Electric two-wheeled vehicle charging method, device and storage medium with intelligent fault diagnosis

Abstract

The present invention discloses a charging method, device and storage medium for electric two-wheeled vehicles with intelligent fault diagnosis, which relates to the technical field of electric vehicle charging fault diagnosis. The method has an intelligent fault diagnosis function, can monitor and analyze various types of data during the charging process in real time, identify potential faults in advance, and improve charging safety and equipment reliability. Through the data collected by the sensor network, such as charging status, short-circuit fluctuations, leakage current and cable temperature, the system can automatically identify faults such as poor plug contact, micro-short circuits, leakage, and abnormal temperature, and generate corresponding fault instructions, and take measures such as power outages and alarms in time to prevent accidents. The self-discharge monitoring function can also determine whether the battery has self-discharge anomalies through battery discharge data, and promptly notify the user to carry out maintenance to avoid battery damage.

Description

Technical Field

[0001] The present invention relates to the technical field of electric vehicle charging fault diagnosis, and in particular to an electric two-wheeled vehicle charging method, device and storage medium with intelligent fault diagnosis. Background Art

[0002] With the increasing popularity of electric two-wheelers, especially in residential areas such as urban communities, the demand for outdoor charging stations is increasing. Although many residential charging stations are equipped with awnings to protect them from direct sunlight, they are still exposed to many potential failure risks due to their outdoor location.

[0003] Charging stations are located outdoors and are susceptible to weather conditions. For example, during rainy days, the charging station connectors and cables may corrode due to water vapor and rainwater penetration, leading to poor contact between the plugs, which in turn affects charging efficiency and may even pose a risk of electrical leakage. Furthermore, in hot weather, the temperature of the charging station and cables can rise dramatically. If not monitored and addressed promptly, abnormal thermal expansion and contraction of the cables can occur, compromising their long-term safety. Summary of the Invention

[0004] In view of the shortcomings of the existing technology, the present invention provides an electric two-wheeled vehicle charging method, device and storage medium with intelligent fault diagnosis to solve the problems mentioned in the background technology.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions: a method for charging an electric two-wheeled vehicle with intelligent fault diagnosis, comprising the following steps:

[0006] Step 1: Charging port abnormality monitoring: Use convolutional neural network to build a fault recognition model to identify the initial charging time period when the i-th charging pile is connected to the j-th electric vehicle charging port. Collect the first charging state data in the initial charging time period t1 and build the plug contact coefficient C p and metal corrosion index M c , preset contact resistance threshold C p_th and corrosion threshold M c_th , and determine whether the following faults exist:

[0007] When C p >C p_th , it is determined that the plug is not fully inserted and a first fault instruction is generated;

[0008] When M c >M c_th , it is determined that there is water vapor in the identification port, which may cause the plug to corrode and affect the conductivity, and a second fault instruction is generated;

[0009] When C p ≤C p_th and Mc ≤M c_th , then resume charging and go to step 2;

[0010] Step 2: Second stage continuous monitoring: After resuming charging, use the fault identification model to enter the second charging time period, collect the second comprehensive data in the second charging time period t2, and construct the short circuit fluctuation factor S f , leakage current I l and cable temperature gradient coefficient T w , preset short-circuit threshold S f_th , humidity threshold, leakage current threshold I l_th , temperature gradient threshold and charging current threshold I c_th , and determine whether the following faults exist:

[0011] When S f >S f_th , it is determined that there is a micro short circuit fault risk and a third fault instruction is generated;

[0012] When the ambient humidity H env > Set the humidity threshold and leakage current I l >I l_th , it is determined that there is a risk of micro-leakage in rainy days, and a fourth fault instruction is generated;

[0013] When the cable temperature gradient coefficient T w >Set the temperature gradient threshold and charging current I c >I c_th , it is determined that the cable connected to the charging port has abnormal thermal expansion and contraction, and a fifth fault instruction is generated;

[0014] Step 3: Self-discharge monitoring after power failure: After the second charging time period t2 is completed, that is, the user's charging behavior ends or after any previous forced power failure process, the charging pile backup power supply is connected to the electric vehicle BMS to collect the third discharge data of the electric vehicle battery in the third time period t3, and the self-discharge trend index D is constructed. t ;

[0015] Preset self-discharge threshold D t_th , and determine whether the following faults exist: When D t >D t_th , it is determined that the battery has self-discharge abnormality and a sixth fault instruction is generated.

[0016] Preferably, step one includes:

[0017] S11. Install a sensor network inside each charging pile, and the user monitors the status data of each electric vehicle charging process in real time. The status data of each electric vehicle charging process includes: first charging status data in the initial charging time period t1, second comprehensive data in the second charging time period t2, and third discharge data in the third time period t3;

[0018] The sensor network includes: pressure sensor, resistance sensor, electrochemical sensor, current sensor, fiber optic temperature sensor, leakage current sensor and temperature and humidity sensor;

[0019] S12, real-time monitoring of the first charging status data of the i-th charging pile and the corresponding j-th electric vehicle charging port, the first charging status data including: electric vehicle model, battery type, battery rated voltage, battery capacity, contact resistance R cont , actual insertion force F ins and the metal mass loss value M caused by corrosion loss ;

[0020] S13. Use a convolutional neural network to construct an initial convolutional neural network model, and use the state data of each electric vehicle charging process to train and test the initial convolutional neural network model. Use the trained initial convolutional neural network model as a fault recognition model, and use the intermediate layer output of the equipment operation state model as a feature vector to identify feature information. Use the acquired feature information to train and test the fault recognition model, and use the trained fault recognition model as data for running prediction to construct the plug contact coefficient C. p and metal corrosion index M c ;

[0021] S131, extracting the contact resistance R in the first charging state data cont and the actual insertion force F ins After dimensionless processing, the plug contact coefficient C is calculated using the following formula p :

[0022]

[0023] Where, F nom Indicates standard insertion force;

[0024] S132, extracting the metal mass loss value M caused by corrosion in the first charging state data loss and identify the initial metal mass M of the port by identifying the electric vehicle model init After dimensionless processing, the metal corrosion index M is calculated using the following formula: c :

[0025]

[0026] Among them, according to the electric vehicle model, the initial metal mass of the plug port is obtained by referring to the standard of the charging interface.

[0027] Preferably, step one further comprises:

[0028] S14, preset contact resistance threshold C p_th and the plug contact coefficient C p and contact resistance threshold C p_th Compare to determine if there is a risk of poor contact due to incomplete insertion of the plug, including:

[0029] When the plug contact coefficient C p >Contact resistance threshold C p_th , it is determined that the plug is not fully inserted, and the first fault instruction is generated, including: the system sends a charging prompt to the user "the charging plug is not fully inserted, please reinsert it to ensure good contact with the plug", and automatically pauses the charging time, set to 30 seconds, and monitors the plug contact coefficient C again. p , when C p ≥C p_th If the problem persists, the system will completely interrupt charging and prompt the operator to check the charging equipment.

[0030] When the plug contact coefficient C p ≤Contact resistance threshold C p_th , determining that the plug is properly inserted and generating a message to continue charging; the plug contact coefficient reflects the quality of the electrical connection. The lower the coefficient, the better the contact quality, the less resistance to current flow, and the more reliable the electrical connection.

[0031] S15, preset corrosion threshold M c_th and the metal corrosion index m c and corrosion threshold m c_th Compare to determine whether there is a risk of plug corrosion caused by water vapor, including:

[0032] When the metal corrosion index M c >Corrosion threshold M c_th If the plug is exposed to water vapor, it will be judged that there is a risk of corrosion of the plug, affecting the conductivity, and a second fault instruction will be generated, including: the system will send a charging reminder to the user, "There is a risk of corrosion in the charging plug. Please check whether the plug has water vapor or corrosion. Use a lint-free cloth to wipe the plug and port to avoid further charging." The charging time will be automatically paused, set to 60 seconds to 120 seconds, and the metal corrosion index M will be monitored again. c , when M c ≤M c_th, indicating normal, then enter step 2 to resume charging; if there is still a risk of water vapor causing plug corrosion, the system will completely interrupt charging and prompt the operation and maintenance personnel to check the charging equipment;

[0033] When the metal corrosion index M c ≤Corrosion threshold M c_th , determine that the plug has no water vapor and corrosion, is inserted normally, and generates a message to continue charging.

