Charging pile abnormal state identification method

Abstract

The invention relates to the field of charging pile monitoring, and discloses a charging pile abnormal state identification method, which is used for solving the problem of false alarm caused by the fact that an abnormal identification device cannot distinguish instantaneous fluctuation and real abnormity caused by sudden incoming call when the charging pile is suddenly powered on after power failure, and comprises the following steps: collecting circuit aging data of the charging pile; analyzing to obtain a circuit aging coefficient, collecting the number of distributed energy sources connected to the charging pile, collecting electric power instantaneous fluctuation data when the charging pile is suddenly powered on after power failure in real time, analyzing to obtain an electric power change coefficient, obtaining power-on duration after the power is powered on in real time, and setting a duration threshold value; a data smoothing index is obtained through comprehensive evaluation, smoothness reduction processing is carried out on data collected by an abnormal monitoring sensor according to the data smoothing index, abnormal state identification is carried out on the data after smoothing processing by using the abnormal monitoring sensor, instantaneous fluctuation and real abnormity caused by sudden incoming call are effectively distinguished, and the false alarm frequency is reduced.

Description

Technical Field

[0001] The present invention relates to the field of charging pile monitoring, and more specifically to a method for identifying abnormal states of charging piles. Background Art

[0002] In the existing charging pile abnormal state identification methods, the voltage, current, power and other operating data of the charging pile are usually collected, and combined with environmental information such as temperature and humidity for analysis, and multiple thresholds are set to monitor abnormal states. These methods mainly rely on fixed thresholds or simple logical judgments to detect abnormal conditions in the operation of charging piles. However, in the process of implementing the technical solution of the invention in the embodiments of the present application, the present application found that the above technology has at least the following technical problems: The existing methods have some shortcomings in practical applications, especially when a call comes in suddenly after the charging pile is powered off, which can easily cause false alarms. The sudden call may cause the data collected by the abnormality recognition device to fluctuate violently, and the existing abnormality recognition methods are difficult to effectively distinguish such instantaneous fluctuations from real abnormalities, resulting in frequent false alarms and affecting the stability and reliability of the system. Summary of the invention

[0003] In order to overcome the above-mentioned defects of the prior art, the present invention provides a method for identifying abnormal state of a charging pile to solve the problems existing in the above-mentioned background technology.

[0004] To achieve the above object, the present invention provides the following technical solutions: The method for identifying abnormal state of a charging pile comprises the following steps: Step 1: Connect the abnormality monitoring sensor to the charging pile; Step 2: Collect charging pile circuit aging data, analyze the circuit aging data to obtain the circuit aging coefficient, the aging data includes resistance data and temperature data; Step 3: Collect the number of distributed energy sources connected to the charging piles; Step 4: Real-time collection of instantaneous power fluctuation data when power is suddenly restored after the charging pile is powered off, and analysis of the instantaneous power fluctuation data to obtain the power variation coefficient; Step 5: Obtain the power-on time after the call in real time and set the time threshold; the method for obtaining the data smoothing index based on the comprehensive evaluation of the circuit aging coefficient, the number of distributed energy resources, the power variation coefficient and the power-on time is as follows: , where DS represents the data smoothing index; T on Indicated as the power-on time, T td It is represented by the duration threshold; CA is represented by the circuit aging coefficient; NUM f It is expressed as the number of distributed energy resources; PC is expressed as the power variation coefficient, , , It is expressed as the weight coefficient of circuit aging coefficient, number of distributed energy resources, and power variation coefficient; Step 6: Reduce the smoothness of the data collected by the abnormal monitoring sensor according to the data smoothing index; Step 7: Use anomaly monitoring sensors to identify abnormal conditions in the smoothed data and perform early warning processing on the abnormal conditions.

[0005] Preferably, the steps of collecting charging pile circuit aging data and analyzing the circuit aging data are as follows: Collecting resistance data in the circuit and evaluating the degree of resistance increase based on the resistance data, the resistance data in the circuit includes line resistance, insulation resistance, and joint and connector resistance; Collect charging pile circuit temperature data in real time, and evaluate the overall temperature change based on the temperature data; The circuit aging coefficient is calculated based on the degree of resistance increase and the degree of overall temperature change.

