Charging pile danger prediction method and system based on historical charging data analysis

Abstract

The invention discloses a charging pile danger prediction method and system based on historical charging data analysis. The method comprises the following steps: acquiring charging working parameters of a charging pile at a plurality of historical time points; screening out a plurality of dangerous time points from all the historical time points according to the charging working parameters; determining time sequence working data according to the charging working parameters at the dangerous time point; and predicting future fault information of the charging pile based on a trained time sequence prediction model according to the time sequence working data. Therefore, the method can achieve the precise fault prediction and early warning of the charging pile based on the historical abnormal time sequence, improves the operation reliability and maintenance efficiency of the charging pile, and reduces the risk of equipment damage or safety accidents caused by the unpredicted fault.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a method and system for predicting the dangers of charging piles based on historical charging data analysis. Background Technology

[0002] With the rapid expansion of electric vehicle charging infrastructure and the increasing demand for charging, users and testing institutions are paying more and more attention to improving the operational reliability and maintenance efficiency of charging piles through accurate fault prediction. Among these challenges, how to achieve early warning of future faults to reduce the risk of equipment damage or safety accidents has become a key technical issue. Existing technologies typically collect real-time or periodic operating parameters of charging piles, use fixed threshold analysis or simple statistical models to assess equipment status, and issue maintenance alerts based on standard rules to support charging pile operation and maintenance. However, existing solutions lack the ability to accurately predict potential fault development trends due to the lack of time-series operational data generated by filtering multiple historical critical time points and the dynamic prediction capabilities of LSTM neural networks. Commonly used static or threshold-driven monitoring strategies cannot capture abnormal time-series characteristics, resulting in insufficient accuracy in fault prediction. This can easily lead to equipment damage or safety accidents due to unpredicted faults, limiting the operational reliability and maintenance efficiency of charging piles. Therefore, existing technologies have shortcomings that urgently need to be addressed. Summary of the Invention

[0003] The technical problem to be solved by the present invention is to provide a charging pile hazard prediction method and system based on historical charging data analysis, which can realize accurate charging pile fault prediction and early warning based on historical abnormal time sequence, improve the operational reliability and maintenance efficiency of charging piles, and reduce the risk of equipment damage or safety accidents caused by unpredicted faults.

[0004] To address the aforementioned technical problems, the first aspect of this invention discloses a method for predicting the hazards of charging piles based on historical charging data analysis, the method comprising: Obtain charging operation parameters of charging piles at multiple historical time points; Based on the charging operating parameters, multiple dangerous time points are selected from all the historical time points; Based on the charging operation parameters at the critical time points, the timing operation data is determined; Based on the time-series working data and a trained time-series prediction model, the future fault information of the charging pile is predicted.

[0005] As an optional implementation, in the first aspect of the present invention, the charging operating parameters include at least one of charging current, charging voltage, charging temperature, charging power, charging duration, and environmental sensing data; the environmental sensing data includes at least one of ambient temperature, ambient humidity, ambient image, and ambient sound.

[0006] As an optional implementation, in the first aspect of the invention, the step of filtering out multiple dangerous time points from all the historical time points based on the charging operating parameters includes: For each historical point in time, multiple historical work data similar to that historical point in time are matched from the historical database; Based on the multiple historical work data and the charging work parameters corresponding to the historical time point, calculate the danger level parameter corresponding to the historical time point; The historical time points where the danger level parameter is greater than a preset parameter threshold are selected to obtain multiple dangerous time points.

[0007] As an optional implementation, in the first aspect of the present invention, the step of matching multiple historical work data similar to the historical time point from the historical database includes: For each historical charging data in the historical database, determine the charging time parameters and charging equipment parameters corresponding to that historical charging data; Calculate the similarity of characteristics between the charging time parameter and the time characteristics at that historical time point; Calculate the parameter similarity between the parameters of the charging equipment and the parameters of the charging pile; Calculate the product of the characteristic similarity and the parameter similarity to obtain the similarity corresponding to the historical charging data; All historical charging data with a similarity greater than a preset similarity threshold are filtered out to obtain multiple historical working data.

[0008] As an optional implementation, in the first aspect of the present invention, the step of calculating the danger level parameter corresponding to the historical time point based on the plurality of historical working data and the charging working parameters corresponding to the historical time point includes: Calculate the charging operation parameters corresponding to the historical time point and the data similarity of each of the historical operation data; Calculate the reciprocal of the weighted summation average of all the data similarities to obtain the risk level parameter corresponding to the historical time point; wherein, the calculation weight corresponding to each of the data similarities is proportional to the characteristic similarity and / or parameter similarity corresponding to the corresponding historical working data.

