Charging gun life evaluation system and method based on charging pile charging analysis

Abstract

This invention relates to the field of charging gun lifespan assessment technology, and particularly to a charging gun lifespan assessment system and method based on charging pile charging analysis. Based on accessed battery attribute data and power time series data, this invention differentiates power output stability through matching standard curve templates of the same battery type and classifying deviation types, and corrects the power change curve according to the deviation type. Based on the corrected power change curve and user operation behavior time series data, it extracts multi-dimensional positional features of the charging stop time point on the corrected curve and analyzes the urgency of the user's stop. Based on the multi-dimensional positional features and the urgency of the user's stop, it quantifies the loss deviation of the charging gun caused by the charging stop event; thus, it achieves the correction of the remaining lifespan of the charging gun. It overcomes the technical bias of treating all stop events equally in existing technologies, significantly improving the quantitative accuracy of the loss deviation caused by a single stop event.

Description

Technical Field

[0001] This invention relates to the field of charging gun life assessment technology, and in particular to a charging gun life assessment system and method based on charging pile charging analysis. Background Technology

[0002] Traditional charging gun life assessments often rely on type test data from the factory, using the rated number of insertions and removals as the sole life indicator. This ignores the differentiated damage caused by factors such as variations in charging power, periods of inactivity, and environmental changes during actual use, leading to a significant disconnect between life assessment results and actual service performance. In contrast, dynamic assessment methods based on real-time operational data leverage the charging pile's built-in power metering, temperature sensing, and communication modules to continuously collect electrical parameters, thermal parameters, and user operation data during the charging process. This enables online monitoring and dynamic life assessment of the charging gun's service status, overcoming the limitations of traditional methods that rely on static factory data and significantly improving the accuracy and timeliness of life assessments.

[0003] However, existing technologies for assessing the lifespan of charging guns do not correlate charging stop events with their specific location on the charging power curve. For example, when a user suddenly stops charging in the peak power region, the current is at its maximum, the temperature rise is highest, and the arcing energy and thermal stress impact far exceed the damage levels when stopping in the power decline region. Without considering the precise location of the stop on the power curve, it is difficult to accurately quantify the differentiated damage caused to the charging gun by a single stop, leading to the blind neglect of the impact caused by different stop behaviors. Furthermore, existing technologies lack a comprehensive consideration of the urgency of the user's stop and the distribution of historical stop events, making it difficult for the assessment results to accurately reflect the actual remaining lifespan of the charging gun and failing to provide a comprehensive basis for the operation and maintenance decisions of charging facilities. Summary of the Invention

[0004] To overcome the defects and shortcomings of existing technologies, this invention provides a charging gun life assessment system and method based on charging pile charging analysis. By collecting charging power, battery attributes, and user operation behavior data, it identifies the multidimensional positional characteristics of the stopping moment on the inverted U-shaped power curve, quantifies the urgency of the user's stopping, and constructs a quantitative mapping model between the stopping position and the equivalent damage of the charging gun. Furthermore, it introduces position distribution entropy and position risk skewness coefficient as two-parameter correction factors to nonlinearly correct the baseline remaining life, thereby achieving accurate dynamic assessment of the charging gun's remaining life.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: In a first aspect, embodiments of the present invention provide a method for evaluating the lifespan of a charging gun based on charging pile charging analysis, comprising the following steps: Step S100: Through the power metering module, battery management system communication module and user interaction behavior monitoring module deployed on the charging pile, the power time series data of each charging process, the access battery attribute data and user operation behavior time series data are collected synchronously. Step S200: Based on the access battery attribute data and power time series data, the power output stability is differentiated by matching the standard curve template of the same battery type and classifying and identifying the deviation type, and the power change curve is corrected according to the deviation type. Step S300: Based on the corrected power change curve and user operation behavior time series data, extract the multi-dimensional position features of the charging stop time point on the corrected curve and analyze the user's urgency to stop; according to the multi-dimensional position features and the user's urgency to stop, quantify the loss deviation of the charging gun caused by the charging stop event. Step S400: Based on the loss deviation caused by the charging gun stopping the charging event, the remaining lifespan of the charging gun is corrected. Step S500: Based on the remaining lifespan correction result of the charging gun, issue a warning for charging gun maintenance and replacement.

[0006] In conjunction with the first aspect, this embodiment of the invention also provides a first implementation of the first aspect, wherein in step S100, the power time series data includes a charging power time series, the start time of each charging, and the user request to stop time; the accessed battery attribute data includes battery type, battery rated capacity, current battery state of charge, and target state of charge; the user operation behavior time series data includes a timestamp sequence of the user viewing the charging status, the trigger time of the vehicle unlock signal, and the timestamp of the charging gun button press signal.

[0007] In conjunction with the first aspect, the present invention also provides a second implementation of the first aspect, wherein step S200 specifically includes the following: Step S201: Based on the battery type, the current state of charge of the battery and the target state of charge, retrieve the standard power change curve template and the corresponding confidence interval bandwidth of the same type of battery in the same state of charge range from the historical database. Step S202: Compare the current charging power time series with the standard power change curve template point by point, and calculate the deviation at each sampling point; the deviation at each sampling point is defined as the difference between the current charging power and the standard power at each sampling point divided by the confidence interval bandwidth. Step S203: Based on the deviation at each sampling point, construct the deviation sequence of the current charging process; perform time series feature analysis on the deviation sequence to obtain the mean deviation, standard deviation, drift rate, and peak deviation factor of the current charging process. Step S204: Combine the mean deviation, standard deviation, drift rate, and peak deviation factor of the current charging process to form the deviation feature vector of the current charging process; input the deviation feature vector of the current charging process into a pre-trained support vector machine classifier for deviation type identification; Step S205: The pre-trained support vector machine classifier outputs the probability distribution of each deviation type, takes the deviation type corresponding to the highest probability as the deviation type determination result of the current charging process, and corrects the power change curve according to the deviation type.