[0034] Preferably, step 2 includes:

[0035] S21, after resuming charging, determining to enter the second charging time period t2, collecting and acquiring second comprehensive data through the sensor network, the second comprehensive data including: the charging current I c 、Ambient humidity H env , charging pile port cable length ΔL, cable end temperature, live wire current I live and neutral current I neutral ;

[0036] S22, using the fault identification model, after training and testing the convolutional neural network initial model fault identification model with the second comprehensive data, analyze and calculate to obtain the short circuit fluctuation factor S f , leakage current I l and cable temperature gradient coefficient T w :

[0037] S221, short circuit fluctuation factor S f Calculated using the following formula:

[0038]

[0039] Where, I c,k represents the charging current at the kth sampling point in the second charging time period t2, Indicates the average charging current; N a Indicates the total number of charging sampling points;

[0040] S222, cable temperature gradient coefficient T w Calculated using the following formula:

[0041]

[0042] ΔT=T end -T start ;

[0043] Where ΔL represents the length of the charging pile port cable, ΔT represents the temperature difference between the two ends of the cable; T end Indicates the temperature of the cable end connected to the electric vehicle charging port, T startIndicates the temperature at the other end of the cable connected to the charging pile;

[0044] S223, leakage current I l Calculated using the following formula:

[0045] I l =I live -I neutral ;

[0046] I live is the live wire current, I neutral is the neutral current; if there is a difference between the two, the difference is the leakage current.

[0047] Preferably, step 2 further includes:

[0048] S23, preset short circuit threshold S f_th , and the short-circuit fluctuation factor S f With short-circuit threshold S f_th Compare to determine whether there is a risk of micro-short circuit failure, including:

[0049] When the short-circuit fluctuation factor S f >Short circuit threshold S f_th , it is determined that there is a risk of micro-short circuit fault, and a third fault instruction is generated, including: automatically disconnecting the relevant power supply and starting circuit isolation measures;

[0050] When the short-circuit fluctuation factor S f ≤ short circuit threshold S f_th , it is determined that there is no risk of micro short circuit fault and continuous monitoring is carried out.

[0051] Preferably, step 2 further includes:

[0052] S24, preset humidity threshold and leakage current threshold I l_th , and the ambient humidity H env and leakage current I l Respectively with the humidity threshold and leakage current threshold I l_th Compare to determine whether there is a risk of micro-leakage in rainy days, including:

[0053] When the ambient humidity H env >Humidity threshold and leakage current I l >I l_th , it is determined that there is a risk of micro-leakage in rainy days, and the fourth fault instruction is generated, including: the charging pile reduces the current charging current by 15%-30%, and turns on the air drying function during charging to reduce the humidity until the ambient humidity H env ≤humidity threshold, if it returns to normal within 1 minute, continue charging; after the charging pile reduces the current charging current by 30%, the leakage current I lStill exceeds the leakage current threshold I l_th , then the power is automatically cut off + power supply;

[0054] When the ambient humidity H env ≤ set humidity threshold and I l ≤I l_th , indicating the risk of micro-leakage on rainy days, continuous monitoring; when the ambient humidity H env ≤ set humidity threshold and I l >I l_th , indicating that there is a risk of leakage in non-rainy days, the charging pile power path is automatically switched to the backup grounding path. After switching to the backup grounding path, I l ≤I l_th , then resume charging, and prompt the user to maintain the charging pile equipment; if I l >I l_th ×200%, the abnormal alarm of the first charging pile is triggered, including: forced power off + locking the charging pile, and reminding the electric vehicle user not to force charging and to switch to other charging piles to continue charging.

[0055] Preferably, step 2 further includes:

[0056] S25, preset temperature gradient threshold and charging current threshold I c_th and the cable temperature gradient coefficient T w With the charging current I c Respectively with the temperature gradient threshold and the charging current threshold I c_th Compare the cables connected to the charging port to determine whether there are any abnormal thermal expansion and contraction, including:

[0057] When the cable temperature gradient coefficient T w >Set the temperature gradient threshold and charging current I c >I c_th , it is determined that the cable connected to the charging port has abnormal thermal expansion and contraction, and a fifth fault instruction is generated, including: the charging pile reduces the current charging current by 15%-30%, and turns on the temperature control fan or liquid cooling system during charging to reduce the cable temperature gradient coefficient until the cable temperature gradient coefficient T w < set temperature gradient threshold, if it returns to normal within 2 minutes, continue charging; after the charging pile reduces the current charging current by 15%-30%, the charging current I c Still exceeds the charging current threshold I c_th , then the power is automatically cut off + power supply;

[0058] When the cable temperature gradient coefficient T w ≤ set temperature gradient threshold and charging current I c ≤I c_th, it is determined that there is no abnormal thermal expansion and contraction of the cable connected to the charging port, and continuous monitoring is performed; when the cable temperature gradient coefficient T w ≤ set temperature gradient threshold and charging current I c ≤I c_th , then construct the short-circuit fluctuation factor S f Go to step S23 for judgment;

[0059] When the cable temperature gradient coefficient T w >Set the temperature gradient threshold and charging current I c ≤I c_th , it means that the cable temperature is abnormal, not caused by thermal expansion and contraction, triggering the second charging pile abnormal alarm, including: the system will automatically switch the charging path to the backup heat dissipation circuit. If the switch is successful, charging will continue; if the switch is unsuccessful, the power will be forced to cut off and the charging pile will be locked, and the user will be reminded to check the status of the charging pile cable and maintain or replace the cable.

[0060] Preferably, step three includes:

[0061] S31, after the second charging time period t2 is completed, that is, after the user's charging behavior ends or after any previous forced power-off process, the electric vehicle BMS is connected to the charging pile backup power supply to collect the third discharge data of the electric vehicle battery in the third time period t3, and the third discharge data includes: the amount of electricity Q init And the amount of electricity lost per unit time Q loss ;

[0062] S32, using the fault recognition model, after training and testing the convolutional neural network initial model fault recognition model with the third discharge data, analyze and calculate to obtain the self-discharge trend index D t :

[0063] Self-discharge trend index D t Calculated using the following formula:

[0064]

[0065] Where Q init Indicates the amount of power that has been charged, Q loss Indicates the amount of power lost per unit time;

[0066] S33, preset self-discharge threshold D t_th , and the self-discharge trend index D t and self-discharge threshold D t_th Compare and determine whether there is an abnormal risk of electric vehicle battery self-discharge, including:

[0067] When D t >D t_th, it is determined that the battery has self-discharge anomaly and a sixth fault instruction is generated, including: through the charging pile device screen, APP or SMS, the user is informed that "your battery has self-discharge anomaly, it is recommended to stop using it and perform battery maintenance"; when D t ≤D t_th , it is determined that the battery self-discharge is normal and continues to monitor.

