Modularized power supply charging abnormity protection method and system

Abstract

The invention discloses a modular power supply charging abnormity protection method and system, and aims to improve the safety, reliability and efficiency of a power supply system. The system can automatically cut off a power supply under various abnormal conditions, trigger alarm and take necessary protection measures. In addition, the intelligent charging algorithm and the real-time monitoring function improve the charging efficiency, prolong the service life of the battery and enhance the user experience. The system is easy to maintain and expand due to the modular design, and the total cost of ownership is reduced. The modular power supply charging abnormity protection system integrated with the isolated forest algorithm significantly improves the abnormity detection capability and reliability of a power supply system. Abnormity such as voltage and current fluctuation in the charging process can be accurately identified. The system ensures that potential problems can be responded immediately through real-time data stream processing and an automatic alarm mechanism, fault response time is shortened, and the overall reliability of the system is improved.

Description

Technical Field

[0001] The present invention relates to the field of power supply charging protection, and in particular to a modular power supply charging abnormality protection method and system. Background Art

[0002] Modern batteries, such as lithium batteries, are very sensitive to the control of the charging process. Excessive charging current or unstable voltage will affect the health of the battery and shorten the battery life. During the power supply charging process, abnormal conditions such as voltage, current, and temperature may occur. If these abnormalities are not handled in time, it may cause the power supply equipment to overheat, damage, or even cause a fire. Modular power supply systems may experience overcurrent, overvoltage, and other conditions during the charging process. If these conditions are not controlled, they may cause the power supply equipment to overload, damage internal components, or reduce the service life of the system. Charging abnormalities in the power supply system not only affect the safety of the equipment, but may also be dangerous to the user. In order to solve these problems, by implementing an abnormal protection system, potential safety risks can be reduced and users can be protected from electrical failures. Summary of the invention

[0003] A modular power supply charging abnormality protection method comprises the following steps: S1. Design protection circuit: Overvoltage protection: Design an overvoltage protection circuit at the input end of the power supply. When the input voltage exceeds the set range, the protection circuit can automatically cut off the power supply or trigger an alarm. Undervoltage protection: When the power supply voltage is lower than the set value, the undervoltage protection is triggered to prevent the equipment from working under low voltage and being damaged. Overcurrent protection: Design an overcurrent protection circuit. When the charging current exceeds the rated value, the power supply is automatically disconnected or an alarm is triggered. Short-circuit protection: Ensure that the power supply system has a short-circuit protection function, and can quickly disconnect the power supply in the event of a short circuit to protect the circuit from damage; S2. Integrated temperature monitoring: Temperature sensor: Install temperature sensors at key parts of the power system to monitor temperature changes in real time. Temperature control protection: Set temperature control thresholds. When the temperature exceeds the preset value, the power supply is automatically disconnected or the cooling system is started to prevent overheating. S3. Data acquisition: Sensor configuration: First, a series of sensors need to be configured to monitor the key parameters of the power module, such as voltage, current, temperature, and data interface: Ensure that these sensors can be connected to the central control unit (such as microcontroller MCU) through a suitable data interface (such as I²C, SPI, UART, etc.) to transmit data in real time; S4. Implement charging algorithm: Use reinforcement learning algorithm to optimize charging strategy. The algorithm can automatically adjust charging parameters based on historical data and real-time monitoring data, predict battery health status and charging demand, and integrate to obtain the formula: , where SoC(t) is the state of charge at time t, SoC(t-1) is the state of charge at the previous time point, Cnom is the nominal capacity of the battery, I(t) is the charging current at time t, Δt is the time interval, k It is the decay constant, which depends on the battery usage and environmental conditions. The charging parameters are adjusted dynamically according to the battery status: voltage, current, and temperature. The charging is carried out in stages, and a staged charging strategy is adopted: constant current charging and constant voltage charging to ensure a safe and efficient charging process. S5. Testing and calibration: Functional test: Perform functional test on the protection circuit to ensure that it can work normally under various abnormal conditions. Calibration: Calibrate the threshold of the protection circuit to ensure its accuracy and reliability. S6. System monitoring and feedback: Real-time monitoring: Realize real-time monitoring of the charging process, record power status, charging current, voltage and other parameters, Alarm mechanism: Design an alarm mechanism to promptly notify the user or automatically stop charging when an abnormality occurs; S7. Maintenance and Updates: Regular Inspection: Regularly inspect and maintain the protection circuit to ensure its continued effectiveness, Software Update: If the power system uses software control, regularly update the software to fix potential vulnerabilities and improve functionality.