[0008] In conjunction with the first aspect, this invention also provides a third implementation method for the first aspect, wherein the specific construction process of the pre-trained support vector machine classifier is as follows: Obtain a historical charging dataset, in which each data point contains a sequence of deviations with known deviation type labels and a corresponding deviation feature vector. The deviation type labels include systemic deviation, fluctuation deviation, oscillation deviation, and severe anomaly. Using the deviation feature vector of the historical charging dataset as the input feature and the corresponding deviation type label as the classification target, a multi-class support vector machine model is constructed using a one-to-one strategy. The radial basis kernel function is selected as the kernel function, and the penalty parameter and kernel parameter are determined through grid search and cross-validation. Based on the historical charging dataset, the deviation feature vector is used as the input of the support vector machine model, and the deviation type label is used as the output target of the support vector machine model. The support vector machine model is trained, and the corresponding penalty parameter and kernel parameter that maximize the average classification accuracy of cross-validation are determined by grid search, thus obtaining the initial support vector machine classifier. The initial support vector machine classifier is validated using a reserved validation set. The classification accuracy on the validation set is calculated, and the initial support vector machine classifier whose classification accuracy on the validation set is greater than or equal to a preset accuracy threshold is output as the support vector machine classifier.

[0009] In conjunction with the first aspect, this embodiment of the invention also provides a fourth implementation of the first aspect, the specific process of correcting the power change curve according to the type of deviation is as follows: When the deviation type of the current charging process is a systemic deviation, obtain the difference between the actual charging power of each sampling point in the current charging power time series and the charging power at the corresponding position in the standard power change curve template; use the median of the corresponding differences of all sampling points as the translation correction amount of the current charging process; subtract the translation correction amount of the current charging process from the actual charging power of each sampling point in the current charging power time series to obtain the corrected charging power of the corresponding sampling point, and obtain the corrected power change curve of the current charging process. When the deviation type of the current charging process is a fluctuating deviation, isolated abnormal points with deviations exceeding the preset threshold are marked and removed, and local smoothing correction is performed using cubic spline interpolation to obtain the corrected power change curve of the current charging process. When the deviation type of the current charging process is an oscillating deviation, the continuous oscillation interval in the deviation quantity sequence is smoothed and corrected by using a sliding window mean filter with a fixed window length to obtain the corrected power change curve of the current charging process. When the deviation type of the current charging process is severe abnormality, the charging data is marked as invalid data, will not be included in subsequent life assessments, and the life assessment process for the current charging process will be terminated.

[0010] In conjunction with the first aspect, this embodiment of the invention also provides a fifth implementation of the first aspect, wherein step S300 specifically includes: Step S301: Based on the power change curve corrected in step S205, identify the peak power of the curve and the corresponding occurrence time and minimum power, wherein the minimum power is the power value corresponding to the charging start time. Step S302: When the user issues a stop request, extract the multi-dimensional position feature vector of the stop time point on the corrected power change curve. The multi-dimensional position feature vector includes the relative power position, normalized time position, normalized power change rate, normalized curve curvature, and stage label encoding. Step S303: Calculate the urgency of the user to stop based on the time sequence data of user operation behavior; Step S304: Based on the multi-dimensional location feature vector and the user's urgency to stop, quantify the loss deviation of the charging gun caused by the charging stop event during the current charging process; the loss deviation is obtained by multiplying the base loss coefficient and the loss amplification factor.

[0011] In conjunction with the first aspect, this embodiment of the invention also provides a sixth implementation of the first aspect, wherein step S400 specifically includes: Step S401: Retrieve the sequence of loss deviations corresponding to all historical charging stop events prior to this charging from the historical database; Step S402: Based on the loss deviation sequence, analyze the cumulative loss deviation of the current charging pile; Step S403: Divide the [0,1] interval into K sub-intervals at equal intervals, and count the frequency of the base value loss coefficient of all historical charging stop events falling into each sub-interval; Step S404: Based on the frequency of the base loss coefficient falling into each sub-interval, analyze the current location distribution entropy and location risk skewness coefficient of the charging pile; Step S405: Based on the current cumulative loss deviation of the charging pile, the location distribution entropy and the location risk skewness coefficient, the baseline remaining life of the charging gun is corrected.

[0012] In conjunction with the first aspect, this embodiment of the invention also provides a seventh implementation of the first aspect, wherein step S500 involves providing a charging gun maintenance and replacement early warning based on the remaining lifespan correction result of the charging gun; specifically including: Step S501: Preset the remaining lifespan threshold. When the corrected remaining lifespan of the charging gun is greater than or equal to the remaining lifespan threshold, maintain normal monitoring status and store the charging data and the loss deviation of the charging gun caused by the charging stop event during the current charging process into the historical database for continuous calibration. Step S502: When the remaining lifespan of the corrected charging gun is less than the remaining lifespan threshold, generate a charging gun replacement warning notification and push it.

[0013] Secondly, embodiments of the present invention also provide a charging gun life assessment system based on charging pile charging analysis, including: The data acquisition module is used to synchronously collect power time-series data, access battery attribute data, and user operation behavior time-series data for each charging process through the power metering module, battery management system communication module, and user interaction behavior monitoring module deployed in the charging pile. The power deviation correction module is used to make differential judgments on power output stability based on the access battery attribute data and power time series data, through matching the standard curve template of the same battery type and classifying and identifying the deviation type, and correct the power change curve according to the deviation type. The loss deviation correction module is used to extract the multi-dimensional position features of the charging stop time point on the correction curve based on the corrected power change curve and user operation behavior time series data, and analyze the user's urgency to stop; based on the multi-dimensional position features and the user's urgency to stop, the loss deviation of the charging gun caused by the charging stop event is quantified. The lifespan correction module is used to correct the remaining lifespan of the charging gun based on the wear deviation caused by the charging gun stopping the charging event. The charging gun maintenance module is used to provide early warnings for charging gun maintenance and replacement based on the remaining lifespan correction results of the charging gun.