[0006] Preferably, the step of evaluating the resistance increase degree according to the resistance data is: Collect initial resistance, which includes line initial resistance, insulation initial resistance, and joint and connector initial resistance. The initial resistance value is the value when the circuit is used for the first time. Collect real-time resistance, including real-time line resistance, real-time insulation resistance, and real-time joint and connector resistance; The resistance increase degree is calculated based on the initial resistance and the real-time resistance.

[0007] Preferably, the step of evaluating the overall temperature change degree according to the temperature data is: Divide the entire charging pile circuit into n parts, record them as subcircuits, and set temperature sensors at key positions in each subcircuit; The temperature data of each sub-circuit is collected in real time by a temperature sensor, and the degree of temperature rise of each sub-circuit is calculated; The temperature rise degree of each sub-circuit is clustered, and the overall temperature change degree is calculated based on the clustering processing result.

[0008] Preferably, the step of clustering the temperature rise degree of each sub-circuit is: Step 2.1: Take the temperature rise degree as a clustering feature, take the temperature rise degree of all sub-circuits as a data set, and each temperature rise degree in the data set as a data point; Step 2.2: Use the GAP statistic method to determine the optimal number of clusters K for the data set; Step 2.3: Randomly select K data points in the data set as initial cluster centers. For each data point, calculate its Euclidean distance to each initial cluster center, traverse the K initial cluster centers, and assign it to the cluster corresponding to the nearest initial cluster center; Step 2.4: After traversing all data points, the initial clusters are obtained. For each initial cluster, the mean of the data points in it is calculated to obtain a new cluster center. Step 2.5: Repeat steps 2.3 and 2.4 until the cluster center no longer changes, and obtain the final cluster and the final cluster center.

[0009] Preferably, the step of calculating the overall temperature change degree according to the clustering processing result is: The weight of each final cluster is calculated by calculating the ratio of the number of data points in each final cluster to the total number of data points in the data set; The weight of each final cluster is weighted and summed with the final cluster center to get the overall temperature change degree, which is calculated as follows: , where TC represents the overall temperature change, Q j Represented as the weight of the jth final cluster, It is represented as the jth final cluster center, and K is the optimal number of clusters of the data set.

[0010] Preferably, the step of analyzing the instantaneous power fluctuation data is: After the charging pile receives a call, starting from time t0, the time series data of the voltage, current and power of the charging pile circuit are collected, the sampling time interval is set, and the voltage, current and power at each sampling point are recorded; At each time series point, calculate the instantaneous rate of change of voltage, the instantaneous rate of change of current and the instantaneous rate of change of power; The method of calculating the instantaneous rate of change of voltage, current and power by calculus is as follows: , where PC is the power variation coefficient, T is the length of the time interval, that is, from t0 to t, Expressed as voltage recovery speed, Expressed as the degree of current fluctuation, Expressed as the rate of change of power.

[0011] Preferably, the step of reducing the smoothness of the data collected by the abnormality monitoring sensor according to the data smoothness index is: Set the smoothed data at the initial moment, and use the smoothed data at the initial moment as the data point initially collected by the abnormality monitoring sensor; Normalize the data smoothing index; The method for dynamically smoothing the data collected by the abnormal monitoring sensor according to the normalized data smoothing index is as follows: ,in It is represented as the raw data collected by the anomaly monitoring sensor at the current time t, Represented as smoothed data at the previous time point, It represents the smoothed data result at the current time point, and DS represents the data smoothing index after normalization.

[0012] Technical effects and advantages of the present invention: Collect charging pile circuit aging data, analyze and obtain the circuit aging coefficient, collect the number of distributed energy sources connected to the charging pile, collect real-time power instantaneous fluctuation data when a call suddenly comes in after the charging pile is powered off, analyze and obtain the power variation coefficient, obtain the power-on time after the call in real time, and set the time threshold; comprehensively evaluate to obtain the data smoothing index, reduce the smoothness of the data collected by the abnormal monitoring sensor according to the data smoothing index, use the abnormal monitoring sensor to identify the abnormal state of the smoothed data, effectively distinguish the instantaneous fluctuation caused by the sudden call from the real abnormality, and reduce the number of false alarms. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 It is the overall flow chart of the present invention. DETAILED DESCRIPTION

[0014] The technical solution of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the present invention. In addition, the forms of the various structures recorded in the following embodiments are merely illustrative, and the abnormal state identification method of the charging pile involved in the present invention is not limited to the various structures recorded in the following embodiments. All other implementations obtained by ordinary technicians in this field without making creative work belong to the scope of protection of the present invention.

