Monitoring and early warning management system and method for automobile charging pile

Abstract

The invention relates to the technical field of automobile charging pile monitoring and early warning, in particular to an automobile charging pile monitoring and early warning management system and method.The automobile charging pile monitoring and early warning management system comprises an operation state monitoring module, an influence analysis module, an influence intensity analysis module, an early warning management module and a cloud database. Abnormal data frames are screened, an abnormal data sequence of parameter dimensions is formed, graphical analysis is facilitated, abnormal charging piles are judged, when the abnormal charging piles are detected, a topological model is constructed based on the electrical topological relation of abnormal sources and the charging piles, a trained machine learning model is utilized to predict the core result of the abnormal charging piles, and the core result of the abnormal charging piles is obtained. And the comprehensive risk value is calculated based on the swept core result, and grading early warning is performed, so that abnormity identification, swept prediction and quantitative risk intervention are realized, and the operation safety and management efficiency of the charging pile are improved.

Description

Technical Field

[0001] This invention relates to the field of monitoring and early warning technology for electric vehicle charging stations, specifically to a monitoring and early warning management system and method for electric vehicle charging stations. Background Technology

[0002] With the rapid expansion of the new energy vehicle market, the number and distribution of electric vehicle charging piles, as a key infrastructure supporting the popularization of electric vehicles, are growing rapidly. However, charging piles face a variety of potential risks during operation, such as hardware failures, software anomalies, external power grid fluctuations, and human error. These anomalies may not only cause the function of a single charging pile to fail, but may also spread the abnormal state to other charging piles on the same line or associated through power lines, communication networks, or electromagnetic coupling, causing a wider range of failures and safety risks, and seriously affecting the reliability and stability of charging services. Currently, although some charging pile systems have certain monitoring functions, most of them only collect and display the basic operating parameters of a single charging pile. They lack in-depth analysis of the charging pile's operating status and the ability to provide early warning of anomalies. When a charging pile malfunctions, it is difficult to quickly and accurately identify the abnormal charging pile, and it is also impossible to effectively predict the scope, time, intensity, and other key information that the abnormal state may affect. Furthermore, it is impossible to assess the risk level of the affected events, thus making it impossible to take targeted early warning and management measures.

[0003] To address the aforementioned shortcomings, a technical solution is provided. Summary of the Invention

[0004] In view of the above-mentioned shortcomings of the existing technology, the present invention provides a monitoring and early warning management system and method for electric vehicle charging piles, which can effectively solve the problems of difficulty in timely identification of charging pile anomalies, difficulty in predicting the scope of anomalies, and difficulty in quantifying and warning of risk levels in the existing technology.

[0005] To achieve the above objectives, the present invention can be implemented through the following technical solutions: This invention provides a monitoring and early warning management system for vehicle charging stations, comprising: The operation status monitoring module is used to collect the working parameters of charging piles in the target area according to the set collection cycle, generate data frames and upload them to the data analysis terminal. The data analysis terminal performs abnormal data frame filtering processing on the data frames, performs time-series processing on the filtered abnormal data frames, and forms an abnormal data sequence under the parameter dimension, which is convenient for graphical analysis and thereby determines abnormal charging piles. The impact analysis module is used to construct a topological model of the abnormal charging pile based on the abnormal source of the abnormal charging pile and the electrical topology relationship when the operation status monitoring module determines that an abnormal charging pile has appeared in the target area. It also uses a trained machine learning model to predict the core results of the impact, which include the probability, time, location, intensity and propagation speed of the impact. The impact intensity analysis module is used to determine the risk level of the impact based on the core results affected by abnormal charging piles in the target area; The early warning management module is used to implement graded early warning and management measures based on the level of risk involved.

[0006] Furthermore, the data analysis terminal performs abnormal data frame filtering processing on the data frame, the specific process of which is as follows: The cumulative runtime of charging piles within the target area is obtained and compared with multiple preset operating level ranges to determine the operating level of the charging piles within the target area. Each operating level corresponds to a different basic weight of working parameters. Based on the basic weights, a real-time anomaly correction coefficient is introduced. For each working parameter, if its collected value exceeds the preset normal threshold range, the anomaly correction coefficient is calculated according to its deviation. The value of the working parameter is then multiplied by the basic weights and the anomaly correction coefficient to obtain the corrected working parameter value. When the data analysis terminal detects that the value of any corrected working parameter exceeds the preset normal threshold range, the corrected working parameter is marked as abnormal data. The number of parameters marked as abnormal data in each data frame is accumulated. When the number of parameters exceeds the preset threshold, the data frame is automatically marked as an abnormal data frame.

