which will affect the performance and stability of the equipment.

[0070] In the embodiments of the present application, the equipment data and environmental data of the target electric vehicle charging and battery swapping facility are obtained to provide comprehensive and accurate basis for subsequent operation and maintenance decisions.

[0071] It should be noted that the present application does not limit the specific equipment data and environmental data obtained, and related technical personnel can select according to actual needs.

[0072] It should also be noted that because the obtained equipment data and environmental data may have problems such as noise, missing values or abnormal values, it is necessary to preprocess them.

[0073] In some specific embodiments, the preprocessing step includes data cleaning, data normalization, and data dimension reduction operations. Data cleaning can remove noise and outliers in the data, ensuring the accuracy and reliability of the data; data normalization can unify data of different ranges and magnitudes to the same scale, avoiding the influence of large differences in data scale on the training effect of the model; data dimension reduction can reduce the dimension of the data, reduce the computational complexity, and improve the running efficiency of the model. Through these preprocessing operations, high-quality data support can be provided for subsequent model establishment.

[0074] In the embodiments of the present application, in order to better adapt the established data to subsequent steps, the present application proposes an innovative data preprocessing method. In data dimension reduction, the present application introduces a method based on the combination of principal component analysis and feature selection. Principal component analysis can extract the main components of the data, reducing the redundant information of the data, while feature selection can select the most representative and relevant features from numerous features, thereby reducing the dimension of the data while retaining the useful information of the data to the greatest extent.

[0075] Specifically, the steps of introducing the method based on the combination of principal component analysis and feature selection can be as follows:

[0076] Step 1.1: Standardize the obtained equipment data and environmental data, and convert the data to a standard normal distribution with a mean of 0 and a standard deviation of 1. This can eliminate the dimensional influence between different features and make the data comparable.

[0077] Step 1.1: Use the principal component analysis method to process the standardized data. Principal component analysis will calculate the covariance matrix of the data, and by solving the eigenvalues and eigenvectors of the covariance matrix, the principal components of the data are obtained. According to the size of the eigenvalues, select the principal components whose cumulative contribution rate reaches a certain threshold (such as 80% or 90%). These principal components contain most of the information of the original data, thereby achieving preliminary dimension reduction of the data.

[0078] Step 1.2: Feature selection. Filter-based feature selection methods can be used to calculate the correlation between each feature and the target variable (such as equipment health status), and select features with high correlation.

[0079] Wrapper-based feature selection methods can also be used, which construct a simple model (such as a logistic regression model) and select the optimal feature subset according to the performance indicators (such as accuracy, recall rate, etc.) of the model.

[0080] Embedded feature selection methods can also be used to automatically select important features during model training, such as using Lasso regression or Ridge regression for feature selection.

[0081] Step 1.3: After completing principal component analysis and feature selection, the resulting data is used as input for subsequent model establishment. Such data not only reduces dimensionality and computational complexity, but also retains the most representative and relevant features, which can improve the performance and accuracy of subsequent models.

[0082] It should be noted that preprocessing improves data quality and provides a foundation for subsequent model establishment, enabling the first state evaluation model to accurately assess the current state of the facility and timely detect potential problems. For example, accurate data can allow the model to capture subtle changes and provide early warning of failures. The innovative preprocessing method, such as the combination of principal component analysis and feature selection, reduces data dimensionality while retaining useful information, allowing the model to better fit reality and develop reasonable operation and maintenance strategies, improving operation and maintenance efficiency and quality, reducing costs, and enhancing industry operation efficiency.

[0083] S102, establishing a first state evaluation model according to the preprocessed equipment data and environment data, wherein:

[0084] It should be noted that after obtaining the preprocessed equipment data and environment data, the first state evaluation model can be constructed. In the present application, the first state evaluation model aims to comprehensively and accurately evaluate the current state of the electric vehicle charging and battery swapping facility.

[0085] In some specific embodiments, during the construction of the first state evaluation model, various machine learning algorithms can be used. For example, the support vector machine algorithm can find the optimal classification hyperplane in high-dimensional space to accurately classify different states of the equipment, thereby determining whether the facility is in a normal operating state, a sub-healthy state, or a failure state. The decision tree algorithm can also be used, which divides the data into layers and determines the state of the equipment based on the values of different features. The random forest algorithm is also a good choice, which consists of multiple decision trees and integrates the results of multiple decision trees to improve the accuracy and stability of the model.

[0086] In some specific embodiments, when using the support vector machine algorithm for the first state evaluation model, the possible steps are as follows:

[0087] Step 2.1: Select an appropriate kernel function, such as linear kernel, polynomial kernel, radial basis kernel, etc. Different kernel functions are suitable for different data distribution situations. For example, the radial basis kernel performs well in handling non-linear data and can better fit complex data patterns.

[0088] Step 2.2: Divide the preprocessed device data and environment data into training set and test set, usually in the ratio of 7:3 or 8:2. Use the training set to train the support vector machine model, and adjust the parameters of the model, such as penalty factor C and kernel coefficient γ, so that the classification accuracy of the model on the training set reaches a high level.

