Charging pile operation state evaluation method, system and device and storage medium

Abstract

The invention discloses a charging pile operation state evaluation method, system and device and a storage medium, and the method comprises the steps: S1, collecting the operation monitoring data of a charging pile, and calculating the operation state data of the charging pile; s2, on the basis of operation state data in a set time period, constructing parameter sequences according to a time sequence, sorting entropy values of the parameter sequences from large to small, performing classification and feature extraction on the parameter sequences in a triple manner, and constructing a feature fusion model for feature fusion; s3, performing nonlinear function mapping on the fusion features of the charging pile samples to obtain similar features and difference features, and performing optimization on the transformation dimensions of the features by adopting an improved zebra optimization algorithm; and S4, evaluating the operation state of the charging pile based on the state evaluation model by taking the similar characteristics or the difference characteristics as input. The problem of inaccurate evaluation caused by introduction of subjective factors in previous charging pile operation state research is solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of online monitoring of electric power metering, and in particular relates to a charging pile operation status evaluation method, system, device and storage medium. Background Art

[0002] In recent years, as the sales of electric vehicles have increased year by year, the number of charging piles has become increasingly large, and the number of charging failures has also increased. Due to the complex factors of charging pile failures and the wide distribution of equipment, traditional inspection and maintenance methods are inefficient and costly, and currently cannot meet the operation and maintenance needs of large-scale charging facilities, which has seriously affected the development of the new energy vehicle industry. Therefore, this patent proposes a charging pile operation status evaluation method based on an online evaluation method to provide support for the intelligent operation and maintenance of electric vehicle charging facilities.

[0003] In He Simiao's "Study on Intelligent Charging Pile State Classification and Fault Prediction Based on Big Data", characteristic indicators of state evaluation are extracted, and each indicator is analyzed to obtain its threshold and confidence level; then the operating state of the charging pile is evaluated in real time, and the charging piles to be identified are divided into healthy charging piles, sub-healthy charging piles, abnormal charging piles and serious charging piles by combining fuzzy clustering algorithm and dynamic degradation degree. In the charging pile state evaluation system described in this technical solution, the weight of each evaluation indicator plays a decisive role in the accuracy of the state result evaluation, and the weight assignment is based on the expert consultation system and has a certain degree of subjectivity.

[0004] The Chinese patent with the publication number CN115146948A discloses a method for evaluating the health status of an electric vehicle charging pile based on a subjective and objective comprehensive fuzzy evaluation method, which includes the following steps: using a set value statistical analysis method to convert multiple qualitative index values ​​in the charging pile to obtain a score value; according to the charging pile mechanism, the subjective weight values ​​of the charging pile performance indicators are determined in turn according to the hierarchical analysis method; the score value is used to obtain the objective weight value of the corresponding performance indicator according to the entropy weight method; according to the additive integrated weighting method, the subjective weight value and the objective weight value are determined according to the relative importance of the indicator to determine the weight coefficient suitable for comprehensive evaluation; combined with the weight coefficient of the comprehensive evaluation, the fuzzy comprehensive evaluation method is used to finally obtain the health level of the charging pile. The set value statistical analysis method in the invention converts qualitative indicators into score values, simplifies complex data, and may cause information loss; and the hierarchical analysis method relies on subjective judgment to determine the weight, and the membership function and fuzzy rule setting of the fuzzy comprehensive evaluation method are highly dependent on expert experience, and there will also be a certain degree of subjectivity.

[0005] A Chinese patent with publication number CN111680924A discloses a method for evaluating the operating status of a charging pile, including: collecting the operating data of the target charging pile during the charging process, and extracting the actual output voltage, actual output current and historical fault information of the target charging pile from the operating data; obtaining the cabinet temperature information of the cabinet where the target charging pile is located during the charging process; evaluating the operating status of the target charging pile according to the actual output voltage, actual output current, historical fault information and cabinet temperature information of the target charging pile. The invention evaluates the operating status of the charging pile based on the static weight calculation method, but the fixed weight cannot be adjusted according to changes in actual conditions. In actual operation, the importance of indicators such as voltage, current, and fault frequency may change over time, and static weights cannot capture these dynamic factors. Moreover, this method cannot be optimized based on historical data or operating experience, and may miss opportunities to improve the evaluation method. Summary of the invention

[0006] The present invention provides a charging pile operating status evaluation method, system, device and storage medium, aiming to solve the problem of inaccurate evaluation caused by the introduction of subjective factors in previous charging pile operating status research.

[0007] In order to solve the above technical problems, the present invention provides a charging pile operation status evaluation method, comprising the following steps: S1: Collect the operation monitoring data of the charging pile, and calculate the operation status data of the charging pile through the operation monitoring data.