[0007] Furthermore, the selected abnormal data frames are processed in a time-series manner, as follows: All the marked anomalous data frames are arranged in chronological order according to their timestamps to obtain a time-ordered sequence of anomalous data frames. From the abnormal data frame sequence, the abnormal data corresponding to each working parameter is extracted to form an abnormal data sequence under the parameter dimension.

[0008] Furthermore, after time-series processing, graphical analysis is performed, the specific process of which is as follows: On a given base time-direction map, the values ​​of the abnormal data corresponding to each working parameter are mapped to the abnormal points on the map in chronological order. Connect adjacent outliers in the time series sequentially to form an outlier data graph for each working parameter; Pairwise comparisons were performed on the abnormal data graphs of each operating parameter to analyze the logic of the concavity and convexity changes of the curves over time. If the abnormal data graphs of different operating parameters show a reasonable synergistic relationship in the time series, then the charging pile's operating status is determined to be a normal trend. If significant inconsistencies or violations of preset logical relationships occur, the charging pile's operating status is determined to be an abnormal trend. Based on the judgment results, charging piles that are judged to have abnormal trends will be marked as abnormal charging piles.

[0009] Furthermore, the topology model of the abnormal charging pile is constructed, and the specific process is as follows: Centered on the abnormal charging pile, a topology network for visualization calculation is constructed based on the actual electrical connection relationship. On the topology network for visualization calculation, the connection method of all charging pile nodes and power distribution nodes in the line is clearly marked. Then, the physical parameters of the line are integrated into the model to form the topology model of the abnormal charging pile.

[0010] Furthermore, the machine learning model is trained, and the specific process is as follows: Historical operational data of events involving charging piles of the same model as the abnormal charging piles are extracted and used to train a machine learning model. The machine learning model can be a convolutional neural network model or a decision tree model, thus obtaining the trained machine learning model.

[0011] Furthermore, the trained machine learning model is used to predict the core outcomes of the spillover effect. The specific process is as follows: The real-time operation data of abnormal charging piles is collected and input into a trained machine learning model. The model will output whether an abnormal charging pile has caused an event and the pattern of the event. When an abnormal charging pile has caused an event, the time, location and intensity of the event will be output simultaneously. After predicting the ripple patterns of abnormal charging pile incidents, the impact data of historical abnormal charging pile incidents is extracted. This impact data reflects the impact range caused by abnormal charging pile incidents with different ripple patterns, and the impact range includes the impact range at different times. The propagation speed is determined based on the ratio of the impact range to the impact time.

[0012] Further, the risk level of the impact is determined, and the specific process is as follows: The comprehensive risk value is calculated based on the core results affected by abnormal charging piles in the target area. Then, the risk is classified according to the preset threshold range to determine the risk level, which includes low risk level, medium risk level and high risk level. The formula for calculating its overall risk value is:

[0013] In the formula, For the comprehensive risk value, For time feature values, For location feature values, For intensity characteristic value, For the characteristic value of propagation speed, , , and These are the weight coefficients for the corresponding feature values, and , These are constant coefficients used for normalization. These are correction coefficients used to compensate for model bias. The label of the abnormal charging pile within the target area; Among them, by extracting the time point of occurrence of abnormal charging pile impact events, and combining real-time monitoring data with historical event data, the duration of impact events is predicted by machine learning models, thereby determining time feature values; Geographic coordinates of abnormal charging piles are obtained based on geographic information systems. Combined with urban planning and functional zoning, different levels of weight values ​​are assigned to the locations of abnormal charging piles to determine location characteristic values. The operating parameters of the affected charging piles are collected by sensors, and these operating parameters are compared with those under normal conditions. The magnitude of the change in the operating parameters is calculated, and the average magnitude of the change in the operating parameters is taken as the intensity characteristic value. The propagation speed characteristic value is determined by calculating the rate of change of the area of ​​influence per unit time.