[0015] The present invention provides a method for identifying abnormal state of a charging pile, comprising the following steps: Step 1: Connect the abnormality monitoring sensor to the charging pile, and the abnormality monitoring sensor is used to monitor the abnormality of the charging pile; Step 2: Collect charging pile circuit aging data, analyze the circuit aging data to obtain the circuit aging coefficient, the aging data includes resistance data and temperature data; In this embodiment, it should be specifically explained that the steps of collecting charging pile circuit aging data and analyzing the circuit aging data are as follows: Collecting resistance data in the circuit and evaluating the degree of resistance increase based on the resistance data, the resistance data in the circuit includes line resistance, insulation resistance, and joint and connector resistance; Collect charging pile circuit temperature data in real time, and evaluate the overall temperature change based on the temperature data; The method for obtaining the circuit aging coefficient based on the degree of resistance increase and the degree of overall temperature change is: , where CA represents the circuit aging coefficient, IR represents the degree of resistance increase, and TC represents the degree of overall temperature change.

[0016] In this embodiment, it should be specifically explained that the steps of evaluating the resistance increase degree according to the resistance data are as follows: Collect initial resistance, which includes line initial resistance, insulation initial resistance, and joint and connector initial resistance. The initial resistance value is the value when the circuit is used for the first time. Collect real-time resistance, including real-time line resistance, real-time insulation resistance, and real-time joint and connector resistance; The method for obtaining the degree of resistance increase based on the initial resistance and the real-time resistance is: ,in It is expressed as the degree of resistance increase, Expressed as real-time line resistance, Expressed as the initial resistance of the line, Expressed as real-time insulation resistance, Expressed as the initial insulation resistance, Expressed as real-time joint and connector resistance, Expressed as initial resistance of joints and connectors.

[0017] In this embodiment, it should be specifically explained that the steps of collecting the charging pile circuit temperature data in real time and evaluating the overall temperature change degree according to the temperature data are as follows: Divide the entire charging pile circuit into n parts, record them as subcircuits, and set temperature sensors at key positions in each subcircuit; Key locations can be the connectors or terminals of the circuit; temperature sensors can be selected from thermistors, thermocouples, and resistance temperature detectors, which can monitor the circuit temperature in real time and reflect the temperature rise of the circuit. Thermistors are highly sensitive and suitable for accurately monitoring local temperature changes; thermocouples are suitable for wide temperature range applications and have strong adaptability; resistance temperature detectors have high accuracy and stability. By installing these temperature sensors at key locations, the temperature changes of each section of the line can be captured in a timely manner, providing reliable data support for line aging analysis.

[0018] The temperature data of each sub-circuit is collected in real time by a temperature sensor, and the method for obtaining the temperature rise degree of each sub-circuit according to the temperature data of each sub-circuit is as follows: , where tr represents the degree of temperature rise, t1 represents the real-time temperature of the sub-circuit, and t2 represents the temperature data at the initial stage of the sub-circuit; The temperature rise degree of each sub-circuit is clustered, and the overall temperature change degree is calculated based on the clustering processing result.

[0019] Clustering can divide circuit segments with different temperature rise patterns into several groups, thereby identifying areas with different degrees of temperature rise. Areas with obvious temperature rise often indicate that the circuit is abnormal or aged. Clustering can help accurately locate these areas, making it easier to discover potential problems in advance and arrange local maintenance. After clustering, subcircuits with different temperature rise characteristics are distinguished, and the overall temperature change can be more accurately calculated through the temperature changes of different clustering categories. This method avoids the simple processing of direct average temperature, improves the accuracy of aging assessment, and can more comprehensively reflect the actual aging of the circuit.