[0009] As an optional implementation, in the first aspect of the invention, determining the timing data based on the charging operating parameters at the critical time point includes: For each of the aforementioned critical time points, calculate the time difference between that critical time point and the reference time point; Calculate the time weight that is inversely proportional to the time difference; The time weight is labeled as an additional feature of the dangerous time point; Based on the dangerous time points from morning to night, the charging operation parameters and additional features of all the dangerous time points are sorted to obtain time-series operation data.

[0010] As an optional implementation, in the first aspect of the invention, the reference time point is calculated through the following steps: Cluster all the aforementioned dangerous time points to obtain a set of time points; the time difference between any two of the dangerous time points in the set of time points is less than any external time difference; the external time difference is the time difference between any one of the dangerous time points and any dangerous time point not in the set of time points. The average value of all the dangerous time points in the set of time points is calculated to obtain the reference time point.

[0011] As an optional implementation, in the first aspect of the present invention, predicting future fault information of the charging pile based on the time-series working data and a trained time-series prediction model includes: The time-series working data is input into a trained LSTM neural network to obtain the future fault time point and fault parameters of the charging pile; the LSTM neural network is trained using a training dataset that includes multiple working parameter time series and corresponding fault point annotations. Determine the control server corresponding to the charging pile within the working time period to which the future failure time point belongs; the control server is an edge server or a cloud server; The future fault time point and fault parameters are sent to the control server to provide early warning.

[0012] A second aspect of this invention discloses a charging pile hazard prediction system based on historical charging data analysis, the system comprising: The acquisition module is used to acquire the charging operation parameters of the charging pile at multiple historical time points; The filtering module is used to filter out multiple dangerous time points from all the historical time points based on the charging operating parameters. The determination module is used to determine the timing operation data based on the charging operation parameters at the critical time points; The prediction module is used to predict future fault information of the charging pile based on the time-series working data and a trained time-series prediction model.

[0013] As an optional implementation, in a second aspect of the present invention, the charging operating parameters include at least one of charging current, charging voltage, charging temperature, charging power, charging duration, and environmental sensing data; the environmental sensing data includes at least one of ambient temperature, ambient humidity, ambient image, and ambient sound.

[0014] As an optional implementation, in a second aspect of the invention, the specific method by which the screening module filters out multiple dangerous time points from all the historical time points based on the charging operating parameters includes: For each historical point in time, multiple historical work data similar to that historical point in time are matched from the historical database; Based on the multiple historical work data and the charging work parameters corresponding to the historical time point, calculate the danger level parameter corresponding to the historical time point; The historical time points where the danger level parameter is greater than a preset parameter threshold are selected to obtain multiple dangerous time points.

[0015] As an optional implementation, in the second aspect of the present invention, the specific method by which the filtering module matches multiple historical work data similar to the historical time point from the historical database includes: For each historical charging data in the historical database, determine the charging time parameters and charging equipment parameters corresponding to that historical charging data; Calculate the similarity of characteristics between the charging time parameter and the time characteristics at that historical time point; Calculate the parameter similarity between the parameters of the charging equipment and the parameters of the charging pile; Calculate the product of the characteristic similarity and the parameter similarity to obtain the similarity corresponding to the historical charging data; All historical charging data with a similarity greater than a preset similarity threshold are filtered out to obtain multiple historical working data.

[0016] As an optional implementation, in the second aspect of the present invention, the specific method by which the screening module calculates the danger level parameter corresponding to the historical time point based on the plurality of historical working data and the charging working parameters corresponding to the historical time point includes: Calculate the charging operation parameters corresponding to the historical time point and the data similarity of each of the historical operation data; Calculate the reciprocal of the weighted summation average of all the data similarities to obtain the risk level parameter corresponding to the historical time point; wherein, the calculation weight corresponding to each of the data similarities is proportional to the characteristic similarity and / or parameter similarity corresponding to the corresponding historical working data.

[0017] As an optional implementation, in a second aspect of the invention, the determining module determines the specific method for determining the timing data based on the charging operating parameters at the critical time point, including: For each of the aforementioned critical time points, calculate the time difference between that critical time point and the reference time point; Calculate the time weight that is inversely proportional to the time difference; The time weight is labeled as an additional feature of the dangerous time point; Based on the dangerous time points from morning to night, the charging operation parameters and additional features of all the dangerous time points are sorted to obtain time-series operation data.

