Charging pile safety monitoring method and system

Abstract

The invention relates to the technical field of monitoring or controlling a charging station, in particular to a charging pile safety monitoring method and system, and the method comprises the following steps: step 1, monitoring: step 1.1, electric leakage monitoring: installing a high-sensitivity electric leakage sensor in a charging pile and a charging gun line, continuously monitoring the electric leakage current in real time, and monitoring the electric leakage current in real time; the monitoring frequency is that data are collected once every 10 milliseconds; step 1.2, overload and short circuit monitoring: real-time current change of a charging circuit is monitored by using a current transformer, and the acquisition frequency is once every 5 milliseconds; step 1.3, monitoring the temperature; according to the scheme, through monitoring of electric leakage, overload and short circuit, temperature, smog and flame, charging gun safety, vision and the like in combination with intelligent analysis, the operation state and the surrounding environment condition of the charging pile can be comprehensively, timely and accurately mastered, electrical faults, temperature anomalies and potential safety hazards can be rapidly detected and judged, and the safety and reliability of the charging pile are improved. And all-around real-time monitoring of the potential safety hazard of the charging pile is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of monitoring or controlling charging stations, and in particular to a charging pile safety monitoring method and system. Background Art

[0002] Charging pile safety monitoring is a technical means used to ensure the safety of charging piles during operation, monitor the various operating parameters of the charging piles and the surrounding environmental conditions in real time, promptly discover and deal with potential safety hazards, and ensure the stable and safe operation of the charging piles and provide users with reliable charging services.

[0003] Traditional monitoring methods and systems have many shortcomings. They can often only monitor simple electrical parameters such as current and voltage. The monitoring of key safety factors such as temperature, smoke, flame, and charging gun insertion and extraction force is not comprehensive or accurate enough. For example, the leakage monitoring accuracy is low, and it may not be possible to detect small leakage risks in time; the temperature monitoring is not detailed, and the temperature changes of key parts inside the charging pile cannot be accurately grasped.

[0004] To this end, the present invention proposes a charging pile safety monitoring method and system, which realize all-round and accurate monitoring through multi-type high-precision sensors, use intelligent image analysis technology to identify and manage vehicles, personnel and foreign objects, establish a complete feedback mechanism to push information to operation and maintenance personnel and users in a timely manner, and automatically or manually intervene to execute corresponding protection and processing measures according to the monitoring and analysis results, thereby realizing comprehensive, efficient and intelligent monitoring of the safety of charging piles, and improving the safety, reliability and user experience of charging pile operation. Summary of the invention

[0005] Technical problems to be solved: Often only simple electrical parameter monitoring can be performed, and the monitoring of key safety factors such as temperature, smoke, flame, and charging gun plugging and pulling force is not comprehensive or accurate enough.

[0006] In view of the deficiencies in the prior art, the present invention provides a charging pile safety monitoring method and system, thereby solving the technical problems mentioned in the background technology.

[0007] To achieve the above objectives, the present invention is implemented through the following technical solutions: A charging pile safety monitoring method comprises the following steps: Step 1: Monitoring steps: Step 1.1, leakage monitoring, install high-sensitivity leakage sensors in the charging pile and charging gun lines to monitor the leakage current in real time and continuously, with a monitoring frequency of collecting data every 10 milliseconds; Step 1.2, overload and short circuit monitoring, using current transformer to monitor the real-time current change of the charging line, with a collection frequency of once every 5 milliseconds; Step 1.3: Temperature monitoring: Install temperature sensors at the key heating parts of the power module and transformer inside the charging pile, and collect temperature data once a minute; Step 1.4, smoke and flame monitoring, the smoke sensor installed inside the equipment detects the smoke concentration every second, and the infrared flame detector scans the flame spectrum every 0.5 seconds; Step 1.5, charging gun safety monitoring, the plugging and unplugging force sensor in the charging gun records the plugging and unplugging force data every 0.1 second during the plugging and unplugging process, and the charging gun locking state monitor continuously monitors the locking state; Step 1.6: Visual monitoring: Install high-definition cameras above and around key locations of the charging piles to monitor and film the charging piles and their surrounding areas 24 hours a day, with a frame rate of 25fps. Through image storage and transmission, the captured video images are stored in real time. The storage duration is set to at least 2 weeks of monitoring video. At the same time, the real-time monitoring images are transmitted to the remote monitoring center at a rate of 10 frames per second. Step 2: Analysis steps: Step 2.1, leakage status analysis, when the leakage current exceeds the set threshold, record the abnormal data point; Step 2.2: overload and short circuit status analysis: determine the overload or short circuit situation based on the current transformer monitoring results and mark the abnormal data; Step 2.3: Temperature status analysis: record data when the temperature exceeds the normal operating temperature range or the danger threshold; Step 2.4, smoke and flame status analysis, when the smoke concentration reaches the set value or the flame detector recognizes the flame spectrum characteristics, record the relevant data; Step 2.5: Analyze the safety status of the charging gun. If the insertion and extraction force exceeds the normal range or the locking state is abnormal, mark it as abnormal. Step 2.6: Vehicle identification and management analysis: Use convolutional neural networks to identify vehicles entering the monitoring area, record vehicle information and associate it with charging records, and issue warning signals for unauthorized vehicles. Step 2.7: Personnel behavior analysis: extract and classify the characteristics of personnel behavior through the frame difference method, and issue an alarm when the personnel behavior is abnormal; Step 2.8: Foreign object detection and analysis: extract foreign object features and compare and identify them with pre-defined foreign object type features. If it is determined that charging safety is affected, an alarm will be issued. Step 3: Feedback steps: Step 3.1: Feedback from maintenance personnel: When a serious fault is detected, the fault information will be sent to the maintenance personnel. Charging pile operation data reports will be generated for the maintenance personnel on a regular basis. After the charging pile is repaired or upgraded, the equipment status update information will be fed back. Step 3.2: User feedback: Send charging pile safety information to users, update charging pile charging status information on user APP in real time, and promptly feedback user remote control operation results; Step 4: Execution steps: Step 4.1: Automatically execute protection measures. According to the monitoring and analysis results, the system automatically executes corresponding protection measures, such as cutting off the power supply, reducing the charging power, and starting the heat dissipation or fire extinguishing device; Step 4.2: Manual intervention is performed. On-site staff conducts manual intervention operations, and remote management personnel issue operation instructions. After the abnormal situation is handled, the fault handling process and results are recorded.

