Charging pile fault identification method based on multi-modal data fusion

Abstract

The invention discloses a charging pile fault identification method based on multi-modal data fusion, and the method comprises the following steps: S1, constructing charging pile group-power distribution network topological mapping, monitoring and collecting the multi-dimensional data of charging pile equipment in real time, and recording the abnormal data; s2, key performance indexes of the charging pile equipment are calculated, anomaly analysis is carried out, and the fault type of the charging pile equipment is judged for the first time; s3, calling the electric energy quality index data of the abnormal charging pile equipment from the abnormal data, performing abnormal analysis, and judging the fault type of the charging pile equipment for the second time; s4, constructing a fault probability comprehensive evaluation function of the charging pile equipment, and judging the fault type of the charging pile equipment for the third time according to a calculation result; and S5, according to the three fault type judgment results of the charging pile equipment, corresponding protection measures are executed, identification of the fault type of the charging pile is realized, and the accuracy of identification of the fault type of the charging pile is effectively improved.

Description

Technical Field

[0001] This invention relates to the field of charging pile fault identification technology, and in particular to a charging pile fault identification method using multimodal data fusion. Background Technology

[0002] With the rapid expansion of the electric vehicle industry, charging piles, as key nodes supporting the smart grid, directly impact the stable supply of the power system and the sustainable development of green transportation through their operational reliability and fault prediction capabilities. In real-world operation scenarios, charging piles often face multiple complex challenges, including but not limited to abnormal operating conditions such as grid voltage dips, current harmonic interference, and power module overheating. If these potential risks are not identified and addressed in a timely manner through real-time monitoring systems, they can not only disrupt charging services but also trigger chain reactions—for example, harmonic pollution spreading and causing distribution network protection devices to malfunction, or localized overheating leading to equipment insulation failure, ultimately threatening the safe operation of the entire regional power grid. Therefore, building an intelligent identification system with multi-parameter collaborative diagnostic capabilities has become a core element in ensuring the resilience of the charging network.

[0003] Therefore, how to effectively identify the types of charging pile faults and take corresponding measures has become an important research issue in this field. Summary of the Invention

[0004] The purpose of this invention is to provide a charging pile fault identification method based on multimodal data fusion. By using key performance indicators and power quality indicators of the charging pile equipment to collaboratively determine the type of charging pile fault multiple times, the accuracy of identifying charging pile fault types is improved.

[0005] To solve the above technical problems, the technical solution of the present invention is: a charging pile fault identification method based on multimodal data fusion, comprising the following steps:

[0006] S1: Construct a topology mapping between charging pile clusters and power distribution networks, monitor and collect multi-dimensional data of charging pile equipment in real time, and record abnormal data;

[0007] S2: Based on real-time collected multi-dimensional data, calculate the key performance indicators of the charging pile equipment and perform anomaly analysis; based on the anomaly analysis results, determine the fault type of the charging pile equipment for the first time.

[0008] If a single charging pile device displays an abnormal key performance indicator, it is determined to be a single charging pile fault, and step S5 is executed;

[0009] If multiple charging pile devices show abnormal key performance indicators, proceed to step S3;

[0010] S3: Retrieve the power quality index data of the abnormal charging pile equipment from the abnormal data in step S1 and perform abnormal analysis; Based on the abnormal analysis results, determine the fault type of the charging pile equipment for the second time.

[0011] If all power quality indicators are abnormal, the charging pile is determined to be faulty on the grid side, and step S5 is executed.

[0012] If some of the power quality indicators are abnormal, proceed to step S4.

[0013] If all indicators in the power quality data are normal, then multiple charging piles are determined to be faulty, and step S5 is executed.

[0014] S4: Based on power quality index data, construct a comprehensive evaluation function for the failure probability of charging pile equipment, as shown in the following expression:

[0015]

[0016] in, This represents the probability of failure. The weights of each power quality indicator, For voltage deviation, Harmonic distortion rate, This refers to the voltage sag / surge time density.

[0017] The range of the failure probability is set to [0.3, 0.7]. Based on the calculation results of the failure probability, the failure type of the charging pile equipment is determined for the third time.

[0018] when When the fault occurs, it is determined to be a fault on the power grid side of the charging pile.

[0019] when At that time, it was determined to be a complex fault of the charging pile;

[0020] when At that time, it was determined that multiple charging piles were faulty;

[0021] S5: Based on the fault type determination results of the three charging pile devices, implement corresponding protection measures to achieve the identification of charging pile fault types.