[0014] Compared with the prior art, the present invention has the following advantages and beneficial effects: 1. This invention refines and quantifies the multidimensional positional features on the inverted U-shaped power curve at the moment of charging stop, and constructs a quantitative mapping model between the stop position and the base value damage coefficient of the charging gun. This overcomes the technical bias of treating all stop events as the same in the prior art, and significantly improves the quantitative accuracy of the loss deviation caused by a single stop event. 2. This invention solves the technical problem that existing technologies cannot perceive the impact of sudden user stopping behavior on the transient impact of charging gun by introducing a user stopping urgency index and constructing a loss amplification factor to dynamically weight and correct the base value damage coefficient, making the loss quantification results closer to the actual physical damage process. 3. This invention performs nonlinear correction on the baseline remaining life of the charging gun from a new dimension of the distribution of stopping positions. The position distribution entropy characterizes the dispersion of historical stopping position distribution to reflect the damage dispersion effect or cumulative enhancement effect, and the position risk skewness coefficient characterizes whether the historical stopping position is biased towards the peak area to reflect the concentration of severe working conditions. The two work together to correct the problem, providing a reliable basis for the precise operation and maintenance decision of charging facilities.

[0015] Other features and advantages of the invention will be set forth in the following description, or some features and advantages may be inferred from the description or determined without doubt, or may be learned by practicing the techniques described above.

[0016] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0017] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0018] Figure 1 This is a schematic diagram of the overall process of the charging gun life assessment method based on charging pile charging analysis in an embodiment of the present invention. Figure 2 This is a flowchart of step S2 in the charging gun life assessment method based on charging pile charging analysis according to an embodiment of the present invention. Figure 3 This is a flowchart of step S3 in the charging gun life assessment method based on charging pile charging analysis according to an embodiment of the present invention. Figure 4 This is a schematic diagram of the charging gun life assessment system based on charging pile charging analysis according to the present invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0020] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention. Example

[0021] To facilitate understanding, the charging gun life assessment method based on charging pile charging analysis provided in this embodiment of the invention will be described first, referring to... Figure 1 Specifically, it includes the following steps: Step S100: Through the power metering module, battery management system communication module and user interaction behavior monitoring module deployed on the charging pile, the power time series data of each charging process, the access battery attribute data and user operation behavior time series data are collected synchronously. Step S200: Based on the access battery attribute data and power time series data, the power output stability is differentiated by matching the standard curve template of the same battery type and classifying and identifying the deviation type, and the power change curve is corrected according to the deviation type. Step S300: Based on the corrected power change curve and user operation behavior time series data, extract the multi-dimensional position features of the charging stop time point on the corrected curve and analyze the user's urgency to stop; according to the multi-dimensional position features and the user's urgency to stop, quantify the loss deviation of the charging gun caused by the charging stop event. Step S400: Based on the loss deviation caused by the charging gun stopping the charging event, the remaining lifespan of the charging gun is corrected. Step S500: Based on the remaining lifespan correction result of the charging gun, issue a warning for charging gun maintenance and replacement.

[0022] In a specific implementation, in step S100, the power time series data includes the charging power time series, the start time of each charge, and the user request to stop time; the accessed battery attribute data includes the battery type, the battery rated capacity, the current state of charge of the battery, and the target state of charge; the user operation behavior time series data includes the timestamp sequence of the user viewing the charging status, the trigger time of the vehicle unlock signal, and the timestamp of the charging gun button pressing signal.

[0023] In practical implementation, the core function of this embodiment is to eliminate the influence of inherent morphological differences in the power curve caused by different battery types and different states of charge ranges on subsequent stopping position identification, ensuring that the stopping position features extracted in step S300 can truly reflect the relative position on the inverted U-shaped curve at the stopping moment, rather than the curve morphological shift caused by differences in battery characteristics. Specifically, as... Figure 2 As shown, step S200 specifically includes the following: Step S201: Based on the battery type, current state of charge (SOC) and target SOC, retrieve the standard power change curve template and corresponding confidence interval bandwidth of the same type of battery under the same SOC range from the historical database; The construction process of the standard power change curve template is as follows: Collect a large number of historical charging power curves of the same type of battery under the same SOC range, normalize the time axis of each curve, and calculate the mean and standard deviation of the corresponding sampling point in all curves to obtain the standard power change curve template, where the mean is the standard power of the corresponding sampling point in the standard power change curve template, and the standard deviation is the confidence interval bandwidth of the corresponding sampling point; Step S202: Compare the current charging power time series with the standard power change curve template point by point, and calculate the deviation at each sampling point; the deviation at each sampling point is defined as the difference between the current charging power and the standard power at each sampling point divided by the confidence interval bandwidth. Step S203: Based on the deviation at each sampling point, construct the deviation sequence of the current charging process; perform time series feature analysis on the deviation sequence to obtain the mean deviation, standard deviation, drift rate, and peak deviation factor of the current charging process. The mean deviation is the average of the deviations at all sampling points; The squared difference between the deviation of each sampling point and the mean deviation is summed, the sum is divided by the difference between the number of sampling points and 1, and the square root of the division result is taken to obtain the standard deviation of the deviation. The offset of each sampling point is subtracted from the offset of the corresponding adjacent previous sampling point; the sum of the subtraction results of all sampling points is divided by the difference between the number of sampling points and 1 to obtain the deviation drift rate. Extract the maximum value of the absolute value of the deviation from all sampling points, and divide the maximum value by the standard deviation of the deviation to obtain the deviation peak factor; Step S204: Combine the mean deviation, standard deviation, drift rate, and peak deviation factor of the current charging process to form the deviation feature vector of the current charging process; input the deviation feature vector of the current charging process into a pre-trained support vector machine classifier for deviation type identification; Step S205: The pre-trained support vector machine classifier outputs the probability distribution of each deviation type, takes the deviation type corresponding to the highest probability as the deviation type determination result of the current charging process, and corrects the power change curve according to the deviation type.