[0018] As an optional implementation, in a second aspect of the invention, the reference time point is calculated through the following steps: Cluster all the aforementioned dangerous time points to obtain a set of time points; the time difference between any two of the dangerous time points in the set of time points is less than any external time difference; the external time difference is the time difference between any one of the dangerous time points and any dangerous time point not in the set of time points. The average value of all the dangerous time points in the set of time points is calculated to obtain the reference time point.

[0019] As an optional implementation, in a second aspect of the invention, the specific method by which the prediction module predicts future fault information of the charging pile based on the time-series working data and a trained time-series prediction model includes: The time-series working data is input into a trained LSTM neural network to obtain the future fault time point and fault parameters of the charging pile; the LSTM neural network is trained using a training dataset that includes multiple working parameter time series and corresponding fault point annotations. Determine the control server corresponding to the charging pile within the working time period to which the future failure time point belongs; the control server is an edge server or a cloud server; The future fault time point and fault parameters are sent to the control server to provide early warning.

[0020] A third aspect of this invention discloses another charging pile hazard prediction system based on historical charging data analysis, the system comprising: Memory containing executable program code; A processor coupled to the memory; The processor calls the executable program code stored in the memory to execute some or all of the steps in the charging pile hazard prediction method based on historical charging data analysis disclosed in the first aspect of the present invention.

[0021] The fourth aspect of the present invention discloses a computer storage medium storing computer instructions, which, when invoked, are used to execute some or all of the steps in the charging pile hazard prediction method based on historical charging data analysis disclosed in the first aspect of the present invention.

[0022] Compared with the prior art, the embodiments of the present invention have the following beneficial effects: This invention acquires charging operation parameters from multiple historical time points of charging piles and filters dangerous time points to form time-series operation data. This data is then input into a trained LSTM neural network to predict future fault information. This enables accurate charging pile fault prediction and early warning based on historical abnormal time sequences, improving the operational reliability and maintenance efficiency of charging piles and reducing the risk of equipment damage or safety accidents caused by unpredicted faults. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This is a flowchart illustrating a method for predicting the dangers of charging piles based on historical charging data analysis, as disclosed in an embodiment of the present invention.

[0025] Figure 2 This is a schematic diagram of a charging pile hazard prediction system based on historical charging data analysis disclosed in an embodiment of the present invention.

[0026] Figure 3 This is a schematic diagram of another charging pile hazard prediction system based on historical charging data analysis disclosed in an embodiment of the present invention. Detailed Implementation

[0027] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0028] The terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this invention are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, apparatus, product, or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or devices.

[0029] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0030] This invention discloses a method and system for predicting charging pile hazards based on historical charging data analysis. By acquiring charging operation parameters from multiple historical time points of the charging pile and filtering for hazardous time points to form time-series operation data, this data is input into a trained LSTM neural network to predict future fault information. This enables accurate charging pile fault prediction and early warning based on historical abnormal time sequences, improving the operational reliability and maintenance efficiency of charging piles, and reducing the risk of equipment damage or safety accidents caused by unpredicted faults. Detailed explanations follow.

[0031] Example 1 Please see Figure 1 , Figure 1 This is a flowchart illustrating a charging pile hazard prediction method based on historical charging data analysis, as disclosed in an embodiment of the present invention. Figure 1 The described method for predicting charging pile hazards based on historical charging data analysis can be applied to data processing systems / data processing equipment / data processing servers (wherein, the server includes local processing servers or cloud processing servers). Figure 1 As shown, the charging pile hazard prediction method based on historical charging data analysis may include the following operations: 101. Obtain the charging operation parameters of the charging pile at multiple historical time points.

[0032] Optionally, the charging operating parameters include at least one of the following: charging current, charging voltage, charging temperature, charging power, charging duration, and environmental sensor data.

[0033] Optionally, the environmental sensing data may include at least one of ambient temperature, ambient humidity, ambient image, and ambient sound.

[0034] 102. Based on the charging operation parameters, select multiple dangerous time points from all historical time points. Optionally, the critical time point can be the time point when overcurrent, overtemperature, low insulation, or abnormal fluctuations occur; this invention does not limit this.

[0035] 103. Determine the timing data based on the charging parameters at critical time points. 104. Based on the time-series working data and the trained time-series prediction model, predict the future fault information of the charging pile.

[0036] Optionally, the future fault information may include the expected fault time, fault type, or fault severity, which is not limited in this invention.

[0037] As can be seen, the above-mentioned embodiments of the invention obtain charging operation parameters of charging piles at multiple historical time points and filter dangerous time points to form time-series operation data. The data is then input into a trained LSTM neural network to predict future fault information. This enables accurate charging pile fault prediction and early warning based on historical abnormal time sequences, improving the operational reliability and maintenance efficiency of charging piles, and reducing the risk of equipment damage or safety accidents caused by unpredicted faults.

