A charging pile fault diagnosis and safety protection control method and system

Abstract

The application provides a charging pile fault diagnosis and safety protection control method and system, relates to the field of charging piles, and solves the technical problem that the prior art cannot effectively distinguish between transient high-resistance connection faults and normal surge currents in the charging starting stage, which easily leads to fault misreporting or false reporting and is difficult to realize early safety warning with high reliability. The method comprises the following steps: collecting charging gun charging information in the charging starting stage timing at a first preset frequency; based on the charging information, calculating the contact resistance sequence of the charging gun connection loop through a resistance identification model; extracting a first feature vector, a second feature vector and a third feature vector, and inputting them into a preset multi-level weighted fuzzy decision tree for timing fusion analysis, and outputting a comprehensive fault confidence; comparing the comprehensive fault confidence with a preset confidence threshold, and performing safety protection control on the charging pile. The application is used in the charging pile fault diagnosis and safety protection process.

Description

Technical Field

[0001] This application relates to the field of charging piles, and in particular to a method and system for fault diagnosis and safety protection control of charging piles. Background Technology

[0002] During DC fast charging of electric vehicles, the quality of electrical contact between the charging gun and the vehicle socket directly determines charging safety. When the main relay closes, the system typically experiences a brief current surge, a normal phenomenon caused by the charging of the vehicle-side bus capacitor or the pre-charging circuit, characterized by high amplitude, short duration, and repeatability. However, if the charging connection interface has defects such as oxidation, contamination, or mechanical loosening, a momentary high-resistance connection fault may occur in the initial stage of current flow: at this time, the contact resistance abnormally increases, which, although not immediately triggering an arc, will generate significant Joule heating locally, becoming a potential overheating or even fire hazard.

[0003] Existing charging stations generally rely on overcurrent protection or fixed temperature thresholds for fault diagnosis. However, these methods struggle to effectively distinguish between these two fundamentally different transient processes: setting the protection threshold too low to detect high-resistance faults can easily misjudge normal surges as faults, leading to frequent false trips and impacting user experience; conversely, raising the threshold to avoid false alarms may miss early high-resistance degradation, resulting in a lost early warning window. Especially in the first 1–3 seconds of charging, current, voltage, and temperature signals are highly coupled and undergo dramatic dynamic changes. Traditional monitoring strategies based on steady-state conditions or single parameters lack a deep understanding of fault mechanisms, failing to achieve a balance between high sensitivity and low false alarm rate.

[0004] Therefore, there is an urgent need for a technical solution that can accurately distinguish between normal surge current and instantaneous high-resistance connection faults during the initial transient phase of charging through multi-physical quantity fusion and dynamic feature identification, in order to support reliable and intelligent early safety warning and graded protection control. Summary of the Invention This application provides a method and system for fault diagnosis and safety protection control of charging piles, which solves the technical problem that the existing technology cannot effectively distinguish between instantaneous high-resistance connection faults and normal surge currents at the beginning of charging, which easily leads to missed or false fault reports and makes it difficult to achieve high-reliability early safety warnings.

[0005] To achieve the above objectives, this application adopts the following technical solution: Firstly, a method for fault diagnosis and safety protection control of charging piles is provided, including: The charging information of the charging gun during the charging start-up phase is collected at a first preset frequency. The charging information includes the output voltage, charging current, DC bus voltage inside the charging pile, and charging gun connector temperature. Based on the output voltage, charging current, and DC bus voltage inside the charging pile, the contact resistance sequence of the charging gun connection circuit is calculated using a resistance identification model. A first feature vector is extracted from the contact resistance sequence, a second feature vector is extracted from the charging current, and a third feature vector is extracted from the charging gun connector temperature; the first feature vector includes the initial rate of change and fluctuation variance of the contact resistance sequence in a preset time interval, the second feature vector includes the rising slope and amplitude envelope area of ​​the charging current waveform in the preset time interval, and the third feature vector is the initial temperature rise rate in the preset time interval. The first feature vector, the second feature vector, and the third feature vector are input into a preset multi-level weighted fuzzy decision tree for time-series fusion analysis; wherein, the multi-level weighted fuzzy decision tree assigns different decision weights to different feature vectors based on different sub-time periods of the charging start stage, and outputs a comprehensive fault confidence score. The overall fault confidence level is compared with a preset confidence threshold. If the overall fault confidence level is greater than or equal to the confidence threshold, a connection abnormality fault is determined to exist, and a fault protection command is generated. The fault protection command is executed to perform safety protection control on the charging pile.

[0006] Based on the above technical solutions, the technical problem urgently needs to be solved in the charging pile fault diagnosis and safety protection control method provided in this application. This is because, in the initial stage of DC fast charging, instantaneous high-resistance connection faults (such as contact oxidation or loosening) and normal surge currents caused by vehicle-side capacitor charging are highly similar in current or temperature performance. Traditional overcurrent / overtemperature protection based on fixed thresholds cannot distinguish between the two, leading to either missed detection of high-resistance faults causing overheating or even fire risks, or misjudgment of surges causing frequent tripping, affecting safety and user experience. This solution constructs a resistance identification model by synchronously collecting output voltage, bus voltage, current, and connector temperature, accurately inverting the contact resistance sequence, fundamentally bypassing the misjudgment trap of relying solely on current amplitude; furthermore, it extracts multi-dimensional early features such as contact resistance change rate and fluctuation variance, current rise slope and envelope area, and initial temperature rise rate, and uses a multi-level weighted fuzzy decision tree to dynamically fuse them according to time segments. In the early stage, it focuses on current waveform identification of surge characteristics, and in the later stage, it focuses on resistance and temperature rise to capture deterioration trends. This design enables a physical distinction between the two types of transient processes, significantly improving the sensitivity and anti-interference capability of early fault identification. It can provide timely warnings of hidden high-resistance faults and avoid malfunctions in response to normal surges, thereby improving system reliability and user experience while ensuring charging safety.

[0007] In conjunction with the first aspect above, in one possible implementation, the calculation of the contact resistance sequence of the charging gun connection circuit is constructed based on the sliding squares method, including: For the current moment During the time window Internally, by minimizing the objective function To solve for the contact resistance sequence; ; in, For dummy variables in the summation, For the width of the sliding window, The regularization coefficient is . for The voltage drop in the charging gun connection circuit at all times. for The charging current at any given moment, for Contact resistance at any given time The contact resistance The rate of change; By analyzing the objective function Differentiate and set it to zero to obtain the contact resistance. The estimated value.