[0024] In practice, the specific construction process of the pre-trained support vector machine classifier is as follows: Obtain a historical charging dataset, in which each data point contains a sequence of deviations with known deviation type labels and a corresponding deviation feature vector. The deviation type labels include systemic deviation, fluctuation deviation, oscillation deviation, and severe anomaly. Using the deviation feature vectors from the historical charging dataset as input features and the corresponding deviation type labels as classification targets, a multi-class support vector machine model is constructed using a one-to-one strategy. The radial basis function kernel is selected, and the penalty and kernel parameters are determined through grid search and cross-validation. The search range for the penalty parameter is... to The search range for kernel parameters is to The step size is always an exponential step size. ; Based on historical charging datasets, the deviation feature vectors are used as input to the Support Vector Machine (SVM) model, and the deviation type labels are used as the output target. The SVM model is trained using a five-fold cross-validation approach. The training data is randomly divided into five subsets, with four subsets used for training and one for validation alternately, repeated five times. The average classification accuracy is used as the performance metric for this set of parameters. A grid search is used to determine the penalty and kernel parameters that maximize the average cross-validation accuracy, resulting in the initial SVM classifier. During training, the support vectors and their corresponding Lagrange multipliers are determined by solving a dual optimization problem, constructing the decision function. For multi-class problems, a one-to-one strategy is adopted, training a binary SVM between every two classes, for a total of six classifiers. A voting method is used to determine the final class during classification.

[0025] The initial support vector machine classifier is validated using a validation set. The classification accuracy on the validation set is calculated, and the initial support vector machine classifier whose classification accuracy on the validation set is greater than or equal to a preset accuracy threshold is output as the support vector machine classifier.

[0026] In practice, the specific process of correcting the power change curve based on the type of deviation is as follows: When the deviation type of the current charging process is a systemic deviation, obtain the difference between the actual charging power of each sampling point in the current charging power time series and the charging power at the corresponding position in the standard power change curve template; use the median of the corresponding differences of all sampling points as the translation correction amount of the current charging process; subtract the translation correction amount of the current charging process from the actual charging power of each sampling point in the current charging power time series to obtain the corrected charging power of the corresponding sampling point, and obtain the corrected power change curve of the current charging process; when the median is positive, it indicates that the actual power is generally too high and needs to be shifted downward; when the median is negative, it indicates that the actual power is generally too low and needs to be shifted upward.

[0027] When the deviation type of the current charging process is a fluctuating deviation, isolated outliers with deviations exceeding a preset threshold are marked and removed. Cubic spline interpolation is used for local smoothing correction to obtain the corrected power change curve of the current charging process. In a specific embodiment, the cubic spline interpolation is implemented as follows: after removing outliers, the remaining normal points are used as interpolation nodes to construct a cubic polynomial piecewise function that satisfies the condition of second-order continuous differentiability, and the power value at the removed points is interpolated and filled.

[0028] When the deviation type of the current charging process is an oscillating deviation, the continuous oscillation interval in the deviation quantity sequence is smoothed and corrected by using sliding window mean filtering with a fixed window length (the window length is preset to 10 sampling points). That is, the correction value of each point is the power average of 5 points before and after that point (a total of 11 points), and the power change curve after correction of the current charging process is obtained. When the deviation type of the current charging process is severe abnormality, the charging data is marked as invalid data, will not be included in subsequent life assessments, and the life assessment process for the current charging process will be terminated.

[0029] Unlike existing technologies that use uniform smoothing or filtering to process charging data, this embodiment employs differentiated correction strategies based on the type of deviation. Specifically, systemic deviations are corrected by translation to preserve the curve shape, fluctuation-type deviations are corrected by spline interpolation to eliminate isolated spikes, and oscillation-type deviations are corrected by mean filtering to suppress continuous oscillations. This eliminates interference while preserving the true shape and characteristics of the original power curve to the greatest extent possible, laying the foundation for accurate identification of the subsequent stopping position.