[0020] In this embodiment, it should be specifically explained that the steps of clustering the temperature rise degree of each sub-circuit are: Step 2.1: Take the temperature rise degree as a clustering feature, take the temperature rise degree of all sub-circuits as a data set, and each temperature rise degree in the data set as a data point; Step 2.2: Use the GAP statistic method to determine the optimal number of clusters K for the data set; The GAP statistic is a method for determining the optimal number of clusters. Its principle is to compare the deviation between the cluster distribution of actual data and the cluster distribution of random data. Specifically, the GAP statistic measures the clustering effect by calculating the difference in cluster dispersion between actual data and uniformly distributed random data at different cluster numbers. The larger the GAP value, the more significant the structure of the actual data clustering, indicating that the current number of clusters can better describe the data distribution. The optimal number of clusters usually corresponds to the number of clusters where the GAP statistic reaches the maximum value; Step 2.3: Randomly select K data points in the data set as the initial cluster centers. For each data point, calculate the Euclidean distance from each initial cluster center. The calculation formula is: ,in Represented as data points To the kth cluster center The Euclidean distance of Represented as the value of the ith data point, It is represented as the value of the kth initial cluster center. For each data point, the K initial cluster centers are traversed and assigned to the cluster corresponding to the nearest initial cluster center. Euclidean distance is the straight-line distance between two points, and the calculation method is based on the difference in the coordinates of the two points in each dimension. In K-means clustering, Euclidean distance is used to measure the distance between each data point and each cluster center to determine which cluster center the data point is closer to, thereby guiding the clustering process.

[0021] Step 2.4: After traversing all data points, the initial clusters are obtained. For each initial cluster, the mean of the data points in it is calculated to obtain a new cluster center. Step 2.5: Repeat steps 2.3 and 2.4 until the cluster center no longer changes, and obtain the final cluster and the final cluster center.

[0022] In this embodiment, it should be specifically explained that the steps of calculating the overall temperature change degree according to the clustering processing result are: The weight of each final cluster is calculated by calculating the ratio of the number of data points in each final cluster to the total number of data points in the data set; The weight of each final cluster is weighted and summed with the final cluster center to get the overall temperature change degree, which is calculated as follows: , where TC represents the overall temperature change, Q j Represented as the weight of the jth final cluster, It is represented as the jth final cluster center, and K is the optimal number of clusters of the data set.

[0023] Step 3: Collect the number of distributed energy sources connected to the charging piles; The main purpose of connecting distributed energy to charging piles is to improve the reliability and stability of power supply, reduce dependence on traditional power grids, and make full use of clean energy. Distributed energy (such as solar energy, wind energy, energy storage systems, etc.) is deployed locally near charging piles to reduce transmission losses and improve energy efficiency. Especially during peak power consumption or when the power grid is unstable, distributed energy can provide additional power support for charging piles to ensure continuous power supply to charging piles. At the same time, this approach can promote the application of renewable energy, reduce carbon emissions, and achieve a greener charging infrastructure.

[0024] Step 4: Real-time collection of instantaneous power fluctuation data when power is suddenly restored after the charging pile is powered off, and analysis of the instantaneous power fluctuation data to obtain the power variation coefficient; In this embodiment, it should be specifically explained that the steps for analyzing the instantaneous power fluctuation data are as follows: After the charging pile receives a call, starting from time t0, the time series data of the voltage, current and power of the charging pile circuit are collected, the sampling time interval is set, and the voltage, current and power at that moment are recorded at each sampling point. It should be noted that the sampling time interval is manually set and can be 0.1s or 0.5s; At each time series point, calculate the instantaneous rate of change of voltage , the instantaneous rate of change of current And the instantaneous rate of change of power ,in , as well as Respectively represent the time t h The voltage, current and power values ​​under , as well as Respectively represent the time t h+1 The voltage, current and power values ​​under It is expressed as the sampling time interval; The method of calculating the instantaneous rate of change of voltage, current and power by calculus is as follows: , where PC is the power variation coefficient, T is the length of the time interval, that is, from t0 to t, Expressed as voltage recovery speed, Expressed as the degree of current fluctuation, Expressed as the rate of change of power.