[0008] In a possible implementation, a charging pile safety monitoring system is implemented in the above-mentioned charging pile safety monitoring method, and the system includes a sensor monitoring module, a visual monitoring module, a central processing module, and a user APP module: The sensor monitoring module is mainly responsible for real-time monitoring of the electrical, temperature and environmental key parameters of the charging pile; The sensor monitoring module includes a leakage detection unit, which uses a high-sensitivity leakage sensor to accurately monitor the leakage current in real time, and trigger an alarm and cut off the power supply when the threshold is exceeded; The sensor monitoring module includes an overload and short-circuit protection unit, which uses a current transformer to monitor current changes and cuts off the circuit through an intelligent circuit breaker when overloaded or short-circuited. The circuit breaker has an automatic reclosing function; The sensor monitoring module includes a temperature monitoring and control unit, which uses a temperature sensor to monitor the temperature, starts heat dissipation control when the temperature exceeds the normal range, and reduces the charging power or stops charging and alarms when the temperature exceeds the dangerous threshold; The sensor monitoring module includes a smoke and flame monitoring unit, which uses a smoke sensor and an infrared flame detector for real-time monitoring and triggers an alarm when an abnormality is detected; The sensor monitoring module includes a charging gun safety monitoring unit, which uses an insertion force sensor and a locking state monitor to monitor the state of the charging gun and sends an alarm signal when an abnormality occurs; The visual monitoring module is used to conduct comprehensive monitoring and intelligent analysis of the charging pile and its surrounding areas; The visual monitoring module includes a high-definition camera unit, which is installed at a key position around the charging pile to perform all-round monitoring and shooting of the charging pile and its surrounding area; The visual monitoring module includes an image storage and transmission unit, which is connected to a high-definition camera to store video images in real time and transmit them to a remote monitoring center; The visual monitoring module includes an intelligent image analysis unit, which is used to perform vehicle identification and management, personnel behavior analysis, and foreign object detection to determine whether it affects charging safety; The central processing module is used to receive and integrate monitoring data, establish a database, manage system communications, and communicate with the user APP module; The user APP module is used to facilitate users to display the charging status in real time, receive safety information, and remotely control the charging pile.

[0009] In a possible implementation, the vehicle identification and management uses a normalized cross-correlation method to calculate the similarity between the features of the vehicle to be identified and the features of the template in the database, and the formula is: ,in, is the template image, is the image to be matched, is the matching position, The closer the value is to 1, the higher the matching degree is; For neural network recognition, a convolutional neural network is used, assuming that the input feature map is , where the height ,width , Number of channels , the convolution kernel is , where the convolution kernel size , number of output channels , output feature map ,in ,in is the step size, and the formula for convolution calculation is: During the training process, the parameters of the convolution kernel are updated through the back propagation algorithm, using the cross entropy loss function ,in is the true label, It is the prediction probability. By training a large amount of vehicle image data, the network learns the feature representation of the vehicle and directly outputs the vehicle category and license plate number information in the recognition stage.

[0010] In a possible implementation, the personnel behavior analysis is implemented by obtaining the movement trajectory of the personnel through the frame difference method. and It is two adjacent frames, the difference image between frames Can be Get, through Morphological processing and connected domain analysis can be performed to obtain the movement area and trajectory of the person, and the person's behavior can be classified according to the extracted behavioral features, such as normal walking, staying, and approaching a vehicle.

[0011] In a possible implementation, the foreign body detection is implemented by calculating contrast, correlation, and energy parameters to describe texture features. Specifically, it is assumed that the training data set ,in is the eigenvector, is the category label, and the goal of SVM is to find a hyperplane , so that the interval between the two types of data is the largest, the interval is defined as , by solving the optimization problem , the constraints are , , for nonlinear problems, through the kernel function Map the data to a high-dimensional space so that it can be linearly separated in the high-dimensional space. Compare and identify the extracted foreign body features with the pre-defined foreign body type features. Separate different categories of data by finding an optimal hyperplane. The decision function is: ,in is the Lagrange multiplier, is the sample category label, is the offset. The parameters of the classifier are obtained through training sample learning, so as to classify foreign objects and determine whether there are foreign objects that affect charging safety.