[0068] An electric two-wheeled vehicle charging device with intelligent fault diagnosis includes:

[0069] A configuration unit is configured to install a sensor network inside each charging pile, so that a user can monitor the status data of each electric vehicle's charging process in real time. The status data of each electric vehicle's charging process includes: first charging status data in an initial charging time period t1, second comprehensive data in a second charging time period t2, and third discharge data in a third time period t3;

[0070] Construct a fault identification model unit, which is used to use a convolutional neural network to build an initial convolutional neural network model, and train and test the initial convolutional neural network model based on the status data of each electric vehicle charging process. The trained initial convolutional neural network model is used as the fault identification model. At the same time, the intermediate layer output of the equipment operation status model is used as a feature vector to identify feature information. The fault identification model is trained and tested based on the acquired feature information, and the trained fault identification model is used as data for operation prediction;

[0071] The initial charging time period abnormality monitoring unit is used to identify that when the i-th charging pile is connected to the j-th electric vehicle charging port, it is the initial charging time period, collect the first charging state data in the initial charging time period t1, and construct the plug contact coefficient C p and metal corrosion index M c , preset contact resistance threshold C p_th and corrosion threshold M c_th , and determine whether the following faults exist:

[0072] When C p <C p_th , it is determined that the plug is not fully inserted and a first fault instruction is generated;

[0073] When M c >M c_th , it is determined that there is water vapor in the identification port, which may cause the plug to corrode and affect the conductivity, and a second fault instruction is generated;

[0074] When C p ≥C p_th and M c ≤M c_th , then resume charging;

[0075] The second stage continuous monitoring unit is used to use the fault identification model to enter the second charging time period, collect the second comprehensive data in the second charging time period t2, and construct the short circuit fluctuation factor S f , leakage current I l and cable temperature gradient coefficient T w , preset short-circuit threshold S f_th , humidity threshold, leakage current threshold I l_th , temperature gradient threshold and charging current threshold I c_th , and determine whether the following faults exist:

[0076] When S f >S f_th , it is determined that there is a micro short circuit fault risk and a third fault instruction is generated;

[0077] When the ambient humidity H env > Set the humidity threshold and leakage current I l >I l_th , it is determined that there is a risk of micro-leakage in rainy days, and a fourth fault instruction is generated;

[0078] When the cable temperature gradient coefficient T w >Set the temperature gradient threshold and charging current I c >I c_th , it is determined that the cable connected to the charging port has abnormal thermal expansion and contraction, and a fifth fault instruction is generated;

[0079] The self-discharge monitoring unit is used to collect the third discharge data of the electric vehicle battery in the third time period t3 through the backup power supply of the charging pile after the second charging time period t2 is completed, that is, after the user's charging behavior ends or after any previous forced power-off process, to construct the self-discharge trend index D t ;

[0080] Preset self-discharge threshold D t_th , and determine whether the following faults exist: When D t >D t_th , it is determined that the battery has self-discharge abnormality and a sixth fault instruction is generated.

[0081] A storage medium includes a computer processor for loading and executing any of the steps and fault instructions of the above-mentioned electric two-wheeled vehicle charging method with intelligent fault diagnosis.

[0082] The present invention provides an electric two-wheeled vehicle charging method, device, and storage medium with intelligent fault diagnosis. It has the following beneficial effects:

[0083] (1) The invention can promptly detect potential fault risks during the charging process, such as incomplete insertion of the plug, metal corrosion caused by water vapor, and leakage, by monitoring abnormalities at the charging pile port and collecting real-time data. This intelligent fault diagnosis mechanism can effectively avoid reduced charging efficiency or leakage risks caused by poor contact or metal corrosion, thereby significantly improving the safety of the charging process. As outdoor equipment, charging piles are greatly affected by weather factors.

[0084] (2) The invention can detect in real time the impact of external factors such as moisture in rainy days and temperature changes on charging piles and cables by setting multiple monitoring indicators such as ambient humidity and temperature. In the event of moisture intrusion or excessive cable temperature, it can automatically detect and issue an alarm, and take timely measures to reduce the probability of failure. For example, when the risk of micro-leakage in rainy days is determined, the charging pile can adjust the charging status or take other safety measures to minimize failures caused by ambient humidity or high temperature.

[0085] (3) When a fault risk occurs in a certain part of the charging pile, the system can automatically generate a fault instruction and send an alarm message to the user in real time through the charging pile device screen, APP or SMS, etc., to inform the user of the possible fault risk in advance. This active early warning mechanism allows users to detect problems in time and take effective measures, reducing the risk of equipment damage or accidents during charging.

[0086] (4) After charging is completed, the present invention also has a function for monitoring battery self-discharge. By collecting discharge data from electric vehicle batteries and analyzing the self-discharge trend index, it is possible to promptly detect abnormal battery self-discharge and avoid battery performance degradation caused by long-term self-discharge. With this function, users can perform maintenance or replacement in advance when battery abnormalities occur, thereby extending the battery life. BRIEF DESCRIPTION OF THE DRAWINGS

[0087] Figure 1 A schematic diagram of the steps of the electric two-wheeled vehicle charging method with intelligent fault diagnosis according to the present invention;

[0088] Figure 2 The figure is a flow chart of the electric two-wheeled vehicle charging device with intelligent fault diagnosis according to the present invention. DETAILED DESCRIPTION

[0089] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0090] Example 1

[0091] See also Figure 1 The present invention provides an electric two-wheeled vehicle charging method with intelligent fault diagnosis, comprising the following steps:

[0092] Step 1: Charging port abnormality monitoring: Use convolutional neural network to build a fault recognition model to identify the initial charging time period when the i-th charging pile is connected to the j-th electric vehicle charging port. Collect the first charging state data in the initial charging time period t1 and build the plug contact coefficient C p and metal corrosion index M c , preset contact resistance threshold C p_th and corrosion threshold M c_th , and determine whether the following faults exist:

[0093] When C p <C p_th , it is determined that the plug is not fully inserted and a first fault instruction is generated;

[0094] When M c >M c_th , it is determined that there is water vapor in the identification port, which may cause the plug to corrode and affect the conductivity, and a second fault instruction is generated;

[0095] When C p ≥C p_th and M c ≤M c_th , then resume charging and go to step 2;

[0096] Step 2: Second stage continuous monitoring: After resuming charging, use the fault identification model to enter the second charging time period, collect the second comprehensive data in the second charging time period t2, and construct the short circuit fluctuation factor S f , leakage current I l and cable temperature gradient coefficient T w , preset short-circuit threshold S f_th , humidity threshold, leakage current threshold I l_th , temperature gradient threshold and charging current threshold I c_th , and determine whether the following faults exist:

[0097] When S f >S f_th , it is determined that there is a micro short circuit fault risk and a third fault instruction is generated;

[0098] When the ambient humidity H env > Set the humidity threshold and leakage current I l >I l_th , it is determined that there is a risk of micro-leakage in rainy days, and a fourth fault instruction is generated;

[0099] When the cable temperature gradient coefficient T w >Set the temperature gradient threshold and charging current I c >I c_th , it is determined that the cable connected to the charging port has abnormal thermal expansion and contraction, and a fifth fault instruction is generated;

[0100] Step 3: Self-discharge monitoring after power failure: After the second charging time period t2 is completed, that is, the user's charging behavior ends or after any previous forced power failure process, the charging pile backup power supply is connected to the electric vehicle BMS to collect the third discharge data of the electric vehicle battery in the third time period t3, and the self-discharge trend index D is constructed. t ;

[0101] Preset self-discharge threshold D t_th , and determine whether the following faults exist: When D t >D t_th , it is determined that the battery has self-discharge abnormality and a sixth fault instruction is generated.

[0102] In this embodiment, the present invention can timely detect possible fault risks during the charging process, such as problems such as the plug not being fully inserted, metal corrosion caused by water vapor, and leakage, through abnormal monitoring and real-time data collection of the charging pile port. This intelligent fault diagnosis mechanism can effectively avoid the risk of reduced charging efficiency or leakage due to poor contact or metal corrosion, thereby significantly improving the safety of the charging process. As an outdoor device, the charging pile is greatly affected by weather factors. By setting multiple monitoring indicators such as ambient humidity and temperature, the present invention can detect in real time the impact of external factors such as moisture in rainy days and temperature changes on the charging pile and cable. In the event of moisture intrusion or excessive cable temperature, it can automatically detect and issue an alarm, and take timely measures to reduce the probability of faults. For example, when determining the risk of micro-leakage on rainy days, the charging pile can adjust the charging status or take other safety measures to minimize faults caused by ambient humidity or high temperature.

[0103] When a fault risk arises in a charging station, the system automatically generates a fault indicator and sends a real-time alert to the user via the charging station device screen, app, or text message, notifying the user of the potential fault risk in advance. This proactive early warning mechanism allows users to identify problems promptly and take effective measures, reducing the risk of equipment damage or accidents during charging.