[0022] Preferably, in step S1, the multi-dimensional data includes time-dimensional data, spatial-dimensional data, and physical-dimensional data; the time-dimensional data includes the charging pile's operating time, the time point of fault alarm occurrence, the timestamp of shutdown and maintenance, and daily peak load; the spatial-dimensional data includes the charging pile's installation location information, the charging pile's station number, the charging pile's relative position within the same station, the charging pile's surrounding environment data, and the group relationship between the charging pile and neighboring charging piles; the physical-dimensional data includes the power module's input / output voltage and current, insulation resistance value, module operating status, electricity meter reading, and switch status; the abnormal data includes the time point of the charging pile's first abnormal state and the duration of the abnormal state in the time dimension, the spatial location of the abnormal charging pile and whether there is a spatial relationship in the spatial dimension, and the power quality indicators of the abnormal charging pile in the physical dimension.

[0023] Preferably, the key performance indicators of the charging pile equipment include power conversion efficiency, IGBT junction temperature, and BMS message packet loss rate.

[0024] Preferably, in step S3, the power quality index data includes voltage deviation, voltage sag / surge, harmonic distortion rate, and three-phase imbalance.

[0025] Preferably, in step 2, the energy conversion efficiency is calculated using the following formula:

[0026]

[0027] in, For power conversion efficiency, This is the rectified DC power. This refers to the AC power before rectification.

[0028] The IGBT junction temperature is calculated using the following formula:

[0029]

[0030] in, for The temperature at any given moment, for NTC sensor temperature at any given time Let the power loss function be . This is the transient thermal impedance function. From 0 to Integral at time step;

[0031] The BMS packet loss rate is automatically generated by the BMS communication software within a set statistical time window.

[0032] Preferably, in step 3, the voltage deviation The calculation formula is as follows:

[0033]

[0034] in, This is the effective value of the voltage. Rated voltage;

[0035] The voltage sag / boost is calculated using the following formula:

[0036]

[0037] in, This is a voltage sag / surge indicator. For the total observation time window, , These represent the start and end times of the voltage sag / surge, respectively. for The effective value of the voltage at time t. Rated voltage;

[0038] The harmonic distortion rate The calculation formula is as follows:

[0039]

[0040] in, The highest harmonic order, For the first Effective value of subharmonic voltage;

[0041] The three-phase unbalance is calculated using the following formula:

[0042]

[0043] in, These are the effective values ​​of the three-phase voltages A, B, and C, respectively. This represents the average value of the three-phase voltage.

[0044] Preferably, step 4 further includes adaptively allocating weights for voltage deviation, harmonic distortion rate, and voltage sag / surge time density based on power quality index data, specifically implemented as follows:

[0045] Step 4.1: Construct the original data matrix based on power quality index data. The expression is as follows:

[0046]

[0047] in, For voltage deviation, Harmonic distortion rate, For voltage sag / surge time density, The number of samples;

[0048] Step 4.2: For the matrix The original data in the data is normalized using the following formula:

[0049]

[0050] in, These are the normalized data values. The original data values, For the first The minimum value of each indicator. For the first The maximum value of each indicator;

[0051] Step 4.3: Based on the normalized data values, calculate the information entropy of the power quality index, using the following formula:

[0052]

[0053]

[0054] in, The proportion of the normalized data values. For the sample size, These are the normalized data values. For project indexing, For the first Information entropy of each indicator;

[0055] Step 4.4: Calculate the weights of the power quality indicators. The calculation formula is as follows:

[0056]

[0057] in, For the first The final weight of each indicator, For the summation index, For the first Information entropy of each indicator This represents the total number of power quality indicators that participate in the weighting allocation.

[0058] Preferably, in step 5, based on the fault type determination result of the three-stage charging pile equipment, corresponding protection measures are implemented, including:

[0059] (1) Single charging pile failure: Locate the faulty charging pile, replace the power module of the faulty charging pile, and then track the data flow of the faulty charging pile in real time through the CAN bus;

[0060] (2) Multiple charging pile failures: Locate multiple faulty charging piles, construct a topology diagram between the faulty charging piles, analyze the spatial distance between multiple faulty charging piles, and after eliminating external interference factors, replace the power modules of the faulty charging piles one by one according to the degree of failure. During the replacement process, track the data flow and status of other faulty charging piles in real time through the CAN bus until all faulty charging piles are restored to normal.

[0061] (3) Faults on the power grid side of charging piles: Based on the power demand and actual situation of the charging pile group and surrounding users, the backup power supply is automatically put into operation or the power is shut down for maintenance. Then, the fault point is located and the fault is eliminated through ground fault location technology.