[0004] Furthermore, a modular power supply charging abnormality protection method is provided. In step S4, a reinforcement learning algorithm is used to optimize the charging strategy. The algorithm can automatically adjust the charging parameters based on historical data and real-time monitoring data, predict the battery health status and charging demand, and dynamically adjust the charging parameters based on the battery status: voltage, current, and temperature. The specific steps are as follows: S41. Problem definition and goal setting: clarify the goal, optimize the charging strategy through reinforcement learning to improve charging efficiency, extend battery life and reduce charging anomalies, indicators: including charging time, battery health, energy efficiency, temperature control, define the state and action space: including battery voltage, current, temperature, charging stage, historical charging data, adjustment range and strategy of charging parameters (such as current, voltage), staged charging settings (such as switching between constant current stage and constant voltage stage); S42. Data collection and preprocessing: Collect historical data through sensor data, system logs, and user records: battery charging voltage, current, temperature, charging time, and health status, and perform data preprocessing including removing outliers and noise, filling missing values, extracting meaningful features, battery aging characteristics, and charging efficiency indicators; S43. Training reinforcement learning model: Select the reinforcement learning algorithm Q-Learning, train the model in a simulation environment, and integrate the formula: , where SoC(t) is the state of charge at time t, SoC(t-1) is the state of charge at the previous time point, Cnom is the nominal capacity of the battery, I(t) is the charging current at time t, Δt is the time interval, k is the decay constant, which depends on the battery usage and environmental conditions. Simulate various charging scenarios, adjust charging parameters, and design a reward function that reflects the goals: charging efficiency, extending battery life, and reducing abnormalities. The reward function may include charging time, temperature control, and battery health score after charging is completed. Initialize model parameters, set hyperparameters such as learning rate and discount factor, and train in a simulated environment. The model learns optimization strategies by interacting with the environment. Adjust strategies based on the performance of the model and use policy gradient and Q value update techniques for optimization. S44. Validation and testing: Test the performance of the model on the validation set to ensure its generalization ability on unseen data, deploy the model in the actual environment, conduct small-scale tests, and evaluate its actual effect; S45. Deployment and real-time adjustment: Integrate the trained model into the actual charging system, adjust the charging parameters in real time during the actual charging process, optimize the charging strategy according to the model's suggestions, continuously monitor the voltage, current, temperature and other data during the charging process, and provide real-time feedback to the model. Adjust the model based on the real-time feedback to optimize the charging strategy. S46. Feedback and optimization: Collect performance data of the model in actual operation, including charging efficiency and abnormality incidence, collect user feedback on the charging process, understand the problems and needs in actual use, retrain the model based on feedback data, improve strategies, and continuously optimize the model to adapt to new charging needs and changes in battery status.