[0089] Step 2.3: In the training process, cross-validation method can be used to further divide the training set into multiple subsets, and training and validation are carried out in turn to evaluate the generalization ability of the model.

[0090] Step 2.4: After training, use the test set to evaluate the model, calculate the accuracy, recall, F1 value and other indicators of the model, and judge whether the performance of the model meets the requirements.

[0091] Step 2.5: If the performance of the model is not ideal, you can adjust the kernel function or parameters and retrain and evaluate until you get satisfactory results.

[0092] In this way, the first state evaluation model constructed by support vector machine algorithm can accurately classify the current state of electric vehicle charging and swapping facilities, providing reliable basis for subsequent fault prediction and operation decision.

[0093] When using decision tree algorithm to construct the first state evaluation model, its steps are different from those of support vector machine algorithm.

[0094] Step 3.1: Determine the growth strategy of decision tree, such as ID3, C4.5 or CART algorithm. Take CART algorithm as an example, which can handle continuous and discrete data and can construct binary decision tree. In the process of constructing decision tree, it is necessary to select appropriate features as the basis for dividing nodes, usually using information gain, information gain ratio or Gini index to measure the importance of features.

[0095] Step 3.2: By continuously dividing the data, until the termination condition is met, such as the number of samples in the node is less than a certain threshold or the purity of the node reaches a certain level.

[0096] Step 3.3: After construction, pruning operation is carried out on the decision tree to avoid overfitting phenomenon. Pre-pruning or post-pruning methods can be used, pre-pruning stops dividing in advance during the growth of decision tree, while post-pruning deletes some unnecessary nodes after the construction of decision tree is completed.

[0097] Step 3.4: Use the test set to evaluate the decision tree model, and optimize and adjust the model according to the evaluation results. The first state evaluation model constructed by decision tree algorithm can intuitively show the judgment process of device state, and has strong interpretability.

[0098] However, the above two methods of establishing the first state evaluation model cannot adapt well to the complex and changeable actual situation. Therefore, the present invention proposes a new method of establishing the first state evaluation model.

[0099] In an embodiment of the present invention, establishing a first state assessment model based on preprocessed device data and environmental data includes:

[0100] A preset health indicator set, which includes several health indicators of the target electric vehicle charging and swapping facilities;

[0101] Select health indicators based on pre-processed equipment data and environmental data;

[0102] The selected health indicators are characterized by a health scoring formula, and the health score of the target electric vehicle charging and swapping facility is calculated according to the scoring formula;

[0103] The health status of the equipment is graded according to the health score to form a first status assessment model.

[0104] It should be noted that the preset set of health indicators is derived from a large amount of data research and expert experience summary, covering indicators in multiple aspects such as equipment performance, operational stability, and environmental adaptability, in order to comprehensively and accurately evaluate the health status of electric vehicle charging and swapping facilities.

[0105] Specifically, the steps for establishing the first state assessment model may be as follows:

[0106] Step 4.1: It is necessary to define a health indicator set H = {h1,h2,...,h n}, where n is the number of health indicators. Each health indicator h i It reflects different aspects of the target electric vehicle charging and swapping facilities, such as voltage, current, temperature, etc.

[0107] For example, h1 is designed to represent device voltage, h2 is designed to represent device current, h3 is designed to represent ambient temperature, and h4 is designed to represent humidity level. The selection of these health indicators should be based on professional knowledge and analysis of historical data to ensure that they can comprehensively reflect the status of the device.

[0108] Step 4.2: Select health indicators based on pre-processed device data and environmental data. After obtaining the device data and environmental data, it is necessary to preliminarily screen these data and select features related to health indicators.

[0109] For example, it is assumed that the present invention has obtained a pre-processed data set D = {d1, d2, ..., d m}, where m is the number of data points. For each data point d j , the present application needs to determine whether it contains valid health indicator information. For example, if d j contains voltage and current data, it can be used to calculate health indicators h1 and h2.

[0110] Step 4.3: Characterize the selected health indicators with a health score formula. Next, the present application needs to characterize each health indicator h i with a score function f(h i ) that converts the raw data into a normalized health score.

[0111] For example, this score function can be represented in the following form:

[0112] S i = f(h i ) = w i · g(h i )

[0113] where S i is the score for the i-th health indicator. w i is the weight coefficient for the i-th health indicator, reflecting its importance. g(h i ) is a normalization function that maps the raw data to a standard range (e.g., between 0 and 1).

[0114] Suppose h1 is the device voltage, the present application can define g(h1) as:

[0115]

[0116] where v actual is the actual measured voltage value. v min and v max are the allowed minimum and maximum voltage values, respectively.

[0117] Step 4.4: Calculate the health score of the target electric vehicle charging and swapping facility according to the score formula. After obtaining the scores of all health indicators, the present application can weight-sum them to obtain the comprehensive health score H total of the entire device:

[0118]

[0119] Here H total is a comprehensive score that reflects the overall health of the device.

[0120] Step 4.5: Grade the health status of the device according to the health score, which can be achieved by setting different thresholds.