[0012] Beneficial effects compared with the prior art: 1. In this solution, through the monitoring of leakage, overload and short circuit, temperature, smoke and flame, charging gun safety and vision, combined with intelligent analysis, the operating status and surrounding environment of the charging pile can be fully, timely and accurately grasped. Whether it is an electrical fault, abnormal temperature, or potential safety hazards, it can be quickly detected and judged, realizing all-round real-time monitoring of safety hazards of charging piles. Once an abnormality is found, the corresponding protection measures are automatically executed immediately, such as cutting off the power supply, starting the fire extinguishing device, etc., effectively avoiding the expansion of the fault, greatly improving the safety of the operation of the charging pile, and ensuring the safety of users, equipment and the surrounding environment; 2. This solution provides operation and maintenance personnel with a wealth of data reports, fault alarm push notifications, and equipment status update reminders, so that they can understand the equipment status in a timely manner, reasonably plan inspection and maintenance work, and prevent potential problems in advance, thereby improving operation and maintenance efficiency, reducing equipment failure rates, and ensuring the continuous and stable operation of charging piles. At the same time, users can obtain charging pile status information in real time, receive safety prompts and remote control feedback, facilitate itinerary planning, enhance users' sense of control and security over the charging process, improve user satisfaction, and optimize the overall charging service experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] The above description is only an overview of the technical solution of the present invention. In order to more clearly understand the technical means of the present invention and implement it according to the contents of the specification, the following is a detailed description of the preferred embodiments of the present invention in conjunction with the accompanying drawings.

[0014] Figure 1 A flowchart of a charging pile safety monitoring method; Figure 2 This is a system framework diagram of a charging pile safety monitoring system. DETAILED DESCRIPTION

[0015] Preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. However, the present invention can also be implemented in various forms, so the present invention is not limited to the embodiments described below. The technical solution in the embodiment of the present application is to solve the problems of the above-mentioned background technology, and the overall idea is as follows: Example 1

[0016] Please refer to Figure 1 As shown, this embodiment introduces a charging pile safety monitoring method, including the following steps: Step 1. Monitoring: Step 1.1, leakage monitoring, install a high-sensitivity leakage sensor (for example, model LD-01, with an accuracy of up to ±0.1mA) in the charging pile and charging gun line, and monitor the leakage current in real time and continuously. The monitoring frequency is to collect data every 10 milliseconds; Step 1.2: Overload and short circuit monitoring: Use a current transformer (ratio of 2000:5) to monitor the real-time current change of the charging line, with a collection frequency of once every 5 milliseconds. Step 1.3: Temperature monitoring: Install temperature sensors (such as PT100, measuring range -50℃-150℃, accuracy ±0.5℃) at key heating parts such as power modules and transformers inside the charging pile, and collect temperature data once a minute; Step 1.4, smoke and flame monitoring, the smoke sensor installed inside the equipment (such as an ion smoke sensor that can detect a change in smoke concentration of 0.1% obs / m) detects the smoke concentration every second, and the infrared flame detector scans the flame spectrum every 0.5 seconds; Step 1.5, charging gun safety monitoring, the plugging and unplugging force sensor (such as strain gauge type, accuracy ±0.3N) in the charging gun records the plugging and unplugging force data every 0.1 second during the plugging and unplugging process, and the charging gun locking state monitor (such as Hall sensor) continuously monitors the locking state; Step 1.6: Visual monitoring. Install high-definition cameras (such as Hikvision, model DS-2CD3T47WD-L, with wide dynamic range and low-light imaging functions) above and around key locations of the charging piles. Monitor and shoot the charging piles and their surrounding areas 24 hours a day with a frame rate of 25fps to ensure that image information can be clearly captured under different lighting conditions (such as direct strong light, dim night, etc.). Through image storage and transmission, the captured video images are stored in real time. The storage time is set to save at least 2 weeks of monitoring video. At the same time, the real-time monitoring images are transmitted to the remote monitoring center at a rate of 10 frames per second via wired Ethernet (RJ45 interface) or wireless network (such as WiFi-6, supporting 2.4GHz and 5GHz frequency bands), so that managers can check the on-site situation of the charging piles at any time.

[0017] Step 2: Analysis Step 2.1, leakage status analysis, assuming that at 14:30:20 on November 15, 2023, the sensor collects a leakage current value of 33mA (exceeding the set threshold of 30mA), and records this time and leakage current value as abnormal data points.