[0008] S2: Based on the operating status data within the set time period, a parameter sequence is constructed according to the time series, the entropy value of each parameter sequence is calculated, the entropy values ​​are sorted from large to small, the parameter sequence is classified and feature extracted as triplets, and a feature fusion model is constructed for feature fusion to obtain fusion features.

[0009] S3: The fusion features of each charging pile sample are mapped through nonlinear functions to obtain similar features and difference features, and the improved zebra optimization algorithm is used to optimize the transformation dimensions of similar features and difference features.

[0010] S4: Taking the optimized similar features or difference features as input, based on the state evaluation model, the operation state of the charging pile is evaluated to obtain the classification probability of the operation state of the charging pile and output the operation state label.

[0011] Preferably, the step S2 is specifically: S21: Based on the operation status data within the set time period, the operation monitoring data includes: input current / voltage, output voltage / current, input power, output power, pile body temperature, ambient temperature, and the operation status data includes: input voltage / current error, output voltage / current error, output voltage / current voltage regulation accuracy, power factor, whole machine power, temperature difference, and construct a parameter sequence according to the time sequence, specifically:

[0012]

[0013]

[0014]

[0015]

[0016]

[0017]

[0018]

[0019]

[0020] In the formula, The input voltage error parameter sequence of the charging pile during the process of charging the car; For charging pile The input voltage error during the charging process of a car, ; The total number of cars charged; The output voltage error parameter sequence of the charging pile during the process of charging the car; For charging pile Output voltage error during charging of a car; The input current error parameter sequence of the charging pile during the process of charging the car; For charging pile Input current error during charging of a car; The output current error parameter sequence of the charging pile during the process of charging the car; For charging pile Output current error during charging of a car; It is the output voltage regulation accuracy parameter sequence of the charging pile during the process of charging the car; For charging pile Output voltage regulation accuracy during the charging process of a car; It is the output current voltage regulation accuracy parameter sequence of the charging pile during the process of charging the car; For charging pile Output current regulation accuracy of a car during charging; is the charging power factor parameter sequence of the charging pile; For charging pile The charging power factor of the vehicle; is the whole power parameter sequence of the charging pile; For charging pile The total power of the vehicle; is the temperature difference parameter sequence between the charging pile body temperature and the ambient temperature; For charging pile The temperature difference of the car.

[0021] S22: Calculate the entropy value of each parameter sequence, sort the entropy values ​​from large to small, classify the parameter sequence into triplets, convert the time series signal into an image, and construct a grouped image of the charging pile sample data based on the red, green, and blue channels , , .

[0022] S23: Perform convolution operations on the grouped images , , Perform feature extraction and obtain , , feature.

[0023] S24: Construct a feature fusion model to perform feature fusion and obtain fusion features. The feature fusion model is specifically:

[0024] In the formula, For fusion features; for Feature weights; for Feature weights; for Feature weights.

[0025] Preferably, in step S22, the timing signal is converted into an image, and a grouped image of the charging pile sample data is constructed based on the red, green and blue channels. , , Specifically: A1: For the converted RGB image, the red channel pixel value is filled by the triplet with the largest entropy value in each group, the green channel pixel value is filled by the triplet with the second largest entropy value in each group, and the blue channel pixel value is filled by the triplet with the smallest entropy value in each group. The data range conversion formula is as follows:

[0026] In the formula, is the pixel of the image, is the rounding function; The first The first charging pile sample Parameters; is the first The minimum value of a parameter; is the first The maximum value of a parameter.

[0027] A2: Constructing a grouped image of charging pile sample data based on the red, green and blue channels , specifically:

[0028]

[0029]

[0030] In the formula, , , They are the first, second and third group of images of charging pile samples respectively; For the The red channel pixels of the group image; For the Green channel pixels of the group image; For the The blue channel pixels of the group image; .

[0031] Preferably, in step S24 , , The feature weights of are processed by the multi-scale attention mechanism, specifically:

[0032] In the formula, Features , , The initial fusion features obtained by direct addition; for , The initial fusion features are obtained by directly adding the features; is the initial fusion feature of the input multi-scale attention mechanism, ; , , , , , , is the intermediate parameter; It is the global average pooling process; It is a 1×1 point-by-point convolution; It is a standardization layer, which aims to make the output become a standard normal distribution and reduce sample differences; is the activation function; The function output is the feature weight; For the general Input the feature weights obtained by the multi-scale attention mechanism, As The feature weights are: .