[0008] In conjunction with the first aspect above, in one possible implementation, the regularization coefficient... ,include: The regularization coefficient According to the charging current Dynamic adjustment of amplitude: when When the amplitude is lower than the first current threshold, a preset first current threshold is used. Value; when When the amplitude is higher than or equal to the first current threshold, a preset second current threshold is used. Value; of which, the second Value less than the first value.

[0009] In conjunction with the first aspect above, in one possible implementation, the multi-level weighted fuzzy decision tree includes at least two levels of decision points, including: The first-level decision node, corresponding to the first sub-time period of the charging start stage, has its decision weights configured as follows: the weight of the second feature vector is greater than the weights of the first feature vector and the third feature vector. The second-level decision node corresponds to the second sub-time period of the charging start stage. It starts after the first sub-time period, and its decision weight is configured as follows: the weights of the first feature vector and the third feature vector are greater than the weight of the second feature vector. Wherein, the sum of the first sub-time period and the second sub-time period is equal to the preset time interval.

[0010] In conjunction with the first aspect described above, in one possible implementation, extracting the first feature vector from the contact resistance sequence further includes: Perform wavelet transform on the contact resistance sequence, calculate its wavelet energy within a preset fault characteristic frequency band, and use the wavelet energy as a component of the first feature vector.

[0011] In conjunction with the first aspect above, in one possible implementation, after determining that a connection anomaly fault exists, a fault classification step is further included: If the overall fault confidence level is within the first confidence level range, a first-level alarm command and a power adjustment command are generated. If the overall fault confidence level is in a second confidence level range that is higher than the first confidence level range, then a level 2 alarm command and a cut-off command are generated.

[0012] In conjunction with the first aspect above, in one possible implementation, the first-level alarm instruction is used to trigger an audible and visual alarm and record the event; the power adjustment instruction is used to control the charging pile to reduce the output power to below the rated power; the second-level alarm instruction is used to trigger a higher-frequency audible and visual alarm and record the fault; and the output cut-off instruction is used to control the main relay of the charging pile to disconnect.

[0013] In conjunction with the first aspect above, in one possible implementation, prior to collecting the charging information, the following steps are included: In response to the charging handshake completion signal, the main relay of the charging pile is closed, and the charging start-up phase monitoring mode is started. After entering the charging start-up phase monitoring mode, charging information is collected; wherein, the total time of the charging start-up phase monitoring mode is preset to 2-3 seconds.

[0014] In conjunction with the first aspect above, in one possible implementation, the determination that there is no connection anomaly includes: After determining that there is no connection abnormality, the system exits the charging start-up monitoring mode and switches to the regular charging monitoring mode based on fixed temperature and current thresholds.

[0015] Secondly, this application provides a charging pile fault diagnosis and safety protection control system, including: a data acquisition module, a fault analysis module, and an early warning module; wherein, the data acquisition module is used to acquire charging information of the charging gun during the charging start-up phase at a first preset frequency; the fault analysis module is used to calculate the contact resistance sequence of the charging gun connection circuit based on the output voltage, charging current, and DC bus voltage inside the charging pile, through a resistance identification model; extract a first feature vector from the contact resistance sequence, extract a second feature vector from the charging current, and extract a third feature vector from the charging gun connector temperature; input the first feature vector, the second feature vector, and the third feature vector into a preset multi-level weighted fuzzy decision tree for time-series fusion analysis, and output a comprehensive fault confidence level; the early warning module is used to compare the comprehensive fault confidence level with a preset confidence level threshold, execute the fault protection command, and perform safety protection control on the charging pile.

[0016] Thirdly, a charging pile fault diagnosis and safety protection control device is provided, comprising: a communication unit and a processing unit; the communication unit is used to collect charging information of the charging gun during the charging start-up phase at a first preset frequency; the processing unit is used to calculate the contact resistance sequence of the charging gun connection circuit based on the output voltage, charging current, and DC bus voltage inside the charging pile, using a resistance identification model; extract a first feature vector from the contact resistance sequence, extract a second feature vector from the charging current, and extract a third feature vector from the charging gun connector temperature; input the first feature vector, the second feature vector, and the third feature vector into a preset multi-level weighted fuzzy decision tree for time-series fusion analysis, and output a comprehensive fault confidence score; compare the comprehensive fault confidence score with a preset confidence score threshold, execute the fault protection command, and perform safety protection control on the charging pile.

[0017] Fourthly, this application provides a charging pile fault diagnosis and safety protection control device, comprising: a processor and a storage medium; the storage medium includes instructions, and the processor is used to execute the instructions to implement the method described in the first aspect and any possible implementation thereof. This charging pile fault diagnosis and safety protection control device can be an electronic device or a chip within an electronic device.

[0018] Fifthly, this application provides a computer-readable storage medium storing instructions that, when executed on a charging pile fault diagnosis and safety protection control device, cause the charging pile fault diagnosis and safety protection control device to perform the method described in the first aspect and any possible implementation thereof.

[0019] Sixthly, this application provides a computer program product containing instructions that, when the computer program product is run on a charging pile fault diagnosis and safety protection control device, causes the charging pile fault diagnosis and safety protection control device to perform the method described in the first aspect and any possible implementation of the first aspect.

[0020] This application provides a method and system for fault diagnosis and safety protection control of charging piles, which can achieve high-precision, early identification and intelligent response to connection abnormality faults in the initial charging stage. Addressing the problem that existing technologies struggle to distinguish between instantaneous high-resistance connection faults and normal surge currents, this solution simultaneously collects the charging gun output voltage, DC bus voltage, charging current, and connector temperature to construct a resistance identification model, accurately retrieving the contact resistance sequence. It then extracts three types of physically meaningful early feature vectors: the initial rate of change and variance of contact resistance, the rising slope and amplitude envelope area of ​​the current waveform, and the initial temperature rise rate. These features are input into a multi-level weighted fuzzy decision tree, dynamically adjusting the decision weights of each feature according to different sub-time periods in the initial charging stage, fully conforming to the fault evolution law and significantly improving the robustness and accuracy of diagnosis. Furthermore, the system implements graded protection based on comprehensive fault confidence: reducing power and issuing an alarm during medium-risk periods, and immediately cutting off output during high-risk periods, balancing safety and availability. This method requires no additional hardware and can complete early warning within 2-3 seconds using only the existing sensor data of the charging pile. It effectively prevents overheating, arcing or fire accidents caused by poor contact, and greatly improves charging safety, system reliability and user satisfaction.