[0104] After charging is complete, the present invention also features a self-discharge monitoring function. By collecting discharge data from electric vehicle batteries and analyzing the self-discharge trend index, it can promptly detect abnormal self-discharge conditions and prevent performance degradation caused by prolonged self-discharge. This function allows users to perform maintenance or replacement in advance when battery anomalies occur, extending the battery's service life.

[0105] A fault identification model built using convolutional neural networks can automatically analyze and diagnose various potential faults during the charging process, reducing the need for manual intervention and improving the efficiency and accuracy of charging station maintenance. This intelligent operation and maintenance system not only enhances equipment safety but also provides data support for charging station operators to make quick decisions. Users do not need to focus on the complex details of the charging process; the system automatically detects and handles anomalies. Through real-time fault alerts and intelligent adjustments, users can charge with greater peace of mind, without worrying about safety hazards caused by external environmental influences on charging equipment. Furthermore, real-time feedback on fault information and battery health status enables users to react quickly, improving the convenience and safety of the charging experience.

[0106] Example 2

[0107] This embodiment is explained in Example 1. Specifically, step 1 includes:

[0108] S11. Install a sensor network inside each charging pile, and the user monitors the status data of each electric vehicle charging process in real time. The status data of each electric vehicle charging process includes: first charging status data in the initial charging time period t1, second comprehensive data in the second charging time period t2, and third discharge data in the third time period t3;

[0109] The sensor network includes: pressure sensor, resistance sensor, electrochemical sensor, current sensor, fiber optic temperature sensor, leakage current sensor and temperature and humidity sensor;

[0110] S12, real-time monitoring of the first charging status data of the i-th charging pile and the corresponding j-th electric vehicle charging port, the first charging status data including: electric vehicle model, battery type, battery rated voltage, battery capacity, contact resistance R cont , actual insertion force F ins and the metal mass loss value M caused by corrosion loss ; It can provide detailed basis for fault diagnosis, so as to timely identify and deal with various abnormal situations in the charging process.

[0111] S13. Use a convolutional neural network to construct an initial convolutional neural network model, and use the state data of each electric vehicle charging process to train and test the initial convolutional neural network model. Use the trained initial convolutional neural network model as a fault recognition model, and use the intermediate layer output of the equipment operation state model as a feature vector to identify feature information. Use the acquired feature information to train and test the fault recognition model, and use the trained fault recognition model as data for running prediction to construct the plug contact coefficient C. p and metal corrosion index Mc ;

[0112] S131, extracting the contact resistance R in the first charging state data cont and the actual insertion force F ins After dimensionless processing, the plug contact coefficient C is calculated using the following formula p :

[0113]

[0114] Data example:

[0115] R cont =1.0mΩ,F nom =20N,F ins =18N;

[0116] Calculating plug contact coefficient

[0117] S132, extracting the metal mass loss value M caused by corrosion in the first charging state data loss and identify the initial metal mass M of the port by identifying the electric vehicle model init After dimensionless processing, the metal corrosion index M is calculated using the following formula: c :

[0118]

[0119] Among them, according to the electric vehicle model, the initial metal mass of the plug port is obtained by referring to the standard of the charging interface.

[0120] Data example:

[0121] M init =100mg,M loss =3mg;

[0122] Calculating Metal Corrosion Index

[0123] S14, preset contact resistance threshold C p_th and the plug contact coefficient C p and contact resistance threshold Co _th Compare to determine if there is a risk of poor contact due to incomplete insertion of the plug, including:

[0124] When the plug contact coefficient C p >Contact resistance threshold C p_th, it is determined that the plug is not fully inserted, and the first fault instruction is generated, including: the system sends a charging prompt to the user "the charging plug is not fully inserted, please reinsert it to ensure good contact with the plug", and automatically pauses the charging time, set to 30 seconds, and monitors the plug contact coefficient C again. p , when C p ≥C p_th If the problem persists, the system will completely interrupt charging and prompt the operator to check the charging equipment.

[0125] When the plug contact coefficient C p ≤Contact resistance threshold C p_th , determining that the plug is properly inserted and generating a message to continue charging; the plug contact coefficient reflects the quality of the electrical connection. The lower the coefficient, the better the contact quality, the less resistance to current flow, and the more reliable the electrical connection.

[0126] S15, preset corrosion threshold M c_th and the metal corrosion index M c and corrosion threshold M c_th Compare to determine whether there is a risk of plug corrosion caused by water vapor, including:

[0127] When the metal corrosion index M c >Corrosion threshold M c_th If the plug is exposed to water vapor, it will be judged that there is a risk of corrosion of the plug, affecting the conductivity, and a second fault instruction will be generated, including: the system will send a charging reminder to the user, "There is a risk of corrosion in the charging plug. Please check whether the plug has water vapor or corrosion. Use a lint-free cloth to wipe the plug and port to avoid further charging." The charging time will be automatically paused, set to 60 seconds to 120 seconds, and the metal corrosion index M will be monitored again. c , when M c ≤M c_th , indicating normal, then enter step 2 to resume charging; if there is still a risk of water vapor causing plug corrosion, the system will completely interrupt charging and prompt the operation and maintenance personnel to check the charging equipment;

[0128] When the metal corrosion index M c ≤Corrosion threshold M c_th , determine that the plug has no water vapor and corrosion, is inserted normally, and generates a message to continue charging.

[0129] In this embodiment, by detecting whether the plug contact coefficient is below the contact resistance threshold, the system intelligently determines whether the plug is fully inserted. If the plug is not fully inserted, it generates a fault indication and sends a prompt to the user, reminding them to reinsert the charging plug. This intelligent processing mechanism effectively prevents the impact of poor plug contact on charging efficiency and safety, and promptly interrupts charging if the problem is not resolved, preventing further damage to the device.

[0130] By comparing the metal corrosion index with the corrosion threshold, the system can identify whether the charging port is at risk of metal corrosion caused by water vapor. If corrosion risk is detected, the system will pause charging and send a prompt to the user to check and recommend measures to clean the plug to prevent further corrosion. If the problem persists, the system will interrupt charging and prompt maintenance personnel to inspect the equipment, effectively preventing corrosion-related conductivity degradation or other electrical failures.

[0131] Example 3

[0132] This embodiment is explained in Example 1. Specifically, step 2 includes:

[0133] S21, after resuming charging, determining to enter the second charging time period t2, collecting and acquiring second comprehensive data through the sensor network, the second comprehensive data including: the charging current I c 、Ambient humidity H env , charging pile port cable length ΔL, cable end temperature, live wire current I live and neutral current I neutral ;

[0134] S22, using the fault identification model, after training and testing the convolutional neural network initial model fault identification model with the second comprehensive data, analyze and calculate to obtain the short circuit fluctuation factor S f , leakage current I l and cable temperature gradient coefficient T w :

[0135] S221, short circuit fluctuation factor S f Calculated using the following formula:

[0136]

[0137] Where, I c,k represents the charging current at the kth sampling point in the second charging time period t2, Indicates the average charging current; N indicates the total number of charging sampling points;

[0138] Data example:

[0139] Assume that the total number of sampling points is N a=5, charging current I c,k =1.0A; 1.1A; 1.2A; 1.1A; 1.0A;

[0140]

[0141] Calculate the kth sampling point in the second charging time period t2

[0142] (1.0-1.08) 2 =0.0064;

[0143] (1.1-1.08) 2 =0.0004;

[0144] (1.2-1.08) 2 =0.0144;

[0145] (1.1-1.08) 2 =0.0004;

[0146] (1.0-1.08) 2 =0.0064;

[0147] Calculate the short-circuit fluctuation factor S f , sum and calculate, specifically:

[0148]

[0149] S222, cable temperature gradient coefficient T w Calculated using the following formula:

[0150]

[0151] ΔT=T end -T start ;

[0152] Where ΔL represents the length of the charging pile port cable, ΔT represents the temperature difference between the two ends of the cable; T end Indicates the temperature of the cable end connected to the electric vehicle charging port, T start Indicates the temperature at the other end of the cable connected to the charging pile;

[0153] Data example: Assume the cable length ΔL is 30m and ΔT = 15°C;

[0154] Calculate the cable temperature gradient coefficient

[0155] S223, leakage current I l Calculated using the following formula:

[0156] Il =I live -I neutral ;

[0157] I live is the live wire current, I neutral is the neutral current; if there is a difference between the two, the difference is the leakage current.

