Charger and monitoring method to avoid overcharging

Abstract

This invention discloses a charger and monitoring method to avoid overcharging, relating to the field of charger safety protection technology. It includes a charger hardware architecture based on gallium nitride (GaN) power devices, supported by dual-redundancy control and multi-dimensional monitoring, a data processing method deeply coupled with GaN characteristics, and a multi-stage charging strategy. The method collects real-time parameters covering charging state parameters and GaN device state parameters, preprocesses these parameters using an improved Kalman filter algorithm and a BP neural network fitting model to obtain usable data, constructs a multi-stage charging strategy, monitors and controls the charger based on the usable data, and incorporates GaN device junction temperature parameters to construct a composite risk factor R for overcharge risk warning and graded control. This application uses the above method to achieve safe and fast charging of lithium / lead-acid batteries, applicable to various scenarios such as consumer electronics and energy storage, balancing charging efficiency and overcharge protection safety.

Description

Technical Field

[0001] This invention relates to the field of charger safety protection technology, and in particular to a charger and monitoring method for preventing overcharging. Background Technology

[0002] With the rapid development of consumer electronics, energy storage, and electric vehicles, the application of lithium batteries and lead-acid batteries is becoming increasingly widespread. As a core supporting device for battery power supply, the charging efficiency and overcharge protection safety of chargers have become a core focus of the industry. Overcharging can lead to battery capacity degradation, bulging, and leakage, and in severe cases, even cause safety accidents such as combustion and explosion. Therefore, overcharge protection is a core technical requirement for charger design.

[0003] Currently, chargers on the market are mainly divided into two categories: traditional silicon-based chargers and gallium nitride (GaN) chargers. Among them, traditional silicon-based chargers use silicon-based MOSFET power modules, which have inherent defects such as low conversion efficiency, slow switching response speed, high power consumption, and poor heat dissipation performance. Moreover, their overcharge protection mostly adopts a single voltage monitoring or single-path cutoff design, which lacks sufficient protection redundancy and is prone to problems such as monitoring misjudgment and cutoff lag, and cannot meet the charging requirements of high power and high safety.

[0004] To address the shortcomings of traditional silicon-based chargers, gallium nitride (GaN) chargers have gradually emerged. As a third-generation semiconductor material, GaN possesses core advantages such as high switching frequency (reaching 1MHz-2MHz), high power density, high conversion efficiency, and excellent heat dissipation. These advantages effectively improve charger charging efficiency and reduce size, making it a technological trend in the charger industry. However, existing overcharge protection technologies for GaN chargers still have significant shortcomings and fail to fully adapt to the inherent characteristics of GaN devices. Specific deficiencies are as follows: First, gallium nitride (GaN) devices are prone to high-frequency noise during high-frequency switching, and their junction temperature is easily affected by power changes. High power density leads to increased coupling between charging parameters and device state parameters. However, the data preprocessing algorithms of existing GaN chargers are mostly general-purpose filtering and correction algorithms that are not deeply coupled with the characteristics of GaN devices. They cannot effectively suppress high-frequency noise and are difficult to eliminate the acquisition deviation caused by parameter coupling, resulting in insufficient accuracy in overcharge judgment and easy to cause misjudgment or missed judgment.

[0005] Second, the existing hardware architecture of gallium nitride chargers lacks targeted redundancy protection design. Most of them adopt single-path control or single protection mechanism. The gallium nitride power modules are mostly simple device replacements without structural optimization for overcharge protection requirements. They also lack precise docking with dual redundant control units, which can easily lead to overcharge protection failure due to control interface failure, device overheating failure, etc.

[0006] Third, the overcharge monitoring system is incomplete, focusing only on basic parameters such as battery voltage and current, without including specific parameters such as junction temperature and switching status of gallium nitride devices in the monitoring scope. It is unable to capture the early signs of overcharge caused by abnormalities in gallium nitride devices in a timely manner, and lacks graded early warning and flexible power adjustment mechanisms, making it difficult to intervene in overcharge risks in advance and resulting in insufficient timeliness of protection.

[0007] Fourth, in the existing technology, the filtering algorithm of gallium nitride chargers mostly adopts a fixed parameter design and does not dynamically adjust the filtering parameters according to the real-time operating status (junction temperature, switching frequency) of gallium nitride devices. This results in poor adaptability of the filtering effect to the device operating conditions. Either the filtering is too strong and loses the overcharge precursor signal, or the filtering is insufficient and cannot effectively suppress noise, which further affects the reliability of overcharge protection.