[0005] Furthermore, a modular power supply charging abnormality protection method is provided. In step S6, the anomaly detection algorithm isolation forest is integrated to monitor the operation data of the power system in real time. By learning the data pattern of normal charging operation, abnormal conditions are detected and marked, more accurate alarms and responses are provided, and real-time monitoring of the charging process is achieved. The power status, charging current, and voltage parameters are recorded, and an alarm mechanism is designed. The specific steps are as follows: S61. Data acquisition and preprocessing: Real-time monitoring of power system data, including voltage, current, and temperature parameters. The data sources are sensor data, system logs, and monitoring interfaces. Data preprocessing is performed to remove noise, outliers, and missing data. Features important for anomaly detection are selected: voltage fluctuations and current change rates. Data is standardized and scaled to the same range. S62. Isolation forest model establishment: Select the isolation forest algorithm as the anomaly detection tool and use normal charging process data for training. The isolation forest model will learn the pattern of normal data. Use the data set to train the isolation forest model and set appropriate parameters: the number of trees and sample size; S63. Anomaly detection and marking: Real-time data stream processing, real-time collection of power system operation data, and data stream processing, input real-time data into the isolation forest model for anomaly detection, determine whether the data point is abnormal based on the output of the isolation forest model, and record the detected abnormal data for further analysis and processing; S64. Alarm mechanism design: define the threshold for anomaly detection, the anomaly score of the isolation forest model, set a specific threshold to decide whether to trigger an alarm, set different types of alarms, including sound alarms, SMS notifications, and email reminders, integrate the alarm mechanism into the power monitoring system, and trigger the alarm mechanism to notify relevant personnel when the abnormal data reaches the set threshold; S65. Real-time monitoring system: Design a user interface to display the real-time status, charging current, and voltage parameters of the power system, and graphically display data trends and abnormal conditions for intuitive understanding by users; record system operation data, abnormal detection results, and alarm events, generate log files, and regularly analyze historical data to understand the frequency and pattern of abnormal conditions; S66. Feedback and optimization: Collect user feedback on the monitoring system and alarm mechanism, understand the problems and needs in actual use, monitor the operating effect of the system, evaluate the accuracy and efficiency of anomaly detection and alarm mechanism, and adjust model parameters and alarm thresholds according to system operation status and feedback.

[0006] A modular power supply charging abnormality protection system, the modular power supply charging abnormality protection system is used to implement any one of the modular power supply charging abnormality protection methods; the modular power supply charging abnormality protection system comprises: a data acquisition module, a data preprocessing module, an abnormality detection module, an alarm mechanism module, a real-time monitoring module, and a feedback and optimization module; The data acquisition module collects the real-time data of the battery’s voltage, current, and temperature, and is responsible for collecting and transmitting sensor data to the processing system; Data preprocessing module: handles missing values, outliers and noise, ensures data accuracy, and extracts meaningful features for anomaly detection; Anomaly detection module: detects abnormal situations based on normal charging data patterns, analyzes data streams in real time, detects and marks anomalies; Alarm mechanism module: defines abnormal thresholds and alarm conditions, triggers alarms when abnormalities are detected, and notifies relevant personnel; Real-time monitoring module: displays real-time data and abnormal information, records system operation data, abnormal events and alarm history; Feedback and optimization module: collects feedback information from users and systems, and optimizes the anomaly detection model based on the feedback data.

[0007] The beneficial effects of the present invention are as follows: preventing equipment damage or fire accidents caused by excessive voltage at the power input terminal. Avoiding performance degradation or permanent damage caused by excessive discharge of the battery. Reducing overheating or other electrical failures caused by excessive current. Quickly detecting and cutting off short circuits to protect the system from damage. Through effective temperature monitoring and control, batteries and other electronic components can be prevented from aging or failure due to overheating. Intelligent charging algorithms ensure that the battery is charged in the best condition, which helps to extend the battery life. The staged charging strategy makes the charging process more efficient and reduces unnecessary energy loss. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Figure 1 A flowchart of a modular power supply charging abnormality protection method; DETAILED DESCRIPTION