[0121] For example, the application can define the following levels: excellent H total > 0.9, good 0.7 < H total ≤ 0.9, general 0.5 < H total ≤ 0.7, poor H total ≤ 0.5.

[0122] It should be noted that in this way, the application can convert the continuous health score into discrete health state levels, which is convenient for subsequent operation and maintenance decisions.

[0123] S103, establishing a second fault probability model according to the first state evaluation model, wherein:

[0124] It should be noted that the first state evaluation model can be used to evaluate and classify the current health state of the electric vehicle charging and battery swapping facility, but it is not enough to accurately predict the future possible faults of the equipment. Therefore, it is necessary to establish a second fault probability model based on the first state evaluation model to further predict the possibility of equipment failure in the future.

[0125] In a specific embodiment, the second fault probability model can combine the equipment health state classification result output by the first state evaluation model and introduce a time series analysis method. Because the occurrence of equipment failure often has a certain time correlation, the past health state will affect the future fault probability.

[0126] In another specific embodiment, the second fault probability model can also combine machine learning algorithms, such as using deep learning models such as recurrent neural network (RNN) or long short-term memory network (LSTM). These models can process data with time series characteristics and learn the change law of equipment health state over time. By inputting the historical health state classification results output by the first state evaluation model and the corresponding time information, the model can train the complex relationship between equipment fault probability, time, and health state.

[0127] However, the above two modeling methods cannot well cope with all complex situations in equipment fault probability prediction. Therefore, the application adopts a more comprehensive method to establish the second fault probability model.

[0128] In an embodiment of the application, establishing a second fault probability model according to the first state evaluation model includes:

[0129] A set of preset fault types is provided, and the set of fault types includes a plurality of different fault types and different fault characteristics corresponding to the plurality of different fault types;

[0130] Performing feature extraction on the output of the first state assessment model to extract feature information related to fault features in a preset fault type set;

[0131] Match and compare the extracted features with different fault features in the preset fault type set, and calculate the matching degree corresponding to each fault type;

[0132] Based on the matching degree and historical fault data, the probability of each fault type occurring in the current equipment state is determined, and a second fault probability model is constructed.

[0133] Specifically, the specific steps for constructing the second fault probability model may be as follows:

[0134] Step 5.1: Define a set of fault types F = {f1,f2,...,f k}, where k is the number of fault types. Each fault type f j Contains several characteristics that reflect the specific manifestations of this fault type.

[0135] For example, f1 can be set as overheating fault; f2 as short circuit fault; f3 as aging fault, and each fault type f j A set of eigenvectors V j ={v j1 ,v j2 ,...,v jl}, where l is the number of features for each fault type.

[0136] Step 5.2: Extract characteristic information related to the fault characteristics in the preset fault type set from the first state assessment model. Assume that the present invention has obtained the health score vector H = {h1, h2, ..., h n}, the present invention needs to extract the part related to the fault characteristics.

[0137] For example, it is assumed that the characteristic vector of the overheating fault f1 is V1={v 11 ,v 12 ,...,v 1l}, where v 11 represents temperature, v 12 The present invention can extract the corresponding eigenvalue X={x1,x2,...,x l}, where x i Corresponding to v ji .

[0138] Step 5.3: Match and compare the extracted features with different fault features in the preset fault type set. Next, calculate the extracted feature X and each fault type fj The eigenvector V j The matching degree M j The matching degree can be calculated by the following formula:

[0139]

[0140] Among them, sim(x i ,v ji ) is the feature x i and fault characteristics v ji Similarity function between them. Common similarity functions include Euclidean distance, cosine similarity, etc.

[0141] w i is the weight coefficient of the i-th feature, reflecting its importance in fault identification. Assuming that the present invention uses Euclidean distance as the similarity function, then:

[0142]

[0143] in:

[0144]

[0145] Step 5.4: Based on the matching degree and historical fault data, determine the probability of each fault type occurring in the current equipment state and the matching degree M j and historical fault data, each fault type f can be calculated j The probability P(f j ).

[0146] For example, this probability can be calculated using Bayes' theorem:

[0147]

[0148] Among them, P(X|f j ) is given a fault type f j The probability of feature X appearing under the condition of j To make an approximate estimate. P(f j ) is the fault type f j The prior probability can be obtained based on historical fault data statistics. P(X) is the total probability of feature X appearing, which can be obtained by weighted summation of all fault types:

[0149]

[0150] Step 5.5: Set the probability P(f j ) are combined into a failure probability model P(F), which can be used to predict future equipment failures.

[0151] For example, assuming that there are three fault types f1, f2, and f3 in the present invention, the fault probability model can be expressed as:

[0152] P(F)={P(f1),P(f2),P(f3)}

[0153] It's important to note that building a second failure probability model based on the first condition assessment model can link the current health of the equipment with the likelihood of future failures, providing a more comprehensive and accurate basis for operation and maintenance decisions. This second failure probability model can predict the type and probability of potential equipment failures in advance, allowing for targeted maintenance planning and spare parts stockpiling, thus avoiding operational disruptions and financial losses caused by sudden equipment failures.