[0033] based on , Get Feature weights , specifically:

[0034] In the formula, for The weight of the feature; for Input the feature weights obtained by the multi-scale attention mechanism; for Input the feature weights obtained by the multi-scale attention mechanism.

[0035] Said Feature weights The specific calculation formula is:

[0036] In the formula, for The weight of the feature.

[0037] Preferably, in step S3, the fusion features of each charging pile sample are mapped by nonlinear function to obtain similar features and difference features, specifically:

[0038]

[0039] In the formula, For the In the first operating state Similar characteristics of charging pile samples; For the In the first operating state Different characteristics of charging pile samples; , The number of different operating states of the set charging pile; For the In the first operating state Characteristics of charging pile samples; , Respectively in The mapping weights of similar features and difference features of charging pile samples under different states; , Respectively in The bias of the two features in the state; , Respectively for Nonlinear function of the two feature maps in the state; , ; , They are , Dimension information of For the The number of charging pile samples under different operating conditions; , They are the dimensions of two characteristics respectively; .

[0040] Preferably, in step S3, the improved zebra optimization algorithm is used to optimize the transformation dimensions of similar features and difference features as follows: B1: Initialize the zebra optimization algorithm parameters and randomly initialize the population in the optimization space.

[0041] B2: Based on the number of optimization searches The span of the position update is dynamically adjusted according to the size of the fitness function change rate, so as to update the new position of the zebra in the foraging stage. Specifically:

[0042]

[0043]

[0044]

[0045] In the formula, is the set number of iterations The span of change affects the parameter; is the current iteration number; is the maximum number of iterations; For the The rate of change of the best fitness function for each iteration; For the The optimal fitness function at the iteration; For the The optimal fitness function at the iteration; and is the set weight coefficient; For the first stage Zebra in the The new location of the dimension; For the Zebra in the The location of the or , correspond , correspond ; For the The fitness function of a zebra; For the first stage The new fitness function value of the zebra; For the pioneer zebra position; For the Dimensional Pioneer Zebra Position; for A random number between for A random value between For the The location of the zebras; For the first stage The new position of the zebra.

[0046] B3: Simulate the zebra's defense strategy against predators. When a lion attacks a zebra, the zebra runs away; when other predators attack the zebra, the zebra fights back; this updates the zebra's new position in the optimization space.

[0047] B4: Repeat steps B2 and B3 until the maximum number of iterations is met, and output the optimal solution of the dimension obtained by the zebra optimization algorithm.

[0048] Preferably, in step S4, a state evaluation model is constructed based on a softmax function, specifically:

[0049] In the formula, For the In this state Similar features of charging pile samples The probability of each category obtained after the softmax function; For the In this state Different characteristics of charging pile samples The probability of each category obtained after the softmax function; , are weight matrices respectively.

[0050] The state assessment model calculates the probability that the charging pile sample belongs to various states, specifically:

[0051] In the formula, For charging pile samples The probability of belonging to each state; is the weight coefficient of the feature, For the Status A charging station sample.

[0052] On the other hand, the present invention provides a charging pile operation status evaluation system, including a data acquisition module, a feature extraction and fusion module, a feature optimization module and an operation status evaluation module.

[0053] The data acquisition module is used to collect the operation monitoring data of the charging pile and calculate the operation status data of the charging pile through the operation monitoring data.

[0054] The feature extraction and fusion module is used to construct a parameter sequence according to the time series based on the operating status data within a set time period, calculate the entropy value of each parameter sequence, sort the entropy values ​​from large to small, classify the parameter sequence as a triplet and extract features, build a feature fusion model for feature fusion, and obtain fusion features.

[0055] The feature optimization module is used to obtain similar features and difference features of the fused features of each charging pile sample through nonlinear function mapping, and use the improved zebra optimization algorithm to optimize the transformation dimensions of similar features and difference features.

[0056] The operating status evaluation module is used to evaluate the operating status of the charging pile based on the status evaluation model, using the optimized similar features or difference features as input, obtain the classification probability of the charging pile operating status and output the operating status label.

[0057] On the other hand, the present invention also provides an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the charging pile operating status evaluation method as described in any embodiment of the present invention is implemented.

[0058] On the other hand, the present invention further provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the method for evaluating the operating status of a charging pile as described in any embodiment of the present invention is implemented.