[0008] In summary, existing chargers (whether traditional silicon-based chargers or conventional gallium nitride chargers) suffer from low overcharge protection accuracy, insufficient redundancy, and slow response. Furthermore, conventional gallium nitride chargers fail to fully adapt to the characteristics of gallium nitride devices and do not fully leverage their technological advantages, thus failing to meet the high efficiency and safety requirements of battery charging in various scenarios.

[0009] Therefore, developing a gallium nitride charger and monitoring method that can deeply couple the characteristics of gallium nitride devices and has high redundancy and high-precision overcharge protection capabilities has become an urgent technical problem to be solved in this field. Summary of the Invention

[0010] To address the aforementioned problems, this application proposes a charger that avoids overcharging, comprising: The main control unit includes a 32-bit MCU and an independent overcharge protection IC. The MCU is used to receive data collected by the multi-dimensional monitoring module, execute the overcharge protection algorithm, and output control commands to the gallium nitride power module and the shutdown circuit. An independent overcharge protection IC independently monitors specific parameters and directly performs a hard cutoff when the threshold is triggered. A gallium nitride power module, wherein the gallium nitride power module is integrated into the main charging power circuit; A multi-dimensional monitoring module works in conjunction with the gallium nitride power module to detect voltage parameters, current parameters, temperature parameters, and cell balancing parameters and transmit them to the MCU. At the same time, it synchronously transmits specific data to an independent protection IC. Power and protection circuits, including power circuits, protection circuits, and heat dissipation circuits; The communication and interaction module includes a communication module that receives charging parameters, device status, warning and fault data transmitted by the MCU and uploads them to the host computer, and receives parameter configuration instructions from the host computer and transmits them to the MCU, which then performs the corresponding parameter adjustments. The interactive module receives status commands from the MCU, controls the indicator lights to turn on and off and the buzzer to sound, providing feedback on the charger's working status.

[0011] Preferably, the gallium nitride power module includes a GaN device, a driving circuit, an overcurrent detection resistor, and a junction temperature acquisition sensor integrated within a ceramic package structure; The GaN device is integrated on a ceramic packaged substrate, including a high thermal conductivity SiC substrate, an AlN buffer layer, a GaN epitaxial layer, a gate structure, and source / drain electrodes. The gate adopts a MIS-HEMT gate structure, and the gate length is controlled to be 0.15μm; Gate pitch is reduced by combining ALD ultrathin gate dielectric with LP-SiN passivation layer; The source and drain electrodes use ohmic contact electrodes; The substrate has reserved connection interfaces with the MCU, protection IC and heat dissipation module, and the ceramic package structure is directly embedded in the charger power circuit.

[0012] Preferably, the power circuit adopts a dual gallium nitride MOSFET series structure to form a dual cutoff path. One path is controlled by the MCU and is used for power regulation during normal charging and cutoff after full charge. The other path is controlled by an independent protection IC and is used for hard cutoff in emergency situations. The dual cutoff paths are independent of each other. The protection circuit monitors the input / output voltage and current in real time. When an abnormality is detected, it immediately outputs a signal to the MCU and the independent protection IC. The MCU and the independent protection IC output control commands to the dual GaN MOS shutdown circuit to execute the cut-off action. The heat dissipation circuit receives temperature data from the temperature monitoring module, triggers the fan to start and stop, and simultaneously feeds back the heat dissipation status to the MCU. It includes a heat-conducting plate and a temperature-controlled fan. The heat-conducting plate is attached to the gallium nitride power module to quickly conduct the heat generated by the device during operation to the heat sink. The temperature-controlled fan is linked with the temperature monitoring module. When the temperature of the gallium nitride device is detected to be higher than the first preset temperature or the internal temperature of the charger is higher than the first preset temperature, the fan is started to dissipate heat. When the temperature drops below the second preset temperature, the fan is turned off.

[0013] On the other hand, this application proposes a method for monitoring chargers to avoid overcharging, comprising the following steps: S1. Collect real-time parameters covering charging state parameters and GaN device state parameters, and preprocess the real-time parameters using an improved Kalman filter algorithm and a BP neural network fitting model to obtain usable data; S2. Construct a multi-stage charging strategy and monitor and control the charger based on available data; S3. A composite risk factor R is constructed by incorporating the junction temperature parameters of gallium nitride devices to conduct overcharge risk warning and graded control.