[0154] S104, establishing a third cost-benefit evaluation model based on the second failure probability model and the operation and maintenance cost parameter;

[0155] It's important to note that simply knowing the types and probabilities of possible future equipment failures isn't enough; in actual O&M, the balance between O&M costs and benefits must also be considered. Therefore, the present invention establishes a third cost-benefit evaluation model based on the second failure probability model and O&M cost parameters to facilitate more economical and reasonable O&M decisions.

[0156] In an embodiment of the present invention, establishing a third cost-benefit evaluation model based on the second failure probability model and the operation and maintenance cost parameter includes:

[0157] Determine the operation and maintenance cost parameters of the target electric vehicle charging and swapping facilities;

[0158] A third cost-benefit evaluation model is established to represent the goals of minimizing operation and maintenance costs and maximizing operation and maintenance benefits.

[0159] Specifically, the steps for establishing the third cost-benefit evaluation model can be as follows:

[0160] Step 6.1: It is necessary to clarify the composition of operation and maintenance costs. Operation and maintenance costs usually include multiple aspects, such as labor costs, material costs, equipment downtime losses, etc. The present invention can represent these costs as a vector C = {c1, c2, ..., c m}, where m is the number of different cost types.

[0161] For example, c1 represents labor cost (which may be the salary of technicians in the present invention), c2 represents material cost (which may be the cost of replacing parts in the present invention), and c3 represents equipment downtime loss (which may be the service interruption loss caused by equipment downtime in the present invention);

[0162] Each cost type c i Further, it can be further divided into different operation and maintenance behaviors. Assuming that the present application has n operation and maintenance behaviors A = {a1, a2,..., an}, the operation and maintenance cost corresponding to each operation and maintenance behavior can be represented as a matrix C n ij , where i represents the ith operation and maintenance behavior, and j represents the jth cost type.

[0163] Step 6.2: Establish a third cost-benefit evaluation model characterized by minimizing operation and maintenance cost and maximizing operation and maintenance benefit Next, the present application needs to establish a comprehensive cost-benefit evaluation model, the goal of which is to minimize operation and maintenance cost and maximize operation and maintenance benefit at the same time.

[0164] Step 6.2.1: Define the operation and maintenance cost function The operation and maintenance cost function TC(A) represents the total cost of executing a set of operation and maintenance behaviors A. It can be defined as:

[0165]

[0166] , where w j is the weight coefficient of the jth cost type, reflecting its importance in the total cost. C ij (a i ) is the specific value of the jth cost corresponding to the execution of the ith operation and maintenance behavior a i .

[0167] Step 6.2.2: Define the operation and maintenance benefit function TB(A) represents the total benefit of executing a set of operation and maintenance behaviors A. It can be defined as:

[0168]

[0169] , where p is the number of different benefit types. v k is the weight coefficient of the kth benefit type, reflecting its importance in the total benefit. B ik (a i ) is the specific value of the kth benefit corresponding to the execution of the ith operation and maintenance behavior a i . Common benefit types include improving device reliability, reducing failure frequency, and improving service quality.

[0170] Step 6.2.3: Combine the failure probability model By combining the failure probability P(f j ) in the second failure probability model, the present application can adjust the operation and maintenance benefit function TB(A) and the operation and maintenance cost function TC(A).

[0171] Specifically, the higher the failure probability, the greater the benefit brought by the corresponding operation and maintenance behavior, but it may also increase the operation and maintenance cost. A correction factor a j ​, for reflecting the influence of failure probability on operation and maintenance benefit:

[0172]

[0173] where, a j is a correction factor, which can be set according to historical data or expert experience. P(f j ) is the probability of the jth failure type. Similarly, for the operation and maintenance cost function TC(A), a similar correction factor b j can be introduced:

[0174]

[0175] Step 6.2.4: Construct a comprehensive cost-benefit evaluation model CE(A), which aims to minimize operation and maintenance costs and maximize operation and maintenance benefits. It can be defined as:

[0176] CE(A) = TB'(A) - TC'(A)

[0177] It should be noted that the third cost-benefit evaluation model based on the second failure probability model and the operation and maintenance cost parameter has the advantage of considering the device failure probability and operation and maintenance cost benefit comprehensively, providing a scientific basis for operation and maintenance decision-making. Through the third cost-benefit evaluation model, the cost and benefit of different operation and maintenance schemes can be quantified, so as to select the optimal operation and maintenance strategy.

[0178] S105, based on the first state evaluation model, the second failure probability model and the third cost-benefit evaluation model, a fourth improved Markov decision model is constructed, wherein the state space contains device health state and environmental factors, the action space contains several operation and maintenance behaviors, and the reward function introduces a failure risk weighting mechanism, wherein:

[0179] It should be noted that the detailed steps of the fourth improved Markov decision model can be as follows:

[0180] Step 7.1: The state space S contains device health state and environmental factors. The present application can represent it as a vector s = (s1, s2,..., s n ), where each element represents a different state feature.

[0181] For example, s1 represents the device health score, s2 represents the environmental temperature, s3 represents the humidity level, and s4 represents the failure frequency in the past period of time. Assuming that the present application has obtained the device health score H total through the first state evaluation model, and combined with environmental data, the state space S can be defined.