[0030] In its specific implementation, this embodiment, for the first time, quantifies the precise position of the charging stop moment on the inverted U-shaped power curve as a multi-dimensional continuous numerical feature, and introduces the user's urgency of stopping the charging as a damage amplification factor, thus achieving refined quantification of the charging gun loss deviation caused by a single stop event. Specifically, as follows... Figure 3 As shown, step S300 specifically includes: Step S301: Based on the power change curve corrected in step S205, identify the peak power of the curve and the corresponding occurrence time and minimum power, wherein the minimum power is the power value corresponding to the charging start time. Step S302: When the user issues a stop request, extract the multi-dimensional position feature vector of the stop time point on the corrected power change curve. The multi-dimensional position feature vector includes the relative power position, normalized time position, normalized power change rate, normalized curve curvature, and stage label encoding. Obtain the difference between the power value and the minimum power at the moment the user stops, based on the corrected power change curve; obtain the difference between the peak power and the minimum power; divide the difference between the power value and the minimum power at the moment the user stops by the difference between the peak power and the minimum power to obtain the relative power position; the relative power position ranges from 0 to 1, where a relative power position of 1 indicates that the user stops at the peak point, and a relative power position of 0 indicates that the user stops at the initial minimum point; Compare the user stop time with the peak occurrence time; when the user stop time is less than or equal to the peak occurrence time, obtain the difference between the user stop time and the charging start time, obtain the difference between the peak occurrence time and the charging start time, divide the difference between the user stop time and the charging start time by the difference between the peak occurrence time and the charging start time, and use it as the normalized time position. When the user stops at a time greater than the peak occurrence time, the difference between the user stops at a time greater than the peak occurrence time is obtained, and the difference between the charging end time and the peak occurrence time is also obtained. The difference between the user stops at a time greater than the peak occurrence time is divided by the difference between the charging end time and the peak occurrence time to obtain the normalized time position. The normalized time position ranges from 0 to 1, representing the relative time position of the stopping point in the rising or falling half of the process. Obtain the absolute value of the first derivative of the corrected power change curve at the moment the user stops charging; traverse the entire charging process and obtain the maximum value of the absolute value of the first derivative of the corrected power change curve; divide the absolute value of the first derivative at the moment the user stops charging by the maximum value of the absolute value of the first derivative of the corrected power change curve to obtain the normalized power change rate; the normalized power change rate ranges from 0 to 1, and the larger the normalized power change rate, the more drastic the power change at the moment of stopping. Obtain the absolute value of the second derivative of the corrected power change curve at the moment the user stops charging; traverse the entire charging process and obtain the maximum value of the absolute value of the second derivative of the corrected power change curve; divide the absolute value of the second derivative at the moment the user stops charging by the maximum value of the absolute value of the second derivative of the corrected power change curve, and use it as the normalized curve curvature; the normalized curve curvature ranges from 0 to 1, representing the degree of curvature of the power curve at the stopping point, and is used to distinguish between two different scenarios: the peak has just passed and the power curve is rapidly decreasing, and the power curve is gradually flattening at the end of the decrease. When the power value of the corrected power change curve at the moment the user stops is greater than or equal to 95% of the peak power, it is determined to be in the peak region, and the stage label code is set to 1; when the moment the user stops is less than the moment the peak occurs and the power value of the corrected power change curve at the moment the user stops is less than 95% of the peak power, it is determined to be in the rising region, and the stage label code is set to 0.5; when the moment the user stops is greater than the moment the peak occurs and the power value of the corrected power change curve at the moment the user stops is less than 95% of the peak power, it is determined to be in the falling region, and the stage label code is set to 0.

[0031] It should be further explained that in this embodiment, the normalized power change rate and the normalized curve curvature are introduced as independent positional dimensions. Their deeper physical significance lies in the collaborative differentiation of the different damage mechanisms caused by the stopping moment in the peak region and the descent region.

[0032] Specifically, the normalized rate of change of power (absolute value of the first derivative) characterizes the drastic transient change of current at the stopping point, while the normalized curve curvature (absolute value of the second derivative) characterizes the degree of curvature and acceleration characteristics of the power curve at the stopping point. Combining these two methods can effectively decouple the following two differentiated failure modes: (1) When the stopping time is in the peak region (the stage label is coded as 1, at which time the first derivative approaches zero and the absolute value of the second derivative is large), the current is in a high-level quasi-steady state, the power change rate is very small, the arc energy generated at the moment of cutting off is relatively low, but the Joule heating effect accumulated by the large current reaches the peak, which leads to a sharp increase in temperature at the contact point of the contact, and the damage is mainly material softening and micro-melting welding tendency dominated by thermal stress. (2) When the stopping time is in the falling region (the stage label code is 0, at which time the absolute value of the first derivative is large and the absolute value of the second derivative gradually decreases), the current is in the rapid falling stage, the power change rate is large, the derivative of the current with respect to time at the moment of disconnection increases significantly, the induced electromotive force is enhanced, which leads to a sharp increase in the arcing energy at the break point, and the damage is mainly due to the vaporization of the contact material and the deterioration of the surface roughness dominated by electrical wear. (3) When the stopping time is in the transition zone where the peak value has just passed and the power drops rapidly, the normalized power change rate and the curvature of the normalized curve are both at a high level. Stopping at this position will simultaneously bear the double impact of thermal stress accumulation and strong arc ablation, and the base loss coefficient will reach a local maximum value.

[0033] Based on the above-mentioned differences in physical mechanisms, the finite element simulation in step S304 can assign thermal-mechanical coupling boundary conditions and electromagnetic-thermal boundary conditions to the dominant failure modes under different stage label codes, so that the calibrated base loss coefficient can more accurately reflect the actual physical damage equivalent, effectively avoiding the systematic bias introduced by the traditional method of treating the peak region and the descent region stopping events as equal.

[0034] Step S303: Calculate the urgency of the user to stop based on the time sequence data of user operation behavior; The formula for calculating the urgency level of the user's cessation is as follows:

[0035] In the formula, U represents the user's urgency to stop. Here, T is the weighting coefficient, and Ts is the user's stopping time. The timestamp of the last time the user viewed the charging status (in this embodiment, the maximum value can be taken from the timestamp sequence of the charging status viewed on the charging APP). To examine the time decay constant of the behavior (in this embodiment, it is preset to 300 seconds, i.e., 5 minutes). The value range is (0,1]. When the user has checked the charging status in a short period of time before stopping, this value is close to 1, indicating that the user has actively paid attention to the charging status. The trigger time for the vehicle unlock signal. The unlock signal has an effective time window (default 60 seconds). When the user unlocks the vehicle within 60 seconds before stopping, the indicator function is set to 1, indicating that the user intends to leave. Timestamp for the charging gun button press signal. This is the effective time window for the button press signal (default is 10 seconds). When the user presses the charging gun button within 10 seconds before stopping, the indicator function is set to 1, indicating that the user is ready to unplug the charging gun. This is an indicator function. It takes the value 1 when the condition in parentheses is true, and 0 otherwise. The closer U is to 1, the more sudden the user's intention to stop and the less pre-buffering time there is. The closer U is to 0, the more relaxed the stopping process is and the more adequate the pre-buffering time there is. Step S304: Based on the multi-dimensional location feature vector and the user's urgency to stop, quantify the loss deviation of the charging gun caused by the charging stop event during the current charging process; the loss deviation is obtained by multiplying the base loss coefficient and the loss amplification factor. The base loss coefficient has a value range of [0,1] and is pre-calibrated through multi-physics finite element simulation of the charging gun. The loss amplification factor is obtained by adding the urgency amplification factor to the value and the user's urgency to stop. Among them, the higher the user's urgency to stop, the greater the loss deviation at the same position. The specific process of simulation calibration is as follows: an electro-thermal-mechanical multi-physics coupled model of the charging gun is established in finite element simulation software (such as COMSOL Multiphysics or ANSYS). The contact arcing energy, temperature rise distribution and mechanical stress at the time of stopping charging are simulated under different combinations of positional features (relative power position, normalized time position, normalized power change rate, normalized curve curvature, and different values ​​of stage label code). The contact material damage equivalent under each working condition is calculated and normalized to obtain the base loss coefficient. The basic logic for calibrating the base loss coefficient in simulation is as follows: the base loss coefficient monotonically increases with relative power position, normalized power change rate, and normalized curve curvature (a larger relative power position indicates that the power is closer to the peak; a larger normalized power change rate indicates that the power change is more drastic; and a larger normalized curve curvature indicates that the curve bends more drastically). When the stage label is coded as the peak region, the range of the base loss coefficient is higher than that when the stage label is coded as the rising region. When the stage label is coded as the rising region, the range of the base loss coefficient is higher than that when the stage label is coded as the falling region. Furthermore, the transition region where the peak has just passed and both the first and second derivatives are relatively large is the peak region of the base loss coefficient. The urgency amplification factor is a positive number greater than zero. In this embodiment, it can be determined by fitting historical data, and the preferred value range is 1.5 to 3.0. The larger the urgency amplification factor, the more significant the amplification effect of the user's urgency on the loss. When U=0, the loss amplification factor is 1 (no amplification). When U=1, the loss amplification factor is the sum of the value and the urgency amplification factor (maximum amplification). The higher the user's urgency to stop, the greater the loss deviation at the same location.

[0036] In its specific implementation, step S400 includes: Step S401: Retrieve the sequence of loss deviations corresponding to all historical charging stop events prior to this charging from the historical database; Step S402: Based on the loss deviation sequence, analyze the cumulative loss deviation of the current charging pile; sum the loss deviations corresponding to all historical charging stop events in the loss deviation sequence to obtain the cumulative loss deviation of the current charging pile. Step S403: Divide the [0,1] interval into K sub-intervals at equal intervals, and count the frequency of the base value loss coefficient of all historical charging stop events falling into each sub-interval. In a specific embodiment, the advantage of the equal interval division method is that it does not introduce artificial bias and can objectively and unbiasedly reflect the statistical distribution characteristics of the base value loss coefficient in actual use. When K=10, the width of each sub-interval is 0.1, which can ensure sufficient statistical resolution to characterize the distribution pattern (avoiding the loss of distribution details due to K being too small), and also avoid the problem of increased frequency estimation variance and sparse samples in some sub-intervals due to K being too large (avoiding the decrease in statistical reliability due to K being too large), thus achieving a balance between statistical resolution and estimation reliability.

[0037] Step S404: Based on the frequency of the base loss coefficient falling into each sub-interval, analyze the current location distribution entropy and location risk skewness coefficient of the charging pile; The formula for calculating the location distribution entropy is as follows: ; In the formula, H is the location distribution entropy, and pj is the frequency at which the base loss coefficient of all historical charging stop events falls into the j-th sub-interval; The formula for calculating the location risk skewness coefficient is as follows: ; In the formula, S is the location risk skewness coefficient. Let Ldi be the stage label code for the i-th charging stop event, N be the total number of charging stop events, Ldi be the loss deviation caused by the i-th charging stop event, and Lt be the cumulative loss deviation of the current charging station. Step S405: Based on the current cumulative loss deviation of the charging pile, the location distribution entropy, and the location risk skewness coefficient, the baseline remaining life of the charging gun is corrected. The correction formula for the baseline remaining lifespan of the charging gun is as follows: ; In the formula, RUL represents the corrected remaining lifespan of the charging gun. The remaining life of the charging gun is the baseline lifespan based on the standard insertion and removal life type test, provided by the type test report at the time of manufacture. Lr is the historical average loss deviation rate, obtained by dividing the current cumulative loss deviation of the charging pile by the service time of the charging gun, where the service time of the charging gun is the total number of insertions and removals that the charging gun has undergone. This is a comprehensive correction coefficient based on the location distribution pattern. It is a dimensionless continuous value, determined by bilinear interpolation on the HS two-dimensional plane. Specifically, when H is greater than a first preset entropy threshold and S is less than a first preset skewness threshold... When H is less than the second preset entropy threshold and S is greater than the second preset skewness threshold, In other cases, .

[0038] In its specific implementation, step S500 uses the remaining lifespan correction result of the charging gun to provide a charging gun maintenance and replacement warning; specifically, it includes: Step S501: Preset the remaining lifespan threshold. When the corrected remaining lifespan of the charging gun is greater than or equal to the remaining lifespan threshold, maintain normal monitoring status and store the charging data and the loss deviation of the charging gun caused by the charging stop event during the current charging process into the historical database for continuous calibration. Step S502: When the remaining lifespan of the corrected charging gun is less than the remaining lifespan threshold, generate a charging gun replacement warning notification and push it.

[0039] It should be noted that the weights and thresholds not specifically described in this embodiment are determined as follows: Power time-series data, access battery attribute data, and user operation behavior time-series data are synchronously collected during the actual operation of 20,000 charging piles. The deviation, deviation feature vector, and deviation type judgment result output by the support vector machine classifier for each sampling point are calculated according to the aforementioned steps. Furthermore, the corrected power change curve, multi-dimensional position feature vector, user urgency of stopping, loss deviation, cumulative loss deviation, position distribution entropy, and position risk skewness coefficient are calculated to obtain the corrected remaining lifespan of the charging gun. Simultaneously, the judgment results of whether abnormal events such as charging gun contact welding, insulation breakdown, excessive contact resistance, or excessive temperature rise actually occurred under the corresponding operating conditions for 20,000 sets of charging piles are obtained, as well as the total number of insertions and removals when the charging gun actually reaches the end of its lifespan. The calculation results of various indicators and corresponding abnormal event judgment results of 20,000 sets of running processes were imported into the MATLAB Curve Fitting Toolbox fitting software. Binary logistic regression was used to perform maximum likelihood estimation fitting for each single indicator. The values ​​of unspecified weights and thresholds were determined with the goal of maximizing the determination coefficient on the validation set.