[0014] Preferably, the state of charge parameters include the individual cell voltage V. cell Total voltage of the entire battery pack V pack Charging circuit current I, cell surface temperature, gallium nitride device surface temperature, voltage change rate, current change rate, gallium nitride device junction temperature T j .

[0015] Preferably, the specific content of preprocessing real-time parameters using the improved Kalman filter algorithm includes: Obtain raw data, including: real-time parameters and GaN device junction temperature T. j Charging current I and GaN device-specific state data, i.e., real-time switching frequency f GaN GaN switching state; To dynamically optimize the filter bandwidth for high-frequency current noise generated by GaN device switching, the filter cutoff frequency is adjusted to the real-time acquired GaN switching frequency f. GaN 1 / 5; The acquired GaN switch state and real-time GaN device junction temperature T j Incorporate Kalman filter parameter adjustment logic; Based on the GaN state, the process noise covariance Q and observation noise covariance R are dynamically adjusted, and then iterative calculations are performed. The specific steps are as follows: State prediction: Substituting the GaN switch state weighting coefficient B, where B=0.8 during switching and B=0.2 during turning off, the formula is used... By combining the filtered data from the previous moment, the estimated parameter values ​​for the current moment are predicted. in, Here is the state transition matrix. These are the weighting coefficients for the GaN switching states. The data after filtering at time k-1 The filtered parameter estimate at time k refers to the GaN junction temperature T. j Charging current I, This is the control input at time k-1; Covariance prediction: using the formula The covariance is calculated to determine the degree of error in the predicted data. in, The covariance of the state at time k is estimated. Here is the state transition matrix. For transpose, Estimate the covariance of the state at time k-1; Adaptive gain adjustment: via formula Calculate the filter gain K k R and Q are synchronized and dynamically adjusted according to the GaN state; in, This is the observation matrix, which has no unit and takes a value of 1.0. Status update: via formula Combined with the original collected data Correct the predicted value, and at the same time when T j At >120℃, increase K k Weighting accelerates the update speed of junction temperature data; After the above processing, the filtered, accurate data at time k is output. That is, the optimized GaN device junction temperature T j Charging current I.

[0016] Preferably, the specific content of the preprocessing of real-time parameters by the BP neural network fitting model includes: The input data includes real-time data and denoised data after Kalman filtering, including the individual unit voltage V. cell Total voltage V pack Optimized charging current I, cell temperature T, voltage change rate t, and current change rate; GaN device characteristic parameters, including GaN switching frequency f GaN GaN junction temperature T j GaN switching loss P loss GaN driving voltage V GaN ; The two types of data, totaling 10 parameters, serve as the input layer data for the BP neural network. For the 10 acquired input parameters, a three-step targeted processing method is used, taking into account the characteristics of GaN, to complete the parameter coupling correction: Model structure adaptation: The 3-layer BP neural network structure was optimized by connecting 10 input parameters to the input layer, setting 12 neurons in the hidden layer, using the Sigmoid activation function, and setting the output layer to have 3 core parameters V. cell T j 、I; The model was trained using measured data from GaN devices under all operating conditions. GaN characteristic constraints are added to the model output layer to perform targeted optimization on the corrected parameters: When the GaN junction temperature T j When the temperature is above 120℃, the correction amplitude of the charging current I is forcibly increased to avoid current acquisition deviation caused by excessive junction temperature. When the GaN switching frequency f GaN When the frequency is >1.5MHz, the single-unit voltage V cell The correction error is controlled within ±3mV to suppress voltage fluctuation deviation caused by high-frequency switching; After model training and constraint correction, the precise correction values ​​of three core parameters are output, namely the corrected unit voltage V. cell GaN junction temperature T j Charging current I.