[0182] Step 7.2: Action space A contains several kinds of operation and maintenance behaviors. Each operation and maintenance behavior can be represented as a vector a = (a1, a2,..., an), where each element represents a different operation and maintenance operation. m

[0183] For example, a1 represents regular inspection, a2 represents replacement of parts, a3 represents adjustment of operating parameters, and a4 represents increasing maintenance frequency. These operation and maintenance behaviors can be evaluated for their costs and benefits by a third cost-benefit evaluation model.

[0184] Step 7.3: The reward function R(s, a) represents the immediate reward obtained by performing action a in state s. In order to introduce a failure risk weighting mechanism, the present application can design the reward function as:

[0185] R(s, a) = a·TB'(a) - β·TC'(a) + γ·D(s)

[0186] where TB'(a) is the modified operation and maintenance benefit function, TC'(a) is the modified operation and maintenance cost function, D(s) is the failure risk function reflecting the risk degree of failure in the current state, and a, β, γ are the weight coefficients of operation and maintenance benefit, operation and maintenance cost, and failure risk, respectively.

[0187] Step 7.4: The failure risk function can be calculated according to the second failure probability model. Assuming that the present application has obtained the probability P(f j ) of each failure type through the second failure probability model, the failure risk function can be represented as:

[0188]

[0189] where P(f j ) is the probability of the jth failure type, and w j is the weight coefficient of the jth failure type, reflecting its severity.

[0190] Step 7.5: The transition probability P(s ′ |s, a) represents the probability of transitioning to state s ′ after performing action a in state s. This probability can be estimated by historical data or simulation models.

[0191] For example, assuming that the present application performs operation and maintenance behavior a in state s, according to historical data or simulation results, the probability P(s ′ |s, a) of transitioning to state s ′ can be estimated.

[0192] ​Step 7.5: Based on the state space, action space, reward function, and transition probabilities defined above, the present application can construct an improved Markov Decision Model MDP (S, A, P, R).

[0193] For example, set the initial state s0 and the initial policy π0. Use value iteration or policy iteration methods to solve the optimal policy π. The goal is to minimize the long-term cumulative cost or maximize the long-term cumulative reward. The value iteration algorithm is:

[0194]

[0195] where V(s) is the value function of state s. γ is the discount factor, reflecting the discount rate of future rewards.

[0196] Policy iteration algorithm can include policy evaluation, that is, given a policy π, calculate its corresponding value function V π (s). Policy improvement is to find a new policy π' according to the current value function V π (s).

[0197] Step 7.6: Once the optimal policy π * is determined, specific operation instructions can be generated according to the policy and sent to the corresponding execution equipment or personnel to implement the operation.

[0198] For example, assume the optimal policy π * suggests performing the operation a * in state s, then the generated operation instructions may include regularly checking the voltage and current of the equipment, replacing aging parts, adjusting the running parameters of the equipment to improve efficiency.

[0199] S106, call the fourth improved Markov decision model through the preset upper policy function to generate the optimal operation policy, and trigger the corresponding lower execution function to implement the operation.

[0200] In an embodiment of the present application, calling the fourth improved Markov decision model through the preset upper policy function includes:

[0201] presetting different policy rules under the combination of equipment health status and environmental factors;

[0202] The upper policy function selects the corresponding rule from the preset policy rule according to the current obtained equipment health status and environmental factors;

[0203] The selected rule is passed to the fourth improved Markov decision model as an input condition;

[0204] Comprehensive evaluation is performed on different operation and maintenance behaviors to find an operation and maintenance strategy that can achieve the optimal cost-benefit ratio in the current state.

[0205] In the embodiments of the present application, triggering the corresponding lower layer execution function to implement the operation and maintenance operation includes:

[0206] When the optimal operation and maintenance strategy is determined, the upper layer strategy function triggers the corresponding lower layer execution function;

[0207] The lower layer execution function generates specific operation and maintenance operation instructions according to the optimal operation and maintenance strategy;

[0208] The instructions include, but are not limited to, maintenance of equipment, replacement of parts, and adjustment of operating parameters;

[0209] The instructions are sent to the corresponding execution equipment or personnel to implement specific operation and maintenance operations.

[0210] In the embodiments of the present application, operation and maintenance operation execution result data is obtained;

[0211] The operation and maintenance operation execution result data is fed back to the first state evaluation model, the second failure probability model, the third cost-benefit evaluation model, and the fourth improved Markov decision model for parameter iteration update;

[0212] Until the iteration condition is met.

[0213] In summary, the present application proposes an operation and maintenance method based on electric vehicle charging and battery swapping facilities, which obtains and preprocesses equipment data and environmental data of target electric vehicle charging and battery swapping facilities, ensuring the accuracy and reliability of data for subsequent model establishment. Equipment data reflects the operating condition of the charging and battery swapping facility itself, while environmental data takes into account the influence of external factors on the facility, laying a foundation for comprehensive evaluation of the facility state.

[0214] The first state evaluation model is established according to the preprocessed data, which can accurately evaluate the current state of the charging and battery swapping facility. This helps to discover potential problems in the facility in a timely manner and prepares for the problems in advance.