[0009] A modular power supply charging abnormality protection method comprises the following steps: S1. Design protection circuit: Overvoltage protection: Design an overvoltage protection circuit at the input end of the power supply. When the input voltage exceeds the set range, the protection circuit can automatically cut off the power supply or trigger an alarm. Undervoltage protection: When the power supply voltage is lower than the set value, the undervoltage protection is triggered to prevent the equipment from working under low voltage and being damaged. Overcurrent protection: Design an overcurrent protection circuit. When the charging current exceeds the rated value, the power supply is automatically disconnected or an alarm is triggered. Short-circuit protection: Ensure that the power supply system has a short-circuit protection function. When a short circuit occurs, the power supply can be quickly disconnected to protect the circuit from damage. S2. Integrated temperature monitoring: Temperature sensor: Install temperature sensors at key parts of the power system to monitor temperature changes in real time. Temperature control protection: Set temperature control thresholds. When the temperature exceeds the preset value, the power supply is automatically disconnected or the cooling system is started to prevent overheating. S3. Data acquisition: Sensor configuration: First, a series of sensors need to be configured to monitor the key parameters of the power module, such as voltage, current, temperature, and data interface: Ensure that these sensors can be connected to the central control unit (such as microcontroller MCU) through a suitable data interface (such as I²C, SPI, UART, etc.) to transmit data in real time; S4. Implement charging algorithm: Use reinforcement learning algorithm to optimize charging strategy. The algorithm can automatically adjust charging parameters based on historical data and real-time monitoring data, predict battery health status and charging demand, and integrate to obtain the formula: , where SoC(t) is the state of charge at time t, SoC(t-1) is the state of charge at the previous time point, Cnom is the nominal capacity of the battery, I(t) is the charging current at time t, Δt is the time interval, k It is the decay constant, which depends on the battery usage and environmental conditions. The charging parameters are adjusted dynamically according to the battery status: voltage, current, and temperature. The charging is carried out in stages, and a staged charging strategy is adopted: constant current charging and constant voltage charging to ensure a safe and efficient charging process. S5. Testing and calibration: Functional test: Perform functional test on the protection circuit to ensure that it can work normally under various abnormal conditions. Calibration: Calibrate the threshold of the protection circuit to ensure its accuracy and reliability. S6. System monitoring and feedback: Real-time monitoring: Realize real-time monitoring of the charging process, record power status, charging current, voltage and other parameters, Alarm mechanism: Design an alarm mechanism to promptly notify the user or automatically stop charging when an abnormality occurs; S7. Maintenance and Updates: Regular Inspection: Regularly inspect and maintain the protection circuit to ensure its continued effectiveness, Software Update: If the power system uses software control, regularly update the software to fix potential vulnerabilities and improve functionality.

[0010] Furthermore, a modular power supply charging abnormality protection method is provided. In step S4, a reinforcement learning algorithm is used to optimize the charging strategy. The algorithm can automatically adjust the charging parameters based on historical data and real-time monitoring data to predict the battery health status and charging demand. The specific steps are as follows: S41. Problem definition and goal setting: clarify the goal, optimize the charging strategy through reinforcement learning to improve charging efficiency, extend battery life and reduce charging anomalies, indicators: including charging time, battery health, energy efficiency, temperature control, define the state and action space: including battery voltage, current, temperature, charging stage, historical charging data, adjustment range and strategy of charging parameters (such as current, voltage), staged charging settings (such as switching between constant current stage and constant voltage stage); S42. Data collection and preprocessing: Collect historical data through sensor data, system logs, and user records: battery charging voltage, current, temperature, charging time, and health status, and perform data preprocessing including removing outliers and noise, filling missing values, extracting meaningful features, battery aging characteristics, and charging efficiency indicators; S43. Training reinforcement learning model: Select the reinforcement learning algorithm Q-Learning, train the model in a simulation environment, and integrate the formula: , where SoC(t) is the state of charge at time t, SoC(t-1) is the state of charge at the previous time point, Cnom is the nominal capacity of the battery, I(t) is the charging current at time t, Δt is the time interval, k is the decay constant, which depends on the battery usage and environmental conditions. Simulate various charging scenarios, adjust charging parameters, and design a reward function that reflects the goals: charging efficiency, extending battery life, and reducing abnormalities. The reward function may include charging time, temperature control, and battery health score after charging is completed. Initialize model parameters, set hyperparameters such as learning rate and discount factor, and train in a simulated environment. The model learns optimization strategies by interacting with the environment. Adjust strategies based on the performance of the model and use policy gradient and Q value update techniques for optimization. S44. Validation and testing: Test the performance of the model on the validation set to ensure its generalization ability on unseen data, deploy the model in the actual environment, conduct small-scale tests, and evaluate its actual effect; S45. Deployment and real-time adjustment: Integrate the trained model into the actual charging system, adjust the charging parameters in real time during the actual charging process, optimize the charging strategy according to the model's suggestions, continuously monitor the voltage, current, temperature and other data during the charging process, and provide real-time feedback to the model. Adjust the model based on the real-time feedback to optimize the charging strategy. S46. Feedback and optimization: Collect performance data of the model in actual operation, including charging efficiency and abnormality incidence, collect user feedback on the charging process, understand the problems and needs in actual use, retrain the model based on feedback data, improve strategies, and continuously optimize the model to adapt to new charging needs and changes in battery status.