[0017] The preferred multi-stage charging strategy includes the following details: Constant current fast charging stage: Relying on the high efficiency of gallium nitride devices, charging is performed at a constant current of 0.5C~1C until the single-cell voltage reaches 4.1V. The voltage rise rate is monitored. If dV / dt>0.1V / min, an abnormality is determined. The MCU immediately controls the gallium nitride device to reduce the current to 0.2C. Constant voltage and current limiting phase: Maintain a constant voltage of 4.2V, with gallium nitride devices precisely regulating the output current to ensure a gradual decrease in current; Triggering full charge condition: A full charge is determined when the current drops to 0.03C and remains there for 10 minutes. A full charge is determined when the voltage remains stable at 4.2V±10mV for more than 30 minutes. Gallium nitride device junction temperature If the temperature exceeds 120℃, the power will be reduced simultaneously. If the temperature continues to rise, the full charge shutdown will be triggered. Trickle maintenance or shutdown phase: After full charge, the main charging circuit is cut off to control the gallium nitride device to turn off, and only the <0.01C micro-current charging option is retained. The gallium nitride device provides a stable micro-current to reduce power consumption. Overtime protection: If the total charging time exceeds 8 hours, the gallium nitride device will be forcibly shut down to avoid overcharging caused by prolonged trickle charging.

[0018] Preferably, the expression for the composite risk factor R in S3 is: ; in, This is the voltage risk weighting coefficient. This is the current risk weighting coefficient. This is the cell temperature risk weighting coefficient. This is the risk weighting coefficient for GaN junction temperature. This represents the real-time voltage of a single battery cell. The full charge voltage threshold for lithium batteries. For real-time charging circuit current, The threshold for trickle charging current. The safe temperature threshold for the battery cell. For real-time cell surface temperature, Real-time junction temperature for GaN devices The safe junction temperature threshold for GaN devices, For overfilling composite risk factors; The specific content of overcharge risk warning and graded control includes: When R>a, an early warning is issued and the power of the gallium nitride device is reduced to 50%, the APP alarm is triggered, and the working status of the gallium nitride is simultaneously indicated. When R>b and b>a, protection measures are implemented to control the gallium nitride device to shut down, initiate heat dissipation, and cut off the charging circuit. When any of the total voltage, charger temperature, or gallium nitride device junction temperature exceeds the corresponding preset value, a hard cut-off is triggered. The protection IC independently shuts down the gallium nitride MOS, which cannot be automatically restored and requires power-off reset to completely prevent overcharging.

[0019] Preferably, the monitoring method is further configured with hardware-level redundancy protection, specifically including: Independent protection IC: Independent of software, the charging circuit is immediately cut off when the voltage is >4.35V or the junction temperature of the gallium nitride device is >150℃, forming dual redundancy with MCU control; Physical protection: Cell expansion triggers a mechanical switch, forcibly disconnecting the charging interface and simultaneously shutting off the power supply to the gallium nitride device to prevent the device from being damaged under no-load conditions.

[0020] In summary, the charger and monitoring method for preventing overcharging according to the present invention have the following advantages compared with traditional technologies: 1. Gallium nitride power devices improve response speed and efficiency. Combined with software strategies and independent hardware ICs, they provide double protection, solving the problems of low efficiency and poor heat dissipation of traditional chargers, while also strengthening overcharge protection. 2. Added junction temperature monitoring for gallium nitride devices, combining voltage, current, temperature, and rate of change to avoid misjudgment based on a single parameter, while ensuring the safety of the device itself; 3. Gallium nitride devices reduce losses and heat, minimizing the risk of overcharging caused by high temperatures, while their fast response characteristics can intervene in advance of overcharging. 4. Active balancing suppresses overcharging of individual cells, and gallium nitride devices provide stable micro-current, improving balancing efficiency and the lifespan of the entire cell group; 5. Proactively intervenes in overcharging risks, reducing the frequency of hard cut-off; supports lithium batteries and lead-acid batteries, with configurable thresholds, adaptable to multiple scenarios, and the miniaturization characteristics of gallium nitride make it suitable for portable devices.

[0021] The technical method of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0022] Figure 1 This is a diagram of a charger architecture designed to avoid overcharging.

[0023] Figure 2 This is a flowchart of a monitoring method for a charger to avoid overcharging. Detailed Implementation

[0024] The technical method of the present invention will be further described below with reference to the accompanying drawings and embodiments. It should be noted that, unless otherwise specifically stated, the relative arrangement, numerical expressions, and values ​​of the components and steps described in these embodiments do not limit the scope of this application.

[0025] The following description of at least one exemplary embodiment is merely illustrative and is in no way intended to limit the scope of this application and its application or use.

[0026] Techniques, systems, and equipment known to those skilled in the art may not be discussed in detail, but where appropriate, they should be considered part of the instruction manual.