[0038] As an optional embodiment, in the above steps, multiple dangerous time points are selected from all historical time points based on charging operating parameters, including: For each historical point in time, multiple historical work data similar to that historical point in time are matched from the historical database; Based on multiple historical work data and the charging work parameters corresponding to that historical time point, calculate the danger level parameter corresponding to that historical time point; Historical time points where the risk level parameter is greater than the preset parameter threshold are selected to obtain multiple risk time points.

[0039] Optionally, the historical working data can be charging records under similar working conditions; this invention does not impose any limitations.

[0040] Optionally, the risk level parameter can be a quantitative risk value of 0-1, which is not limited in this invention.

[0041] As can be seen, through the above optional embodiments, by matching similar historical work data at historical time points in the historical database and calculating the degree of danger parameters to filter dangerous time points, accurate identification of dangerous moments based on similar historical comparisons can be achieved, improving the accuracy and pertinence of dangerous time point screening, providing high-quality abnormal samples for time series data construction, and reducing the risk of fault prediction deviation caused by abnormal omissions.

[0042] As an optional embodiment, the step described above, matching multiple historical work data points similar to the historical time point from the historical database, includes: For each historical charging data in the historical database, determine the charging time parameters and charging equipment parameters corresponding to that historical charging data; Calculate the similarity of characteristics between the charging time parameter and the time characteristics at that historical time point; Calculate the parameter similarity between the parameters of the charging equipment and the parameters of the charging pile; The similarity of the historical charging data is obtained by multiplying the feature similarity and parameter similarity. All historical charging data with a similarity greater than a preset similarity threshold are filtered out to obtain multiple historical working data.

[0043] Optionally, the charging time parameter may include season, time period, and ambient temperature, and the charging equipment parameter may include charging pile model, power level, and service life; however, this invention does not impose any limitations on these parameters.

[0044] Optionally, this feature similarity can be multidimensional vector cosine similarity, which is not limited in this invention.

[0045] Optionally, the similarity parameter can be the reciprocal of the weighted Euclidean distance, but this invention does not limit it.

[0046] As can be seen, through the above optional embodiments, by calculating the similarity between the charging time parameters and charging equipment parameters of each historical charging data in the historical database and each historical time point, and by calculating the product of the time characteristics and parameter similarity of the historical charging data, historical working data can be filtered, thereby achieving accurate risk assessment based on multi-dimensional similarity and improving the comprehensiveness and reliability of risk identification at historical time points.

[0047] As an optional embodiment, the step described above, calculating the danger level parameter corresponding to a historical time point based on multiple historical operating data and the charging operating parameters corresponding to that historical time point, includes: Calculate the charging operation parameters corresponding to the historical time point and the data similarity of each historical operation data; Calculate the reciprocal of the weighted summation of the similarities of all data points to obtain the risk level parameter corresponding to that historical time point.

[0048] The weight of each data similarity is proportional to the characteristic similarity and / or parameter similarity of the corresponding historical working data.

[0049] Optionally, this proportional weight implements the evaluation logic of "the more similar the historical data, the greater the contribution", which is not limited in this invention.

[0050] Optionally, the data similarity can be the cosine similarity after standardization of multi-parameter vectors; this invention does not impose any limitation on this.

[0051] As can be seen, through the above optional embodiments, the risk level parameter is calculated based on the reciprocal of the weighted data similarity, thereby achieving accurate risk assessment based on multi-dimensional similarity, improving the comprehensiveness and reliability of risk identification at historical time points, and reducing the risk of misjudgment of risks due to single similarity calculation.

[0052] As an optional embodiment, the step above, determining the timing data based on the charging operating parameters at the critical time point, includes: For each critical time point, calculate the time difference between that critical time point and the reference time point; Calculate the time weight that is inversely proportional to the time difference; Add time weights as additional features for this dangerous time point; Based on the dangerous time points from morning to night, the charging operation parameters and additional characteristics of all dangerous time points are sorted to obtain time-series operation data.

[0053] Optionally, the time weight can be weight = 1 / (time_diff + 1000) to avoid division by zero, where time_diff is the time difference, which is not limited in this invention.

[0054] Optionally, the sorting can form a multi-channel time series (parameter channel + weight channel), which is not limited in this invention.

[0055] As can be seen, through the above optional embodiments, by calculating the time difference between the dangerous time point and the reference time point as an inverse time weight as an additional feature, the charging working parameters and additional features are sorted by time to form time-series working data, thereby achieving accurate time-series feature enhancement based on time weighting, improving the ability of time-series data to represent fault trends, and reducing the risk of prediction model input deviation caused by ignoring time weight.