[0215] The second failure probability model is established based on the first state evaluation model, which can predict the possibility of failure of the facility. Through analysis and modeling of a large amount of data, the probability of failure can be judged more scientifically, so that operation and maintenance work can be arranged in a targeted manner.

[0216] The third cost-benefit evaluation model is established by combining the second failure probability model and the operation and maintenance cost parameter, which can optimize the operation and maintenance cost under the premise of ensuring the normal operation of the facility. This avoids resource waste caused by excessive operation and maintenance, and also prevents additional costs caused by frequent failures of the facility due to insufficient operation and maintenance.

[0217] A fourth improved Markov decision model is constructed, which includes the device health status and environmental factors into the state space, multiple operation and maintenance behaviors as the action space, and introduces a failure risk weighted mechanism reward function, so that the model is more in line with the actual situation, and a more reasonable operation and maintenance strategy can be formulated.

[0218] The fourth improved Markov decision model is called by the preset upper strategy function to generate the optimal operation and maintenance strategy, and the lower execution function is triggered to implement the operation and maintenance operation, realizing seamless connection from strategy formulation to actual execution, and improving the efficiency and quality of operation and maintenance work.

[0219] The application realizes intelligent and precise operation and maintenance of electric vehicle charging and battery swapping facilities. Not only can it effectively reduce the failure rate of the facilities and improve the reliability and stability of the facilities, but also can optimize the operation and maintenance cost and improve the operation efficiency. At the same time, considering the environmental factors and multiple operation and maintenance behaviors, the operation and maintenance strategy is more comprehensive and flexible, which can better adapt to different actual scenes and provide strong guarantee for the healthy development of electric vehicle charging and battery swapping industry.

[0220] In a preferred embodiment, the detailed steps of calling the fourth improved Markov decision model by the preset upper strategy function to generate the optimal operation and maintenance strategy and triggering the corresponding lower execution function to implement the operation and maintenance operation can be as follows:

[0221] Firstly, the application needs to define a series of preset strategy rules, which are based on different combinations of device health status and environmental factors. Each rule corresponds to a specific state and operation and maintenance behavior.

[0222] For example, rule 1: when the device health score Htotal>0.9 and the environmental temperature is within the normal range, perform regular inspection. total

[0223] Rule 2: when the device health score 0.7<Htotal≤0.9 and the humidity level is high, replace the aging parts. total

[0224] Rule 3: when the device health score Htotal≤0.5 and the failure frequency in the past period of time is high, perform comprehensive maintenance and adjust the operation parameters. total

[0225] These rules need to be set according to historical data, expert experience and the second failure probability model.

[0226] Secondly, the upper strategy function is responsible for selecting the most appropriate strategy rule according to the current obtained device health status and environmental factors. Assuming that the application has obtained the device health score Htotal and environmental data (such as temperature, humidity, etc.) through the first state evaluation model, the preset strategy rule can be matched according to these data.​​​

[0227] For example, assuming that the current device health score is 0.85, the ambient temperature is 25° C., and the humidity is 60%, the upper-layer policy function will select rule 2, that is, replace aging parts.

[0228] Once the most appropriate policy rule is selected, it is passed as input to the fourth improved Markov decision model, which calculates the optimal cost-benefit ratio based on the current state and the selected operation and maintenance behavior.

[0229] For example, assuming Rule 2 is selected, the input condition "replace aging parts" is passed to the fourth improved Markov decision model. The model will evaluate the costs and benefits of this action and compare it with other possible actions to determine the optimal operation and maintenance strategy.

[0230] Again, the fourth improved Markov decision model will comprehensively evaluate all possible operation and maintenance behaviors, calculate the cost and benefit of each behavior, and find the operation and maintenance strategy that can achieve the optimal cost-benefit ratio under the current state.

[0231] For example, the model evaluates the following maintenance behaviors: regular inspection, low cost, average benefit; replacement of parts, high cost, but significant benefit; adjustment of operating parameters, low cost, good benefit;

[0232] Ultimately, the model might recommend a “replacement part” because, while costly, it would significantly improve the reliability and performance of the equipment.

[0233] Trigger the corresponding lower-level execution function to implement operation and maintenance operations:

[0234] 1. Once the optimal O&M strategy is determined, the upper-layer strategy function triggers the corresponding lower-layer execution function. The lower-layer execution function is responsible for generating specific O&M instructions and sending them to the executing device or personnel.

[0235] For example, assuming the optimal operation and maintenance strategy is "replace parts", the upper-level strategy function will trigger the lower-level execution function to generate the following instructions: check the current status of the equipment, prepare the required parts, and arrange for technicians to replace them.

[0236] 2. The lower-level execution function generates specific operation and maintenance instructions based on the optimal operation and maintenance strategy. These instructions include but are not limited to equipment maintenance, component replacement, and operating parameter adjustment.

[0237] For example, assuming the optimal O&M strategy is "replace parts," the specific instructions generated may include: checking whether the voltage and current of the equipment are normal, replacing aging or damaged parts, and adjusting the equipment's operating parameters to optimize performance;

[0238] 3. The generated operation instructions are sent to the corresponding execution devices or personnel to implement specific maintenance operations. These operations can be completed by automated systems or performed manually by technicians.