[0014] Furthermore, tiered early warning and management measures are implemented based on the level of risk affected, as follows: When the risk level is low, the detailed information of the abnormal charging pile is transmitted to the target area management terminal and recorded and archived in the background. When the risk level is medium, an early warning message will be sent to the management terminal of the target area; When the risk level is high, an alarm mechanism will be triggered immediately, and coordinated protection measures will be implemented.

[0015] Furthermore, a method for monitoring and early warning management of car charging stations includes the following steps: Step 1: Collect the working parameters of charging piles in the target area according to the set collection cycle, generate data frames and upload them to the data analysis terminal. The data analysis terminal performs abnormal data frame filtering processing on the data frames, performs time-series processing on the filtered abnormal data frames to form an abnormal data sequence under the parameter dimension, which is convenient for graphical analysis and thus determines abnormal charging piles. Step 2: When the operation status monitoring module determines that an abnormal charging pile has appeared in the target area, based on the abnormal source of the abnormal charging pile and combined with the electrical topology relationship, a topological model of the abnormal charging pile is constructed, and the trained machine learning model is used to predict the core results of the impact. The core results cover the possibility, time, location, intensity and propagation speed of the impact. Step 3: Based on the core results affected by abnormal charging piles in the target area, determine the risk level of the impact; Step 4: Implement tiered early warning and management measures based on the level of risk affected.

[0016] Compared with the prior art, the technical solution provided by the present invention has the following beneficial effects: 1. This invention periodically collects the operating parameters of charging piles within the target area and performs anomaly screening and graphical analysis on the collected data frames to achieve rapid and accurate identification of abnormal charging piles, effectively avoiding the misjudgment or missed judgment problems that may be caused by traditional monitoring methods that rely solely on the display of a single parameter. 2. This invention constructs a topology model based on the electrical topology relationship of charging piles and combines it with a trained machine learning model to predict the possibility, time, location, intensity and propagation speed of abnormal charging pile impacts, providing maintenance personnel with early prediction of potential risks and a scientific basis for decision-making. 3. This invention calculates the comprehensive risk value of abnormal charging pile-related events by comprehensively analyzing key characteristic indicators such as time, location, intensity, and propagation speed, and classifies the risks according to a preset threshold range. Under different risk levels, corresponding graded early warning and management measures can be implemented to achieve targeted intervention and effectively reduce the possibility of the occurrence and spread of abnormal events. Attached Figure Description

[0017] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0018] Figure 1 This is an overall module block diagram of the present invention.

[0019] Figure 2 This is a flowchart illustrating the overall process of the present invention. Detailed Implementation

[0020] To make the purpose, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0021] like Figure 1 As shown, a monitoring and early warning management system for electric vehicle charging piles includes: an operation status monitoring module, a ripple effect analysis module, an impact intensity analysis module, an early warning management module, and a cloud database; The cloud database is used to store historical operational data on charging pile-related events and impact data on historical abnormal charging pile-related events.

[0022] The operation status monitoring module is used to collect the operating parameters of charging piles in the target area through built-in sensors. The specific collection process is as follows: Multiple types of sensors are deployed inside the charging piles within the target area, including but not limited to voltage sensors, current sensors, temperature sensors, humidity sensors, insulation resistance detectors, and ambient light sensors. Each sensor collects corresponding operating parameters at a set acquisition cycle (such as 1 second or 5 seconds). These operating parameters include voltage, current, temperature, humidity, insulation resistance, and ambient light. A timestamp and a unique device identifier are added to each set of collected operating parameters, generating a data frame and uploading it to the data analysis terminal, thereby enabling real-time monitoring and anomaly prediction of the charging pile's operating status.