[0011] Furthermore, a modular power supply charging abnormality protection method is provided. In step S6, the anomaly detection algorithm isolation forest is integrated to monitor the operation data of the power system in real time. By learning the data pattern of normal charging operation, abnormal conditions are detected and marked, more accurate alarms and responses are provided, and real-time monitoring of the charging process is achieved. The power status, charging current, and voltage parameters are recorded, and an alarm mechanism is designed. The specific steps are as follows: S61. Data acquisition and preprocessing: Real-time monitoring of power system data, including voltage, current, and temperature parameters. The data sources are sensor data, system logs, and monitoring interfaces. Data preprocessing is performed to remove noise, outliers, and missing data. Features important for anomaly detection are selected: voltage fluctuations and current change rates. Data is standardized and scaled to the same range. S62. Isolation forest model establishment: Select the isolation forest algorithm as the anomaly detection tool and use normal charging process data for training. The isolation forest model will learn the pattern of normal data. Use the data set to train the isolation forest model and set appropriate parameters: the number of trees and sample size; S63. Anomaly detection and marking: Real-time data stream processing, real-time collection of power system operation data, and data stream processing, input real-time data into the isolation forest model for anomaly detection, determine whether the data point is abnormal based on the output of the isolation forest model, and record the detected abnormal data for further analysis and processing; S64. Alarm mechanism design: define the threshold for anomaly detection, the anomaly score of the isolation forest model, set a specific threshold to decide whether to trigger an alarm, set different types of alarms, including sound alarms, SMS notifications, and email reminders, integrate the alarm mechanism into the power monitoring system, and trigger the alarm mechanism to notify relevant personnel when the abnormal data reaches the set threshold; S65. Real-time monitoring system: Design a user interface to display the real-time status, charging current, and voltage parameters of the power system, and graphically display data trends and abnormal conditions for intuitive understanding by users; record system operation data, abnormal detection results, and alarm events, generate log files, and regularly analyze historical data to understand the frequency and pattern of abnormal conditions; S66. Feedback and optimization: Collect user feedback on the monitoring system and alarm mechanism, understand the problems and needs in actual use, monitor the operating effect of the system, evaluate the accuracy and efficiency of anomaly detection and alarm mechanism, and adjust model parameters and alarm thresholds according to system operation status and feedback.

[0012] A modular power supply charging abnormality protection system, the modular power supply charging abnormality protection system is used to implement any one of the modular power supply charging abnormality protection methods; the modular power supply charging abnormality protection system comprises: a data acquisition module, a data preprocessing module, an abnormality detection module, an alarm mechanism module, a real-time monitoring module, and a feedback and optimization module; The data acquisition module collects the real-time data of the battery’s voltage, current, and temperature, and is responsible for collecting and transmitting sensor data to the processing system; Data preprocessing module: handles missing values, outliers and noise, ensures data accuracy, and extracts meaningful features for anomaly detection; Anomaly detection module: detects abnormal situations based on normal charging data patterns, analyzes data streams in real time, detects and marks anomalies; Alarm mechanism module: defines abnormal thresholds and alarm conditions, triggers alarms when abnormalities are detected, and notifies relevant personnel; Real-time monitoring module: displays real-time data and abnormal information, records system operation data, abnormal events and alarm history; Feedback and optimization module: collects feedback information from users and systems, and optimizes the anomaly detection model based on the feedback data.