[0056] As an optional embodiment, the reference time point in the above steps is calculated through the following steps: Cluster all hazardous time points to obtain a set of time points; calculate the average value of all hazardous time points in the set of time points to obtain a reference time point.

[0057] Optionally, the time difference between any two dangerous time points in the time point set is less than any external time difference.

[0058] Optionally, the external time difference is the time difference between any dangerous time point and any dangerous time point in any set of non-time points.

[0059] Optionally, the clustering can be performed using hierarchical clustering based on time difference, with a clustering threshold of 1 hour; however, this invention does not impose any limitations on this method.

[0060] As can be seen, through the above optional embodiments, by using the average value of the cluster calculation of dangerous time points as a reference time point, a precise time benchmark based on the cluster center is constructed, which improves the scientificity and stability of time weight calculation, provides a reliable reference for the generation of time series working data, and reduces the risk of weight distortion caused by improper selection of reference points.

[0061] As an optional embodiment, the step above, predicting future fault information of charging piles based on time-series working data and a trained time-series prediction model, includes: The time-series working data is input into the trained LSTM neural network to obtain the future fault time point and fault parameters of the charging pile; optionally, the LSTM neural network is trained using a training dataset that includes multiple working parameter time series and corresponding fault point annotations. Determine the control server corresponding to the charging pile during the working time period to which the future failure time point belongs; optionally, the control server can be an edge server or a cloud server. Send future fault time points and fault parameters to the control server to provide early warning.

[0062] Optionally, the LSTM neural network can be a 4-layer bidirectional long short-term memory network with 256 hidden units per layer. Layer normalization and residual connections are added, and the mean squared error loss function is used. It is trained for 150 training epochs on 200,000 charging pile time-series labeled data, with the learning rate decreasing from 0.001 to 0.00001. The final prediction error is less than 12 minutes. This invention does not limit the scope of the invention.

[0063] Optionally, the warning may trigger remote power reduction, scheduled maintenance, or user APP push notifications; this invention does not impose any limitations on this.

[0064] As can be seen, through the above optional embodiments, by inputting time-series working data into a trained LSTM neural network to predict future fault time points and parameters, and sending them to the corresponding control server to realize early warning, accurate fault prediction and hierarchical push based on long-term time-series modeling are realized, improving the timeliness and pertinence of charging pile fault early warning, and reducing the risk of equipment downtime caused by fault prediction lag.

[0065] Let's illustrate the above technical solution with a specific example: The 60kW DC charging station, numbered CP-047, is currently in operation.

[0066] The system obtains charging operation parameters sampled every minute over the past 30 days, including an average current of 42A, a voltage of 385V, a peak battery temperature of 48℃, and a minimum insulation resistance of 1.2MΩ.

[0067] By matching similarity, 12 dangerous time points were selected (such as the current suddenly rising to 68A at 14:22 on December 15, and the temperature reaching 52℃ at 09:15 on December 18).

[0068] Calculate the reference time point: After clustering, three time point sets are obtained. The average time point of the largest set is around 14:30 every day, which is determined as the reference time point.

[0069] Calculate a time weight for each critical time point (the closer to 14:30, the higher the weight), and append it to the parameter sequence.

[0070] After sorting, time-series working data is formed (length 12×8-dimensional parameters + 1-dimensional weight).

[0071] Inputting a 4-layer bidirectional long short-term memory network, the predicted future fault time is around 14:18 on December 24th, with fault parameters of "power module overheating + insulation resistance dropping to 0.8MΩ".

[0072] The system immediately sends the warning information to the edge server of the DC charging pile, and the server automatically reduces the power of the pile to 30kW and pushes a maintenance work order.

[0073] On December 23, maintenance personnel discovered severe dust accumulation on the power module's cooling fan. Timely cleaning prevented a potential malfunction, and the 100% success rate validated the accuracy of the prediction.

[0074] Example 2 Please see Figure 2 , Figure 2 This is a schematic diagram of a charging pile hazard prediction system based on historical charging data analysis, as disclosed in an embodiment of the present invention. Figure 2 The described charging pile hazard prediction system based on historical charging data analysis can be applied to data processing systems / data processing equipment / data processing servers (wherein, the server includes local processing servers or cloud processing servers). For example... Figure 2 As shown, the charging pile hazard prediction system based on historical charging data analysis may include: The acquisition module 201 is used to acquire the charging operation parameters of the charging pile at multiple historical time points.