[0040] Example 2 This embodiment provides a charging gun life assessment system based on charging pile charging analysis, referring to... Figure 4 ,include: The data acquisition module is used to synchronously collect power time-series data, access battery attribute data, and user operation behavior time-series data for each charging process through the power metering module, battery management system communication module, and user interaction behavior monitoring module deployed in the charging pile. The power deviation correction module is used to make differential judgments on power output stability based on the access battery attribute data and power time series data, through matching the standard curve template of the same battery type and classifying and identifying the deviation type, and correct the power change curve according to the deviation type. The loss deviation correction module is used to extract the multi-dimensional position features of the charging stop time point on the correction curve based on the corrected power change curve and user operation behavior time series data, and analyze the user's urgency to stop; based on the multi-dimensional position features and the user's urgency to stop, the loss deviation of the charging gun caused by the charging stop event is quantified. The lifespan correction module is used to correct the remaining lifespan of the charging gun based on the wear deviation caused by the charging gun stopping the charging event. The charging gun maintenance module is used to provide early warnings for charging gun maintenance and replacement based on the remaining lifespan correction results of the charging gun.

[0041] The parameters and steps of each unit module in the charging gun life assessment system based on charging pile charging analysis of the present invention described above can be referred to the parameters and steps in the embodiments of the charging gun life assessment method based on charging pile charging analysis above, and will not be repeated here.

[0042] The various embodiments in this invention are described in a progressive manner. Similar or identical parts between embodiments can be referred to interchangeably. Each embodiment focuses on its differences from other embodiments. In particular, the embodiments for IoT devices and media are relatively simple in description because they are fundamentally similar to the method embodiments; relevant parts can be referred to the descriptions in the method embodiments.

[0043] The systems, media, and methods provided in the embodiments of the present invention are in one-to-one correspondence. Therefore, the systems and media also have similar beneficial technical effects as their corresponding methods. Since the beneficial technical effects of the methods have been described in detail above, the beneficial technical effects of the systems and media will not be repeated here.

[0044] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied 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.

[0045] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, as well as 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, generate instructions 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.

[0046] 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.

[0047] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0048] 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.

[0049] Computer-readable media include both permanent and non-permanent, removable and non-removable media that can store information by 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.

[0050] It should also be noted that the terms include, encompass, or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Without further limitations, the inclusion of an element by a statement that includes a… does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes the element.

[0051] The above are merely embodiments of the present invention and are not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of the present invention should be included within the scope of the claims of the present invention.

Claims

1. A method for evaluating the lifespan of a charging gun based on charging pile charging analysis, characterized in that, Includes the following steps: Step S100: Through the power metering module, battery management system communication module and user interaction behavior monitoring module deployed on the charging pile, the power time series data of each charging process, the access battery attribute data and user operation behavior time series data are collected synchronously. Step S200: Based on the access battery attribute data and power time series data, the power output stability is differentiated by matching the standard curve template of the same battery type and classifying and identifying the deviation type, and the power change curve is corrected according to the deviation type. Step S300: Based on the corrected power change curve and user operation behavior time series data, extract the multi-dimensional position features of the charging stop time point on the corrected curve, and analyze the user's urgency to stop charging. Based on multi-dimensional location characteristics and the urgency of the user's charging stop, the loss deviation of the charging gun caused by the charging stop event is quantified; Step S400: Based on the loss deviation caused by the charging gun stopping the charging event, the remaining lifespan of the charging gun is corrected. Step S500: Based on the remaining lifespan correction result of the charging gun, issue a warning for charging gun maintenance and replacement.

2. The charging gun life assessment method based on charging pile charging analysis according to claim 1, characterized in that, In step S100, the power time series data includes the charging power time series, the start time of each charge, and the user request to stop time; the accessed battery attribute data includes the battery type, the battery rated capacity, the current state of charge of the battery, and the target state of charge; the user operation behavior time series data includes the timestamp sequence of the user viewing the charging status, the trigger time of the vehicle unlock signal, and the timestamp of the charging gun button pressing signal.

3. The charging gun life assessment method based on charging pile charging analysis according to claim 2, characterized in that, Step S200 specifically includes the following: Step S201: Based on the battery type, the current state of charge of the battery and the target state of charge, retrieve the standard power change curve template and the corresponding confidence interval bandwidth of the same type of battery in the same state of charge range from the historical database. Step S202: Compare the current charging power time series with the standard power change curve template point by point, and calculate the deviation at each sampling point; the deviation at each sampling point is defined as the difference between the current charging power and the standard power at each sampling point divided by the confidence interval bandwidth. Step S203: Based on the deviation at each sampling point, construct the deviation sequence of the current charging process; perform time series feature analysis on the deviation sequence to obtain the mean deviation, standard deviation, drift rate, and peak deviation factor of the current charging process. Step S204: Combine the mean deviation, standard deviation, drift rate, and peak deviation factor of the current charging process to form the deviation feature vector of the current charging process; input the deviation feature vector of the current charging process into a pre-trained support vector machine classifier for deviation type identification; Step S205: The pre-trained support vector machine classifier outputs the probability distribution of each deviation type, takes the deviation type corresponding to the highest probability as the deviation type determination result of the current charging process, and corrects the power change curve according to the deviation type.