[0059] Compared with the prior art, the present invention has the following technical effects: The present invention proposes a charging pile operation status evaluation method. The method first completes parameter classification by entropy calculation to effectively screen out important features. Then, a multi-scale attention mechanism is used to complete feature fusion, capture key information at different scales, and improve the expression ability of features. Furthermore, similarities and differences are extracted for the fused features to ensure sensitivity to subtle changes in fused features. Finally, the evaluation of the charging pile operation status is completed through a status evaluation model, ensuring the high accuracy and reliability of the charging pile operation status evaluation. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 It is an overall flow chart of the charging pile operation status evaluation method of the present invention; Figure 2 It is the overall structural diagram of the multi-scale attention mechanism described in the present invention. DETAILED DESCRIPTION

[0061] In order to make the objectives, technical solutions and advantages of the present invention more clear, the technical solutions of the present invention will be clearly and completely described below in combination with specific embodiments of the present application and with reference to the accompanying drawings.

[0062] Embodiment 1 This embodiment provides a charging pile operation status evaluation method, see Figure 1 As shown, the following steps are included: S1: Collect the operation monitoring data of the charging pile, and calculate the operation status data of the charging pile through the operation monitoring data. The operation monitoring data includes: input current / voltage, output voltage / current, input power, output power, pile body temperature, ambient temperature, and the operation status data includes: input voltage / current error, output voltage / current error, output voltage / current voltage regulation accuracy, power factor, whole machine power, temperature difference. The operation status of the charging pile can be divided into different states: healthy, sub-healthy, abnormal, severe, etc. In order to accurately evaluate the status of the charging pile, the charging of a car by the charging pile is taken as a time point, and the input / output voltage, current error, output voltage / current stability, charging power factor and whole machine efficiency, temperature difference and other data are calculated respectively.

[0063] Specifically, the input / output voltage and current of a certain car during charging by the charging pile within a set time period are collected, and the input / output voltage and current errors are calculated based on the single charging data:

[0064]

[0065]

[0066]

[0067] In the formula, , They respectively represent the input / output voltage errors during the charging process of a certain car; , Respectively represent the input / output voltage limit values; , Respectively represent the standard value of input / output voltage. , Respectively represent the input / output current error; , Respectively represent the input / output current limit value; , Respectively represent the standard value of input / output current.

[0068] Based on the collected output voltage and current, calculate the stability of the charging pile output voltage / current:

[0069]

[0070] In the formula, Indicates the output voltage regulation accuracy during the charging process of a certain car; Indicates the output voltage fluctuation limit; Indicates the mean output voltage; Indicates the output current stabilization accuracy; Indicates the output current fluctuation limit; Indicates the average output current.

[0071] Calculate the charging power factor of the charging pile And the overall efficiency :

[0072]

[0073] in, Indicates the active power input by the charging pile. Indicates the reactive power input by the charging pile. Indicates the output power.

[0074] The difference between the pile temperature and the ambient temperature is one of the manifestations of the charging pile status. :

[0075] in, It indicates the average temperature of the charging pile during the charging process of a certain car. Indicates the average temperature of the surrounding environment.

[0076] S2: Based on the operating status data within the set time period, a parameter sequence is constructed according to the time series, the approximate entropy of each parameter sequence is calculated, the entropy values ​​are sorted from large to small, the parameter sequence is classified and feature extracted as triplets, a feature fusion model is constructed for feature fusion, and fusion features are obtained.

[0077] As a preferred implementation of this embodiment, step S2 is specifically as follows: S21: Based on the running status data within the set time period, a parameter sequence is constructed according to the time sequence, specifically:

[0078]

[0079]

[0080]

[0081]

[0082]

[0083]

[0084]

[0085]

[0086] In the formula, The input voltage error parameter sequence of the charging pile during the process of charging the car; For charging pile The input voltage error during the charging process of a car, ; The total number of cars charged; The output voltage error parameter sequence of the charging pile during the process of charging the car; For charging pile Output voltage error during charging of a car; The input current error parameter sequence of the charging pile during the process of charging the car; For charging pile Input current error during charging of a car; The output current error parameter sequence of the charging pile during the process of charging the car; For charging pile Output current error during charging of a car; It is the output voltage regulation accuracy parameter sequence of the charging pile during the process of charging the car; For charging pile Output voltage regulation accuracy during the charging process of a car; It is the output current voltage regulation accuracy parameter sequence of the charging pile during the process of charging the car; For charging pile Output current regulation accuracy of a car during charging; is the charging power factor parameter sequence of the charging pile; For charging pile The charging power factor of the vehicle; is the whole power parameter sequence of the charging pile; For charging pile The total power of the vehicle; is the temperature difference parameter sequence between the charging pile body temperature and the ambient temperature; For charging pile The temperature difference of the car.

[0087] S22: Calculate the entropy value of each parameter sequence, sort the entropy values ​​from large to small, and classify the parameter sequences into triplets. The triplet classification may be:

[0088]

[0089]

[0090] In the formula, Indicates that the entropy value is ranked A parameter sequence of , For the The first parameter Time series data at the moment, , , Represents the classified triples.