[0075] The filtering module 202 is used to filter out multiple dangerous time points from all historical time points based on the charging operation parameters. The determination module 203 is used to determine the timing operation data based on the charging operation parameters at the critical time point. The prediction module 204 is used to predict future fault information of charging piles based on time-series working data and a trained time-series prediction model.

[0076] As can be seen, the above-mentioned embodiments of the invention obtain charging operation parameters of charging piles at multiple historical time points and filter dangerous time points to form time-series operation data. The data is then input into a trained LSTM neural network to predict future fault information. This enables accurate charging pile fault prediction and early warning based on historical abnormal time sequences, improving the operational reliability and maintenance efficiency of charging piles, and reducing the risk of equipment damage or safety accidents caused by unpredicted faults.

[0077] As an optional embodiment, the charging operating parameters include at least one of the following: charging current, charging voltage, charging temperature, charging power, charging duration, and environmental sensing data; the environmental sensing data includes at least one of the following: ambient temperature, ambient humidity, ambient image, and ambient sound.

[0078] As can be seen, the above optional embodiments limit the content of charging operation parameters so that the data can comprehensively characterize the state characteristics of the charging pile during the charging process, assist in realizing accurate charging pile fault prediction and early warning based on historical abnormal time series, improve the operational reliability and maintenance efficiency of the charging pile, and reduce the risk of equipment damage or safety accidents caused by unpredictable faults.

[0079] As an optional embodiment, the specific method by which the filtering module filters out multiple dangerous time points from all historical time points based on charging operating parameters includes: For each historical point in time, multiple historical work data similar to that historical point in time are matched from the historical database; Based on multiple historical work data and the charging work parameters corresponding to that historical time point, calculate the danger level parameter corresponding to that historical time point; Historical time points where the risk level parameter is greater than the preset parameter threshold are selected to obtain multiple risk time points.

[0080] As can be seen, through the above optional embodiments, by matching similar historical work data at historical time points in the historical database and calculating the degree of danger parameters to filter dangerous time points, accurate identification of dangerous moments based on similar historical comparisons can be achieved, improving the accuracy and pertinence of dangerous time point screening, providing high-quality abnormal samples for time series data construction, and reducing the risk of fault prediction deviation caused by abnormal omissions.

[0081] As an optional embodiment, the specific method by which the filtering module matches multiple historical work data points similar to that historical time point from the historical database includes: For each historical charging data in the historical database, determine the charging time parameters and charging equipment parameters corresponding to that historical charging data; Calculate the similarity of characteristics between the charging time parameter and the time characteristics at that historical time point; Calculate the parameter similarity between the parameters of the charging equipment and the parameters of the charging pile; The similarity of the historical charging data is obtained by multiplying the feature similarity and parameter similarity. All historical charging data with a similarity greater than a preset similarity threshold are filtered out to obtain multiple historical working data.

[0082] As can be seen, through the above optional embodiments, by calculating the similarity between the charging time parameters and charging equipment parameters of each historical charging data in the historical database and each historical time point, and by calculating the product of the time characteristics and parameter similarity of the historical charging data, historical working data can be filtered, thereby achieving accurate risk assessment based on multi-dimensional similarity and improving the comprehensiveness and reliability of risk identification at historical time points.

[0083] As an optional embodiment, the specific method by which the filtering module calculates the danger level parameter corresponding to a historical time point based on multiple historical working data and the charging working parameters corresponding to that historical time point includes: Calculate the charging operation parameters corresponding to the historical time point and the data similarity of each historical operation data; Calculate the reciprocal of the weighted summation of the similarities of all data to obtain the risk level parameter corresponding to that historical time point; wherein, the calculation weight corresponding to each data similarity is proportional to the characteristic similarity and / or parameter similarity corresponding to the historical working data.

[0084] As can be seen, through the above optional embodiments, the risk level parameter is calculated based on the reciprocal of the weighted data similarity, thereby achieving accurate risk assessment based on multi-dimensional similarity, improving the comprehensiveness and reliability of risk identification at historical time points, and reducing the risk of misjudgment of risks due to single similarity calculation.

[0085] As an optional embodiment, the determining module determines the specific method of timing operation data based on the charging operation parameters at the critical time point, including: For each critical time point, calculate the time difference between that critical time point and the reference time point; Calculate the time weight that is inversely proportional to the time difference; Add time weights as additional features for this dangerous time point; Based on the dangerous time points from morning to night, the charging operation parameters and additional characteristics of all dangerous time points are sorted to obtain time-series operation data.