[0025] The calculus method can calculate the instantaneous rate of change of voltage, current and power through differentiation. This instantaneous rate of change can more accurately reflect the specific time and intensity of the fluctuation, thereby identifying the impact of instantaneous fluctuations; through integration, the fluctuation data can be accumulated over a period of time to calculate the average change of voltage, current and power fluctuations. This cumulative calculation can capture the fluctuation trend of the power system over the entire time interval, thereby more comprehensively evaluating the overall stability of the system; compared with the fluctuation evaluation of a single instantaneous value, the cumulative effect of integration is more stable, can avoid the misleading caused by instantaneous spikes, and is suitable for evaluating aging and performance changes over a longer period of time.

[0026] Step 5: Obtain the power-on time after the call in real time and set the time threshold. As the power-on time increases, the system gradually becomes stable and the data fluctuation will gradually decrease. The time threshold represents the time required for stabilization after power-on under normal conditions. The method for obtaining the data smoothing index based on the comprehensive evaluation of the circuit aging coefficient, the number of distributed energy resources, the power variation coefficient and the power-on time is as follows: , where DS represents the data smoothing index; T on Indicated as the power-on time, T tdExpressed as a duration threshold, in the initial period after the charging pile is powered on again, the current and voltage usually experience large instantaneous fluctuations, and a higher smoothing index is required to filter out these fluctuations to avoid false alarms. However, as the power-on time increases, the system gradually returns to a stable state, and the power fluctuations gradually decrease, so the demand for data smoothing also decreases accordingly. An exponential decay function is used to ensure that the smoothing index decreases with the increase of power-on time. When the power-on time is short, the smoothing index is large, indicating that a higher smoothing is required; when the power-on time gradually approaches the threshold, the exponential term tends to zero, indicating that the system no longer requires smoothing; CA is expressed as the circuit aging coefficient. The more serious the line aging, the slower the circuit response speed may be, and the greater the instantaneous fluctuations and noise may be. Therefore, a higher smoothing index is required to filter out instantaneous fluctuations, thereby avoiding false alarms caused by line aging and ensuring that the charging pile system still maintains stable monitoring performance under aging line conditions; NUM f It is expressed as the number of distributed energy sources. When the number of distributed energy sources supplying power to charging piles increases, the combination of different energy sources can better balance power fluctuations, improve the stability and continuity of the overall power supply, and thus reduce instantaneous fluctuations such as voltage and current. Since the increase in the number of distributed energy sources can reduce the volatility and instability of the system, the data smoothing index can also be reduced, that is, the required degree of smoothing is reduced. This design can make the system more sensitive when the power supply is more stable, so as to more accurately identify actual anomalies without excessively filtering out normal data changes; PC is expressed as the power variation coefficient. The power variation coefficient reflects the degree of fluctuation of parameters such as voltage, current, and power. When the power variation coefficient is large, it means that the system has strong fluctuations and high instability after the instantaneous call, so a higher smoothing index is required to filter these instantaneous changes to avoid false alarms. When the power variation coefficient increases, the data smoothing index also increases accordingly, so that the system can perform stronger smoothing processing on the data under severe fluctuations to ensure that the monitoring data is more stable and reliable; when the power variation coefficient decreases, the fluctuation decreases, the smoothing demand decreases, and the system can detect actual anomalies more sensitively to ensure the accuracy and real-time nature of monitoring. , , It is expressed as the weight coefficient of circuit aging coefficient, distributed energy quantity and power variation coefficient, and , the specific value is set by professionals according to the actual situation, for example, , , It can be 0.2, 0.4, 0.4. It should be explained that the circuit aging coefficient, the number of distributed energy sources, the power variation coefficient and the power-on time are all de-unitized data.

[0027] Step 6: Reduce the smoothness of the data collected by the abnormal monitoring sensor according to the data smoothing index.