[0018] Step 2.2: Overload and short-circuit status analysis: If at 14:30:20 on November 15, 2023, the current transformer detects that the current instantly rises to 48A (exceeding the short-circuit current setting value of 45A), record the time-varying data and time point; or when the current continuously exceeds 120% of the rated current of the charging pile (assuming the rated current is 15A, that is, exceeds 18A) for 6 seconds (such as during 14:30:15-14:30:21), it is determined to be an overload condition and the corresponding abnormal data is marked; Step 2.3, temperature status analysis, assuming that the power module temperature is 72°C (exceeding the normal operating temperature range of 70°C) at 14:31:00 on November 15, 2023, this data is recorded. If the temperature continues to rise to 93°C (exceeding the dangerous threshold of 90°C), such as at 14:32:00, it is marked as high temperature abnormal data; Step 2.4: Smoke and flame status analysis. For example, at 14:30:30.500 on November 15, 2023, the smoke sensor detects that the smoke concentration reaches 0.6% obs / m, and the flame detector recognizes similar flame spectral characteristics, and records the time and related data; Step 2.5: Analyze the safety status of the charging gun. If the plugging and unplugging force reaches 22N (exceeding the normal range of 5-20N) at 14:30:05.300 during the plugging and unplugging operation starting at 14:30:05 on November 15, 2023, it is marked as abnormal. If the Hall sensor detects an abnormal locking state (locking signal disappears) at 14:31:30 during the charging process, record this situation. Step 2.6, vehicle identification and management analysis. When a vehicle enters the monitoring area, the high-definition camera captures the vehicle image, and the intelligent image analysis system uses a convolutional neural network (CNN) to identify the vehicle. For example, at 14:35:00 on November 15, 2023, a car with a license plate of Guangdong A12345 drove into the charging station. After the camera captured the image, the system used the license plate recognition module in the trained CNN model to segment and extract features of the license plate characters, and finally accurately identified the license plate number, and recorded the vehicle model (such as SUV), color (white) and other information, and associated this information with the charging record. If the vehicle is not in the authorized vehicle list, the system automatically issues a warning signal and displays "Unauthorized vehicle Guangdong A12345 enters the charging area" on the display screen of the remote monitoring center; Step 2.7: Personnel behavior analysis: extract the characteristics of personnel behavior in the image by using the inter-frame difference method. Assume and It is two adjacent frames, the difference image between frames Can be Get, through Morphological processing (such as dilation and erosion) and connected domain analysis are performed to obtain the movement area and trajectory of the person. For example, during the period of 14:40:00-14:40:10, a person is detected wandering around the vehicle, and his / her movement trajectory is displayed on the monitoring screen. The person's behavior is classified according to the extracted behavior features. If the person stays around the vehicle for more than the set threshold (such as 3 minutes) and is close to the vehicle charging port (less than 0.5 meters away), it is judged as abnormal behavior and the system issues an alarm; Step 2.8, foreign body detection and analysis, extract the shape, texture, color and other features of the foreign body. The shape features are described by calculating the geometric parameters such as the perimeter, area, and circularity of the object's contour. The texture features are extracted through the gray-level co-occurrence matrix (GLCM). The frequency of different gray value combinations in the image in a certain direction and distance is counted. The texture features are described by calculating parameters such as contrast, correlation, and energy. For example, at 14:45:00, the camera captured an unidentified object on the ground near the charging pile. After the system extracts its features, it compares and identifies them with the pre-defined foreign body type features (such as flammable objects, obstacles, etc.). If it is determined that the object may affect charging safety (such as being identified as a flammable object), an alarm prompt is issued; Step 3: Feedback Step 3.1: Feedback from maintenance personnel Step 3.1.1, fault alarm push. When serious faults such as leakage, overload, short circuit, high temperature, smoke, flame, etc. are detected, the central processing module pushes notifications through SMS gateway, mail server or dedicated operation and maintenance APP, and sends fault information to the mobile phone or computer of the operation and maintenance personnel. For example, a short circuit fault of the charging pile is detected at 14:30:20 on November 15, 2023. The central processing module immediately sends a text message to Mr. Li who is responsible for the maintenance of the charging piles in the area, with the content "[Specific charging station name] Charging pile No. 001 had a short circuit fault at 2023-11-15 14:30:20, please go to handle it as soon as possible." At the same time, detailed fault information is displayed in a pop-up window on the operation and maintenance APP, including the short-circuit current value (such as 48A), the location of the fault (such as the power module inside the charging pile) and possible causes of the fault (such as component short circuit, line damage, etc.); Step 3.1.2: Provide data reports. Generate charging pile operation data reports for maintenance personnel on a regular basis (such as daily or weekly). The content covers charging capacity, charging time, number of various faults and time distribution, temperature change trend, current and voltage fluctuation, etc. For example, after logging into the operation and maintenance management system every Monday morning, maintenance personnel Zhang Gong can view the data reports of charging piles in the area he was responsible for last week, understand the frequency of use of charging piles by analyzing charging capacity data, formulate reasonable inspection plans based on the number of faults and time distribution, and predict possible overheating problems in advance based on temperature change trends, so as to carry out preventive maintenance on the equipment; Step 3.1.3, device status update reminder. When the charging pile is repaired or the system is upgraded, the central processing module will feedback the device status update information to the maintenance personnel to ensure that they understand the current available status of the equipment. For example, after the charging pile with a short circuit fault was repaired on November 16, 2023, the central processing module sent a notification to Mr. Li, informing him that "the charging pile No. 001 has been repaired and the equipment is running normally after testing and can be put into use." At the same time, the status of the charging pile is marked as "normal operation" in the operation and maintenance management system to facilitate subsequent management by the operation and maintenance personnel; Step 3.2: User feedback Step 3.2.1, push safety information, push messages or SMS notifications through mobile phone APP, send charging pile safety information to users. When the charging pile has leakage, overload, short circuit and other fault alarms, or detects abnormal operation of the charging gun, abnormal surrounding environment (foreign objects or people are abnormally close), the user's mobile phone APP will immediately receive a pop-up reminder, accompanied by a sound prompt. For example, when user Xiao Wang is using the charging pile to charge, if the charging gun has abnormal plugging and unplugging force, his mobile phone APP will pop up a prompt "The charging pile charging gun you are using has abnormal plugging and unplugging force. Please check whether the operation is correct. If you have any questions, please contact the staff." If smoke is detected near the charging pile, the APP will push "[Specific charging station name] Smoke is detected near the charging pile. Please pay attention to safety. If necessary, please stop charging and stay away from the scene." Step 3.2.2, charging status update, real-time update of the charging status information of the charging pile displayed on the user's APP, including whether it is idle, charging current, voltage, estimated full charge time, charged amount, etc. For example, user Xiao Zhang binds the commonly used charging pile on the APP. When he is ready to go to charge, he can check in advance whether the charging pile is idle. If it is charging (may be used by other family members), he can see that the current charging current is 15A, the voltage is 220V, the charged amount is 30kWh, and the estimated full charge time is 1 hour and 30 minutes, which is convenient for him to arrange his trip reasonably; Step 3.2.3, remote control feedback. When the user remotely controls the charging pile through the APP (such as remotely stopping charging, triggering the emergency stop button, etc.), the APP will promptly feedback the operation results to the user. For example, user Xiao Li receives a reminder that charging is complete but cannot arrive at the site in time. After he remotely stops charging through the APP, the APP will display "The remote stop charging command has been sent, and the charging pile has stopped charging." If the user finds an abnormality during the charging process and remotely triggers the emergency stop button, the APP will prompt "The emergency stop command has been executed, and the charging pile has stopped working. To ensure safety, please use it again after troubleshooting." At the same time, the central processing module will record the user's remote control operation information for subsequent query and analysis; Step 4: Execution Step 4.1, automatically execute protection measures. According to the monitoring and analysis results, the system automatically executes corresponding protection measures. For electrical faults such as leakage, overload, short circuit, etc., the intelligent circuit breaker (such as Schneider brand, model IC65N) immediately automatically cuts off the power supply of the charging pile to prevent the fault from further expanding and protect the charging pile and household circuit from damage. For high temperature conditions, when the temperature exceeds the dangerous threshold (such as 90°C), the system automatically reduces the charging power or stops charging, and starts the cooling fan at full speed (such as the cooling fan speed is increased from 1500rpm to 3000rpm), and opens the solenoid valve of the cooling air duct to enhance the heat dissipation effect and prevent the equipment from being damaged or even causing a fire due to overheating. When the smoke or flame sensor detects an abnormality, the fire extinguishing device (such as a dry powder fire extinguisher or a gas fire extinguishing system, depending on the configuration of the charging station) is automatically started to perform fire extinguishing operations, and an alarm signal is sent to the fire department (such as connecting to the fire alarm platform through the network); Step 4.2: Manual intervention (depending on the situation). After checking the on-site situation, the on-site staff will perform manual intervention operations according to actual needs. For example, after eliminating the leakage fault, manually reset the intelligent circuit breaker to restore the power supply of the charging pile; clean up foreign objects around the charging pile to ensure a safe charging environment; inspect and repair the charging gun (such as replacing a damaged plug-in force sensor or locking device), etc. The remote management personnel will issue further operation instructions based on the information viewed, such as notifying the maintenance personnel to bring professional tools to the site for equipment maintenance; adjust the operating parameters of the charging pile (such as charging power limit, temperature threshold setting, etc.) to optimize the performance of the charging pile; arrange staff to conduct a comprehensive inspection of the charging station to find other possible safety hazards, etc. After the abnormal situation is handled, the management personnel will confirm that the fault is eliminated through the monitoring center system and the system has returned to normal operation. The fault handling process and results will be recorded to form a fault handling report to provide a reference for subsequent equipment maintenance and management. Example 2