[0027] In all the examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not as limitations. Therefore, other examples of exemplary embodiments may have different values.

[0028] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning as understood by one of ordinary skill in the art to which this invention pertains.

[0029] Example 1 A charger designed to prevent overcharging employs a dual architecture of a 32-bit MCU and an independent overcharge protection IC, forming a dual core for overcharge prevention control. These two components operate independently yet in conjunction with each other to ensure control reliability. Figure 1 As shown, it includes: The main control unit includes a 32-bit MCU and an independent overcharge protection IC. The MCU is used to receive data collected by the multi-dimensional monitoring module, execute the overcharge protection algorithm, and output control commands to the gallium nitride power module and the shutdown circuit.

[0030] The independent overcharge protection IC serves as the core of hardware-level protection. It does not rely on the MCU, independently monitors specific parameters, and directly performs a hard cut-off when the threshold is triggered, forming a dual protection of software control and hardware fallback.

[0031] Independent protection ICs (such as S-8200 and R5425): They have a high-precision voltage detection capability of ±10mV, a fast cut-off response speed of <10ms, and a static power consumption of <1μA. They can independently monitor the charging output voltage and the junction temperature of gallium nitride devices. When the parameters are detected to be out of control, they can directly output a cut-off command to the dual GaN MOS shutdown circuit without going through the MCU. As the last hardware line of defense for overcharge protection, they improve the reliability and response speed of the protection.

[0032] The MCU receives real-time data from voltage, current, and temperature sampling modules, and also receives operating status feedback from the gallium nitride (GaN) power module (such as junction temperature and switching status). The MCU outputs control commands to the GaN power module (to adjust power and switching status) and the cell balancing module (to start / stop balancing). An independent protection IC collects output voltage and GaN device junction temperature data in real time. When these values ​​exceed the limits, it directly outputs a cut-off command to the dual GaN MOS shutdown circuit and simultaneously feeds back a fault signal to the MCU, achieving bidirectional data interaction and redundant control.

[0033] Gallium nitride (GaN) power modules replace traditional silicon-based MOSFET power modules. They adopt third-generation gallium nitride (GaN) power devices (such as GaN Systems GS66508B and Navitas NV6127) in an integrated design and are integrated into the main charging power circuit. Unlike the single structure of traditional silicon-based power modules, they are adapted to the redundancy protection and fast response requirements of overcharge protection systems.

[0034] Furthermore, the gallium nitride power module includes GaN devices, driving circuits, overcurrent detection resistors, and junction temperature acquisition sensors integrated within a ceramic package structure.

[0035] The GaN device is integrated on a ceramic packaged substrate, including a high thermal conductivity SiC substrate, an AlN buffer layer, a GaN epitaxial layer, a gate structure, and source / drain electrodes.

[0036] The gate adopts a MIS-HEMT gate structure, and the gate length is controlled to be 0.15μm.

[0037] Gate pitch is reduced by combining ALD ultrathin gate dielectric with LP-SiN passivation layer.

[0038] By combining the ALD ultrathin gate dielectric with the LP-SiN passivation layer, the Schottky leakage problem of traditional GaN devices is solved. At the same time, the gate pitch is reduced to achieve fast transmission of switching signals, providing structural support for the rapid cut-off of overcharge precursors. Unlike the planar gate structure of traditional silicon-based modules, the response delay is significantly reduced.

[0039] The source and drain electrodes use ohmic contact electrodes.

[0040] The substrate has reserved connection interfaces with the MCU, protection IC and heat dissipation module, and the ceramic package structure is directly embedded in the charger power circuit.

[0041] The package size is reduced by more than 30% compared to traditional silicon-based power modules; the package uses an integrated heat dissipation substrate that is seamlessly attached to the external heat-conducting plate, which solves the problem of uneven heat dissipation and large device spacing caused by traditional modules, thus addressing the issue of response lag. It adapts to the miniaturization requirements of chargers while ensuring timely overcharge protection.

[0042] The multi-dimensional monitoring module works in conjunction with the gallium nitride power module to detect voltage parameters, current parameters, temperature parameters, and cell balancing parameters and transmit them to the MCU. At the same time, it synchronously transmits specific data to the independent protection IC.