[0091] Convert the time series signal into an image and construct a grouped image of the charging pile sample data based on the red, green and blue channels , , .

[0092] As a preferred implementation of this embodiment, in step S22, the timing signal is converted into an image, and a grouped image of the charging pile sample data is constructed based on the red, green and blue channels. , , Specifically: A1: For the converted RGB image, the red channel pixel value is filled by the triplet with the largest entropy value in each group, the green channel pixel value is filled by the triplet with the second largest entropy value in each group, and the blue channel pixel value is filled by the triplet with the smallest entropy value in each group, so that the data range is transformed to The data range transformation formula is as follows:

[0093] In the formula, is the pixel of the image, is the rounding function; The first The first charging pile sample Parameters; is the first The minimum value of a parameter; is the first The maximum value of a parameter.

[0094] A2: Constructing a grouped image of charging pile sample data based on the red, green and blue channels , specifically:

[0095]

[0096]

[0097] In the formula, , , They are the first, second and third group of images of charging pile samples respectively; For the The red channel pixels of the group image; For the Green channel pixels of the group image; For the The blue channel pixels of the group image; .

[0098] S23: Perform convolution operations on the grouped images , , Perform feature extraction and obtain , , Specifically, the convolution kernel size can be set to =3*5*5 or 3*3*3, where Represents the number of channels, Represents height, Represents width.

[0099] S24: Construct a feature fusion model to perform feature fusion and obtain fusion features. The feature fusion model is specifically:

[0100] In the formula, For fusion features; for Feature weights; for Feature weights; for Feature weights.

[0101] As a preferred implementation of this embodiment, in step S24 , , The feature weights of are processed by the multi-scale attention mechanism, see Figure 2 As shown, specifically:

[0102] In the formula, Features , , The initial fusion features obtained by direct addition; for , The initial fusion features are obtained by directly adding the features; is the initial fusion feature of the input multi-scale attention mechanism, ; , , , , , , is the intermediate parameter; It is the global average pooling process; It is a 1×1 point-by-point convolution; It is a standardization layer, which aims to make the output become a standard normal distribution and reduce sample differences; is the activation function; The function output is the feature weight; For the general Input the feature weights obtained by the multi-scale attention mechanism, As The feature weights are: .

[0103] based on , Get Feature weights , specifically:

[0104] In the formula, for The weight of the feature; for Input the feature weights obtained by the multi-scale attention mechanism; for Input the feature weights obtained by the multi-scale attention mechanism.

[0105] Said Feature weights The specific calculation formula is:

[0106] In the formula, for The weight of the feature.

[0107] As a preferred implementation of this embodiment, considering that the characteristics of charging pile samples in the same state are highly similar and have small differences, nonlinear transformation is used to obtain similar characteristics and difference characteristics of the same type of samples. In step S3, the fusion characteristics of each charging pile sample are mapped by a nonlinear function to obtain similar characteristics and difference characteristics, specifically:

[0108]

[0109] In the formula, For the In the first operating state Similar characteristics of charging pile samples; For the In the first operating state Different characteristics of charging pile samples; , The number of different operating states of the set charging pile; For the In the first operating state Characteristics of charging pile samples; , Respectively in The mapping weights of similar features and difference features of charging pile samples under different states; , Respectively in The bias of the two features in the state; , Respectively for Nonlinear function of the two feature maps in the state; , ; , They are , Dimension information of For the The number of charging pile samples under different operating conditions; , They are the dimensions of two characteristics respectively; .

[0110] S3: The fusion features of each charging pile sample are mapped through nonlinear functions to obtain similar features and difference features, and the improved zebra optimization algorithm is used to optimize the transformation dimensions of similar features and difference features.

[0111] Dimensions of similar and different features , It is a key parameter that affects the evaluation results. A low dimension may not be able to fully extract information, while a high dimension will generate redundant information. As a preferred implementation of this embodiment, in step S3, the improved zebra optimization algorithm is used to optimize the transformation dimensions of similar features and difference features. Specifically: B1: Initialize the zebra optimization algorithm parameters and randomly initialize the population in the optimization space:

[0112] In the formula, For the Zebra in the The location of the dimension; To find the optimal lower bound; To find the upper boundary; for A random number between .