[0239] For example, assuming the generated instruction is "replace parts", these instructions will be sent to on-site technicians who will follow the instructions to ensure the equipment returns to normal operation.

[0240] Wherein, the operation execution result data is obtained and fed back to each model for parameter iteration update, and data feedback and parameter update are continuously performed until the model converges or reaches the predetermined iteration condition. This usually means that the error between the model's predicted result and the actual result is small enough, or the maximum number of iterations has been reached.

[0241] For example, assuming the maximum number of iterations is set to 10, the model parameters are updated according to the new operation result at each iteration. If the error between the model's predicted result and the actual result is less than the preset threshold at the 8th iteration, the iteration is stopped and the model is considered to have converged.

[0242] The present application can ensure the effectiveness and accuracy of the maintenance strategy, improve the reliability and efficiency of the equipment, and reduce the maintenance cost. These steps rely on accurate data processing and reasonable mathematical models to ensure the accuracy and reliability of the evaluation results.

[0243] Embodiment 3, refer to Figure 2 In this embodiment, a maintenance system based on electric vehicle charging and battery swapping facilities is also provided, comprising:

[0244] A data acquisition and processing module is used to acquire equipment data and environmental data of the target electric vehicle charging and battery swapping facility, and to preprocess the equipment data and environmental data;

[0245] A first model establishment module is used to establish a first state evaluation model based on the preprocessed equipment data and environmental data;

[0246] A second model establishment module is used to establish a second fault probability model based on the first state evaluation model;

[0247] A third model establishment module is used to establish a third cost-benefit evaluation model based on the second fault probability model and the maintenance cost parameters;

[0248] A fourth model establishing module is configured to establish a fourth improved Markov decision model based on the first state evaluation model, the second failure probability model and the third cost-benefit evaluation model, wherein the state space comprises device health states and environmental factors, the action space comprises a plurality of operation and maintenance behaviors, and the reward function introduces a failure risk weighting mechanism.

[0249] An operation and maintenance module is configured to call the fourth improved Markov decision model through a preset upper-layer strategy function, generate an optimal operation and maintenance strategy, and trigger a corresponding lower-layer execution function to implement operation and maintenance operations.

[0250] The above modules can be embedded in or independent of a processor in the electronic device in hardware form, or stored in a memory in the electronic device in software form, so as to be called and executed by the processor to perform the operations corresponding to the above modules.

[0251] The embodiment also provides an electronic device, which can be a terminal. An internal structure diagram of the electronic device can be as shown in Figure 2 The electronic device comprises a processor, a memory, a communication interface, a display screen and an input device connected through a system bus. The processor of the electronic device is configured to provide computing and control capabilities. The memory of the electronic device comprises a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operating system and the computer program in the non-volatile storage medium to run. The communication interface of the electronic device is configured to perform wired or wireless communication with an external terminal. The wireless communication can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The computer program is executed by the processor to implement an operation and maintenance method based on an electric vehicle charging and battery swapping facility. The display screen of the electronic device can be a liquid crystal display screen or an electronic ink display screen. The input device of the electronic device can be a touch layer overlaid on the display screen, or a key, trackball or touchpad arranged on the shell of the electronic device, or an external keyboard, touchpad or mouse, etc.

[0252] The embodiment also provides a computer readable storage medium having a computer program stored thereon. The computer program is executed by the processor to implement the following steps:

[0253] Obtaining device data and environmental data of a target electric vehicle charging and battery swapping facility, and preprocessing the device data and the environmental data;

[0254] Establishing a first state evaluation model according to the preprocessed device data and environmental data;

[0255] Establishing a second failure probability model according to the first state evaluation model;

[0256] According to the second failure probability model and the operation and maintenance cost parameter, a third cost-benefit evaluation model is established;

[0257] Based on the first state evaluation model, the second failure probability model and the third cost-benefit evaluation model, a fourth improved Markov decision model is constructed, wherein the state space contains the equipment health state and the environmental factors, the action space contains several operation and maintenance behaviors, and the reward function introduces a failure risk weighting mechanism;

[0258] The fourth improved Markov decision model is called by a preset upper layer strategy function to generate an optimal operation and maintenance strategy, and a corresponding lower layer execution function is triggered to implement the operation and maintenance operation.

[0259] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present application and are not limiting. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced equivalently without departing from the spirit and scope of the present application, and all modifications and replacements should be included in the scope of the claims of the present application.

[0260] Although the preferred embodiments of the present application have been described, those skilled in the art can make additional changes and modifications to the embodiments once they understand the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all modifications and modifications falling within the scope of the present application.

[0261] Obviously, those skilled in the art can make various modifications and variations to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalent technologies, the present application also intends to include these modifications and variations.