[0043] Specifically, voltage monitoring: A high-precision voltage divider sampling circuit and ADC analog-to-digital conversion module are used to collect the voltage of individual cells and the total voltage of the entire battery pack in real time. The full charge threshold of the lithium battery is set to 4.2V±0.05V, with a sampling accuracy of ±5mV. The monitoring data is synchronously fed back to the MCU, which in turn coordinates with the gallium nitride device to adjust the output voltage in real time to avoid overcharging due to excessive voltage. At the same time, the voltage data is synchronously transmitted to an independent protection IC as one of the criteria for hardware cut-off.

[0044] Current monitoring: A series high-precision sampling resistor and operational amplifier are used to monitor the charging circuit current in real time. When the current drops to 0.02C~0.05C, it is determined that the charging circuit is close to full charge. The monitoring data is fed back to the MCU, which immediately controls the gallium nitride device to reduce the output current and gradually transition to trickle mode to avoid overcharging caused by continuous high current charging after full charge. At the same time, abnormal current fluctuations during the charging process are monitored and early warnings are triggered in a timely manner.

[0045] Temperature monitoring: NTC thermistors are used, attached to the surface of the battery cell and the surface of the gallium nitride power module respectively, to collect temperature data of both in real time. The safe temperature threshold for the battery cell is set to 0℃~60℃, and the safe junction temperature threshold for the gallium nitride device is set to 0℃~150℃. When the battery cell temperature >60℃, the MCU controls the gallium nitride device to reduce power (to 50% of the rated power). When the temperature >70℃, the gallium nitride device is forcibly shut down, and the protection IC is triggered to hard-cut off. When the junction temperature of the gallium nitride device >120℃, the power is reduced synchronously. When the junction temperature >150℃, a hard cut-off is triggered to avoid device failure and overcharging risks caused by high temperature.

[0046] Cell balancing module: It adopts an active balancing circuit to collect the voltage of each individual cell in real time. When the voltage difference of an individual cell is detected to be >50mV, the balancing operation is started. Cells with high voltage are micro-discharged and cells with low voltage are micro-charged to avoid overcharging of a single cell (overcharging of a single cell can easily cause safety hazards to the entire battery pack). During the balancing process, gallium nitride devices provide a stable micro-current (<0.01C) to improve balancing efficiency and reduce power consumption during the balancing process.

[0047] Each monitoring submodule (voltage, current, temperature, cell balancing) collects corresponding parameters in real time. After filtering, amplification, and analog-to-digital conversion, the precise data is transmitted to the MCU. The MCU analyzes and processes the data to determine whether there are signs of overcharging. At the same time, it transmits key data (such as individual cell voltage and gallium nitride junction temperature) to the independent protection IC as a dual basis for judgment. The cell balancing module receives the start / stop command from the MCU and feeds back the balancing status to the MCU, forming a closed-loop data flow of acquisition-transmission-analysis-feedback.

[0048] Power and protection circuits, including power circuits, protection circuits, and heat dissipation circuits.

[0049] Furthermore, the power circuit adopts a dual gallium nitride MOSFET series structure to form a dual cutoff path. One path is controlled by the MCU for power regulation during normal charging and cutoff after full charge; the other path is controlled by an independent protection IC for hard cutoff in emergency situations (such as MCU failure or severe parameter overrun). The dual cutoff paths are independent of each other, forming a "dual-path independent control" structural redundancy, which solves the problem of traditional silicon-based power modules having only single-path control and being prone to overcharge protection failure due to control interface failure. At the same time, the module has a built-in junction temperature acquisition interface, which directly connects to the temperature monitoring module to realize direct transmission of junction temperature signals without the need for additional adapter circuits.

[0050] The gallium nitride power module receives power adjustment commands (current and voltage parameters) from the MCU, converts the AC input into a stable DC output, and simultaneously collects its own operating status (junction temperature, switching frequency, and output power) in real time and feeds it back to the MCU; it receives shutdown commands from the MCU or independent protection IC to quickly cut off the main charging circuit; and it provides stable micro-current support for the cell balancing module to ensure the stability of the balancing process.

[0051] The protection circuit integrates input overvoltage, output overvoltage, overcurrent, and reverse connection protection circuits. When the input voltage exceeds the standard (such as AC input exceeding 260V) or the output voltage exceeds 4.5V, the MCU immediately controls the gallium nitride device to shut down the output, and at the same time, the independent protection IC triggers hard cutoff, forming dual redundancy protection to avoid overcharging risks caused by abnormal input and output. The reverse connection protection circuit can prevent device damage and overcharging risks caused by reverse battery connection.