[0019] Please refer to Figure 2 As shown, this embodiment introduces a charging pile safety monitoring system implemented in Embodiment 1, which introduces a charging pile safety monitoring method, including the following modules: 1. Sensor monitoring module 1.1. Leakage detection unit High-sensitivity leakage sensors are installed in the charging pile and charging gun circuits to accurately monitor leakage current in real time. Once the leakage current exceeds the set threshold (such as 30mA), an alarm signal is immediately triggered and the power supply of the charging pile is quickly cut off to effectively prevent electric shock accidents. 1.2. Overload and short circuit protection unit Use current transformers to monitor real-time current changes in the charging line. When the current exceeds a certain percentage of the rated current of the charging pile (such as 120%), it is determined to be an overload situation; if the current increases sharply and exceeds the short-circuit current setting value, it is determined to be a short-circuit fault. According to the monitoring results of the current transformer, at the moment of overload or short circuit, the intelligent circuit breaker quickly cuts off the circuit to protect the charging pile and household circuits from damage. This circuit breaker has an automatic reclosing function and can automatically restore power supply after the fault is eliminated, but there is a certain delay time (such as 3-5 minutes) and a reclosing limit (such as 3-5 times) to avoid repeated faults causing dangers; 1.3. Temperature monitoring and control unit Temperature sensors are used in key heating parts of the charging pile, such as power modules and transformers, to monitor temperature changes in real time. When the temperature exceeds the set normal operating temperature range (such as 70°C), heat dissipation control measures are initiated. When the temperature rises, the cooling fan automatically speeds up to increase air circulation to remove heat. The heat sink uses high-efficiency thermal conductive materials to increase the heat dissipation area and assist in heat dissipation. If the temperature continues to rise and exceeds the dangerous threshold (such as 90°C), the charging power is reduced or charging is stopped, and a high temperature alarm signal is issued to prevent equipment damage or even fire due to overheating; 1.4. Smoke and flame monitoring unit The smoke sensor installed inside the device and the infrared flame detector are used to monitor the environment around the charging pile in real time. The smoke concentration changes in the air and the flame spectrum characteristics can be sensed. Once the set characteristics are detected, an alarm signal is immediately triggered to remind the surrounding personnel to pay attention to possible fire hazards. 1.5. Charging gun safety monitoring unit The charging gun plugging and unplugging force sensor installed inside the charging gun is used to accurately monitor the force applied during the plugging and unplugging process. When the plugging and unplugging force exceeds the normal range, an abnormal operation alarm signal is promptly issued to remind the user to operate the charging gun correctly to avoid damage to the charging gun or charging interface due to improper force. In addition, the charging gun locking state monitor is used to detect the working state of the charging gun locking device in real time to ensure that the charging gun is reliably locked during the charging process. If the locking device fails or is accidentally unlocked, charging is immediately stopped and an alarm signal is issued to prevent the charging gun from accidentally falling off during the charging process and ensure charging safety. 2. Visual monitoring module 2.1. High-definition camera unit: installed above the charging pile or at key positions around it, it can conduct all-round monitoring and shooting of the charging pile and its surrounding areas. It has wide dynamic range and low-light imaging functions to ensure that image information can be clearly captured under different lighting conditions (such as direct strong light, dim night, etc.).