Claims

1. A modular power supply charging abnormality protection method, characterized in that: The following steps are involved: S1. Design protection circuit: Overvoltage protection: Design an overvoltage protection circuit at the input end of the power supply. When the input voltage exceeds the set range, the protection circuit can automatically cut off the power supply and trigger an alarm. Undervoltage protection: When the power supply voltage is lower than the set value, the undervoltage protection is triggered. Overcurrent protection: Design an overcurrent protection circuit. When the charging current exceeds the rated value, the power supply is automatically disconnected and an alarm is triggered. Short-circuit protection: The power supply system has a short-circuit protection function. When a short circuit occurs, the power supply can be quickly disconnected to protect the circuit from damage. S2. Integrated temperature monitoring: Install temperature sensors at key locations of the power system to monitor temperature changes in real time and set temperature control thresholds. When the temperature exceeds the preset value, the power supply is automatically disconnected and the cooling system is activated to prevent overheating. S3. Data acquisition: configure a series of sensors to monitor the key parameters of the power module: voltage, current, temperature, and connect to the central control unit: microcontroller MCU through appropriate data interfaces: I²C, SPI, UART, and transmit data in real time; S4. Implement charging algorithm: Use reinforcement learning algorithm to optimize charging strategy. The algorithm can automatically adjust charging parameters based on historical data and real-time monitoring data, predict battery health status and charging demand, and integrate to obtain the formula: , where SoC(t) is the state of charge at time t, SoC(t-1) is the state of charge at the previous time point, Cnom is the nominal capacity of the battery, I(t) is the charging current at time t, Δt is the time interval, k It is the decay constant, which depends on the battery usage and environmental conditions. The charging parameters are adjusted dynamically according to the battery status: voltage, current, and temperature. The charging is carried out in stages, and a staged charging strategy is adopted: constant current charging and constant voltage charging to ensure a safe and efficient charging process. S5. Testing and calibration: Perform functional tests on the protection circuit to test whether it can work normally under various abnormal conditions and calibrate the threshold of the protection circuit; S6. System monitoring and feedback: Integrate the anomaly detection algorithm Isolation Forest to monitor the operating data of the power system in real time. By learning the data pattern of normal charging operation, detect and mark abnormal situations, provide more accurate alarms and responses, realize real-time monitoring of the charging process, record the power status, charging current, and voltage parameters, design an alarm mechanism, and notify the user in time and automatically stop charging when an abnormality occurs; S7. Maintenance and Update: Regular inspection: Regularly inspect and maintain the protection circuit. The power system uses software control and the software is regularly updated to fix potential vulnerabilities and improve functions.

2. A modular power supply charging abnormality protection method as claimed in claim 1, characterized in that: In step S4, a reinforcement learning algorithm is used to optimize the charging strategy. The algorithm can automatically adjust the charging parameters based on historical data and real-time monitoring data to predict the battery health status and charging demand. The specific steps are as follows: S41. Problem definition and goal setting: clarify the goal, optimize the charging strategy through reinforcement learning to improve charging efficiency, extend battery life and reduce charging anomalies, indicators: including charging time, battery health, energy efficiency, temperature control, define the state and action space: including battery voltage, current, temperature, charging stage, historical charging data, adjustment range and strategy of charging parameters, and setting of staged charging; S42. Data collection and preprocessing: Collect historical data through sensor data, system logs, and user records: battery charging voltage, current, temperature, charging time, and health status, and perform data preprocessing including removing outliers and noise, filling missing values, extracting meaningful features, battery aging characteristics, and charging efficiency indicators; S43. Training reinforcement learning model: Select the reinforcement learning algorithm Q-Learning, train the model in a simulation environment, and integrate the formula: , where SoC(t) is the state of charge at time t, SoC(t-1) is the state of charge at the previous time point, Cnom is the nominal capacity of the battery, I(t) is the charging current at time t, Δt is the time interval, k is the decay constant, which depends on the battery usage and environmental conditions. Simulate various charging scenarios, adjust charging parameters, and design a reward function that reflects the goals: charging efficiency, extending battery life, and reducing abnormalities. The reward function may include charging time, temperature control, and battery health score after charging is completed. Initialize model parameters, set learning rate and discount factor hyperparameters, and train in a simulated environment. The model learns optimization strategies by interacting with the environment. Adjust strategies based on the performance of the model and use policy gradient and Q value update techniques for optimization. S44. Validation and testing: Test the performance of the model on the validation set to ensure its generalization ability on unseen data, deploy the model in the actual environment, conduct small-scale tests, and evaluate its actual effect; S45. Deployment and real-time adjustment: Integrate the trained model into the actual charging system, adjust the charging parameters in real time during the actual charging process, optimize the charging strategy according to the model's suggestions, continuously monitor the voltage, current, temperature and other data during the charging process, and provide real-time feedback to the model. Adjust the model based on the real-time feedback to optimize the charging strategy. S46. Feedback and optimization: Collect performance data of the model in actual operation, including charging efficiency and abnormality incidence, collect user feedback on the charging process, understand the problems and needs in actual use, retrain the model based on feedback data, improve strategies, and continuously optimize the model to adapt to new charging needs and changes in battery status.