Claims

1. An operation and maintenance method based on electric vehicle charging and swapping facilities, characterized in that: include: Acquire equipment data and environmental data of the target electric vehicle charging and swapping facility, and preprocess the equipment data and environmental data; Establishing a first state assessment model based on the preprocessed equipment data and environmental data; Establishing a second fault probability model based on the first state assessment model; Establishing a third cost-benefit evaluation model based on the second failure probability model and the operation and maintenance cost parameter; A fourth improved Markov decision model is constructed based on the first state assessment model, the second failure probability model, and the third cost-benefit assessment model, wherein the state space includes the equipment health status and environmental factors, the action space includes several operation and maintenance behaviors, and the reward function introduces a failure risk weighting mechanism; The fourth improved Markov decision model is called by a preset upper-layer strategy function to generate an optimal operation and maintenance strategy, and trigger a corresponding lower-layer execution function to implement the operation and maintenance operation.

2. The operation and maintenance method based on electric vehicle charging and swapping facilities according to claim 1, characterized in that: Also includes: Obtain operation and maintenance operation execution result data; Feeding back the operation and maintenance operation execution result data to the first state assessment model, the second failure probability model, the third cost-benefit assessment model, and the fourth improved Markov decision model for parameter iterative update; Until the iteration condition is met.

3. The operation and maintenance method based on electric vehicle charging and swapping facilities according to claim 2, characterized in that: The establishing of the first state assessment model according to the pre-processed device data and environmental data includes: A preset health indicator set, wherein the health indicator set includes several health indicators of the target electric vehicle charging and swapping facilities; Select health indicators based on pre-processed equipment data and environmental data; The selected health indicators are characterized by a health scoring formula, and the health score of the target electric vehicle charging and swapping facility is calculated according to the scoring formula; The health status of the equipment is graded according to the health score to form a first status assessment model.

4. The operation and maintenance method based on electric vehicle charging and swapping facilities according to claim 3, characterized in that: The establishing of a second fault probability model according to the first state assessment model comprises: A preset fault type set, wherein the fault type set includes several different fault types and different fault characteristics corresponding to the several different fault types; Performing feature extraction on the output of the first state assessment model to extract feature information related to fault features in a preset fault type set; Match and compare the extracted features with different fault features in the preset fault type set, and calculate the matching degree corresponding to each fault type; Based on the matching degree and historical fault data, the probability of each fault type occurring in the current equipment state is determined, and a second fault probability model is constructed.

5. The operation and maintenance method based on electric vehicle charging and swapping facilities according to claim 4, characterized in that: The establishing of the third cost-benefit evaluation model according to the second failure probability model and the operation and maintenance cost parameter includes: Determine the operation and maintenance cost parameters of the target electric vehicle charging and swapping facilities; A third cost-benefit evaluation model is established to represent the goals of minimizing operation and maintenance costs and maximizing operation and maintenance benefits.

6. The operation and maintenance method based on electric vehicle charging and swapping facilities according to claim 5, characterized in that: The calling of the fourth improved Markov decision model by a preset upper layer strategy function includes: Preset policy rules for different combinations of device health status and environmental factors; The upper-layer policy function selects the corresponding rule from the preset policy rules based on the currently acquired device health status and environmental factors; passing the selected rules as input conditions to the fourth improved Markov decision model; Comprehensively evaluate different operation and maintenance behaviors to find the operation and maintenance strategy that can achieve the best cost-benefit ratio under the current status.

7. The operation and maintenance method based on electric vehicle charging and swapping facilities according to claim 6, characterized in that: The triggering of the corresponding lower layer execution function to implement the operation and maintenance operation includes: Once the optimal operation and maintenance strategy is determined, the upper-layer strategy function triggers the corresponding lower-layer execution function; The lower-level execution function generates specific operation and maintenance instructions based on the optimal operation and maintenance strategy; The instructions include but are not limited to equipment maintenance, replacement of parts, and adjustment of operating parameters; These instructions are sent to the corresponding execution equipment or personnel to implement specific operation and maintenance operations.

8. An operation and maintenance system based on electric vehicle charging and swapping facilities, applying the method according to any one of claims 1 to 7, characterized in that: include: A data acquisition and processing module is used to acquire the equipment data and environmental data of the target electric vehicle charging and swapping facilities, and pre-process the equipment data and environmental data; A first model building module, configured to build a first state assessment model based on preprocessed device data and environmental data; A second model building module, configured to build a second fault probability model based on the first state assessment model; A third model building module is used to establish a third cost-benefit evaluation model based on the second failure probability model and the operation and maintenance cost parameter; a fourth model building module, configured to construct a fourth improved Markov decision model based on the first state assessment model, the second failure probability model, and the third cost-benefit assessment model, wherein the state space includes the equipment health state and environmental factors, the action space includes several operation and maintenance behaviors, and the reward function introduces a failure risk weighting mechanism; The operation and maintenance module is used to call the fourth improved Markov decision model through a preset upper-layer strategy function to generate an optimal operation and maintenance strategy, and trigger the corresponding lower-layer execution function to implement the operation and maintenance operation.

9. An electronic 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 an operation and maintenance method based on electric vehicle charging and swapping facilities according to any one of claims 1 to 7 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 an operation and maintenance method based on electric vehicle charging and swapping facilities according to any one of claims 1 to 7 are implemented.

Images