[0158] In this embodiment, this embodiment realizes in-depth monitoring of the charging process during the second charging time period by collecting comprehensive data such as charging current, ambient humidity, cable temperature, current difference, etc. By calculating the short-circuit fluctuation factor, leakage current and cable temperature gradient coefficient, it is possible to identify potential risks in the charging process in real time, prevent the occurrence of safety hazards such as short circuit, leakage or cable overheating, and improve the safety of the charging process. The calculation method of the short-circuit fluctuation factor enables the system to accurately monitor the fluctuation of the charging current. By analyzing the charging current of each sampling point and the deviation from the average current, the system can promptly detect possible short circuit or current instability problems, issue early warnings and take corresponding measures (such as suspending charging or adjusting the current) to avoid damage to the battery or circuit.

[0159] By calculating the cable's temperature gradient coefficient, the system effectively monitors cable temperature changes during charging, ensuring it remains within safety thresholds. Real-time monitoring of temperature differences and cable length provides effective technical safeguards against risks such as excessive cable temperature, which could lead to damage or fire.

[0160] This embodiment calculates the difference between the live and neutral currents to determine the leakage current. This leakage current detection effectively identifies electrical leakage issues in the charging system and prevents them. Promptly detecting leakage current and taking appropriate measures, such as automatic power-off or alarms, significantly improves the safety of charging stations and reduces the potential hazards of electrical failures.

[0161] Example 4

[0162] This embodiment is explained in Example 1. Specifically, step 2 further includes:

[0163] S23, preset short circuit threshold S f_th , and the short-circuit fluctuation factor S f With short-circuit threshold S f_th Compare to determine whether there is a risk of micro-short circuit failure, including:

[0164] When the short-circuit fluctuation factor S f >Short circuit threshold S f_th , it is determined that there is a risk of micro-short circuit fault, and a third fault instruction is generated, including: automatically disconnecting the relevant power supply and starting circuit isolation measures;

[0165] When the short-circuit fluctuation factor S f ≤ short circuit threshold S f_th , it is determined that there is no risk of micro short circuit fault and continuous monitoring is carried out.

[0166] S24, preset humidity threshold and leakage current threshold I l_th , and the ambient humidity H env and leakage current I l Respectively with the humidity threshold and leakage current threshold I l_th Compare to determine whether there is a risk of micro-leakage in rainy days, including:

[0167] When the ambient humidity H env >Humidity threshold and leakage current I l >I l_th , it is determined that there is a risk of micro-leakage in rainy days, and the fourth fault instruction is generated, including: the charging pile reduces the current charging current by 15%-30%, and turns on the air drying function during charging to reduce the humidity until the ambient humidity H env ≤humidity threshold, if it returns to normal within 1 minute, continue charging; after the charging pile reduces the current charging current by 30%, the leakage current I l Still exceeds the leakage current threshold I l_th , then the power is automatically cut off + power supply;

[0168] When the ambient humidity H env ≤ set humidity threshold and I l ≤I l_th , indicating the risk of micro-leakage on rainy days, continuous monitoring; when the ambient humidity H env ≤ set humidity threshold and I l >I l_th , indicating that there is a risk of leakage in non-rainy days, the charging pile power path is automatically switched to the backup grounding path. After switching to the backup grounding path, I l ≤I l_th , then resume charging, and prompt the user to maintain the charging pile equipment; if I l >I l_th ×200%, the abnormal alarm of the first charging pile is triggered, including: forced power off + locking the charging pile, and reminding the electric vehicle user not to force charging and to switch to other charging piles to continue charging.

[0169] S25, preset temperature gradient threshold and charging current threshold I c_th and the cable temperature gradient coefficient T w With the charging current I c Respectively with the temperature gradient threshold and the charging current threshold I c_th Compare the cables connected to the charging port to determine whether there are any abnormal thermal expansion and contraction, including:

[0170] When the cable temperature gradient coefficient T w >Set the temperature gradient threshold and charging current I c >I c_th , it is determined that the cable connected to the charging port has abnormal thermal expansion and contraction, and a fifth fault instruction is generated, including: the charging pile reduces the current charging current by 15%-30%, and turns on the temperature control fan or liquid cooling system during charging to reduce the cable temperature gradient coefficient until the cable temperature gradient coefficient T w < set temperature gradient threshold, if it returns to normal within 2 minutes, continue charging; after the charging pile reduces the current charging current by 15%-30%, the charging current I c Still exceeds the charging current threshold I c_th , then the power is automatically cut off + power supply;

[0171] When the cable temperature gradient coefficient T w ≤ set temperature gradient threshold and charging current I c ≤I c_th , it is determined that there is no abnormal thermal expansion and contraction of the cable connected to the charging port, and continuous monitoring is performed; when the cable temperature gradient coefficient T w ≤ set temperature gradient threshold and charging current I c ≤I c_th , then construct the short-circuit fluctuation factor S f Go to step S23 for judgment;

[0172] When the cable temperature gradient coefficient T w >Set the temperature gradient threshold and charging current I c ≤I c_th , it means that the cable temperature is abnormal, not caused by thermal expansion and contraction, triggering the second charging pile abnormal alarm, including: the system will automatically switch the charging path to the backup heat dissipation circuit. If the switch is successful, charging will continue; if the switch is unsuccessful, the power will be forced to cut off and the charging pile will be locked, and the user will be reminded to check the status of the charging pile cable and maintain or replace the cable.

[0173] This embodiment effectively identifies the risk of micro-short circuit failures by presetting a short-circuit threshold and comparing it with the short-circuit fluctuation factor. When the short-circuit fluctuation factor exceeds the set threshold, the system automatically disconnects the power supply and initiates circuit isolation measures to prevent the short circuit from causing more serious electrical failures. This proactive safety protection mechanism significantly improves charging safety and reduces the risk of damage to electrical equipment.

[0174] By monitoring ambient humidity and leakage current and comparing them with set humidity and leakage current thresholds, the system can effectively identify the risk of micro-leakage during rainy days. If this risk occurs, the system intelligently reduces the charging current and activates the air-drying function to reduce the impact of humidity and ensure a safe charging process. If humidity is too high and leakage current is not effectively controlled, the system automatically shuts off power to avoid potential hazards caused by electrical leakage. This mechanism enhances the adaptability of charging piles in humid environments and avoids safety hazards caused by environmental factors.

[0175] By setting temperature gradient and charging current thresholds, the system can effectively monitor cable thermal expansion and contraction. If the cable temperature gradient exceeds the set threshold and the charging current is too high, the system automatically reduces the charging current and activates a temperature-controlled fan or liquid cooling system to adjust the cable temperature gradient and prevent overheating or damage. If the charging current is not effectively controlled, the system will disconnect and cut off the power supply, reducing safety risks caused by abnormal cable temperatures.