[0020] 2.2 Image storage and transmission unit: connected to a high-definition camera, it can store the captured video images in real time, and the storage duration can be set according to demand (such as saving surveillance videos for at least 1 week). At the same time, it has the ability to transmit real-time surveillance images or video clips triggered by specific events to a remote monitoring center via a wired or wireless network, so that managers can check the on-site conditions of the charging piles at any time.

[0021] 2.3 Intelligent Image Analysis Unit A high-definition camera is used to collect images of the charging pile and its surrounding environment. The high-definition camera is installed inside the charging pile and can monitor and shoot the area in front of the charging pile. It has a wide dynamic range and low-light imaging function to ensure that image information can be clearly captured under different lighting conditions (such as direct strong light, dim night, etc.); 2.3.1 Vehicle identification and management: Automatically identify vehicles entering the monitoring area, record the vehicle's license plate number, model, color and other information, and associate it with the charging record for easy traceability and management. When an unauthorized vehicle enters a specific charging area, the system automatically issues a warning signal and extracts vehicle features from the collected images. Common features include vehicle outlines, license plate positions, vehicle color histograms, etc. For example, for license plate recognition, edge detection algorithms (such as Sobel operators and Canny operators) can be used to locate the edges of license plates. The templates of the Sobel operator in the horizontal and vertical directions are: ; The horizontal and vertical gradients are obtained by convolution with the image, and then the edge position is determined. The vehicle color histogram can be obtained by counting the number of pixels in different color intervals in the image, which is used for vehicle color recognition; The extracted vehicle features are matched with the pre-established vehicle feature database. This database may contain information such as various vehicle models and license plate styles. Template matching, neural network and other methods may be used for identification. For example, in simple template matching, the similarity between the vehicle features to be identified and the template features in the database is calculated, such as using the normalized cross correlation (NCC) method. The formula is: ;in, is the template image, is the image to be matched, is the matching position, The closer the value is to 1, the higher the matching degree is. For neural network-based recognition, a convolutional neural network (CNN) is used. Assuming that the input feature map is (high ,width , Number of channels ), the convolution kernel is (Convolution kernel size , number of output channels ), output feature map ,in ( is the step length), the formula for convolution calculation is: During the training process, the parameters of the convolution kernel are updated through the back propagation algorithm to minimize the loss function (such as the cross entropy loss function , in is the true label, is the prediction probability), which allows the network to learn the characteristic representation of vehicles by training a large amount of vehicle image data, and directly outputs the vehicle category, license plate number and other information in the recognition stage; When a car drives into the charging station, the high-definition camera captures the image of the vehicle. The license plate recognition module in the intelligent image analysis system separates the license plate characters from the background through the segmentation model of the convolutional neural network (CNN). It can learn the difference between the characters and the background and accurately locate the boundary of each character. Then, the segmented characters are extracted. The CNN model analyzes the stroke structure, shape ratio and other feature information of the characters. Finally, these features are matched and recognized with the pre-trained character library to obtain the complete license plate number. 2.3.2. Personnel behavior analysis is implemented by extracting the characteristics of personnel behavior in the image, including the personnel's motion trajectory, posture, movement speed, etc. The motion trajectory of the personnel is obtained by the inter-frame difference method. Assume and It is two adjacent frames, the difference image between frames Can be Get, through Morphological processing (such as dilation and erosion) and connected domain analysis can be performed to obtain the movement area and trajectory of the person. The person's behavior can be classified according to the extracted behavior characteristics, such as normal walking, staying, approaching vehicles, etc. Abnormal behavior detection can be achieved by establishing a behavior rule model; For example, if a person stays around a vehicle for too long and moves close to key parts of the vehicle (such as the charging port), it can be judged as abnormal behavior. This can be judged by setting time thresholds and spatial distance thresholds. Greater than the set threshold , and the distance from the charging port When less than the threshold , determined as abnormal behavior; 2.3.3, foreign body detection implementation, extract the shape, texture, color and other features of the foreign body. The shape features can be described by calculating the geometric parameters such as the contour circumference, area, and circularity of the object. For example, the circularity can be To calculate, is the area of ​​the object, is the circumference of the object. For a circular object , the more irregular the shape, the smaller the texture features can be extracted through the gray-level co-occurrence matrix (GLCM), which counts the frequency of different gray value combinations in the image in a certain direction and distance, and describes the texture features by calculating parameters such as contrast, correlation, and energy. Specifically, it is assumed that the training data set ,in is the eigenvector, is the category label, and the goal of SVM is to find a hyperplane , so that the interval between the two types of data is the largest, the interval is defined as , by solving the optimization problem , the constraints are , for nonlinear problems, through the kernel function Map the data to a high-dimensional space so that it can be linearly separated in the high-dimensional space. Compare and identify the extracted foreign body features with the pre-defined foreign body type features (such as flammable objects, obstacles, etc.). Separate different categories of data by finding an optimal hyperplane. The decision function is: ,in is the Lagrange multiplier, is the sample category label, is the offset. The parameters of the classifier are obtained through training sample learning, so as to classify foreign objects and determine whether there are foreign objects that affect charging safety.