[0023] After receiving a data frame at the data analysis end, an abnormal data frame filtering process is executed: By obtaining the cumulative running time of charging piles in the target area (i.e., the sum of the running time of each charging from the first charging), and comparing it with multiple preset operating level intervals, the operating level of the charging piles in the target area is determined. Each operating level corresponds to a different basic weight of working parameters. As the usage time of the charging pile increases, the weight of safety-related parameters such as temperature and insulation resistance gradually increases, so as to improve the monitoring sensitivity of old charging piles. Based on the basic weights, a real-time anomaly correction coefficient is introduced. For each working parameter, if its collected value exceeds the preset normal threshold range, the anomaly correction coefficient is calculated according to its deviation. The value of the working parameter is then multiplied by the basic weights and the anomaly correction coefficient to obtain the corrected working parameter value. The specific implementation process for calculating the anomaly correction coefficient based on its degree of deviation is as follows: Suppose the collected value of a certain working parameter is The corresponding normal threshold range is ; when When the value of the working parameter is within the normal range, the deviation degree D is recorded as 0. when When the deviation degree D is calculated, it is as follows: ; when When the deviation degree D is calculated, it is as follows: ; Based on the degree of deviation D, the anomaly correction coefficient Q is calculated, and the specific calculation formula is as follows: ,in, This is a preset correction sensitivity coefficient used to adjust the sensitivity of different parameters to deviations; When the data analysis terminal detects that the value of any corrected working parameter exceeds the preset normal threshold range, the corrected working parameter is marked as abnormal data. The number of parameters marked as abnormal data in each data frame is accumulated. When the number of parameters exceeds the preset threshold, the data frame is automatically marked as an abnormal data frame. This not only effectively avoids misjudgment caused by the instantaneous fluctuation of a single parameter, but also enables a comprehensive evaluation of the overall operating status, thereby more accurately reflecting the true operating status of the charging pile.

[0024] After filtering out abnormal data frames, the abnormal data frames are further processed in a time sequence and analyzed graphically to achieve intelligent determination of abnormal status of charging piles in the target area. The specific implementation process is as follows: All the marked anomalous data frames are arranged in chronological order according to their timestamps to obtain a time-ordered sequence of anomalous data frames. From the abnormal data frame sequence, the abnormal data corresponding to each working parameter is extracted to form an abnormal data sequence under the parameter dimension, which facilitates the identification of abnormal evolution patterns and provides reliable data support for subsequent visualization analysis. Using the time direction as the core reference, construct the abnormal data graph of each working parameter as follows: On the given base time direction graph (e.g., a circular time direction graph), map the values ​​of the abnormal data corresponding to each working parameter to the abnormal points on the graph in chronological order. The larger the value of the abnormal data, the further away the plotted point is from the center. Connect adjacent outliers in the time series sequentially to form anomaly data graphs for each operating parameter (such as current anomaly data graph, voltage anomaly data graph, insulation resistance anomaly data graph, etc.). Pairwise comparisons were performed on the abnormal data graphs of each operating parameter to analyze the logic of the concavity and convexity changes of the curves over time. If the abnormal data graphs of different operating parameters show a reasonable cooperative relationship in the time series (e.g., fluctuating in the same direction, with amplitude changes maintaining a certain proportion), then the charging pile's operating status is determined to be a normal trend. If significant inconsistencies or violations of preset logical relationships occur (e.g., an abnormal peak value of a certain parameter is seriously out of sync with the abnormal trends of other parameters), the charging pile's operating status is determined to be an abnormal trend. Based on the assessment results, charging piles that are identified as having abnormal trends will be marked as abnormal charging piles to provide data support for subsequent risk warning and control.

[0025] The impact analysis module is used to determine the core results of the impact of the abnormal charging pile in the target area when the operation status monitoring module determines that an abnormal charging pile has appeared in the target area. Based on the electrical topology relationship between the abnormal charging pile and other charging piles on the same line and historical operation data, the core results cover the possibility, time, location, intensity and propagation speed of the impact. The specific implementation process is as follows: Extract the unique identifier, line number, time of occurrence of anomaly, and characteristics of abnormal parameters (such as sudden changes in current amplitude and voltage drop) of abnormal charging piles in the target area, and use this information as the source of the anomaly. It should be noted that the unique identifier and the line number of the anomaly source can accurately determine the specific location of the abnormal charging pile in the target line. When constructing the topology network, the abnormal charging pile is taken as the center, and its connection relationship with other charging piles and power distribution nodes is clarified. Just like determining a key location on a map, the transportation network of the entire area is drawn based on this, thus clearly showing the relationship of "who is connected to whom and how energy is transferred".