[0021] It should be understood that the descriptions of technical features, technical solutions, beneficial effects, or similar language in this application do not imply that all features and advantages can be achieved in any single embodiment. Rather, it is understood that the description of a feature or beneficial effect means that a specific technical feature, technical solution, or beneficial effect is included in at least one embodiment. Therefore, the descriptions of technical features, technical solutions, or beneficial effects in this specification do not necessarily refer to the same embodiment. Furthermore, the technical features, technical solutions, and beneficial effects described in this embodiment can be combined in any suitable manner. Those skilled in the art will understand that embodiments can be implemented without one or more specific technical features, technical solutions, or beneficial effects of a particular embodiment. In other embodiments, additional technical features and beneficial effects may be identified in specific embodiments that do not embody all embodiments. Attached Figure Description

[0022] Figure 1 A system architecture diagram of a charging pile fault diagnosis and safety protection control system provided in this application embodiment; Figure 2A flowchart illustrating a charging pile fault diagnosis and safety protection control method provided in an embodiment of this application; Figure 3 This is a schematic diagram of the structure of a control device provided in an embodiment of this application; Figure 4 This is a schematic diagram of the hardware structure of a control device provided in an embodiment of this application. Detailed Implementation

[0023] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. 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.

[0024] The charging pile fault diagnosis and safety protection control method provided in this application embodiment can be applied to a charging pile fault diagnosis and safety protection control system, such as... Figure 1 As shown, the communication system includes: a data acquisition module, a fault analysis module, and an early warning module; The data acquisition module is used to acquire charging information of the charging gun during the charging start-up phase at a first preset frequency. The fault analysis module is used to calculate the contact resistance sequence of the charging gun connection circuit based on the output voltage, charging current, and DC bus voltage inside the charging pile, using a resistance identification model. The first feature vector is extracted from the contact resistance sequence, the second feature vector is extracted from the charging current, and the third feature vector is extracted from the charging gun connector temperature. The first feature vector, the second feature vector, and the third feature vector are input into a preset multi-level weighted fuzzy decision tree for time series fusion analysis, and the comprehensive fault confidence is output. The early warning module is used to compare the comprehensive fault confidence level with the preset confidence level threshold, execute fault protection commands, and perform safety protection control on the charging pile.

[0025] To address the technical problem that existing technologies cannot effectively distinguish between instantaneous high-resistance connection faults and normal surge currents during the initial charging phase, leading to missed or false alarms and hindering the achievement of highly reliable early safety warnings, this application provides a charging pile fault diagnosis and safety protection control method. The method includes: collecting charging gun charging information during the initial charging phase at a first preset frequency; the charging information includes output voltage, charging current, internal DC bus voltage of the charging pile, and charging gun connector temperature; calculating the contact resistance sequence of the charging gun connection circuit based on the output voltage, charging current, and internal DC bus voltage of the charging pile using a resistance identification model; extracting a first feature vector from the contact resistance sequence, a second feature vector from the charging current, and a third feature vector from the charging gun connector temperature; the first feature vector includes the initial rate of change and fluctuation variance of the contact resistance sequence within a preset time interval, and the second feature vector includes the charging current fluctuation... The first, second, and third feature vectors are input into a preset multi-level weighted fuzzy decision tree for time-series fusion analysis. The multi-level weighted fuzzy decision tree assigns different decision weights to different feature vectors based on different sub-time periods during the initial charging phase and outputs a comprehensive fault confidence score. The comprehensive fault confidence score is compared with a preset confidence threshold. If the comprehensive fault confidence score is greater than or equal to the confidence threshold, a connection anomaly fault is determined, and a fault protection command is generated. The fault protection command is executed to provide safety protection control for the charging pile. Based on this, this application constructs a contact resistance identification model by frequently acquiring multi-source electrical and thermal signals during the initial charging phase and extracting three types of feature vectors reflecting early degradation, effectively overcoming the shortcomings of traditional methods that rely on a single threshold and struggle to distinguish between surges and real faults. Combined with a multi-level weighted fuzzy decision tree, feature weights are dynamically allocated according to different time periods, achieving high sensitivity and low false alarm identification of instantaneous high-resistance connection faults. Ultimately, by triggering graded protection based on comprehensive fault confidence, it can provide timely warnings of potential overheating risks and avoid malfunctions, significantly improving the safety, reliability, and user experience of charging piles.

[0026] like Figure 2 As shown in the embodiment of this application, a method for fault diagnosis and safety protection control of a charging pile includes: S201. After the charging handshake is completed, the main relay of the charging pile is closed and the charging start-up phase monitoring mode is started. The charging gun output voltage, charging current, internal DC bus voltage of the charging pile and the charging gun connector temperature are collected at the first preset frequency during the charging start-up phase.

[0027] The total time for the initial charging phase monitoring mode is preset to 2-3 seconds.

[0028] In some implementations, in response to the charging handshake completion signal, the control system controls the main relay of the charging pile to close and simultaneously initiates the charging start-up phase monitoring mode. In this monitoring mode, key operating parameters are continuously sampled within a time window of the charging start-up phase (e.g., 0-30 seconds after the main relay closes) at a first preset frequency (e.g., 100Hz or 500Hz). The collected data includes: charging gun output voltage, charging current, DC bus voltage inside the charging pile, and charging gun connector temperature. This multi-source timing data provides the basic input for subsequent contact resistance identification and early fault feature extraction.

[0029] It should be noted that the "charging start-up phase" specifically refers to the initial charging period after the main relay closes. This phase is the most sensitive window for the dynamic response of the contact interface to electrothermal activity. Its start time is based on the actual closure of the main relay, not on the completion of the communication handshake. Although the handshake completion is a prerequisite for relay closure, to ensure that the collected data reflects the true dynamic process of power-on, the start of the monitoring mode and data acquisition are strictly synchronized with the main relay closure event. In addition, the first preset frequency should be high enough to accurately capture key transient characteristics such as the current rise edge, voltage transients, and the initial slope of temperature rise, typically not lower than 50Hz.