[0062] (4) Complex charging pile faults: Isolation is achieved by circuit breaker tripping and SVG activation on the grid side of the charging pile. The faulty charging pile needs to be powered down urgently through hardware DESAT protection and quickly forced to disconnect from the vehicle contactor.

[0063] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) The present invention can diagnose charging pile faults in a multi-dimensional collaborative manner through holographic perception of time-space-physical three dimensions, realize accurate source tracing of charging pile faults, and greatly improve the accuracy of fault type identification; (2) Through intelligent weight allocation, it can effectively adapt to various climate environments and improve the adaptive protection capability of charging pile groups; (3) The good hierarchical protection strategy not only shortens the fault handling time, but also reduces operation and maintenance costs. Attached Figure Description

[0064] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0065] Figure 1 This is a schematic diagram of the charging pile fault identification process based on multimodal data fusion of the present invention. Detailed Implementation

[0066] The present invention will now be described in further detail with reference to the accompanying drawings. The terminal technical solutions of the embodiments of the present invention will be clearly and completely described. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0067] This embodiment takes a charging pile group consisting of three or more charging piles as an example.

[0068] like Figure 1 As shown, this invention provides a charging pile fault identification method based on multimodal data fusion, comprising the following steps:

[0069] S1: Construct a topology mapping between charging pile clusters and power distribution networks, monitor and collect multi-dimensional data of charging pile equipment in real time, and record abnormal data.

[0070] The multi-dimensional data includes time-dimensional data, spatial-dimensional data, and physical-dimensional data. The time-dimensional data includes the charging pile's operating time, the time of fault alarm occurrence, the timestamp of shutdown for maintenance, and the daily peak load. The spatial-dimensional data includes the charging pile's installation location information, the charging pile's station number, the charging pile's relative position within the same station, the charging pile's surrounding environment data, and the group relationship between the charging pile and neighboring charging piles. The physical-dimensional data includes the power module's input / output voltage and current, insulation resistance value, module operating status, electricity meter reading, and switch status.

[0071] The abnormal data includes the time point of the first abnormal state of the charging pile and the duration of the abnormal state in the time dimension, the spatial location of the abnormal charging pile and whether there is a spatial relationship in the spatial dimension, and the power quality indicators of the abnormal charging pile in the physical dimension.

[0072] In this embodiment, the charging pile group also includes a power quality analyzer, a zero-sequence current transformer, a synchronous phasor measurement unit (PMU), an acoustic detector, a partial discharge detector, and a CAN controller with a built-in error frame counter. For special areas with special requirements for temperature, humidity, and corrosiveness, devices such as temperature and humidity transmitters, water accumulation sensors, and salt spray concentration meters can be added to monitor the charging pile data and analyze its status.

[0073] In this embodiment, the charging pile group-distribution network topology mapping establishes a relationship matrix between various nodes by collecting single-line data of the distribution network, cable parameters, equipment nameplate data, and real-time SCADA data. This matrix is ​​used for the analysis of the impact of voltage sags / surges on the charging piles and the analysis of harmonic transmission after the charging piles are injected with harmonic current.

[0074] S2: Based on real-time collected multi-dimensional data, calculate the key performance indicators of the charging pile equipment and perform anomaly analysis; based on the anomaly analysis results, determine the fault type of the charging pile equipment for the first time. The key performance indicators of the charging pile equipment include power conversion efficiency, IGBT junction temperature, and BMS message packet loss rate. If any indicator in the charging pile equipment shows an abnormality, it is considered a charging pile equipment fault.

[0075] If a single charging pile device displays an abnormal key performance indicator, it is determined to be a single charging pile fault, and step S5 is executed;

[0076] If three or more charging pile devices show abnormal key performance indicators, then proceed to step S3.

[0077] The energy conversion efficiency is calculated using the following formula:

[0078]

[0079] in, For power conversion efficiency, This is the rectified DC power. This represents the AC power before rectification.

[0080] when If the value is too high, it indicates an abnormal power conversion efficiency; otherwise, it is normal.

[0081] The IGBT junction temperature is measured using the device's built-in NTC sensor, and then calculated using a thermodynamic model. The calculation formula is as follows:

[0082]

[0083] in, for The temperature at any given moment, for NTC sensor temperature at any given time Let the power loss function be . This is the transient thermal impedance function. From 0 to Integral over time.