[0113] B2: Based on the number of optimization searches The span of the position update is dynamically adjusted according to the size of the fitness function change rate, so as to update the new position of the zebra in the foraging stage. Specifically:

[0114]

[0115]

[0116]

[0117] In the formula, is the set number of iterations The span of change affects the parameter; is the current iteration number; is the maximum number of iterations; For the The rate of change of the best fitness function for each iteration; For the The optimal fitness function at the iteration; For the The optimal fitness function at the iteration; and is the set weight coefficient; For the first stage Zebra in the The new location of the dimension; For the Zebra in the The location of the or , correspond , correspond ; For the The fitness function of the zebra, which can be the loss value of the loss function in the state evaluation model; For the first stage The new fitness function value of the zebra; For the pioneer zebra position; For the Dimensional Pioneer Zebra Position; for A random number between for A random value between For the The location of the zebras; For the first stage The new position of the zebra.

[0118] B3: Simulate the zebra's defense strategy against predators. When a lion attacks a zebra, the zebra flees; when other predators attack the zebra, the zebra fights back. This updates the zebra's new position in the optimization space:

[0119]

[0120] In the formula, is the switching probability; for A random number between When using strategy; others use ; In the zebra position update, if the new position has a better fitness function value, accept the new position; otherwise, stay at the current position; is the number of iterations; is the maximum number of iterations; is a constant of 0.01; for A random value between The state of the zebra being attacked; For the The state of the attacked zebra.

[0121] B4: Repeat steps B2 and B3 until the maximum number of iterations is met, and output the optimal solution of the dimension obtained by the zebra optimization algorithm.

[0122] S4: Taking the optimized similar features or difference features as input, based on the state evaluation model, the operation state of the charging pile is evaluated to obtain the classification probability of the operation state of the charging pile and output the operation state label.

[0123] As a preferred implementation of this embodiment, in step S4, a state assessment model is constructed based on a softmax function, specifically:

[0124] In the formula, For the In this state Similar features of charging pile samples The probability of each category obtained after the softmax function; For the In this state Different characteristics of charging pile samples The probability of each category obtained after the softmax function; , are weight matrices respectively.

[0125] The state assessment model calculates the probability that the charging pile sample belongs to various states, specifically:

[0126] In the formula, For charging pile samples The probability of belonging to each state; is the weight coefficient of the feature, For the Status A charging station sample.

[0127] Furthermore, the state evaluation model selects the cross entropy loss function for training. The loss function Defined as:

[0128] In the formula, For the In the state The real coded labels of the charging pile samples.

[0129] In order to verify the effectiveness and superiority of the method provided in this embodiment, some specific cases are provided below: Based on the method of this patent, the charging pile status is evaluated online, and the charging pile operation status is divided into healthy, sub-healthy, abnormal, and severe. The comparison between some evaluation results and actual verification results is shown in Table 1: Table 1 Comparison of some evaluation results and actual verification results

[0130] The evaluation accuracy of the method described in this embodiment is shown in Table 2.

[0131] Table 2 Evaluation accuracy table

[0132] Embodiment 2 Accordingly, this embodiment provides a charging pile operation status evaluation system, including a data acquisition module, a feature extraction and fusion module, a feature optimization module and an operation status evaluation module.

[0133] The data acquisition module is used to collect the operation monitoring data of the charging pile and calculate the operation status data of the charging pile through the operation monitoring data. This module is used to implement the function of step S1 in the first embodiment and will not be described in detail here.

[0134] The feature extraction and fusion module is used to construct a parameter sequence according to the time series based on the operating status data within a set time period, calculate the entropy value of each parameter sequence, sort the entropy value from large to small, classify the parameter sequence as a triplet and extract features, build a feature fusion model for feature fusion, and obtain fusion features. This module is used to implement the function of step S2 in Example 1 and will not be repeated here.

[0135] The feature optimization module is used to obtain similar features and difference features of the fusion features of each charging pile sample through nonlinear function mapping, and use the improved zebra optimization algorithm to optimize the transformation dimensions of the similar features and difference features. This module is used to implement the function of step S3 in Example 1 and will not be repeated here.

[0136] The operating status evaluation module is used to evaluate the operating status of the charging pile based on the status evaluation model, using the optimized similar features or difference features as input, obtain the classification probability of the charging pile operating status and output the operating status label. This module is used to implement the function of step S4 in Example 1 and will not be repeated here.

[0137] Embodiment 3 This embodiment provides an electronic device, which includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, and when the processor executes the computer program, the charging pile operating status assessment method as described in any embodiment of the present invention is implemented.

[0138] Embodiment 4 This embodiment provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the charging pile operating status evaluation method as described in any embodiment of the present invention is implemented.

[0139] In the embodiments of the present application, "at least one" refers to one or more, and "more than one" refers to two or more. "And / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can represent the existence of A alone, the existence of A and B at the same time, and the existence of B alone. Among them, A and B can be singular or plural. The character " / " generally indicates that the previous and next associated objects are in an "or" relationship. "At least one of the following" and similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one of a, b and c can be represented by: a, b, c, a and b, a and c, b and c, or a and b and c, where a, b, c can be single or multiple.