[0030] It should also be noted that in this application, the total time for the initial charging phase monitoring mode is preset to 2-3 seconds, mainly based on the physical response characteristics of early signs of contact faults during DC charging of electric vehicles. When the main relay closes and a large current is suddenly applied to the charging circuit, if there is poor contact (such as oxidation, loosening, or contamination), the contact resistance will change rapidly within 1-2 seconds after current flow, and the rising edge of the current waveform will also be completed within this window. Although the absolute value of the connector temperature changes only slightly, its initial temperature rise rate can effectively reflect the risk of local overheating. Experiments and engineering practice show that 2-3 seconds is sufficient to fully capture key early characteristics such as the initial rate of change and fluctuation variance of contact resistance, the rising slope and amplitude envelope area of ​​the current, and the temperature rise rate, while avoiding the introduction of steady-state interference or delay in protection action due to an excessively long monitoring window. Therefore, this time range takes into account fault sensitivity, real-time diagnostics, and system reliability, and is the optimal observation window for achieving high-precision early fault warning.

[0031] S202. Based on the output voltage, charging current, and DC bus voltage inside the charging pile, the contact resistance sequence of the charging gun connection circuit is calculated using a resistance identification model.

[0032] It should be noted that the contact resistance is not obtained by direct measurement, but is the equivalent contact impedance indirectly identified through multiple electrical quantities. Its accuracy depends on the synchronous sampling performance of the voltage / current sensor and the model's ability to compensate for parasitic parameters of the line (such as cable resistance and inductance). In addition, to avoid calculation distortion under zero current or extremely low operating conditions, effective resistance identification is only performed during the period when the absolute value of the charging current is greater than a preset threshold (e.g., 10 A). The contact resistance values ​​in other periods are weighted by confidence or marked as invalid to ensure that the generated contact resistance sequence has engineering usability and fault sensitivity.

[0033] S203. Extract the first feature vector from the contact resistance sequence, extract the second feature vector from the charging current, and extract the third feature vector from the charging gun connector temperature.

[0034] The first feature vector includes the initial rate of change and fluctuation variance of the contact resistance sequence within a preset time interval; the second feature vector includes the rising slope and amplitude envelope area of ​​the charging current waveform within a preset time interval; and the third feature vector is the initial temperature rise rate within the preset time interval.

[0035] In some implementations, a first feature vector is extracted from the contact resistance sequence, a second feature vector is extracted from the charging current waveform, and a third feature vector is extracted from the charging gun connector temperature data. Specifically, the first feature vector consists of the initial rate of change of the contact resistance sequence within a preset time interval (e.g., 0–2.5 seconds after the main relay closes) (i.e., the average slope of the resistance value within the first 500 milliseconds) and the fluctuation variance (the standard deviation or peak-to-peak value reflecting the instability of the contact interface); the second feature vector includes the rising slope of the charging current within the same time interval (the rate at which the current rises from 10% to 90% of the target value) and the amplitude envelope area (the integral area enclosed by the current waveform and the time axis, characterizing the energy injection characteristics); the third feature vector is defined as the initial temperature rise rate of the connector temperature in the initial segment of the preset time interval (e.g., 0–1 second), used to capture the early trend of local overheating.

[0036] It should be noted that the three feature vectors mentioned above focus on the transient response process at the beginning of charging, rather than steady-state operating data, as they are more sensitive to early faults such as contact degradation and loose connections. Furthermore, although the calculation window for each feature is collectively referred to as the "preset time interval," it can be fine-tuned according to the response speed of the physical quantity; for example, a shorter window (0–1 second) is used for the temperature rise rate, while the current envelope area covers the entire rise process (0–2 seconds) to maximize feature discrimination capability. In addition, all feature extraction is performed at a high sampling frequency (e.g., ≥100Hz) to ensure accurate capture of rapid dynamic processes and avoid feature distortion due to undersampling.

[0037] S204. Input the first feature vector, the second feature vector, and the third feature vector into a preset multi-level weighted fuzzy decision tree for time series fusion analysis.

[0038] Among them, the multi-level weighted fuzzy decision tree assigns different decision weights to different feature vectors based on different sub-time periods of the initial charging stage, and outputs a comprehensive fault confidence score.

[0039] In some implementations, the first, second, and third feature vectors are input into a pre-defined multi-level weighted fuzzy decision tree for time-series fusion analysis. This multi-level weighted fuzzy decision tree divides the initial charging phase (e.g., 0–3 seconds) into multiple consecutive sub-time periods (e.g., 0–1 seconds, 1–2 seconds, 2–3 seconds), each sub-time period corresponding to different fault evolution sensitivity characteristics. Within each sub-time period, the decision tree dynamically assigns corresponding decision weights to the three feature vectors based on prior knowledge or historical fault data. For example, within 0–1 seconds, the abrupt change in contact resistance is most significant, so the first feature vector is given a higher weight; while within 2–3 seconds, the temperature rise effect gradually becomes prominent, and the weight of the third feature vector is correspondingly increased. The fuzzy rules at each level infer the weighted features based on membership functions, and finally, through defuzzification, output a comprehensive fault confidence level between 0 and 1, used to quantify the possibility of contact abnormalities in the connector.

[0040] It should be noted that the "multi-level" structure is reflected not only in the segmentation of the time dimension but also in the hierarchical nature of the decision-making logic: lower-level nodes handle feature fusion within a single time period, while higher-level nodes integrate confidence trends across time periods to suppress transient interference. Furthermore, the decision weights are not fixed values ​​but can be adaptively adjusted based on the charging pile model, rated current level, or ambient temperature, thereby improving the model's generalization ability under different operating conditions. This fusion mechanism effectively overcomes the shortcomings of single features being susceptible to noise interference or having limited applicability, significantly enhancing the robustness and accuracy of early fault identification.

[0041] S205. Compare the overall fault confidence level with the preset confidence level threshold. If the overall fault confidence level is greater than or equal to the confidence level threshold, it is determined that there is a connection abnormality fault, and a fault protection command is generated. Execute the fault protection command to perform safety protection control on the charging pile.

[0042] In some implementations, the overall fault confidence level is compared with a preset confidence threshold. If the overall fault confidence level is greater than or equal to the threshold (e.g., 0.75), a connection fault is determined to exist between the charging gun and the socket (such as loose contact, oxidation, or risk of localized overheating), and a corresponding fault protection command is immediately generated. This command may include: disconnecting the charging pile main relay, preventing the initiation of subsequent charging sessions, sending a stop signal to the vehicle's BMS, illuminating the local fault indicator light, or uploading alarm logs to the cloud platform. Subsequently, the charging pile control system executes the fault protection command, cuts off the high-voltage output circuit, and implements safety protection controls to prevent safety accidents such as arcing, welding, or thermal runaway.