[0028] In this embodiment, it should be specifically explained that the steps of reducing the smoothness of the data collected by the abnormality monitoring sensor according to the data smoothness index are as follows: The smoothed data S0 at the initial moment is set as the data point X0 initially collected by the abnormal monitoring sensor, and S0 is used as the reference point for subsequent smoothing calculations; Normalize the data smoothness index and convert it to between 0 and 1 to control the dynamic change of smoothness; The method for dynamically smoothing the data collected by the abnormal monitoring sensor according to the normalized data smoothing index is as follows: ,in It is represented as the raw data collected by the anomaly monitoring sensor at the current time t, Represented as smoothed data at the previous time point, It represents the smoothed data result at the current time point, and DS represents the data smoothing index after normalization.

[0029] Using the data smoothing index to control smoothness can dynamically adjust the smoothing intensity according to parameters such as circuit status, fluctuations, and power-on time. This allows the system to enhance smoothness and suppress noise and instantaneous fluctuations when fluctuations are large; and gradually reduce smoothness when the system tends to be stable, so as to timely reflect real data changes; this adaptive smoothing method ensures that the system can self-adjust with changes in the power environment, reducing the need for manual adjustments; dynamic adjustment of smoothness can effectively filter out instantaneous spikes or high-frequency noise at the beginning of the call, reducing the false alarm rate. For example, in the case of violent fluctuations at the beginning of power-on, a higher smoothness can avoid false alarms caused by short-term voltage or current fluctuations; reducing smoothing can improve the accuracy of monitoring data, especially when the system is stable, it can more accurately capture real abnormal changes and improve the validity of data.

[0030] Step 7: Use anomaly monitoring sensors to identify abnormal conditions in the smoothed data, and perform early warning processing on the abnormal conditions to remind relevant staff to repair the charging pile in a timely manner.

[0031] Abnormal state identification and real-time warning can detect potential faults before they worsen, reminding staff to intervene in time. This early detection and warning mechanism can effectively prevent fault expansion, reduce the risk of sudden fault downtime, ensure the continuous operation of charging piles, and improve user experience; the warning mechanism can provide accurate abnormal information, so that staff can obtain clues to the problem at the first time, thereby accurately locating the fault and shortening the troubleshooting time. This targeted maintenance reduces the additional costs caused by frequent inspections or emergency maintenance, while improving maintenance efficiency.

[0032] Timely detection of problems such as line aging and overheating prevents these small problems from accumulating into major failures. Timely maintenance helps extend the service life of charging pile equipment and improve the long-term stability of the system; the automated abnormal monitoring and early warning system reduces the need for manual real-time monitoring. Staff do not need to monitor each charging pile at all times, the system will automatically notify abnormal situations, thereby improving management efficiency and allowing human resources to be more reasonably allocated.

[0033] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

[0034] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for identifying abnormal state of a charging pile, characterized in that: The following steps are involved: Step 1: Connect the abnormality monitoring sensor to the charging pile; Step 2: Collect charging pile circuit aging data, analyze the circuit aging data to obtain the circuit aging coefficient, the aging data includes resistance data and temperature data; Step 3: Collect the number of distributed energy sources connected to the charging piles; Step 4: Real-time collection of instantaneous power fluctuation data when power is suddenly restored after the charging pile is powered off, and analysis of the instantaneous power fluctuation data to obtain the power variation coefficient; Step 5: Obtain the power-on time after the call in real time and set the time threshold; the method for obtaining the data smoothing index based on the comprehensive evaluation of the circuit aging coefficient, the number of distributed energy resources, the power variation coefficient and the power-on time is as follows: , where DS represents the data smoothing index; T on Indicated as the power-on time, T td It is represented by the duration threshold; CA is represented by the circuit aging coefficient; NUM f It is expressed as the number of distributed energy resources; PC is expressed as the power variation coefficient, , , It is expressed as the weight coefficient of circuit aging coefficient, number of distributed energy resources, and power variation coefficient; Step 6: Reduce the smoothness of the data collected by the abnormal monitoring sensor according to the data smoothing index; Step 7: Use anomaly monitoring sensors to identify abnormal conditions in the smoothed data and perform early warning processing on the abnormal conditions.