[0084] when If the temperature is too high, it indicates an abnormal IGBT junction temperature; otherwise, it is normal.

[0085] The BMS packet loss rate is automatically generated by the BMS communication software within a set statistical time window. If the BMS packet loss rate... If the duration exceeds 2 minutes, it indicates an abnormal packet loss rate in the BMS.

[0086] S3: Retrieve the power quality index data of the abnormal charging pile equipment from the abnormal data in step S1 and perform anomaly analysis; Based on the anomaly analysis results, determine the fault type of the charging pile equipment for the second time; The power quality index data includes voltage deviation, voltage sag / surge, harmonic distortion rate and three-phase imbalance.

[0087] If all power quality indicators are abnormal, the charging pile is determined to be faulty on the grid side, and step S5 is executed.

[0088] If some of the power quality indicators are abnormal, proceed to step S4.

[0089] If all power quality indicators are normal, then multiple charging piles are determined to be faulty, and step S5 is executed.

[0090] The voltage deviation The calculation formula is as follows:

[0091]

[0092] in, This is the effective value of the voltage. Rated voltage;

[0093] when If the voltage deviation is too large, it indicates that the indicator is abnormal; conversely, if the voltage deviation is not large, it indicates that the indicator is normal.

[0094] The voltage sag / boost is calculated using the following formula:

[0095]

[0096] in, This is a voltage sag / surge indicator. For the total observation time window, , These represent the start and end times of the voltage sag / surge, respectively. for The effective value of the voltage at time t. This is the rated voltage.

[0097] when or At that time, a voltage dip / surge begins. , The preset minimum and maximum threshold voltage values ​​are... Set to 0.9 , Set to 1.1 .

[0098] In this embodiment, the voltage waveform of the abnormal charging pile equipment is extracted, transformed, and analyzed. If the voltage drops within half a cycle... If the voltage sag lasts for more than 8ms, it indicates an abnormal voltage dip; if the voltage rises within half a cycle... The fact that the voltage spike lasted for more than 8ms indicates an abnormal voltage rise.

[0099] The harmonic distortion rate The calculation formula is as follows:

[0100]

[0101] in, The highest harmonic order, For the first Effective value of subharmonic voltage.

[0102] when If the value is too high, it indicates that the harmonic distortion rate is too high and the indicator is abnormal; conversely, it indicates that the indicator is normal.

[0103] The three-phase unbalance is calculated using the following formula:

[0104]

[0105] in, These are the effective values ​​of the three-phase voltages A, B, and C, respectively. This represents the average value of the three-phase voltage.

[0106] The three-phase imbalance of the charging pile equipment is monitored using a power quality analyzer. When the three-phase imbalance... If the three-phase imbalance is high, it indicates an abnormal indicator; conversely, if the imbalance is low, it indicates a normal indicator.

[0107] S4: Based on power quality index data, construct a comprehensive evaluation function for the failure probability of charging pile equipment, as shown in the following expression:

[0108]

[0109] in, This represents the probability of failure. The weights of each power quality indicator, For voltage deviation, Harmonic distortion rate, This represents the voltage sag / boost time density.

[0110] The range of the failure probability is set to [0.3, 0.7]. Based on the calculation results of the failure probability, the failure type of the charging pile equipment is determined for the third time.

[0111] when When the fault occurs, it is determined to be a fault on the power grid side of the charging pile.

[0112] when At that time, it was determined to be a complex fault of the charging pile;

[0113] when At that time, it was determined that multiple charging piles were faulty.

[0114] This embodiment also includes adaptively allocating weights for voltage deviation, harmonic distortion rate, and voltage sag / surge time density based on power quality index data, as specifically implemented as follows:

[0115] Step 4.1: Construct the original data matrix based on power quality index data. The expression is as follows:

[0116]

[0117] in, For voltage deviation, Harmonic distortion rate, For voltage sag / surge time density, The number of samples;

[0118] Step 4.2: For the matrix The original data in the data is normalized using the following formula:

[0119]

[0120] in, These are the normalized data values. The original data values, For the first The minimum value of each indicator. For the first The maximum value of each indicator;

[0121] Step 4.3: Based on the normalized data values, calculate the information entropy of the power quality index, using the following formula:

[0122]

[0123]

[0124] in, The proportion of the normalized data values. For the sample size, These are the normalized data values. For project indexing, For the first Information entropy of each indicator;

[0125] Step 4.4: Calculate the weights of the power quality indicators. The calculation formula is as follows:

[0126]

[0127] in, For the first The final weight of each indicator, For the summation index, For the first Information entropy of each indicator This represents the total number of power quality indicators that participate in the weighting allocation.