[0052] The system monitors the input / output voltage and current in real time. When an abnormality is detected, it immediately outputs a signal to the MCU and the independent protection IC. The MCU and the independent protection IC then output control commands to the dual GaN MOS shutdown circuit to execute the cut-off action.

[0053] The heat dissipation circuit receives temperature data from the temperature monitoring module, triggers the fan to start and stop, and simultaneously feeds back the heat dissipation status to the MCU. It includes a heat-conducting plate and a temperature-controlled fan. The heat-conducting plate is attached to the gallium nitride power module to quickly conduct the heat generated by the device during operation to the heat sink. The temperature-controlled fan is linked with the temperature monitoring module. When the temperature of the gallium nitride device is detected to be >55℃ or the internal temperature of the charger is >50℃, the fan is activated to dissipate heat. When the temperature drops below 45℃, the fan is turned off. This ensures the stable operation of the gallium nitride device, avoids overcharge protection failure due to overheating, and reduces heat dissipation power consumption.

[0054] The power circuit receives the stable DC output from the gallium nitride power module and transmits it to the battery interface; the protection circuit monitors the input / output voltage and current in real time, and immediately outputs signals to the MCU and the independent protection IC when an abnormality is detected; the MCU and the independent protection IC respectively output control commands to the dual GaN MOS shutdown circuit to execute the cut-off action; the heat dissipation circuit receives the temperature data from the temperature monitoring module, triggers the fan to start and stop, and simultaneously feeds back the heat dissipation status to the MCU.

[0055] The communication and interaction module receives charging parameters, device status, warning and fault data transmitted by the MCU and uploads them to the host computer. It also receives parameter configuration instructions from the host computer and transmits them to the MCU, which then performs the corresponding parameter adjustments.

[0056] Communication function: It can upload charging parameters (single cell voltage, total voltage, charging current, cell temperature), gallium nitride device operating status (junction temperature, conversion efficiency, switching status), overcharge warning and fault information to the APP / host computer in real time; users can send parameter configuration commands (such as full charge voltage threshold, current threshold) through the APP / host computer to realize personalized overcharge protection settings.

[0057] The interactive module receives status commands from the MCU, controls the indicator lights to turn on and off and the buzzer to sound, providing feedback on the charger's working status.

[0058] Interactive features: Integrated indicator lights and buzzer. Different colored indicator lights distinguish the charging status (red light: fast charging, green light: full charging, yellow light: warning). The buzzer sounds (short beep: warning, long beep: fault) to indicate overcharge warnings and fault information (such as gallium nitride device abnormality, voltage over-limit), making it easy for users to quickly identify the charger's working status.

[0059] The communication module receives charging parameters, device status, warning and fault data transmitted by the MCU and uploads them to the APP / host computer; it receives parameter configuration instructions from the APP / host computer and transmits them to the MCU, which then performs the corresponding parameter adjustments; the interaction module receives status instructions from the MCU, controls the indicator lights to turn on and off and the buzzer to sound, and provides feedback on the charger's working status.

[0060] Example 2 This application proposes a method for monitoring chargers to avoid overcharging, such as... Figure 2 As shown, it includes the following steps: To address the core characteristics of gallium nitride (GaN) devices, such as high-frequency noise from high-frequency switching (1MHz-2MHz), volatile junction temperature, and increased parameter coupling due to high power density, this paper proposes a deep coupling improvement between the adaptive Kalman filter algorithm and the BP neural network fitting model, in conjunction with the overcharge prevention monitoring requirements. This solves the technical problems of poor compatibility between traditional algorithms and GaN device characteristics, as well as insufficient data processing accuracy. It ensures that the preprocessed data can accurately match the control response requirements of GaN devices, providing a reliable basis for GaN device control and overcharge determination.

[0061] S1. Collect real-time parameters covering charging state parameters and GaN device state parameters, and preprocess the real-time parameters using an improved Kalman filter algorithm and a BP neural network fitting model to obtain usable data.

[0062] The real-time parameters specifically include: individual cell voltage V cell Total voltage of the entire battery pack V pack Charging circuit current I, temperature T (including cell surface temperature and gallium nitride device surface temperature), voltage change rate dV / dt (unit: V / min), current change rate dI / dt (unit: A / min), gallium nitride device junction temperature T j (Unit: °C) This provides comprehensive and accurate raw data support for subsequent data preprocessing, while focusing on collecting GaN device state parameters to lay the foundation for deep coupling between the algorithm and GaN characteristics.