[0086] As can be seen, through the above optional embodiments, by calculating the time difference between the dangerous time point and the reference time point as an inverse time weight as an additional feature, the charging working parameters and additional features are sorted by time to form time-series working data, thereby achieving accurate time-series feature enhancement based on time weighting, improving the ability of time-series data to represent fault trends, and reducing the risk of prediction model input deviation caused by ignoring time weight.

[0087] As an optional embodiment, the reference time point is calculated through the following steps: Cluster all hazardous time points to obtain a set of time points; optionally, the time difference between any two hazardous time points in the set of time points is less than any external time difference; the external time difference is the time difference between any hazardous time point and any hazardous time point in any non-time point set. Calculate the average value of all dangerous time points in the time point set to obtain the reference time point.

[0088] As can be seen, through the above optional embodiments, by using the average value of the cluster calculation of dangerous time points as a reference time point, a precise time benchmark based on the cluster center is constructed, which improves the scientificity and stability of time weight calculation, provides a reliable reference for the generation of time series working data, and reduces the risk of weight distortion caused by improper selection of reference points.

[0089] As an optional embodiment, the prediction module predicts future fault information of charging piles based on time-series working data and a trained time-series prediction model in the following specific ways: The time-series working data is input into the trained LSTM neural network to obtain the future fault time point and fault parameters of the charging pile; optionally, the LSTM neural network is trained using a training dataset that includes multiple working parameter time series and corresponding fault point annotations. Determine the control server corresponding to the charging pile during the working time period to which the future failure time point belongs; optionally, the control server can be an edge server or a cloud server. Send future fault time points and fault parameters to the control server to provide early warning.

[0090] As can be seen, through the above optional embodiments, by inputting time-series working data into a trained LSTM neural network to predict future fault time points and parameters, and sending them to the corresponding control server to realize early warning, accurate fault prediction and hierarchical push based on long-term time-series modeling are realized, improving the timeliness and pertinence of charging pile fault early warning, and reducing the risk of equipment downtime caused by fault prediction lag.

[0091] Example 3 Please see Figure 3 , Figure 3This is another charging pile hazard prediction system based on historical charging data analysis disclosed in the embodiments of the present invention. Figure 3 The described charging pile hazard prediction system based on historical charging data analysis is applied in a data processing system / data processing equipment / data processing server (wherein, the server includes a local processing server or a cloud processing server). For example... Figure 3 As shown, the charging pile hazard prediction system based on historical charging data analysis may include: Memory 301 storing executable program code; Processor 302 coupled to memory 301; The processor 302 calls the executable program code stored in the memory 301 to execute the steps of the charging pile hazard prediction method based on historical charging data analysis described in Embodiment 1.

[0092] Example 4 This invention discloses a computer read storage medium that stores a computer program for electronic data exchange, wherein the computer program causes a computer to execute the steps of the charging pile hazard prediction method based on historical charging data analysis described in Embodiment 1.

[0093] Example 5 This invention discloses a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program, and the computer program is operable to cause a computer to perform the steps of the charging pile hazard prediction method based on historical charging data analysis described in Embodiment 1.

[0094] The foregoing has described specific embodiments of this specification; other embodiments are within the scope of the appended claims. In some cases, the actions or steps described in the claims may be performed in a different order than those shown in the embodiments and may still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily have to follow the specific or sequential order shown to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0095] The systems, devices, modules, or units described in the above embodiments can be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, a computer can be, for example, a personal computer, laptop computer, cellular phone, camera phone, smartphone, personal digital assistant, media player, navigation device, email device, game console, tablet computer, wearable device, or any combination of these devices.

[0096] For ease of description, the above devices are described in terms of function, divided into various units. Of course, in implementing this specification, the functions of each unit can be implemented in one or more software and / or hardware.

[0097] Those skilled in the art will understand that the embodiments of this specification can be provided as methods, systems, or computer program products. Therefore, the embodiments of this specification can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the embodiments of this specification can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0098] This specification is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this specification. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create a machine for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0099] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0100] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0101] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0102] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0103] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0104] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0105] This specification can be described in the general context of computer-executable instructions that are executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform a specific task or implement a specific abstract data type. This specification can also be practiced in distributed computing environments, where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0106] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to interchangeably. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments.

[0107] Finally, it should be noted that the charging pile hazard prediction method and system based on historical charging data analysis disclosed in the embodiments of the present invention are merely preferred embodiments of the present invention and are only used to illustrate the technical solutions of the present invention, not to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for predicting the hazards of charging piles based on historical charging data analysis, characterized in that, The method includes: Obtain charging operation parameters of charging piles at multiple historical time points; Based on the charging operating parameters, multiple dangerous time points are selected from all the historical time points; Based on the charging operation parameters at the critical time points, the timing operation data is determined; Based on the time-series working data and a trained time-series prediction model, the future fault information of the charging pile is predicted.