[0022] 3. Central processing module Receive real-time data from the sensor monitoring module and the visual monitoring module, including leakage current value, line current, temperature data, smoke concentration, flame signal, charging gun plugging and pulling force and locking status, visual monitoring image analysis results, etc., and integrate and process them to establish a complete charging pile operation status database. At the same time, it is responsible for the communication management between the modules in the entire system to ensure stable, efficient and safe data transmission. It coordinates the working sequence between different monitoring modules and execution units to avoid data conflicts and command confusion, ensure the overall operation smoothness and reliability of the system, and communicate with the user APP module to realize data sharing and remote control functions; 5. User APP module Users register an account and log in to the APP through their mobile phone number. They can bind multiple commonly used charging piles for convenient management. The charging status of the charging piles that users are concerned about is displayed in real time, including whether it is idle, charging current, voltage, expected full time, and charged amount, etc., which is convenient for users to plan their trips in advance and understand the charging progress. They receive safety information pushed from the central processing module, such as charging pile failure alarms (leakage, overload, short circuit, high temperature, smoke, flame, etc.), abnormal operation prompts of charging guns, abnormal surrounding environment (foreign objects or people approaching abnormally), etc., to ensure that users are aware of the safety status during the charging process in a timely manner. Users can view detailed safety event information and historical records in the APP. In certain circumstances, users can remotely control the charging pile through the APP. For example, when receiving a charging completion reminder but unable to reach the charging pile site in time, charging can be stopped remotely; if an abnormal situation is found during the charging process and there is no personnel on site to handle it, the emergency stop button can be triggered remotely.

[0023] Finally, it should be noted that: Obviously, the above embodiments are only examples for clearly explaining the present invention, and are not intended to limit the implementation methods. For ordinary technicians in the relevant field, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the implementation methods here. The obvious changes or modifications derived from this are still within the scope of protection of the present invention.

Claims

1. A charging pile safety monitoring method, characterized in that: The following steps are involved: Step 1: Monitoring step, conduct leakage, overload and short circuit, temperature, smoke and flame, charging gun safety, and visual multi-faceted monitoring; Step 2: Analysis steps: Step 2.1, leakage status analysis, when the leakage current exceeds the set threshold, record the abnormal data point; Step 2.2: overload and short circuit status analysis: determine the overload or short circuit situation based on the current transformer monitoring results and mark the abnormal data; Step 2.3: Temperature status analysis: record data when the temperature exceeds the normal operating temperature range or the danger threshold; Step 2.4, smoke and flame status analysis, when the smoke concentration reaches the set value or the flame detector recognizes the flame spectrum characteristics, record the relevant data; Step 2.5: Analyze the safety status of the charging gun. If the insertion and extraction force exceeds the normal range or the locking state is abnormal, mark it as abnormal. Step 2.6: Vehicle identification and management analysis: Use convolutional neural networks to identify vehicles entering the monitoring area, record vehicle information and associate it with charging records, and issue warning signals for unauthorized vehicles. Step 2.7: Personnel behavior analysis: extract and classify the characteristics of personnel behavior through the frame difference method, and issue an alarm when the personnel behavior is abnormal; Step 2.8: Foreign object detection and analysis: extract foreign object features and compare and identify them with pre-defined foreign object type features. If it is determined that charging safety is affected, an alarm will be issued. Step 3: Feedback step: provide feedback to operation and maintenance personnel and users; Step 4: Execute the steps to perform automatic protection and manual intervention.

2. A charging pile safety monitoring method according to claim 1, characterized in that: The step 1 specifically includes the following sub-steps: Step 1.1, leakage monitoring, install high-sensitivity leakage sensors in the charging pile and charging gun lines to monitor the leakage current in real time and continuously, with a monitoring frequency of collecting data every 10 milliseconds; Step 1.2, overload and short circuit monitoring, using current transformer to monitor the real-time current change of the charging line, with a collection frequency of once every 5 milliseconds; Step 1.3: Temperature monitoring: Install temperature sensors at the key heating parts of the power module and transformer inside the charging pile, and collect temperature data once a minute; Step 1.4, smoke and flame monitoring, the smoke sensor installed inside the equipment detects the smoke concentration every second, and the infrared flame detector scans the flame spectrum every 0.5 seconds; Step 1.5, charging gun safety monitoring, the plugging and unplugging force sensor in the charging gun records the plugging and unplugging force data every 0.1 second during the plugging and unplugging process, and the charging gun locking state monitor continuously monitors the locking state; Step 1.6: Visual monitoring. Install high-definition cameras above and around key locations of the charging piles to monitor and film the charging piles and their surrounding areas 24 hours a day. The frame rate is 25fps. Through image storage and transmission, the captured video images are stored in real time. The storage time is set to save monitoring videos for at least 2 weeks. At the same time, the real-time monitoring images are transmitted to the remote monitoring center at a rate of 10 frames per second.