[0176] Example 5

[0177] This embodiment is explained in Example 4. Specifically, step 3 includes:

[0178] S31, after the second charging time period t2 is completed, that is, after the user's charging behavior ends or after any previous forced power-off process, the electric vehicle BMS is connected to the charging pile backup power supply to collect the third discharge data of the electric vehicle battery in the third time period t3, and the third discharge data includes: the amount of electricity Q init And the amount of electricity lost per unit time Q loss ;

[0179] S32, using the fault recognition model, after training and testing the convolutional neural network initial model fault recognition model with the third discharge data, analyze and calculate to obtain the self-discharge trend index D t :

[0180] Self-discharge trend index D t Calculated using the following formula:

[0181]

[0182] Where Q init Indicates the amount of power that has been charged, Q loss Indicates the amount of power lost per unit time;

[0183] Data example:

[0184] Q init =5.0mAh; Q loss =0.2mAh;

[0185] Calculate the self-discharge trend index D t :

[0186]

[0187] S33, preset self-discharge threshold D t_th , and the self-discharge trend index D t and self-discharge threshold D t_th Compare and determine whether there is an abnormal risk of electric vehicle battery self-discharge, including:

[0188] When D t >D t_th , it is determined that the battery has self-discharge anomaly and a sixth fault instruction is generated, including: through the charging pile device screen, APP or SMS, the user is informed that "your battery has self-discharge anomaly, it is recommended to stop using it and perform battery maintenance"; when D t ≤D t_th , it is determined that the battery self-discharge is normal and continues to monitor.

[0189] In this embodiment, through the real-time collection and analysis of the third discharge data, the system can accurately identify whether the electric vehicle battery has self-discharge anomalies. The calculation of the self-discharge trend index and the preset self-discharge threshold comparison mechanism can promptly issue an alarm when the battery self-discharge anomaly occurs. In this way, users can discover and deal with the problem in the early stages, avoiding damage or failure of the battery due to self-discharge anomalies, thereby improving the battery life and charging safety. Once the battery self-discharge anomaly is detected, the system will promptly notify the user of the anomaly information through the charging station screen, APP or SMS. This proactive notification method can effectively enhance the user's safety awareness and remind the user to take measures to repair the battery as soon as possible. In this way, users can identify and avoid potential safety hazards earlier.

[0190] The following is a sample data example chart 1:

[0191]

[0192]

[0193] The contact resistance threshold is: C p_th =1.10mΩ; corrosion threshold M c_th =3.2% short circuit threshold S f_th =0.0058A 2 ; Humidity threshold is 85%; Leakage current threshold I l_th =30mA; the temperature gradient threshold is 2.3m / ℃; the charging current threshold I c_th =8A; self-discharge threshold D t_th=4.5%;

[0194] The following is the sample data example chart 1 corresponding to the fault instruction judgment is as follows:

[0195] Charging port abnormal monitoring, plug not fully inserted (first fault instruction): triggered when the plug contact coefficient is greater than 1.10mΩ. Triggered battery vehicle numbers: 002 (1.15mΩ), 007 (1.14mΩ), 009 (1.13mΩ);

[0196] Plug corrosion risk (second fault instruction): triggered when the metal corrosion index is greater than 3.2%.

[0197] Triggering battery car numbers: 003 (3.2%), 005 (3.5%), 007 (3.3%), 008 (3.4%)

[0198] Meet the conditions to resume charging: When the plug contact coefficient is ≤1.10mΩ and the metal corrosion index is ≤3.2%, normal charging will resume. Normal vehicles: 001, 004, 006, 009, 010;

[0199] Second stage continuous monitoring: micro short circuit fault (third fault instruction); when the short circuit fluctuation factor is greater than 0.0058A 2 Triggering battery car number: 002 (0.0060A 2 )、005(0.0059A 2 )、010(0.0061A 2 ) Automatically disconnect the relevant power supply and initiate circuit isolation measures;

[0200] Micro-leakage risk in rainy days (fourth fault instruction): Triggered when the ambient humidity is >85% and the leakage current is >30mA. Triggered electric vehicle numbers: 002 (humidity 90%, leakage 35mA), 003 (humidity 88%, leakage 28mA), 005 (humidity 92%, leakage 50mA), 008 (humidity 95%, leakage 30mA);

[0201] Abnormal cable thermal expansion and contraction (fifth fault instruction): triggered when the cable temperature gradient coefficient is greater than 2.3m / ℃ and the charging current is greater than 8A. Triggering battery vehicle numbers: 005 (2.3m / ℃, charging current 10A), 010 (2.3m / ℃, charging current 8A);

[0202] Self-discharge monitoring after full charge: Self-discharge abnormality (sixth fault instruction): triggered when the self-discharge trend index is greater than 4.5%. Triggering battery vehicle number: 002 (4.5%), 008 (4.6%);

[0203] Example 6

[0204] Please refer to Figure 2 , an electric two-wheeled vehicle charging device with intelligent fault diagnosis, including:

[0205] A configuration unit is configured to install a sensor network inside each charging pile, so that a user can monitor the status data of each electric vehicle's charging process in real time. The status data of each electric vehicle's charging process includes: first charging status data in an initial charging time period t1, second comprehensive data in a second charging time period t2, and third discharge data in a third time period t3;

[0206] Construct a fault identification model unit, which is used to use a convolutional neural network to build an initial convolutional neural network model, and train and test the initial convolutional neural network model based on the status data of each electric vehicle charging process. The trained initial convolutional neural network model is used as the fault identification model. At the same time, the intermediate layer output of the equipment operation status model is used as a feature vector to identify feature information. The fault identification model is trained and tested based on the acquired feature information, and the trained fault identification model is used as data for operation prediction;

[0207] The initial charging time period abnormality monitoring unit is used to identify that when the i-th charging pile is connected to the j-th electric vehicle charging port, it is the initial charging time period, collect the first charging state data in the initial charging time period t1, and construct the plug contact coefficient C p and metal corrosion index M c , preset contact resistance threshold C p_th and corrosion threshold M c_th , and determine whether the following faults exist:

[0208] When C p <C p_th , it is determined that the plug is not fully inserted and a first fault instruction is generated;

[0209] When M c >M c_th , it is determined that there is water vapor in the identification port, which may cause the plug to corrode and affect the conductivity, and a second fault instruction is generated;

[0210] When C p ≥C p_th and M c ≤M c_th , then resume charging;

[0211] The second stage continuous monitoring unit is used to use the fault identification model to enter the second charging time period, collect the second comprehensive data in the second charging time period t2, and construct the short circuit fluctuation factor S f , leakage current I l and cable temperature gradient coefficient T w , preset short-circuit threshold S f_th, humidity threshold, leakage current threshold I l_th , temperature gradient threshold and charging current threshold I c_th , and determine whether the following faults exist:

[0212] When S f >S f_th , it is determined that there is a micro short circuit fault risk and a third fault instruction is generated;

[0213] When the ambient humidity H env > Set the humidity threshold and leakage current I l >I l_th , it is determined that there is a risk of micro-leakage in rainy days, and a fourth fault instruction is generated;

[0214] When the cable temperature gradient coefficient T w >Set the temperature gradient threshold and charging current I c >I c_th , it is determined that the cable connected to the charging port has abnormal thermal expansion and contraction, and a fifth fault instruction is generated;

[0215] The self-discharge monitoring unit is used to collect the third discharge data of the electric vehicle battery in the third time period t3 through the backup power supply of the charging pile after the second charging time period t2 is completed, that is, after the user's charging behavior ends or after any previous forced power-off process, to construct the self-discharge trend index D t ;

[0216] Preset self-discharge threshold D t_th , and determine whether the following faults exist: When D t >D t_th , it is determined that the battery has self-discharge abnormality and a sixth fault instruction is generated.

[0217] A storage medium includes a computer processor for loading and executing any of the steps and fault instructions of the above-mentioned electric two-wheeled vehicle charging method with intelligent fault diagnosis.

[0218] The threshold is set to facilitate comparison. The size of the threshold depends on the amount of sample data and the number of bases set by technicians in this field for each set of sample data; as long as it does not affect the proportional relationship between the parameter and the quantized value.