[0043] It should be noted that the confidence threshold is not fixed, but can be dynamically adjusted based on the charging power level, ambient temperature, or historical operating data. For example, a lower threshold can be used in high-power fast charging (≥120kW) scenarios to improve sensitivity, while the threshold can be appropriately increased in low-temperature environments to avoid misjudgments caused by slow temperature rise. In addition, to prevent malfunctions caused by transient interference, the system can also introduce a confidence duration criterion (such as requiring the confidence to continuously exceed the threshold for more than 200ms) or a multi-cycle consistency verification mechanism, thereby improving diagnostic reliability while ensuring safety.

[0044] Based on the above technical solution, the solution to this technical problem is crucial in the charging pile fault diagnosis and safety protection control method provided in this application. This is because, in the initial stage of DC fast charging, the instantaneous high-resistance connection fault caused by contact degradation and the normal surge current generated by vehicle capacitor charging exhibit similar characteristics in traditional monitoring signals (such as current or temperature). This makes it difficult for existing protection mechanisms based on fixed thresholds to effectively distinguish between them: oversensitivity leads to false tripping, affecting user experience; while oversensitivity may miss early high-resistance hazards, causing overheating or even fire risks. This solution constructs a resistance identification model by simultaneously collecting the charging gun output voltage, charging pile DC bus voltage, charging current, and connector temperature. It calculates the contact resistance sequence reflecting the actual contact state in real time and extracts multi-dimensional early features, including the initial rate of change and fluctuation variance of contact resistance, the rising slope and amplitude envelope area of ​​the current waveform, and the initial temperature rise rate. These features are input into a multi-level weighted fuzzy decision tree, dynamically adjusting the decision weights of each feature according to different sub-time periods in the initial charging stage, thereby achieving a distinction between the physical nature of surges and faults. Therefore, while ensuring high sensitivity in detecting early contact anomalies, it effectively suppresses misjudgments of normal surges, significantly improving the accuracy, safety, and reliability of fault warning and system operation.

[0045] In one possible implementation of this application embodiment, the above-mentioned S202 can be specifically implemented by the following S301, S302 and S303, which are described in detail below: S301, Regarding the current moment During the time window Internally, by minimizing the objective function To solve for the contact resistance sequence; ; in, For dummy variables in the summation, For the width of the sliding window, The regularization coefficient is . for The voltage drop in the charging gun connection circuit at all times is the voltage loss caused by the combined effects of line impedance and contact resistance. for The charging current at any given moment, for Contact resistance at any given time Contact resistance The rate of change can effectively suppress drastic fluctuations in the estimated contact resistance, and improve the smoothness and physical rationality of the identification results; It should be noted that the regularization coefficient According to the charging current Dynamic adjustment of amplitude: when When the amplitude is lower than the first current threshold, a preset first current threshold is used. Values ​​are adjusted to enhance sensitivity to subtle contact changes at low currents; when When the amplitude is higher than or equal to the first current threshold, a preset second current threshold is used. The value is adjusted to reduce the excessive constraint on normal thermal drift under high current conditions, thereby achieving more accurate and robust contact resistance estimation at different charging stages; among which, the second Value less than the first value.

[0046] S302, By adjusting the objective function Differentiate and set it to zero to obtain the contact resistance. The estimated value.

[0047] Based on the above technical solution, accurate identification of contact resistance is crucial for determining the connection status during the initial charging stage. However, this process faces two major challenges: first, drastic current fluctuations lead to severe noise interference in the voltage drop; second, the contact resistance itself changes slowly, and without constraint, spurious abrupt changes can easily occur. Traditional methods directly calculate resistance using Ohm's law, neglecting dynamic characteristics and the influence of measurement noise, making it difficult to obtain smooth and reliable estimates. Therefore, this solution employs a minimization objective function based on the sliding squares method. Solve the contact resistance sequence, treat the difference between voltage drop and current as a residual term, and introduce a regularization term. To suppress drastic fluctuations in resistance estimation and improve physical plausibility. Meanwhile, the regularization coefficient... Dynamically adjust based on charging current amplitude: Use a larger current during low current phases (such as the pre-charging period). This enhances robustness to weak disturbances; and allows for the use of smaller [sized components] during high-current phases. This design avoids over-smoothing and loss of true changes. It effectively balances model fitting accuracy and stability, achieving high-precision, low-noise contact resistance identification under different operating conditions. This significantly improves the reliability of early fault feature extraction, provides high-quality input data for subsequent multi-source fusion diagnosis, and thus enhances the overall system's safety early warning capability.

[0048] In one possible implementation of this application embodiment, the above-mentioned S203 further includes performing wavelet transform on the contact resistance sequence, calculating its wavelet energy within a preset fault characteristic frequency band, and using the wavelet energy as a component of the first feature vector.

[0049] In some implementations, the contact resistance sequence is subjected to continuous wavelet transform (CWT) or discrete wavelet transform (DWT) to extract its localized features in the time-frequency domain. Specifically, wavelet basis functions with tight support and good time-frequency focusing capabilities (such as the Daubechies 4th order wavelet or Morlet wavelet) are selected to decompose the contact resistance sequence into multiple scales within a preset time window (e.g., 0–3 seconds). Subsequently, focusing on a preset fault characteristic frequency band (e.g., 5–20 Hz, which corresponds to high-frequency resistance fluctuations caused by instability at the contact interface during the initial current flow) related to typical fault mechanisms such as contact loosening and fretting wear, the sum of squares of the wavelet coefficients at each scale within this frequency band is calculated, which is the wavelet energy. This wavelet energy is used as a new component of the first eigenvector, which, together with the original initial rate of change and fluctuation variance, constitutes a more comprehensive characterization of the contact state.

[0050] It should be noted that the "preset fault characteristic frequency band" is not a universal fixed value, but is derived through spectral analysis of historical fault data or calibration by accelerated aging tests in the laboratory. Different charging pile models or connector types may correspond to different sensitive frequency bands. In addition, to avoid noise interference causing artificially high wavelet energy, the system can apply low-pass filtering or a threshold denoising strategy to the contact resistance sequence before transformation, and perform wavelet analysis only during the time period when the charging current rises steadily and the amplitude exceeds the effective threshold, to ensure that the extracted wavelet energy truly reflects the physical fault rather than measurement disturbance. By introducing the time-frequency domain feature of wavelet energy, the ability to identify early, weak, and non-stationary contact degradation phenomena is significantly enhanced, improving the sensitivity and robustness of the overall diagnostic system.