4. The charging gun life assessment method based on charging pile charging analysis according to claim 3, characterized in that, The specific construction process of the pre-trained support vector machine classifier is as follows: Obtain a historical charging dataset, in which each data point contains a sequence of deviations with known deviation type labels and a corresponding deviation feature vector. The deviation type labels include systemic deviation, fluctuation deviation, oscillation deviation, and severe anomaly. Using the deviation feature vector of the historical charging dataset as the input feature and the corresponding deviation type label as the classification target, a multi-class support vector machine model is constructed using a one-to-one strategy. The radial basis kernel function is selected as the kernel function, and the penalty parameter and kernel parameter are determined through grid search and cross-validation. Based on the historical charging dataset, the deviation feature vector is used as the input of the support vector machine model, and the deviation type label is used as the output target of the support vector machine model. The support vector machine model is trained, and the corresponding penalty parameter and kernel parameter that maximize the average classification accuracy of cross-validation are determined by grid search, thus obtaining the initial support vector machine classifier. The initial support vector machine classifier is validated using a reserved validation set. The classification accuracy on the validation set is calculated, and the initial support vector machine classifier whose classification accuracy on the validation set is greater than or equal to a preset accuracy threshold is output as the support vector machine classifier.

5. The charging gun life assessment method based on charging pile charging analysis according to claim 4, characterized in that, The specific process for correcting the power change curve based on the type of deviation is as follows: When the deviation type of the current charging process is a systemic deviation, obtain the difference between the actual charging power of each sampling point in the current charging power time series and the charging power at the corresponding position in the standard power change curve template; use the median of the corresponding differences of all sampling points as the translation correction amount of the current charging process. Subtract the translation correction amount of the current charging process from the actual charging power of each sampling point in the current charging power time series, and use the corrected charging power of the corresponding sampling point as the result, to obtain the corrected power change curve of the current charging process. When the deviation type of the current charging process is a fluctuating deviation, isolated abnormal points with deviations exceeding the preset threshold are marked and removed, and local smoothing correction is performed using cubic spline interpolation to obtain the corrected power change curve of the current charging process. When the deviation type of the current charging process is an oscillating deviation, the continuous oscillation interval in the deviation quantity sequence is smoothed and corrected by using a sliding window mean filter with a fixed window length to obtain the corrected power change curve of the current charging process. When the deviation type of the current charging process is severe abnormality, the charging data is marked as invalid data, will not be included in subsequent life assessments, and the life assessment process for the current charging process will be terminated.

6. The charging gun life assessment method based on charging pile charging analysis according to claim 5, characterized in that, Step S300 specifically includes: Step S301: Based on the power change curve corrected in step S205, identify the peak power of the curve and the corresponding occurrence time and minimum power, wherein the minimum power is the power value corresponding to the charging start time. Step S302: When the user issues a stop request, extract the multi-dimensional position feature vector of the stop time point on the corrected power change curve. The multi-dimensional position feature vector includes the relative power position, normalized time position, normalized power change rate, normalized curve curvature, and stage label encoding. Step S303: Calculate the urgency of the user to stop based on the time sequence data of user operation behavior; Step S304: Based on the multi-dimensional location feature vector and the user's urgency to stop, quantify the loss deviation of the charging gun caused by the charging stop event during the current charging process; the loss deviation is obtained by multiplying the base loss coefficient and the loss amplification factor.

7. The charging gun life assessment method based on charging pile charging analysis according to claim 6, characterized in that, Step S400 specifically includes: Step S401: Retrieve the sequence of loss deviations corresponding to all historical charging stop events prior to this charging from the historical database; Step S402: Based on the loss deviation sequence, analyze the cumulative loss deviation of the current charging pile; Step S403: Divide the [0,1] interval into K sub-intervals at equal intervals, and count the frequency of the base value loss coefficient of all historical charging stop events falling into each sub-interval; Step S404: Based on the frequency of the base loss coefficient falling into each sub-interval, analyze the current location distribution entropy and location risk skewness coefficient of the charging pile; Step S405: Based on the current cumulative loss deviation of the charging pile, the location distribution entropy and the location risk skewness coefficient, the baseline remaining life of the charging gun is corrected.

8. The charging gun life assessment method based on charging pile charging analysis according to claim 7, characterized in that, In step S500, a charging gun maintenance and replacement warning is issued based on the remaining lifespan correction result of the charging gun. Specifically, it includes: Step S501: Preset the remaining lifespan threshold. When the corrected remaining lifespan of the charging gun is greater than or equal to the remaining lifespan threshold, maintain normal monitoring status and store the charging data and the loss deviation of the charging gun caused by the charging stop event during the current charging process into the historical database for continuous calibration. Step S502: When the remaining lifespan of the corrected charging gun is less than the remaining lifespan threshold, generate a charging gun replacement warning notification and push it.

9. A charging gun life assessment system based on charging pile charging analysis, implemented according to any one of claims 1-8, characterized in that, The system includes: The data acquisition module is used to synchronously collect power time-series data, access battery attribute data, and user operation behavior time-series data for each charging process through the power metering module, battery management system communication module, and user interaction behavior monitoring module deployed in the charging pile. The power deviation correction module is used to make differential judgments on power output stability based on the access battery attribute data and power time series data, through matching the standard curve template of the same battery type and classifying and identifying the deviation type, and correct the power change curve according to the deviation type. The loss deviation correction module is used to extract the multi-dimensional position features of the charging stop time point on the correction curve based on the corrected power change curve and user operation behavior time series data, and analyze the user's urgency to stop; based on the multi-dimensional position features and the user's urgency to stop, the loss deviation of the charging gun caused by the charging stop event is quantified. The lifespan correction module is used to correct the remaining lifespan of the charging gun based on the wear deviation caused by the charging gun stopping the charging event. The charging gun maintenance module is used to provide early warnings for charging gun maintenance and replacement based on the remaining lifespan correction results of the charging gun.

Images