[0140] Those of ordinary skill in the art will appreciate that the various units and algorithm steps described in the embodiments disclosed herein can be implemented in a combination of electronic hardware, computer software, and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0141] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0142] In several embodiments provided in the present application, if any function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory; hereinafter referred to as: ROM), random access memory (Random Access Memory; hereinafter referred to as: RAM), disk or optical disk, and other media that can store program codes.

[0143] The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A charging pile operation status evaluation method, characterized in that: The following steps are involved: S1: Collecting operation monitoring data of the charging pile, and calculating the operation status data of the charging pile through the operation monitoring data; S2: Based on the running status data within the set time period, a parameter sequence is constructed according to the time series, the entropy value of each parameter sequence is calculated, the entropy values ​​are sorted from large to small, the parameter sequence is classified and feature extracted as a triple, a feature fusion model is constructed for feature fusion, and fusion features are obtained; S3: The fusion features of each charging pile sample are mapped through nonlinear functions to obtain similar features and difference features, and the improved zebra optimization algorithm is used to optimize the transformation dimensions of similar features and difference features; S4: Taking the optimized similar features or difference features as input, based on the state evaluation model, the operation state of the charging pile is evaluated to obtain the classification probability of the operation state of the charging pile and output the operation state label.

2. The charging pile operation status evaluation method according to claim 1, characterized in that: The step S2 is specifically as follows: S21: Based on the operation status data within the set time period, the operation monitoring data includes: input current / voltage, output voltage / current, input power, output power, pile body temperature, ambient temperature, and the operation status data includes: input voltage / current error, output voltage / current error, output voltage / current voltage regulation accuracy, power factor, whole machine power, temperature difference, and construct a parameter sequence according to the time sequence, specifically: In the formula, The input voltage error parameter sequence of the charging pile during the process of charging the car; For charging pile The input voltage error during the charging process of a car, ; The total number of cars charged; The output voltage error parameter sequence of the charging pile during the process of charging the car; For charging pile Output voltage error during charging of a car; The input current error parameter sequence of the charging pile during the process of charging the car; For charging pile Input current error during charging of a car; The output current error parameter sequence of the charging pile during the process of charging the car; For charging pile Output current error during charging of a car; It is the output voltage regulation accuracy parameter sequence of the charging pile during the process of charging the car; For charging pile Output voltage regulation accuracy during charging of a vehicle; It is the output current voltage regulation accuracy parameter sequence of the charging pile during the process of charging the car; For charging pile Output current regulation accuracy of a car during charging; is the charging power factor parameter sequence of the charging pile; For charging pile The charging power factor of the vehicle; is the whole power parameter sequence of the charging pile; For charging pile The total power of the vehicle; is the temperature difference parameter sequence between the charging pile body temperature and the ambient temperature; For charging pile The temperature difference between the two cars; S22: Calculate the entropy value of each parameter sequence, sort the entropy values ​​from large to small, classify the parameter sequence into triplets, convert the time series signal into an image, and construct a grouped image of the charging pile sample data based on the red, green, and blue channels , , ; S23: Perform convolution operations on the grouped images , , Perform feature extraction and obtain , , feature; S24: Construct a feature fusion model to perform feature fusion and obtain fusion features. The feature fusion model is specifically: In the formula, For fusion features; for Feature weights; for Feature weights; for Feature weights.

3. The charging pile operation status evaluation method according to claim 2, characterized in that: In step S22, the timing signal is converted into an image, and a grouped image of the charging pile sample data is constructed based on the red, green and blue channels. , , Specifically: A1: For the converted RGB image, the red channel pixel value is filled by the triplet with the largest entropy value in each group, the green channel pixel value is filled by the triplet with the second largest entropy value in each group, and the blue channel pixel value is filled by the triplet with the smallest entropy value in each group. The data range conversion formula is as follows: In the formula, is the pixel of the image, is the rounding function; The first The first charging pile sample Parameters; is the first The minimum value of a parameter; is the first The maximum value of a parameter; A2: Constructing a grouped image of charging pile sample data based on the red, green and blue channels , specifically: In the formula, , , They are the first, second and third group of images of charging pile samples respectively; For the The red channel pixels of the group image; For the Green channel pixels of the group image; For the The blue channel pixels of the group image; .