[0051] In one possible implementation of this application embodiment, the multi-level weighted fuzzy decision tree in S204 above includes at least two levels of decision points, as follows: The first-level decision node corresponds to the first sub-time period of the charging start stage, and its decision weight configuration is as follows: the weight of the second feature vector is greater than the weights of the first feature vector and the third feature vector. The second-level decision node corresponds to the second sub-time period of the charging start stage. It starts after the first sub-time period, and its decision weight configuration is as follows: the weights of the first feature vector and the third feature vector are greater than the weight of the second feature vector. The sum of the first and second sub-time periods equals the preset time interval.

[0052] In some implementations, the multi-level weighted fuzzy decision tree includes at least two levels of time-series decision nodes, each corresponding to one of the two consecutive sub-time periods at the start of charging. The first-level decision node covers the first sub-time period (e.g., 0–1 seconds). During this period, the charging current is rapidly increasing, and its waveform characteristics are highly sensitive to abnormal loop impedance. Therefore, the decision weights for this node are configured such that the weight of the second feature vector (charging current-related features) is greater than the weights of the first feature vector (contact resistance features) and the third feature vector (temperature features). A typical weight allocation could be (first:second:third = 0.2:0.6:0.2). The second-level decision node covers the immediately following second sub-time period (e.g., 1–3 seconds). At this point, the current has stabilized, while the contact resistance begins to show significant changes due to the Joule heating effect, and the connector temperature rise rate gradually increases. Therefore, the weights for this node are adjusted such that the weights of the first and third feature vectors are greater than the weight of the second feature vector, for example, a weight allocation of (0.4:0.2:0.4). The first and second sub-time periods are consecutive and do not overlap, and their total duration is equal to the preset time interval (e.g., 3 seconds), ensuring complete coverage of the entire initial charging phase.

[0053] It should be noted that the above weighting configuration is not arbitrarily set, but is based on statistical analysis of a large amount of measured data: in the initial stage of current flow (0–1 second), the current rise slope and envelope area most sensitively reflect the initial conduction quality of the contact interface; while in the subsequent stage (1–3 seconds), the fluctuation variance of contact resistance and the temperature rise rate become key indicators for judging oxidation, loosening, or local overheating. In addition, the output confidence of the two-level nodes can be further used to generate a final comprehensive fault confidence through high-level fusion rules (such as weighted average or maximum confidence retention), thereby taking into account both transient response and steady-state trend, effectively suppressing the risk of misjudgment in a single time period, and improving the accuracy and timeliness of overall diagnosis.

[0054] Based on the above technical solution, during the initial charging stage, the sensitivity of different physical quantities to contact faults dynamically changes over time: initially (e.g., 0–1 seconds), the current rises rapidly, and its waveform characteristics (e.g., slope, envelope) can effectively reflect whether the circuit is conducting normally, while contact resistance and temperature have not yet fully responded; subsequently (e.g., 1–3 seconds), the current tends to stabilize, but the contact resistance begins to show abnormal fluctuations due to Joule heating, and the connector temperature rise rate gradually becomes a key indicator. If a fixed weight is used to fuse multi-source features, it will not be able to adapt to this time-varying sensitivity, leading to early misjudgment or missed judgment. To this end, this solution designs a multi-level weighted fuzzy decision tree, dividing the preset time interval into two continuous sub-time periods, and configuring differentiated decision weights for each level of nodes: the first level focuses on the second feature vector (current feature) to accurately identify normal surges and abnormal current distortions; the second level increases the weights of the first (contact resistance) and third (temperature rise rate) feature vectors, focusing on the thermal-electric coupling evolution of contact degradation. This timing-adaptive weighting mechanism fully aligns with the physical development of faults, avoiding false alarms caused by excessive focus on temperature rise or resistance noise in the early stages, while preventing missed alarms due to neglecting thermal effects in later stages. Thus, without increasing hardware costs, it significantly improves the early identification accuracy and robustness of instantaneous high-resistance connection faults, achieving synergistic optimization of safety warnings and user experience.

[0055] In one possible implementation of this application embodiment, S205 includes: After determining that a connection fault exists, the process also includes a fault classification step: If the overall fault confidence level is within the first confidence level range, a first-level alarm command and a power adjustment command will be generated. If the overall fault confidence level is in the second confidence level range, which is higher than the first confidence level range, then a level 2 alarm command and a cut-off command will be generated.

[0056] Among them, the first-level alarm command is used to trigger an audible and visual alarm and record the event; the power adjustment command is used to control the charging pile to reduce the output power to below the rated power; the second-level alarm command is used to trigger a higher frequency audible and visual alarm and record the fault; and the output cut-off command is used to control the main relay of the charging pile to disconnect.

[0057] It should be noted that after determining that there is no connection abnormality, the system exits the charging start-up monitoring mode and switches to the regular charging monitoring mode based on fixed temperature and current thresholds.

[0058] In some implementations, after determining that a connection anomaly exists, the system further executes a fault classification handling strategy: If the overall fault confidence level is in the first confidence level range (e.g., 0.60 ≤ confidence level < 0.85), a first-level alarm command and a power adjustment command are generated. The first-level alarm command triggers an audible and visual alarm device (e.g., a buzzer and a yellow warning light) and records the event to the local log or remote platform, while the power adjustment command controls the charging pile to dynamically limit the output power to below 50% of the rated power (e.g., reduce it to 30kW) to reduce the temperature rise at the contact interface and maintain limited charging capacity. If the overall fault confidence level is in the second confidence level range (e.g., confidence level ≥ 0.85), a second-level alarm command and an output cut-off command are generated. The second-level alarm command triggers a more urgent audible and visual alarm (e.g., a high-frequency buzzer and a red flashing light) and records a high-priority fault log, while the output cut-off command immediately controls the main relay to disconnect, forcibly terminating the charging process to prevent the risk of arcing, welding, or thermal runaway.

[0059] It should be noted that if the overall fault confidence level is lower than the preset confidence threshold (e.g., <0.60), indicating that no connection fault exists, the system will automatically exit the high-sensitivity charging initiation monitoring mode and switch to the regular charging monitoring mode. In this regular mode, basic protection relies solely on fixed temperature thresholds (e.g., connector temperature >85°C) and fixed current thresholds (e.g., overcurrent >120% of rated value), and high-frequency feature extraction and fuzzy fusion analysis are no longer performed. This reduces the system's computational load and extends the controller's lifespan while ensuring safety. This dual-mode switching mechanism of "early intelligent diagnosis + later simplified monitoring" balances the foresight of fault warning with the economic efficiency of long-term operation.