[0219] The above formulas are obtained by collecting a large amount of data and performing software simulation, and a formula close to the actual value is selected. The coefficients in the formula are set by those skilled in the art according to actual conditions. The above is only a preferred specific implementation method of the present invention, but the protection scope of the present invention is not limited to this. Any technician familiar with this technical field, within the technical scope disclosed by the present invention, can make equivalent replacements or changes based on the technical solution and inventive concept of the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A method for charging an electric two-wheeled vehicle with intelligent fault diagnosis, characterized in that: The following steps are involved: Step 1: Charging port abnormality monitoring: Use convolutional neural network to build a fault recognition model to identify the initial charging time period when the i-th charging pile is connected to the j-th electric vehicle charging port. Collect the first charging state data in the initial charging time period t1 and build the plug contact coefficient C p and metal corrosion index M c , preset contact resistance threshold C p_th and corrosion threshold M c_th , and determine whether the following faults exist: When C p >C p_th , it is determined that the plug is not fully inserted and a first fault instruction is generated; When M c >M c_th , it is determined that there is water vapor in the identification port, which may cause the plug to corrode and affect the conductivity, and a second fault instruction is generated; When C p ≤C p_th and M c ≤M c_th , then resume charging and go to step 2; Step 2: After resuming charging, use the fault identification model to enter the second charging time period, collect the second comprehensive data in the second charging time period t2, and construct the short circuit fluctuation factor S f , leakage current I l and cable temperature gradient coefficient T w , preset short-circuit threshold S f_th , humidity threshold, leakage current threshold I l_th , temperature gradient threshold and charging current threshold I c_th , and determine whether the following faults exist: When S f >S f_th , it is determined that there is a micro short circuit fault risk and a third fault instruction is generated; When the ambient humidity H env > Set the humidity threshold and leakage current I l >I l_th , it is determined that there is a risk of micro-leakage in rainy days, and a fourth fault instruction is generated; When the cable temperature gradient coefficient T w >Set the temperature gradient threshold and charging current I c >I c_th , it is determined that the cable connected to the charging port has abnormal thermal expansion and contraction, and a fifth fault instruction is generated; Step 3: Self-discharge monitoring after power failure: After the second charging time period t2 is completed, that is, the user's charging behavior ends or after any previous forced power failure process, the charging pile backup power supply is connected to the electric vehicle BMS to collect the third discharge data of the electric vehicle battery in the third time period t3, and the self-discharge trend index D is constructed. t ; Preset self-discharge threshold D t_th , and determine whether the following faults exist: When D t >D t_th , it is determined that the battery has self-discharge abnormality and a sixth fault instruction is generated.

2. The electric two-wheeled vehicle charging method with intelligent fault diagnosis according to claim 1, characterized in that: Step one includes: S11. Install a sensor network inside each charging pile, and the user monitors the status data of each electric vehicle charging process in real time. The status data of each electric vehicle charging process includes: first charging status data in the initial charging time period t1, second comprehensive data in the second charging time period t2, and third discharge data in the third time period t3; The sensor network includes: pressure sensor, resistance sensor, electrochemical sensor, current sensor, fiber optic temperature sensor, leakage current sensor and temperature and humidity sensor; S12, real-time monitoring of the first charging status data of the i-th charging pile and the corresponding j-th electric vehicle charging port, the first charging status data including: electric vehicle model, battery type, battery rated voltage, battery capacity, contact resistance R cont , actual insertion force F ins and the metal mass loss value M caused by corrosion loss ; S13. Use a convolutional neural network to construct an initial convolutional neural network model, and use the state data of each electric vehicle charging process to train and test the initial convolutional neural network model. Use the trained initial convolutional neural network model as a fault recognition model, and use the intermediate layer output of the equipment operation state model as a feature vector to identify feature information. Use the acquired feature information to train and test the fault recognition model, and use the trained fault recognition model as data for running prediction to construct the plug contact coefficient C. p and metal corrosion index M c ; S131, extracting the contact resistance R in the first charging state data cont and the actual insertion force F ins After dimensionless processing, the plug contact coefficient C is calculated using the following formula p : Where, F nom Indicates standard insertion force; S132, extracting the metal mass loss value M caused by corrosion in the first charging state data loss and identify the initial metal mass M of the port by identifying the electric vehicle model init After dimensionless processing, the metal corrosion index M is calculated using the following formula: c : Among them, according to the electric vehicle model, the initial metal mass of the plug port is obtained by referring to the standard of the charging interface.

3. The electric two-wheeled vehicle charging method with intelligent fault diagnosis according to claim 2, characterized in that: Step 1 also includes: S14, preset contact resistance threshold C p_th and the plug contact coefficient C p and contact resistance threshold C p_th Compare to determine if there is a risk of poor contact due to incomplete insertion of the plug, including: When the plug contact coefficient C p >Contact resistance threshold C p_th , it is determined that the plug is not fully inserted, and the first fault instruction is generated, including: the system sends a charging reminder to the user "The charging plug is not fully inserted, please reinsert it to ensure good contact with the plug", and automatically pauses the charging time, setting it to 30 seconds, and monitors the plug contact coefficient C again. p , when C p ≥C p_th If the problem persists, the system will completely interrupt charging and prompt the operator to check the charging equipment. When the plug contact coefficient C p ≤Contact resistance threshold C p_th , determine that the plug is inserted normally and generate a message to continue charging; S15, preset corrosion threshold M c_th and the metal corrosion index M c and corrosion threshold M c_h Compare to determine whether there is a risk of plug corrosion caused by water vapor, including: When the metal corrosion index M c >Corrosion threshold M c_th If the plug is exposed to water vapor, it will be judged that there is a risk of corrosion, which will affect the electrical conductivity. A second fault instruction will be generated, including: the system will send a charging reminder to the user, "There is a risk of corrosion in the charging plug. Please check whether the plug has water vapor or corrosion. Use a lint-free cloth to wipe the plug and port to avoid further charging." The system will automatically pause the charging time, set to 60 seconds to 120 seconds, and monitor the metal corrosion index M again. c , when M c ≤M c_th , indicating normal, then enter step 2 to resume charging; if there is still a risk of water vapor causing plug corrosion, the system will completely interrupt charging and prompt the operation and maintenance personnel to check the charging equipment; When the metal corrosion index M c ≤Corrosion threshold M c_th , determine that the plug has no water vapor and corrosion, is inserted normally, and generates a message to continue charging.

4. The electric two-wheeled vehicle charging method with intelligent fault diagnosis according to claim 1, characterized in that: Step 2 includes: S21, after resuming charging, determining to enter the second charging time period t2, collecting and acquiring second comprehensive data through the sensor network, the second comprehensive data including: the charging current I c 、Ambient humidity H env , charging pile port cable length ΔL, cable end temperature, live wire current I live and neutral current I neutral ; S22, using the fault identification model, after training and testing the convolutional neural network initial model fault identification model with the second comprehensive data, analyze and calculate to obtain the short circuit fluctuation factor S f , leakage current I l and cable temperature gradient coefficient T w : S221, the short-circuit fluctuation factor S f Calculated using the following formula: Where, I c,k represents the charging current at the kth sampling point in the second charging time period t2, Indicates the average charging current; N a Indicates the total number of charging sampling points; S222, the cable temperature gradient coefficient T w Calculated using the following formula: ΔT=T end -T start ; Where ΔL represents the length of the charging pile port cable, ΔT represents the temperature difference between the two ends of the cable; T end Indicates the temperature of the cable end connected to the electric vehicle charging port, T start Indicates the temperature at the other end of the cable connected to the charging pile; S223, the leakage current I l Calculated using the following formula: I l =I live -I neutral ; I live is the live wire current, I neutral is the neutral current; if there is a difference between the two, the difference is the leakage current.

5. The electric two-wheeled vehicle charging method with intelligent fault diagnosis according to claim 4, characterized in that: Step 2 also includes: S23, preset short circuit threshold S f_th , and the short-circuit fluctuation factor S f With short-circuit threshold S f_th Compare to determine whether there is a risk of micro-short circuit failure, including: When the short-circuit fluctuation factor S f >Short circuit threshold S f_th , it is determined that there is a risk of micro-short circuit fault, and a third fault instruction is generated, including: automatically disconnecting the relevant power supply and starting circuit isolation measures; When the short-circuit fluctuation factor S f ≤ short circuit threshold S f_th , it is determined that there is no risk of micro short circuit fault and continuous monitoring is carried out.