4. The charging pile operation status evaluation method according to claim 2, characterized in that: In step S24 , , The feature weights of are processed by the multi-scale attention mechanism, specifically: In the formula, Features , , The initial fusion features obtained by direct addition; for , The initial fusion features are obtained by directly adding the features; is the initial fusion feature of the input multi-scale attention mechanism, ; , , , , , , is the intermediate parameter; It is the global average pooling process; It is a 1×1 point-by-point convolution; It is a standardization layer, which aims to make the output become a standard normal distribution and reduce sample differences; is the activation function; The function output is the feature weight; For the general Input the feature weights obtained by the multi-scale attention mechanism, As The feature weights are: ; based on , Get Feature weights , specifically: In the formula, for The weight of the feature; for Input the feature weights obtained by the multi-scale attention mechanism; for Input the feature weights obtained by the multi-scale attention mechanism; Said Feature weights The specific calculation formula is: In the formula, for The weight of the feature.

5. The charging pile operation status evaluation method according to claim 1, characterized in that: In step S3, the fusion features of each charging pile sample are mapped by nonlinear function to obtain similar features and difference features, specifically: In the formula, For the In the first operating state Similar characteristics of charging pile samples; For the In the first operating state Different characteristics of charging pile samples; , The number of different operating states of the set charging pile; For the In the first operating state Characteristics of charging pile samples; , Respectively in The mapping weights of similar features and difference features of charging pile samples under different states; , Respectively in The bias of the two features in the state; , Respectively Nonlinear function of the two feature maps in the state; , ; , They are , Dimension information of For the The number of charging pile samples under different operating conditions; , They are the dimensions of two characteristics respectively; .

6. The charging pile operation status evaluation method according to claim 5, characterized in that: In step S3, the improved zebra optimization algorithm is used to optimize the transformation dimensions of similar features and difference features. Specifically, B1: Initialize the zebra optimization algorithm parameters and randomly initialize the population in the optimization space; B2: Based on the number of optimization searches The span of the position update is dynamically adjusted according to the size of the fitness function change rate, so as to update the new position of the zebra in the first stage, that is, the foraging stage. Specifically: In the formula, is the set number of iterations The span of change affects the parameter; is the current iteration number; is the maximum number of iterations; For the The rate of change of the best fitness function for each iteration; For the The optimal fitness function at the iteration; For the The optimal fitness function at the iteration; and is the set weight coefficient; For the first stage Zebra in the The new location of the dimension; For the Zebra in the The location of the or , correspond , correspond ; For the The fitness function of a zebra; For the first stage The new fitness function value of the zebra; For the pioneer zebra position; For the Dimensional Pioneer Zebra Position; for A random number between for A random value between For the The location of the zebras; For the first stage the new position of the zebra; B3: Simulate the zebra's defense strategy against predators. When a lion attacks a zebra, the zebra flees; when other predators attack the zebra, the zebra fights back. This updates the zebra's new position in the optimization space. B4: Repeat steps B2 and B3 until the maximum number of iterations is met, and output the optimal solution of the dimension obtained by the zebra optimization algorithm.

7. The charging pile operation status evaluation method according to claim 5, characterized in that: In step S4, a state evaluation model is constructed based on the softmax function, specifically: In the formula, For the In this state Similar features of charging pile samples The probability of each category obtained after the softmax function; For the In this state Different characteristics of charging pile samples The probability of each category obtained after the softmax function; , are weight matrices respectively; The state assessment model calculates the probability that the charging pile sample belongs to various states, specifically: In the formula, For charging pile samples The probability of belonging to each state; is the weight coefficient of the feature, For the Status A charging station sample.

8. A charging pile operation status evaluation system, characterized in that: It includes data acquisition module, feature extraction and fusion module, feature optimization module and operation status evaluation module; A data acquisition module is used to collect the operation monitoring data of the charging pile and calculate the operation status data of the charging pile through the operation monitoring data; The feature extraction and fusion module is used to construct a parameter sequence according to the time sequence based on the operating status data within the set time period, calculate the entropy value of each parameter sequence, sort the entropy values ​​from large to small, classify the parameter sequence into triplets and extract features, build a feature fusion model for feature fusion, and obtain fusion features; The feature optimization module is used to obtain similar features and difference features of the fusion features of each charging pile sample through nonlinear function mapping, and use the improved zebra optimization algorithm to optimize the transformation dimensions of similar features and difference features; The operating status evaluation module is used to evaluate the operating status of the charging pile based on the status evaluation model, using the optimized similar features or difference features as input, obtain the classification probability of the charging pile operating status and output the operating status label.

9. An electronic device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the charging pile operating status assessment method as described in any one of claims 1 to 7 when executing the computer program.

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 charging pile operating status evaluation method according to any one of claims 1 to 7 is implemented.

Images