2. The method for identifying abnormal state of a charging pile according to claim 1, characterized in that: The steps of collecting charging pile circuit aging data and analyzing the circuit aging data are as follows: Collecting resistance data in the circuit and evaluating the degree of resistance increase based on the resistance data, the resistance data in the circuit includes line resistance, insulation resistance, and joint and connector resistance; Collect charging pile circuit temperature data in real time, and evaluate the overall temperature change based on the temperature data; The circuit aging coefficient is calculated based on the degree of resistance increase and the degree of overall temperature change.

3. The method for identifying abnormal state of a charging pile according to claim 2, characterized in that: The step of evaluating the resistance increase degree according to the resistance data is as follows: Collect initial resistance, which includes line initial resistance, insulation initial resistance, and joint and connector initial resistance. The initial resistance value is the value when the circuit is used for the first time. Collect real-time resistance, including real-time line resistance, real-time insulation resistance, and real-time joint and connector resistance; The resistance increase degree is calculated based on the initial resistance and the real-time resistance.

4. The method for identifying abnormal state of a charging pile according to claim 2, characterized in that: The step of evaluating the overall temperature change degree according to the temperature data is as follows: Divide the entire charging pile circuit into n parts, record them as subcircuits, and set temperature sensors at key positions in each subcircuit; The temperature data of each sub-circuit is collected in real time by a temperature sensor, and the degree of temperature rise of each sub-circuit is calculated; The temperature rise degree of each sub-circuit is clustered, and the overall temperature change degree is calculated based on the clustering processing result.

5. The method for identifying abnormal state of a charging pile according to claim 4, characterized in that: The steps of clustering the temperature rise degree of each sub-circuit are as follows: Step 2.1: Take the temperature rise degree as a clustering feature, take the temperature rise degree of all sub-circuits as a data set, and each temperature rise degree in the data set as a data point; Step 2.2: Use the GAP statistic method to determine the optimal number of clusters K for the data set; Step 2.3: Randomly select K data points in the data set as initial cluster centers. For each data point, calculate its Euclidean distance to each initial cluster center, traverse the K initial cluster centers, and assign it to the cluster corresponding to the nearest initial cluster center; Step 2.4: After traversing all data points, the initial clusters are obtained. For each initial cluster, the mean of the data points in it is calculated to obtain a new cluster center. Step 2.5: Repeat steps 2.3 and 2.4 until the cluster center no longer changes, and obtain the final cluster and the final cluster center.

6. The method for identifying abnormal state of a charging pile according to claim 4, characterized in that: The step of calculating the overall temperature change degree according to the clustering processing result is: The weight of each final cluster is calculated by calculating the ratio of the number of data points in each final cluster to the total number of data points in the data set; The weight of each final cluster is weighted and summed with the final cluster center to get the overall temperature change degree, which is calculated as follows: , where TC represents the overall temperature change, Q j Represented as the weight of the jth final cluster, It is represented as the jth final cluster center, and K is the optimal number of clusters of the data set.

7. The method for identifying abnormal state of a charging pile according to claim 1, characterized in that: The steps of analyzing the instantaneous power fluctuation data are as follows: After the charging pile receives a call, starting from time t0, the time series data of the voltage, current and power of the charging pile circuit are collected, the sampling time interval is set, and the voltage, current and power at each sampling point are recorded; At each time series point, calculate the instantaneous rate of change of voltage, the instantaneous rate of change of current and the instantaneous rate of change of power; The method of calculating the instantaneous rate of change of voltage, current and power by calculus is as follows: , where PC is the power variation coefficient, T is the length of the time interval, that is, from t0 to t, Expressed as voltage recovery speed, Expressed as the degree of current fluctuation, Expressed as the rate of change of power.

8. The method for identifying abnormal state of a charging pile according to claim 1, characterized in that: The step of reducing the smoothness of the data collected by the abnormal monitoring sensor according to the data smoothness index is as follows: Set the smoothed data at the initial moment, and use the smoothed data at the initial moment as the data point initially collected by the abnormality monitoring sensor; Normalize the data smoothing index; The method for dynamically smoothing the data collected by the abnormal monitoring sensor according to the normalized data smoothing index is as follows: ,in It is represented as the raw data collected by the anomaly monitoring sensor at the current time t, Represented as smoothed data at the previous time point, It represents the smoothed data result at the current time point, and DS represents the data smoothing index after normalization.

Images