6. The electric two-wheeled vehicle charging method with intelligent fault diagnosis according to claim 5, characterized in that: Step 2 also includes: S24, preset humidity threshold and leakage current threshold I l_th , and the ambient humidity H env and leakage current I l Respectively with the humidity threshold and leakage current threshold I l_th Compare to determine whether there is a risk of micro-leakage in rainy days, including: When the ambient humidity H env >Humidity threshold and leakage current I l >I l_th , it is determined that there is a risk of micro-leakage in rainy days, and the fourth fault instruction is generated, including: the charging pile reduces the current charging current by 15%-30%, and turns on the air drying function during charging to reduce the humidity until the ambient humidity H env ≤humidity threshold, if it returns to normal within 1 minute, continue charging; after the charging pile reduces the current charging current by 30%, the leakage current I l Still exceeds the leakage current threshold I l_th , then the power is automatically cut off + power supply; When the ambient humidity H env ≤ set humidity threshold and I l ≤I l_th , indicating the risk of micro-leakage on rainy days, continuous monitoring; when the ambient humidity H env ≤ set humidity threshold and I l >I l_th , indicating that there is a risk of leakage in non-rainy days, the charging pile power path is automatically switched to the backup grounding path. After switching to the backup grounding path, I l ≤I l_th , then resume charging, and prompt the user to maintain the charging pile equipment; if I l >I l_th ×200%, the abnormal alarm of the first charging pile is triggered, including: forced power off + locking the charging pile, and reminding the electric vehicle user not to force charging and to switch to other charging piles to continue charging.

7. The electric two-wheeled vehicle charging method with intelligent fault diagnosis according to claim 6, characterized in that: Step 2 also includes: S25, preset temperature gradient threshold and charging current threshold I c_th and the cable temperature gradient coefficient T w With the charging current I c Respectively with the temperature gradient threshold and the charging current threshold I c_th Compare the cables connected to the charging port to determine whether there are any abnormal thermal expansion and contraction, including: When the cable temperature gradient coefficient T w >Set the temperature gradient threshold and charging current I c >I c_th , it is determined that the cable connected to the charging port has abnormal thermal expansion and contraction, and a fifth fault instruction is generated, including: the charging pile reduces the current charging current by 15%-30%, and turns on the temperature control fan or liquid cooling system during charging to reduce the cable temperature gradient coefficient until the cable temperature gradient coefficient T w < set temperature gradient threshold, if it returns to normal within 2 minutes, continue charging; after the charging pile reduces the current charging current by 15%-30%, the charging current I c Still exceeds the charging current threshold I c_th , then the power is automatically cut off + power supply; When the cable temperature gradient coefficient T w ≤ set temperature gradient threshold and charging current I c ≤I c_th , it is determined that there is no abnormal thermal expansion and contraction of the cable connected to the charging port, and continuous monitoring is performed; when the cable temperature gradient coefficient T w ≤ set temperature gradient threshold and charging current I c ≤I c_th , then construct the short-circuit fluctuation factor S f Go to step S23 for judgment; When the cable temperature gradient coefficient T w >Set the temperature gradient threshold and charging current I c ≤I c_th , it means that the cable temperature is abnormal, not caused by thermal expansion and contraction, triggering the second charging pile abnormal alarm, including: the system will automatically switch the charging path to the backup heat dissipation circuit. If the switch is successful, charging will continue; if the switch is unsuccessful, the power will be forced to cut off and the charging pile will be locked, and the user will be reminded to check the status of the charging pile cable and maintain or replace the cable.

8. The electric two-wheeled vehicle charging method with intelligent fault diagnosis according to claim 7, characterized in that: Step three includes: S31, after the second charging time period t2 is completed, that is, after the user's charging behavior ends or after any previous forced power-off process, the electric vehicle BMS is connected to the charging pile backup power supply to collect the third discharge data of the electric vehicle battery in the third time period t3, and the third discharge data includes: the amount of electricity Q init And the amount of electricity lost per unit time Q loss ; S32, using the fault recognition model, after training and testing the convolutional neural network initial model fault recognition model with the third discharge data, analyze and calculate to obtain the self-discharge trend index D t : The self-discharge trend index D t Calculated using the following formula: Where Q init Indicates the amount of power that has been charged, Q loss Indicates the amount of power lost per unit time; S33, preset self-discharge threshold D t_th , and the self-discharge trend index D t and self-discharge threshold D t_th Compare and determine whether there is an abnormal risk of electric vehicle battery self-discharge, including: When D t >D t_th , it is determined that the battery has self-discharge anomaly and a sixth fault instruction is generated, including: through the charging pile device screen, APP or SMS, the user is informed that "your battery has self-discharge anomaly, it is recommended to stop using it and perform battery maintenance"; when D t ≤D t_th , it is determined that the battery self-discharge is normal and continues to monitor.

9. An electric two-wheeled vehicle charging device with intelligent fault diagnosis, applied to an electric two-wheeled vehicle charging method with intelligent fault diagnosis according to any one of claims 1 to 8, characterized in that: include: A configuration unit is configured to install a sensor network inside each charging pile, so that a user can monitor the status data of each electric vehicle's charging process in real time. The status data of each electric vehicle's charging process includes: first charging status data in an initial charging time period t1, second comprehensive data in a second charging time period t2, and third discharge data in a third time period t3; Construct a fault identification model unit, which is used to use a convolutional neural network to build an initial convolutional neural network model, and train and test the initial convolutional neural network model based on the status data of each electric vehicle charging process. The trained initial convolutional neural network model is used as the fault identification model. At the same time, the intermediate layer output of the equipment operation status model is used as a feature vector to identify feature information. The fault identification model is trained and tested based on the acquired feature information, and the trained fault identification model is used as data for operation prediction; The initial charging time period abnormality monitoring unit is used to identify that when the i-th charging pile is connected to the j-th electric vehicle charging port, it is the initial charging time period, collect the first charging state data in the initial charging time period t1, and construct the plug contact coefficient C p and metal corrosion index M c , preset contact resistance threshold C p_th and corrosion threshold M c_th , and determine whether the following faults exist: When C p >C p_th , it is determined that the plug is not fully inserted and a first fault instruction is generated; When M c >M c_th , it is determined that there is water vapor in the identification port, which may cause the plug to corrode and affect the conductivity, and a second fault instruction is generated; When C p ≤C p_th and M c ≤M c_th , then resume charging and go to step 2; The second stage continuous monitoring unit is used to use the fault identification model to enter the second charging time period, collect the second comprehensive data in the second charging time period t2, and construct the short circuit fluctuation factor S f , leakage current I l and cable temperature gradient coefficient T w , preset short-circuit threshold S f_th , humidity threshold, leakage current threshold I l_th , temperature gradient threshold and charging current threshold I c_th , and determine whether the following faults exist: When S f >S f_th , it is determined that there is a micro short circuit fault risk and a third fault instruction is generated; When the ambient humidity H enb > Set the humidity threshold and leakage current I l >I l_th , it is determined that there is a risk of micro-leakage in rainy days, and a fourth fault instruction is generated; When the cable temperature gradient coefficient T w >Set the temperature gradient threshold and charging current I c >I c_th , it is determined that the cable connected to the charging port has abnormal thermal expansion and contraction, and a fifth fault instruction is generated; The self-discharge monitoring unit is used to collect the third discharge data of the electric vehicle battery in the third time period t3 through the backup power supply of the charging pile after the second charging time period t2 is completed, that is, after the user's charging behavior ends or after any previous forced power-off process, to construct the self-discharge trend index D t ; Preset self-discharge threshold D t_th , and determine whether the following faults exist: When D t >D t_th , it is determined that the battery has self-discharge abnormality and a sixth fault instruction is generated.

10. A storage medium comprising a computer processor for loading and executing the steps and fault instructions of the electric two-wheeled vehicle charging method with intelligent fault diagnosis according to any one of claims 1 to 8.

Images