[0128] S5: Based on the fault type determination results of the three charging piles, implement corresponding protection measures to achieve fault type identification of the charging piles, including:

[0129] (1) Single charging pile failure: Locate the faulty charging pile, replace the power module of the faulty charging pile, and then track the data flow of the faulty charging pile in real time through the CAN bus.

[0130] (2) Multiple charging pile failures: Locate multiple faulty charging piles, construct a topology diagram between the faulty charging piles, and determine whether there are other fault causes, such as heavy object compression or local water accumulation, based on the spatial distance analysis of multiple faulty charging piles. After eliminating external factors, replace the power modules of the faulty charging piles one by one according to the degree of failure. During the replacement process, track the data flow and status of other faulty charging piles in real time through the CAN bus until all faulty charging piles are restored to normal.

[0131] (3) Charging pile grid-side faults: Based on the power demand and actual situation of the charging pile group and surrounding users, the backup power supply is automatically put into operation or a power outage is carried out for maintenance. Then, the fault point is located and the fault is eliminated through ground fault location technology. Active filters are installed in charging pile groups with frequent accidents to improve the power quality of the charging pile group.

[0132] (4) Complex charging pile faults: Isolation is achieved on the grid side of the charging pile through circuit breaker tripping and SVG activation. The faulty charging pile needs to be urgently powered down through hardware DESAT protection and quickly forcibly disconnected from the vehicle contactor. After troubleshooting, the fault process is reproduced by building a grid-charging pile model in ETAP to verify the feasibility of the handling measures and analyze whether there is a better solution. For high-frequency faulty charging pile groups, dual control board hot backup is adopted, and a dynamic voltage restorer (DVR) is installed on the grid side.

[0133] All parts not covered in this invention are the same as or implemented using existing technologies.

[0134] The above description, in conjunction with specific embodiments, provides a further detailed explanation of the present invention. It should not be construed that the specific implementation of the present invention is limited to these descriptions. For those skilled in the art, various simple deductions or substitutions can be made without departing from the concept of the present invention, and all such modifications and substitutions should be considered within the scope of protection of the present invention.

Claims

1. A method for fault identification of charging piles using multimodal data fusion, characterized in that: Includes the following steps: S1: Construct a topology mapping between charging pile clusters and power distribution networks, monitor and collect multi-dimensional data of charging pile equipment in real time, and record abnormal data; S2: Based on real-time collected multi-dimensional data, calculate the key performance indicators of the charging pile equipment and perform anomaly analysis; based on the anomaly analysis results, determine the fault type of the charging pile equipment for the first time. If a single charging pile device displays an abnormal key performance indicator, it is determined to be a single charging pile fault, and step S5 is executed; If multiple charging pile devices show abnormal key performance indicators, proceed to step S3; S3: Retrieve the power quality index data of the abnormal charging pile equipment from the abnormal data in step S1 and perform abnormal analysis; Based on the abnormal analysis results, determine the fault type of the charging pile equipment for the second time. If all power quality indicators are abnormal, the charging pile is determined to be faulty on the grid side, and step S5 is executed. If some of the power quality indicators are abnormal, proceed to step S4. If all indicators in the power quality data are normal, then multiple charging piles are determined to be faulty, and step S5 is executed. S4: Based on power quality index data, construct a comprehensive evaluation function for the failure probability of charging pile equipment, as shown in the following expression: in, This represents the probability of failure. The weights of each power quality indicator, For voltage deviation, Harmonic distortion rate, Voltage sag / surge time density; The range of the failure probability is set to [0.3, 0.7]. Based on the calculation results of the failure probability, the failure type of the charging pile equipment is determined for the third time. when When the fault occurs, it is determined to be a fault on the power grid side of the charging pile. when At that time, it was determined to be a complex fault of the charging pile; when At that time, it was determined that multiple charging piles were faulty; S5: Based on the fault type determination results of the three charging pile devices, implement corresponding protection measures to achieve the identification of charging pile fault types.