[0026] Based on the electrical topology relationship between the abnormal charging pile and other charging piles on the same line, a topology model of the abnormal charging pile is constructed, which is specifically constructed as follows: S1: Centered on abnormal charging piles, a topology network for visual calculation is constructed based on the actual electrical connection relationship; S2: On the topology network of the visualization calculation, clearly mark the connection method of all charging pile nodes (including abnormal charging piles) and power distribution nodes (such as distribution boxes and transformers) in the line, such as "charging pile A - distribution box 1 - charging pile B" and "charging pile C - transformer 2 - charging pile D", and clarify the "neighbor relationship" and hierarchy of each node. S3: Then incorporate physical parameters such as line resistance (e.g., wire resistance), inductance, and capacitance into the model. These physical parameters directly affect the degree of propagation and attenuation of anomalies. For example, the greater the resistance, the faster the current anomaly attenuates during propagation, and the smaller the affected area may be. Conversely, the smaller the resistance, the farther the affected area may be. Through the above S1-S3 steps, a topological model of the abnormal charging pile is formed, which serves as the basic structure for analyzing the impact path.

[0027] Furthermore, machine learning models are introduced to predict core results, which cover the likelihood, timing, location, intensity, and propagation speed of the impact. The specific prediction process is as follows: Historical operational data of charging pile impact events with the same model as the abnormal charging pile are extracted from the cloud database, and machine learning models are trained using the historical operational data. The extracted historical operational data must include records of affected charging piles and ensure that it covers information such as the time, location, and intensity of each affected charging pile when the incident occurred. The machine learning model is trained using extracted historical operating data. This solution can use convolutional neural network models, decision tree models, etc., and there are no specific limitations on the specific model. Based on the trained machine learning model, the prediction of abnormal charging pile impact events can be performed in the subsequent execution. Furthermore, the system acquires real-time operational data of abnormal charging piles collected by sensors and inputs this data into a trained machine learning model. The model will output whether an abnormal charging pile incident has occurred and the incident pattern. When it is clear that an abnormal charging pile incident has occurred, the system will simultaneously output the time, location, and intensity of the incident. It should be noted that the charging pile spillover event in this solution refers to the event where, when a charging pile malfunctions due to hardware failure, software abnormality, external interference (such as power grid fluctuations), or human error, its abnormal state spreads to other charging piles through physical connections (such as power lines or communication networks) or electromagnetic coupling, resulting in performance degradation, functional failure, or even safety risks to adjacent or related charging piles. Furthermore, after predicting the ripple pattern of abnormal charging pile events, in order to clarify the impact of the ripple pattern of the abnormal charging pile events at different times, the impact data of historical abnormal charging pile events is extracted from the cloud database. This impact data reflects the impact range caused by abnormal charging pile events with different ripple patterns, and the impact range includes the impact range at different times. The propagation speed can be determined based on the ratio of the impact range to the impact time.

[0028] The impact intensity analysis module is used to determine the risk level of the impact based on the core results affected by abnormal charging piles in the target area. Its specific implementation process is as follows: The comprehensive risk value is calculated based on the core results affected by abnormal charging piles within the target area. The formula for calculating the comprehensive risk value is as follows:

[0029] In the formula, For the comprehensive risk value, For time feature values, For location feature values, For intensity characteristic value, For the characteristic value of propagation speed, , , and These are the weight coefficients for the corresponding feature values, and , These are constant coefficients used for normalization. These are correction coefficients used to compensate for model bias. The label for the abnormal charging pile within the target area.

[0030] in: The time feature value represents the duration of the impact event of the abnormal charging pile. Generally speaking, the longer the impact time, the higher the risk. This solution extracts the occurrence time of the abnormal charging pile impact event and compares it with real-time monitoring data and historical event data. It uses a machine learning model to predict the duration of the impact event, thereby determining the time feature value. Location feature values ​​indicate the importance of the location of the incident caused by abnormal charging piles. Charging piles in key areas (such as commercial centers, hospitals, and transportation hubs) have higher location risks. This solution obtains the geographic coordinates of abnormal charging piles based on Geographic Information System (GIS), and combines urban planning and functional zoning to assign different levels of weight values ​​to the locations of abnormal charging piles in order to determine the location feature values. The intensity characteristic value represents the intensity of the impact of an abnormal charging pile ripple event on other charging piles. This solution collects the operating parameters (such as current, voltage, power, etc.) of the affected charging piles through sensors, compares these operating parameters with the operating parameters under normal conditions, calculates the change range of the operating parameters, and takes the average change range of the operating parameters as the intensity characteristic value. The propagation speed characteristic value represents the speed at which the abnormal impact spreads. The faster the propagation speed, the higher the potential risk. This scheme determines the propagation speed characteristic value by recording the impact range (i.e. the number of charging piles affected) of the abnormal charging pile impact event at different times and calculating the rate of change of the impact range per unit time.