[0060] Based on the above technical solutions, the severity of connection faults during charging is continuous and uncertain, not simply a binary "normal / fault" state. If only a single threshold is used to trigger a hard disconnection, moderate degradation (such as slight oxidation) may be misjudged as an emergency fault, leading to unnecessary charging interruptions and impacting user experience. Conversely, insufficient response may allow high-risk hazards to escalate into safety incidents. Therefore, a mechanism that can dynamically classify responses based on fault confidence is urgently needed. This solution divides the system into multiple levels based on comprehensive fault confidence: when the confidence level is in the first level (medium risk), only an audible and visual alarm is triggered, and the output power is actively reduced. This limits contact point temperature rise, delays fault development, maintains limited charging capacity, and improves service continuity. When the confidence level enters the second level (high risk), the main circuit is immediately disconnected, forcibly terminating charging and eliminating the risk of fire or arcing. This tiered strategy precisely matches diagnostic results with control actions, avoiding false stops or delays caused by "one-size-fits-all" protection, and balancing safety and availability. Meanwhile, different levels of audible and visual alarms and event logging provide clear fault level information for operation and maintenance, facilitating subsequent maintenance decisions. Thus, while ensuring personal and equipment safety, the intelligence, reliability, and user satisfaction of the charging pile system are significantly improved.

[0061] The above primarily describes the solutions of the embodiments of this application from the perspective of device implementation. It is understood that each device, such as a charging pile fault diagnosis and safety protection control device, includes at least one of the hardware structures and software modules corresponding to each function in order to achieve the above-mentioned functions. Those skilled in the art should readily recognize that, in conjunction with the units and algorithm steps of the various examples described in the embodiments disclosed herein, this application can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed by hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0062] This application embodiment can divide a charging pile fault diagnosis and safety protection control device into functional units based on the above method example. For example, each function can be divided into separate functional units, or two or more functions can be integrated into one processing unit. The integrated unit can be implemented in hardware or as a software functional unit. It should be noted that the unit division in this application embodiment is illustrative and only represents a logical functional division; other division methods may be used in actual implementation.

[0063] When using integrated units, Figure 3A possible structural schematic diagram of a charging pile fault diagnosis and safety protection control device (referred to as control device 50) involved in the above embodiments is shown. The control device 50 includes a processing unit 502 and a communication unit 501, and may also include a storage unit 503. Figure 3 The structural diagram shown can be used to illustrate the structure of a charging pile fault diagnosis and safety protection control device involved in the above embodiments.

[0064] when Figure 3 The schematic diagram shown illustrates the structure of a charging pile fault diagnosis and safety protection control device involved in the above embodiments. The processing unit 502 is used to control and manage the operation of the charging pile fault diagnosis and safety protection control device, the communication unit 501 is used for the charging pile fault diagnosis and safety protection control device to communicate with other devices, and the storage unit 503 is used to store the program code and data of the charging pile fault diagnosis and safety protection control device.

[0065] For example, communication unit 501 is used to collect charging gun charging information during the charging start phase at a first preset frequency. The processing unit 502 is used to calculate the contact resistance sequence of the charging gun connection circuit based on the output terminal voltage, charging current, and DC bus voltage inside the charging pile through a resistance identification model; extract a first feature vector from the contact resistance sequence, extract a second feature vector from the charging current, and extract a third feature vector from the charging gun connector temperature; input the first feature vector, the second feature vector, and the third feature vector into a preset multi-level weighted fuzzy decision tree for time series fusion analysis, and output a comprehensive fault confidence level.

[0066] In one possible implementation, the processing unit 502 is further configured to compare the comprehensive fault confidence level with a preset confidence threshold, execute a fault protection command, and perform safety protection control on the charging pile.

[0067] The processing unit 502 can be a processor or a controller, and the communication unit 501 can be a communication interface, transceiver, transceiver circuit, transceiver device, etc. The term "communication interface" is a general term and may include one or more interfaces. The storage unit 503 can be a memory. When the control device 50 is a chip, the processing unit 502 can be a processor or a controller, and the communication unit 501 can be an input interface and / or an output interface, pins, or circuits, etc. The storage unit 503 can be a storage unit within the chip (e.g., a register, cache, etc.) or a storage unit located outside the chip (e.g., read-only memory (ROM), random access memory (RAM, etc.).

[0068] The communication unit can also be called a transceiver unit. The antenna and control circuit with transceiver functions in the control device 50 can be considered as the communication unit 501 of the control device 50, and the processor with processing functions can be considered as the processing unit 502 of the control device 50. Optionally, the device in the communication unit 501 that implements the receiving function can be considered as a communication unit, which is used to execute the receiving steps in the embodiments of this application. The communication unit can be a receiver, a receiver circuit, etc. The device in the communication unit 501 that implements the transmitting function can be considered as a transmitting unit, which is used to execute the transmitting steps in the embodiments of this application. The transmitting unit can be a transmitter, a transmitter, a transmitting circuit, etc.

[0069] Figure 3 If the integrated units in the process are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the embodiments of this application, in essence, or the parts that contribute to the prior art, or all or part of the technical solutions, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute all or part of the steps of the methods of the various embodiments of this application. Storage media for storing computer software products include various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory, random access memory, magnetic disks, or optical disks.

[0070] Figure 3 The units in the process can also be called modules; for example, a processing unit can be called a processing module.

[0071] This application embodiment also provides a hardware structure diagram of a charging pile fault diagnosis and safety protection control device (denoted as control device 60), see [link to diagram]. Figure 4 The control device 60 includes a processor 601, and optionally, a memory 602 connected to the processor 601.

[0072] In the first possible implementation, see Figure 4 The control device 60 also includes a transceiver 603. The processor 601, memory 602, and transceiver 603 are connected via a bus. The transceiver 603 is used to communicate with other devices or communication networks. Optionally, the transceiver 603 may include a transmitter and a receiver. The device in the transceiver 603 that implements the receiving function can be considered as a receiver, which is used to perform the receiving steps in the embodiments of this application. The device in the transceiver 603 that implements the transmitting function can be considered as a transmitter, which is used to perform the transmitting steps in the embodiments of this application.