2. The charging pile fault identification method based on multimodal data fusion according to claim 1, characterized in that: In step S1, the multi-dimensional data includes time-dimensional data, spatial-dimensional data, and physical-dimensional data. The time-dimensional data includes the charging pile's operating time, the time of fault alarm occurrence, the timestamp of shutdown for maintenance, and daily peak load. The spatial-dimensional data includes the charging pile's installation location information, the charging pile's station number, the charging pile's relative position within the same station, the charging pile's surrounding environment data, and the group relationship between the charging pile and neighboring charging piles. The physical-dimensional data includes the power module's input / output voltage and current, insulation resistance value, module operating status, electricity meter reading, and switch status. The abnormal data includes the time of the charging pile's first abnormal state and the duration of the abnormal state in the time dimension, the spatial location of the abnormal charging pile and whether there is a spatial relationship in the spatial dimension, and the power quality indicators of the abnormal charging pile in the physical dimension.

3. The charging pile fault identification method based on multimodal data fusion according to claim 1, characterized in that: In step S2, the key performance indicators of the charging pile equipment include power conversion efficiency, IGBT junction temperature, and BMS message packet loss rate.

4. The charging pile fault identification method based on multimodal data fusion according to claim 1, characterized in that: In step S3, the power quality index data includes voltage deviation, voltage sag / surge, harmonic distortion rate, and three-phase imbalance.

5. The charging pile fault identification method based on multimodal data fusion according to claim 3, characterized in that: In step 2, the energy conversion efficiency is calculated using the following formula: in, For power conversion efficiency, This is the rectified DC power. This refers to the AC power before rectification. The IGBT junction temperature is calculated using the following formula: in, for The temperature at any given moment, for NTC sensor temperature at any given time Let the power loss function be . This is the transient thermal impedance function. From 0 to Integral at time step; The BMS packet loss rate is automatically generated by the BMS communication software within a set statistical time window.

6. The charging pile fault identification method based on multimodal data fusion according to claim 4, characterized in that: In step 3, the voltage deviation The calculation formula is as follows: in, This is the effective value of the voltage. Rated voltage; The voltage sag / boost is calculated using the following formula: in, This is a voltage sag / surge indicator. For the total observation time window, , These represent the start and end times of the voltage sag / surge, respectively. for The effective value of the voltage at time t. Rated voltage; The harmonic distortion rate The calculation formula is as follows: in, The highest harmonic order, For the first RMS value of subharmonic voltage; The three-phase unbalance is calculated using the following formula: in, These are the effective values ​​of the three-phase voltages A, B, and C, respectively. This represents the average value of the three-phase voltage.

7. The charging pile fault identification method based on multimodal data fusion according to claim 1, characterized in that: Step 4 also includes adaptively allocating weights for voltage deviation, harmonic distortion rate, and voltage sag / rise time density based on power quality index data, as specifically implemented below: Step 4.1: Construct the original data matrix based on power quality index data. The expression is as follows: in, For voltage deviation, Harmonic distortion rate, For voltage sag / surge time density, The number of samples; Step 4.2: For the matrix The original data in the data is normalized using the following formula: in, These are the normalized data values. The original data values, For the first The minimum value of each indicator. For the first The maximum value of each indicator; Step 4.3: Based on the normalized data values, calculate the information entropy of the power quality index, using the following formula: in, The proportion of the normalized data values. For the sample size, These are the normalized data values. For project indexing, For the first Information entropy of each indicator; Step 4.4: Calculate the weights of the power quality indicators. The calculation formula is as follows: in, For the first The final weight of each indicator, For the summation index, For the first Information entropy of each indicator This represents the total number of power quality indicators that participate in the weighting allocation.

8. The charging pile fault identification method based on multimodal data fusion according to claim 1, characterized in that: In step 5, based on the fault type determination results of the three charging pile devices, corresponding protection measures are implemented, including: (1) Single charging pile failure: Locate the faulty charging pile, replace the power module of the faulty charging pile, and then track the data flow of the faulty charging pile in real time through the CAN bus; (2) Multiple charging pile failures: Locate multiple faulty charging piles, construct a topology diagram between the faulty charging piles, analyze the spatial distance between multiple faulty charging piles, and after eliminating external interference factors, replace the power modules of the faulty charging piles one by one according to the degree of failure. During the replacement process, track the data flow and status of other faulty charging piles in real time through the CAN bus until all faulty charging piles are restored to normal. (3) Faults on the power grid side of charging piles: Based on the power demand and actual situation of the charging pile group and surrounding users, the backup power supply is automatically put into operation or the power is shut down for maintenance. Then, the fault point is located and the fault is eliminated through ground fault location technology. (4) Complex charging pile faults: Isolation is achieved by circuit breaker tripping and SVG activation on the grid side of the charging pile. The faulty charging pile needs to be powered down urgently through hardware DESAT protection and quickly forced to disconnect from the vehicle contactor.

Images