[0031] After calculating the comprehensive risk value, it is based on the preset threshold range. The risks are classified as follows: When the comprehensive risk value < If so, the impact of abnormal charging piles in the target area is determined to be at a low risk level; When the comprehensive risk value ≤ < If so, the impact of abnormal charging piles in the target area is determined to be at a medium risk level; When the comprehensive risk value ≥ If so, the impact of abnormal charging piles within the target area is determined to be at a high risk level.

[0032] The early warning management module is used to implement graded early warning and management measures based on the risk level of abnormal charging piles in the target area. The specific implementation management process is as follows: When the risk level of the abnormal charging pile in the target area is determined to be low, the detailed information of the abnormal charging pile is transmitted to the target area management terminal, recorded and archived in the background, and the operation and maintenance personnel are prompted to perform routine checks. When the risk level of abnormal charging piles in the target area is determined to be medium risk, an early warning message will be pushed to the management terminal of the target area to remind the regional operation and maintenance team to pay close attention and make emergency preparations. When the risk level of abnormal charging piles in the target area is determined to be high, an alarm mechanism is immediately triggered to implement linkage protection measures, including: sending a high-priority alarm to the management terminal to ensure that the management personnel can receive it as soon as possible, automatically switching the power supply path to reduce the impact of charging piles on the overall line, and calling backup power to ensure the continuous operation of critical charging piles to prevent the risk from spreading further.

[0033] like Figure 2 As shown, a monitoring and early warning management method for car charging stations includes the following steps: Step 1: Collect the working parameters of charging piles in the target area according to the set collection cycle, generate data frames and upload them to the data analysis terminal. The data analysis terminal performs abnormal data frame filtering processing on the data frames. Specifically, the filtering is as follows: obtain the cumulative running time of charging piles in the target area and compare it with multiple preset operating level intervals to determine the operating level to which the charging piles in the target area belong. Each operating level corresponds to a different basic weight of working parameters. Based on the basic weights, a real-time anomaly correction coefficient is introduced. For each working parameter, if its collected value exceeds the preset normal threshold range, the anomaly correction coefficient is calculated according to its deviation. The value of the working parameter is then multiplied by the basic weights and the anomaly correction coefficient to obtain the corrected working parameter value. When the data analysis terminal detects that the value of any corrected working parameter exceeds the preset normal threshold range, the corrected working parameter is marked as abnormal data. The number of parameters marked as abnormal data in each data frame is accumulated. When the number of parameters exceeds the preset number threshold, the data frame is automatically marked as an abnormal data frame. The selected abnormal data frames are processed to be time-ordered. Specifically, all the marked abnormal data frames are arranged in chronological order according to their timestamps to obtain a time-ordered sequence of abnormal data frames. From the abnormal data frame sequence, extract the abnormal data corresponding to each working parameter to form an abnormal data sequence under the parameter dimension; On a given base time-direction map, the values ​​of the abnormal data corresponding to each working parameter are mapped to the abnormal points on the map in chronological order. Connect adjacent outliers in the time series sequentially to form an outlier data graph for each working parameter; Pairwise comparisons were performed on the abnormal data graphs of each operating parameter to analyze the logic of the concavity and convexity changes of the curves over time. If the abnormal data graphs of different operating parameters show a reasonable synergistic relationship in the time series, then the charging pile's operating status is determined to be a normal trend. If significant inconsistencies or violations of preset logical relationships occur, the charging pile's operating status is determined to be an abnormal trend. Based on the judgment results, charging piles that are judged to have abnormal trends will be marked as abnormal charging piles. Step Two: When the operation status monitoring module determines that an abnormal charging pile has appeared in the target area, based on the abnormal source of the abnormal charging pile and combined with the electrical topology relationship, a topological model of the abnormal charging pile is constructed, and the trained machine learning model is used to predict the core results of the impact. The specific prediction process is as follows: The real-time operation data of abnormal charging piles is collected and input into a trained machine learning model. The model will output whether an abnormal charging pile has caused an event and the pattern of the event. When an abnormal charging pile has caused an event, the time, location and intensity of the event will be output simultaneously. After predicting the ripple pattern of abnormal charging pile ripple events, the impact data of historical abnormal charging pile ripple events is extracted. This impact data reflects the impact range caused by abnormal charging pile ripple events with different ripple patterns, and the impact range includes the impact range at different times. The propagation speed is determined based on the ratio of the impact range to the impact time. Step 3: Based on the core results of the abnormal charging piles in the target area, calculate the comprehensive risk value, and then classify the risk according to the preset threshold range to determine the risk level of the affected area. The risk levels include low risk level, medium risk level and high risk level. The formula for calculating its overall risk value is:

[0034] In the formula, For the comprehensive risk value, For time feature values, For location feature values, For intensity characteristic value, For the characteristic value of propagation speed, , , and These are the weight coefficients for the corresponding feature values, and , These are constant coefficients used for normalization. These are correction coefficients used to compensate for model bias. The label of the abnormal charging pile within the target area; Step 4: Implement tiered early warning and management measures based on the level of risk affected, specifically as follows: When the risk level is low, the detailed information of the abnormal charging pile is transmitted to the target area management terminal and recorded and archived in the background. When the risk level is medium, an early warning message will be sent to the management terminal of the target area; When the risk level is high, an alarm mechanism will be triggered immediately, and coordinated protection measures will be implemented.

[0035] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to any specific implementation. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A monitoring and early warning management system for electric vehicle charging piles, characterized in that, include: The operation status monitoring module is used to collect the working parameters of charging piles in the target area according to the set collection cycle, generate data frames and upload them to the data analysis terminal. The data analysis terminal performs abnormal data frame filtering processing on the data frames, performs time-series processing on the filtered abnormal data frames, and forms an abnormal data sequence under the parameter dimension, which is convenient for graphical analysis and thereby determines abnormal charging piles. The impact analysis module is used to construct a topological model of the abnormal charging pile based on the abnormal source of the abnormal charging pile and the electrical topology relationship when the operation status monitoring module determines that an abnormal charging pile has appeared in the target area. It also uses a trained machine learning model to predict the core results of the impact, which include the probability, time, location, intensity and propagation speed of the impact. The impact intensity analysis module is used to determine the risk level of the impact based on the core results affected by abnormal charging piles in the target area; The early warning management module is used to implement graded early warning and management measures based on the level of risk involved.

2. The monitoring and early warning management system for car charging piles according to claim 1, characterized in that, The data analysis terminal performs abnormal data frame filtering processing on the data frame, and the specific process is as follows: The cumulative runtime of charging piles within the target area is obtained and compared with multiple preset operating level ranges to determine the operating level of the charging piles within the target area. Each operating level corresponds to a different basic weight of working parameters. Based on the basic weights, a real-time anomaly correction coefficient is introduced. For each working parameter, if its collected value exceeds the preset normal threshold range, the anomaly correction coefficient is calculated according to its deviation. The value of the working parameter is then multiplied by the basic weights and the anomaly correction coefficient to obtain the corrected working parameter value. When the data analysis terminal detects that the value of any corrected working parameter exceeds the preset normal threshold range, the corrected working parameter is marked as abnormal data. The number of parameters marked as abnormal data in each data frame is accumulated. When the number of parameters exceeds the preset threshold, the data frame is automatically marked as an abnormal data frame.

3. The monitoring and early warning management system for car charging piles according to claim 1, characterized in that, The selected abnormal data frames are then processed for time-series analysis, as follows: All the marked anomalous data frames are arranged in chronological order according to their timestamps to obtain a time-ordered sequence of anomalous data frames. From the abnormal data frame sequence, the abnormal data corresponding to each working parameter is extracted to form an abnormal data sequence under the parameter dimension.

4. The monitoring and early warning management system for electric vehicle charging piles according to claim 1, characterized in that, After time-series processing, graphical analysis is performed. The specific process is as follows: On a given base time-direction map, the values ​​of the abnormal data corresponding to each working parameter are mapped to the abnormal points on the map in chronological order. Connect adjacent outliers in the time series sequentially to form an outlier data graph for each working parameter; Pairwise comparisons were performed on the abnormal data graphs of each operating parameter to analyze the logic of the concavity and convexity changes of the curves over time. If the abnormal data graphs of different operating parameters show a reasonable synergistic relationship in the time series, then the charging pile's operating status is determined to be a normal trend. If significant inconsistencies or violations of preset logical relationships occur, the charging pile's operating status is determined to be an abnormal trend. Based on the judgment results, charging piles that are judged to have abnormal trends will be marked as abnormal charging piles.