2. The charging pile hazard prediction method based on historical charging data analysis according to claim 1, characterized in that, The charging operating parameters include at least one of the following: charging current, charging voltage, charging temperature, charging power, charging duration, and environmental sensing data; the environmental sensing data includes at least one of the following: ambient temperature, ambient humidity, ambient image, and ambient sound.

3. The charging pile hazard prediction method based on historical charging data analysis according to claim 1, characterized in that, The step of filtering out multiple dangerous time points from all the historical time points based on the charging operating parameters includes: For each historical point in time, multiple historical work data similar to that historical point in time are matched from the historical database; Based on the multiple historical work data and the charging work parameters corresponding to the historical time point, calculate the danger level parameter corresponding to the historical time point; The historical time points where the danger level parameter is greater than a preset parameter threshold are selected to obtain multiple dangerous time points.

4. The charging pile hazard prediction method based on historical charging data analysis according to claim 3, characterized in that, The process of matching multiple historical work data points similar to that historical time point from the historical database includes: For each historical charging data in the historical database, determine the charging time parameters and charging equipment parameters corresponding to that historical charging data; Calculate the similarity of characteristics between the charging time parameter and the time characteristics at that historical time point; Calculate the parameter similarity between the parameters of the charging equipment and the parameters of the charging pile; Calculate the product of the characteristic similarity and the parameter similarity to obtain the similarity corresponding to the historical charging data; All historical charging data with a similarity greater than a preset similarity threshold are filtered out to obtain multiple historical working data.

5. The charging pile hazard prediction method based on historical charging data analysis according to claim 4, characterized in that, The step of calculating the danger level parameter corresponding to the historical time point based on the multiple historical working data and the charging working parameters corresponding to the historical time point includes: Calculate the charging operation parameters corresponding to the historical time point and the data similarity of each of the historical operation data; Calculate the reciprocal of the weighted summation average of all the data similarities to obtain the risk level parameter corresponding to the historical time point; wherein, the calculation weight corresponding to each of the data similarities is proportional to the characteristic similarity and / or parameter similarity corresponding to the corresponding historical working data.

6. The charging pile hazard prediction method based on historical charging data analysis according to claim 1, characterized in that, The step of determining the timing data based on the charging operation parameters at the critical time points includes: For each of the aforementioned critical time points, calculate the time difference between that critical time point and the reference time point; Calculate the time weight that is inversely proportional to the time difference; The time weight is labeled as an additional feature of the dangerous time point; Based on the dangerous time points from morning to night, the charging operation parameters and additional features of all the dangerous time points are sorted to obtain time-series operation data.

7. The charging pile hazard prediction method based on historical charging data analysis according to claim 6, characterized in that, The reference time point is calculated through the following steps: Cluster all the aforementioned dangerous time points to obtain a set of time points; the time difference between any two of the dangerous time points in the set of time points is less than any external time difference; the external time difference is the time difference between any one of the dangerous time points and any dangerous time point not in the set of time points. The average value of all the dangerous time points in the set of time points is calculated to obtain the reference time point.

8. The charging pile hazard prediction method based on historical charging data analysis according to claim 1, characterized in that, The step of predicting future fault information of the charging pile based on the time-series working data and a trained time-series prediction model includes: The time-series working data is input into a trained LSTM neural network to obtain the future fault time point and fault parameters of the charging pile; the LSTM neural network is trained using a training dataset that includes multiple working parameter time series and corresponding fault point annotations. Determine the control server corresponding to the charging pile within the working time period to which the future failure time point belongs; the control server is an edge server or a cloud server; The future fault time point and fault parameters are sent to the control server to provide early warning.

9. A charging pile hazard prediction system based on historical charging data analysis, characterized in that, The system includes: The acquisition module is used to acquire the charging operation parameters of the charging pile at multiple historical time points; The filtering module is used to filter out multiple dangerous time points from all the historical time points based on the charging operating parameters. The determination module is used to determine the timing operation data based on the charging operation parameters at the critical time points; The prediction module is used to predict future fault information of the charging pile based on the time-series working data and a trained time-series prediction model.

10. A charging pile hazard prediction system based on historical charging data analysis, characterized in that, The system includes: Memory containing executable program code; A processor coupled to the memory; The processor calls the executable program code stored in the memory to execute the charging pile hazard prediction method based on historical charging data analysis as described in any one of claims 1-8.

Images