3. A modular power supply charging abnormality protection method as claimed in claim 1, characterized in that: In step S6, the anomaly detection algorithm isolation forest is integrated to monitor the operation data of the power system in real time. By learning the data pattern of normal charging operation, abnormal conditions are detected and marked, more accurate alarms and responses are provided, and real-time monitoring of the charging process is achieved. The power status, charging current, and voltage parameters are recorded, and an alarm mechanism is designed. The specific steps are as follows: S61. Data acquisition and preprocessing: Real-time monitoring of power system data, including voltage, current, and temperature parameters. The data sources are sensor data, system logs, and monitoring interfaces. Data preprocessing is performed to remove noise, outliers, and missing data. Features important for anomaly detection are selected: voltage fluctuations and current change rates. Data is standardized and scaled to the same range. S62. Isolation forest model establishment: Select the isolation forest algorithm as the anomaly detection tool and use normal charging process data for training. The isolation forest model will learn the pattern of normal data. Use the data set to train the isolation forest model and set appropriate parameters: the number of trees and sample size; S63. Anomaly detection and marking: Real-time data stream processing, real-time collection of power system operation data, and data stream processing, input real-time data into the isolation forest model for anomaly detection, determine whether the data point is abnormal based on the output of the isolation forest model, and record the detected abnormal data for further analysis and processing; S64. Alarm mechanism design: define the threshold for anomaly detection, the anomaly score of the isolation forest model, set a specific threshold to decide whether to trigger an alarm, set different types of alarms, including sound alarms, SMS notifications, and email reminders, integrate the alarm mechanism into the power monitoring system, and trigger the alarm mechanism to notify relevant personnel when the abnormal data reaches the set threshold; S65. Real-time monitoring system: Design a user interface to display the real-time status, charging current, and voltage parameters of the power system, and graphically display data trends and abnormal conditions for intuitive understanding by users; record system operation data, abnormal detection results, and alarm events, generate log files, and regularly analyze historical data to understand the frequency and pattern of abnormal conditions; S66. Feedback and optimization: Collect user feedback on the monitoring system and alarm mechanism, understand the problems and needs in actual use, monitor the operating effect of the system, evaluate the accuracy and efficiency of anomaly detection and alarm mechanism, and adjust model parameters and alarm thresholds according to system operation status and feedback.

4. A modular power supply charging abnormality protection system, characterized in that: The modular power supply charging abnormality protection system is used to implement the modular power supply charging abnormality protection method according to any one of claims 1 to 3; the modular power supply charging abnormality protection system comprises: a data acquisition module, a data preprocessing module, an abnormality detection module, an alarm mechanism module, a real-time monitoring module, and a feedback and optimization module; The data acquisition module collects the real-time data of the battery’s voltage, current, and temperature, and is responsible for collecting and transmitting sensor data to the processing system; Data preprocessing module: handles missing values, outliers and noise, ensures data accuracy, and extracts meaningful features for anomaly detection; Anomaly detection module: detects abnormal situations based on normal charging data patterns, analyzes data streams in real time, detects and marks anomalies; Alarm mechanism module: defines abnormal thresholds and alarm conditions, triggers alarms when abnormalities are detected, and notifies relevant personnel; Real-time monitoring module: displays real-time data and abnormal information, records system operation data, abnormal events and alarm history; Feedback and optimization module: collects feedback information from users and systems, and optimizes the anomaly detection model based on the feedback data.

Images