5. The monitoring and early warning management system for car charging piles according to claim 1, characterized in that, The topology model of the abnormal charging pile is constructed as follows: Centered on the abnormal charging pile, a topology network for visualization calculation is constructed based on the actual electrical connection relationship. On the topology network for visualization calculation, the connection method of all charging pile nodes and power distribution nodes in the line is clearly marked. Then, the physical parameters of the line are integrated into the model to form the topology model of the abnormal charging pile.

6. The monitoring and early warning management system for electric vehicle charging piles according to claim 1, characterized in that, The specific process of training a machine learning model is as follows: Historical operational data of events involving charging piles of the same model as the abnormal charging piles were extracted and used to train a machine learning model, resulting in a trained machine learning model.

7. The monitoring and early warning management system for car charging piles according to claim 1, characterized in that, The core outcome of the spillover is predicted using a trained machine learning model, and the specific process is as follows: The real-time operation data of abnormal charging piles is collected and input into a trained machine learning model. The model will output whether an abnormal charging pile has caused an event and the pattern of the event. When an abnormal charging pile has caused an event, the time, location and intensity of the event will be output simultaneously. After predicting the ripple patterns of abnormal charging pile incidents, the impact data of historical abnormal charging pile incidents is extracted. This impact data reflects the impact range caused by abnormal charging pile incidents with different ripple patterns, and the impact range includes the impact range at different times. The propagation speed is determined based on the ratio of the impact range to the impact time.

8. The monitoring and early warning management system for car charging piles according to claim 1, characterized in that, The specific process for determining the level of risk impact is as follows: The comprehensive risk value is calculated based on the core results affected by abnormal charging piles in the target area. Then, the risk is classified according to the preset threshold range to determine the risk level, which includes low risk level, medium risk level and high risk level. The formula for calculating its overall risk value is: In the formula, For the comprehensive risk value, For time feature values, For location feature values, For intensity characteristic value, For the characteristic value of propagation speed, , , and These are the weight coefficients for the corresponding feature values, and , The constant coefficients, For correction factor, The label of the abnormal charging pile within the target area; Among them, by extracting the time point of occurrence of abnormal charging pile impact events, and combining real-time monitoring data with historical event data, the duration of impact events is predicted by machine learning models, thereby determining time feature values; Geographic coordinates of abnormal charging piles are obtained based on geographic information systems. Combined with urban planning and functional zoning, different levels of weight values ​​are assigned to the locations of abnormal charging piles to determine location characteristic values. The operating parameters of the affected charging piles are collected by sensors, and these operating parameters are compared with those under normal conditions. The magnitude of the change in the operating parameters is calculated, and the average magnitude of the change in the operating parameters is taken as the intensity characteristic value. The propagation speed characteristic value is determined by calculating the rate of change of the area of ​​influence per unit time.

9. The monitoring and early warning management system for vehicle charging piles according to claim 1, characterized in that, Based on the level of risk affected, tiered early warning and management measures will be implemented, and the specific process is as follows: When the risk level is low, the detailed information of the abnormal charging pile is transmitted to the target area management terminal and recorded and archived in the background. When the risk level is medium, an early warning message will be sent to the management terminal of the target area; When the risk level is high, an alarm mechanism will be triggered immediately, and coordinated protection measures will be implemented.

10. A monitoring and early warning management method for electric vehicle charging piles, applied to the monitoring and early warning management system for electric vehicle charging piles as described in claim 1, characterized in that, Includes the following steps: Step 1: Collect the working parameters of charging piles in the target area according to the set collection cycle, generate data frames and upload them to the data analysis terminal. The data analysis terminal performs abnormal data frame filtering processing on the data frames, performs time-series processing on the filtered abnormal data frames to form an abnormal data sequence under the parameter dimension, which is convenient for graphical analysis and thus determines abnormal charging piles. Step 2: When the operation status monitoring module determines that an abnormal charging pile has appeared in the target area, based on the abnormal source of the abnormal charging pile and combined with the electrical topology relationship, a topological model of the abnormal charging pile is constructed, and the trained machine learning model is used to predict the core results of the impact. The core results cover the possibility, time, location, intensity and propagation speed of the impact. Step 3: Based on the core results affected by abnormal charging piles in the target area, determine the risk level of the impact; Step 4: Implement tiered early warning and management measures based on the level of risk affected.

Images