[0073] Based on the first possible implementation method Figure 4 The structural diagram shown can be used to illustrate the structure of a charging pile fault diagnosis and safety protection control device involved in the above embodiments.

[0074] in, Figure 4 This can also be illustrated as a system chip in a charging pile fault diagnosis and safety protection control device. In this case, the actions performed by the aforementioned charging pile fault diagnosis and safety protection control device can be implemented by this system chip. The specific actions performed can be found above and will not be repeated here.

[0075] Some of the data in the above formula are calculated by removing dimensions and taking their numerical values. The formula is the closest to the real situation obtained by software simulation of a large amount of collected data. The preset parameters and preset thresholds in the formula are set by those skilled in the art according to the actual situation or obtained through simulation of a large amount of data.

Claims

1. A method for fault diagnosis and safety protection control of charging piles, characterized in that, include: The charging information of the charging gun during the charging start-up phase is collected at a first preset frequency. The charging information includes the output voltage, charging current, DC bus voltage inside the charging pile, and charging gun connector temperature. Based on the output voltage, charging current, and DC bus voltage inside the charging pile, the contact resistance sequence of the charging gun connection circuit is calculated using a resistance identification model. A first feature vector is extracted from the contact resistance sequence, a second feature vector is extracted from the charging current, and a third feature vector is extracted from the charging gun connector temperature; the first feature vector includes the initial rate of change and fluctuation variance of the contact resistance sequence in a preset time interval, the second feature vector includes the rising slope and amplitude envelope area of ​​the charging current waveform in the preset time interval, and the third feature vector is the initial temperature rise rate in the preset time interval. The first feature vector, the second feature vector, and the third feature vector are input into a preset multi-level weighted fuzzy decision tree for time-series fusion analysis; wherein, the multi-level weighted fuzzy decision tree assigns different decision weights to different feature vectors based on different sub-time periods of the charging start stage, and outputs a comprehensive fault confidence score. The overall fault confidence level is compared with a preset confidence threshold. If the overall fault confidence level is greater than or equal to the confidence threshold, a connection abnormality fault is determined to exist, and a fault protection command is generated. The fault protection command is executed to perform safety protection control on the charging pile.

2. The charging pile fault diagnosis and safety protection control method according to claim 1, characterized in that, The sequence of contact resistances for the charging gun connection circuit is constructed based on the sliding squares method, including: For the current moment During the time window Internally, by minimizing the objective function To solve for the contact resistance sequence; ； in, For dummy variables in the summation, For the width of the sliding window, The regularization coefficient is . for The voltage drop in the charging gun connection circuit at all times. for The charging current at any given moment, for Contact resistance at any given time The contact resistance The rate of change; By analyzing the objective function Differentiate and set it to zero to obtain the contact resistance. The estimated value.

3. The charging pile fault diagnosis and safety protection control method according to claim 2, characterized in that, The regularization coefficient ,include: The regularization coefficient According to the charging current Dynamic adjustment of amplitude: when When the amplitude is lower than the first current threshold, a preset first current threshold is used. Value; when When the amplitude is higher than or equal to the first current threshold, a preset second current threshold is used. Value; of which, the second Value less than the first value.

4. The charging pile fault diagnosis and safety protection control method according to claim 1, characterized in that, The multi-level weighted fuzzy decision tree includes at least two levels of decision points, including: The first-level decision node, corresponding to the first sub-time period of the charging start stage, has its decision weights configured as follows: the weight of the second feature vector is greater than the weights of the first feature vector and the third feature vector. The second-level decision node corresponds to the second sub-time period of the charging start stage. It starts after the first sub-time period, and its decision weight is configured as follows: the weights of the first feature vector and the third feature vector are greater than the weight of the second feature vector. Wherein, the sum of the first sub-time period and the second sub-time period is equal to the preset time interval.

5. The charging pile fault diagnosis and safety protection control method according to claim 1, characterized in that, Extracting the first feature vector from the contact resistance sequence specifically includes: Perform wavelet transform on the contact resistance sequence, calculate its wavelet energy within a preset fault characteristic frequency band, and use the wavelet energy as a component of the first feature vector.

6. The charging pile fault diagnosis and safety protection control method according to claim 1, characterized in that, After determining that a connection fault exists, the process also includes a fault classification step: If the overall fault confidence level is within the first confidence level range, a first-level alarm command and a power adjustment command are generated. If the overall fault confidence level is in a second confidence level range that is higher than the first confidence level range, then a level 2 alarm command and a cut-off command are generated.

7. The charging pile fault diagnosis and safety protection control method according to claim 6, characterized in that, The first-level alarm command is used to trigger an audible and visual alarm and record the event; the power adjustment command is used to control the charging pile to reduce the output power to below the rated power. The secondary alarm command is used to trigger a higher frequency audible and visual alarm and record the fault; the output cut-off command is used to control the main relay of the charging pile to disconnect.

8. The charging pile fault diagnosis and safety protection control method according to claim 1, characterized in that, Before collecting the charging information, the following steps are included: In response to the charging handshake completion signal, the main relay of the charging pile is closed, and the charging start-up phase monitoring mode is started. After entering the charging start-up phase monitoring mode, charging information is collected.

9. The charging pile fault diagnosis and safety protection control method according to claim 8, characterized in that, The determination that there is no connection failure includes: After determining that there is no connection abnormality, the system exits the charging start-up monitoring mode and switches to the regular charging monitoring mode based on fixed temperature and current thresholds.

10. A charging pile fault diagnosis and safety protection control system, operating based on the charging pile fault diagnosis and safety protection control method according to any one of claims 1-9, characterized in that, It includes a data acquisition module, a fault analysis module, and an early warning module; The acquisition module is used to acquire charging information of the charging gun during the charging start phase at a first preset frequency. The fault analysis module is used to calculate the contact resistance sequence of the charging gun connection circuit based on the output voltage, charging current, and DC bus voltage inside the charging pile, using a resistance identification model. A first feature vector is extracted from the contact resistance sequence, a second feature vector is extracted from the charging current, and a third feature vector is extracted from the charging gun connector temperature; The first feature vector, the second feature vector, and the third feature vector are input into a preset multi-level weighted fuzzy decision tree for time series fusion analysis, and the comprehensive fault confidence is output. The early warning module is used to compare the comprehensive fault confidence level with a preset confidence threshold, execute the fault protection command, and perform safety protection control on the charging pile.

Images