3. A charging pile safety monitoring method according to claim 1, characterized in that: The step 3 specifically includes the following sub-steps: Step 3.1: Feedback from maintenance personnel: When a serious fault is detected, the fault information will be sent to the maintenance personnel. Charging pile operation data reports will be generated for the maintenance personnel on a regular basis. After the charging pile is repaired or upgraded, the equipment status update information will be fed back. Step 3.2: User feedback: Send charging pile safety information to users, update charging pile charging status information on user APP in real time, and promptly feedback user remote control operation results.

4. A charging pile safety monitoring method according to claim 1, characterized in that: The step 4 specifically includes the following sub-steps: Step 4.1: Automatically execute protection measures. According to the monitoring and analysis results, the system automatically executes corresponding protection measures, such as cutting off the power supply, reducing the charging power, and starting the heat dissipation or fire extinguishing device; Step 4.2: Manual intervention is performed. On-site staff conducts manual intervention operations, and remote management personnel issue operation instructions. After the abnormal situation is handled, the fault handling process and results are recorded.

5. A charging pile safety monitoring system implemented in a charging pile safety monitoring method according to any one of claims 1 to 4, the system comprising a sensor monitoring module, a visual monitoring module, a central processing module, and a user APP module, characterized in that: The sensor monitoring module is mainly responsible for real-time monitoring of the electrical, temperature and environmental key parameters of the charging pile; The sensor monitoring module includes a leakage detection unit, which uses a high-sensitivity leakage sensor to accurately monitor the leakage current in real time, and trigger an alarm and cut off the power supply when the threshold is exceeded; The sensor monitoring module includes an overload and short-circuit protection unit, which uses a current transformer to monitor current changes and cuts off the circuit through an intelligent circuit breaker when overloaded or short-circuited. The circuit breaker has an automatic reclosing function; The sensor monitoring module includes a temperature monitoring and control unit, which uses a temperature sensor to monitor the temperature, starts heat dissipation control when the temperature exceeds the normal range, and reduces the charging power or stops charging and alarms when the temperature exceeds the dangerous threshold; The sensor monitoring module includes a smoke and flame monitoring unit, which uses a smoke sensor and an infrared flame detector for real-time monitoring and triggers an alarm when an abnormality is detected; The sensor monitoring module includes a charging gun safety monitoring unit, which uses an insertion force sensor and a locking state monitor to monitor the state of the charging gun and sends an alarm signal when an abnormality occurs; The visual monitoring module is used to conduct comprehensive monitoring and intelligent analysis of the charging pile and its surrounding areas; The visual monitoring module includes a high-definition camera unit, which is installed at a key position around the charging pile to perform all-round monitoring and shooting of the charging pile and its surrounding area; The visual monitoring module includes an image storage and transmission unit, which is connected to a high-definition camera to store video images in real time and transmit them to a remote monitoring center; The visual monitoring module includes an intelligent image analysis unit, which is used to perform vehicle identification and management, personnel behavior analysis, and foreign object detection to determine whether it affects charging safety; The central processing module is used to receive and integrate monitoring data, establish a database, manage system communications, and communicate with the user APP module; The user APP module is used to facilitate users to display the charging status in real time, receive safety information, and remotely control the charging pile.

6. A charging pile safety monitoring system as claimed in claim 5, characterized in that: The vehicle identification and management uses the normalized cross-correlation method to calculate the similarity between the features of the vehicle to be identified and the template features in the database. The formula is: ,in, is the template image, is the image to be matched, is the matching position, The closer the value is to 1, the higher the matching degree is; For neural network recognition, a convolutional neural network is used, assuming that the input feature map is , where the height ,width , Number of channels , the convolution kernel is , where the convolution kernel size , number of output channels , output feature map ,in , ,in is the step size, and the formula for convolution calculation is: During the training process, the parameters of the convolution kernel are updated through the back propagation algorithm, using the cross entropy loss function ,in is the true label, It is the prediction probability. By training a large amount of vehicle image data, the network learns the feature representation of the vehicle and directly outputs the vehicle category and license plate number information in the recognition stage.

7. A charging pile safety monitoring system as claimed in claim 5, characterized in that: The personnel behavior analysis is implemented by obtaining the movement trajectory of personnel through the frame difference method. and It is two adjacent frames of images, the difference image between frames Can be Get, through Morphological processing and connected domain analysis can be performed to obtain the movement area and trajectory of the person, and the person's behavior can be classified according to the extracted behavioral features, such as normal walking, staying, and approaching a vehicle.

8. A charging pile safety monitoring system as claimed in claim 5, characterized in that: The foreign body detection is implemented by calculating contrast, correlation, and energy parameters to describe texture features. Specifically, it is assumed that the training data set ,in is the eigenvector, is the category label, and the goal of SVM is to find a hyperplane , so that the interval between the two types of data is the largest, the interval is defined as , by solving the optimization problem , the constraints are , for nonlinear problems, through the kernel function Map the data to a high-dimensional space so that it can be linearly separated in the high-dimensional space. Compare and identify the extracted foreign body features with the pre-defined foreign body type features. Separate different categories of data by finding an optimal hyperplane. The decision function is: ,in is the Lagrange multiplier, is the sample category label, is the offset. The parameters of the classifier are obtained through training sample learning, so as to classify foreign objects and determine whether there are foreign objects that affect charging safety.

Images