[0063] Preferably, the state of charge parameters include the individual cell voltage V. cell Total voltage of the entire battery pack V pack Charging circuit current I, cell surface temperature, gallium nitride device surface temperature, voltage change rate, current change rate, gallium nitride device junction temperature T j .

[0064] Furthermore, the specific details of preprocessing real-time parameters using the improved Kalman filter algorithm include: Obtain raw data, including: real-time parameters and GaN device junction temperature T. j Charging current I and GaN device-specific state data, i.e., real-time switching frequency f GaN (Hz, value 1MHz~2MHz, obtained through the built-in frequency acquisition element of GaN power module), GaN switching state (high frequency switching state during fast charging / off state during trickle charging / low power state during trickle charging).

[0065] To dynamically optimize the filter bandwidth for high-frequency current noise generated by GaN device switching, the filter cutoff frequency is adjusted to the real-time acquired GaN switching frequency f. GaN 1 / 5 of the original value is used to prevent high-frequency noise from penetrating the filtering stage, while retaining the parameter mutation characteristics of overcharge precursors (such as voltage surge and GaN junction temperature surge).

[0066] The acquired GaN switch state and real-time GaN device junction temperature T j Incorporate Kalman filter parameter adjustment logic.

[0067] Based on the GaN state, the process noise covariance Q and observation noise covariance R are dynamically adjusted, and then iterative calculations are performed. The specific steps are as follows: When GaN devices are in a high-frequency switching state (fast charging stage), the switching noise is significant. The device automatically increases the Q value (Q=0.01~0.03) and decreases the R value (R=0.001~0.005) to enhance noise suppression capability.

[0068] When the junction temperature T_j of the GaN device exceeds 100℃ (close to the safety threshold), the junction temperature fluctuation causes an increase in parameter acquisition deviation. The system automatically reduces the Q value (Q=0.005~0.01) and increases the R value (R=0.005~0.01) to improve data smoothness and avoid misjudgment caused by junction temperature fluctuation. When the GaN device is in the off or trickle-down phase (low power state), the noise is relatively small, maintaining the basic values ​​of Q=0.003 and R=0.002, balancing data real-time performance and accuracy.

[0069] State prediction: Substituting the GaN switch state weighting coefficient B, where B=0.8 during switching and B=0.2 during turning off, the formula is used... By combining the filtered data from the previous moment, the estimated parameter values ​​for the current moment are predicted.

[0070] in, This is the state transition matrix, a 1×1 matrix (for adapting to single-parameter filtering), with a value of 1.0, representing the temporal continuity of the data. These are the weighting coefficients for the GaN switching states. The data after filtering at time k-1 The filtered parameter estimate at time k refers to the GaN junction temperature T. j Charging current I, This is the control input at time k-1. Here, it is the GaN device switching control signal, with a value of 1 (switching) or 0 (turning off), and no unit.

[0071] Covariance prediction: using the formula The covariance is calculated to determine the degree of error in the predicted data.

[0072] in, The state estimate covariance at time k reflects the degree of error in the filtered data and is dimensionless. Here is the state transition matrix. For transpose, The covariance of the state at time k-1 is estimated.

[0073] Adaptive gain adjustment: via formula Calculate the filter gain K k The convergence speed and denoising effect of the filter are determined by the synchronous linkage between R and Q, which are dynamically adjusted according to the GaN state.

[0074] in, The observation matrix is ​​a 1×1 matrix with no unit and a value of 1.0, representing the consistency between the observation data and the state data.

[0075] Status update: via formula Combined with the original collected data Correct the predicted value, and at the same time when T j At >120℃, increase K k The weighting (weighting coefficient increased to 1.2) accelerates the junction temperature data update speed and responds promptly to the high temperature status of GaN devices.

[0076] After the above processing, the filtered, accurate data at time k is output. That is, the optimized GaN device junction temperature T j Charging current I.

[0077] Furthermore, the specific preprocessing of real-time parameters by the BP neural network fitting model includes: The input data includes real-time data and denoised data after Kalman filtering, including the individual unit voltage V. cell Total voltage V pack The optimized charging current I, cell temperature T, voltage change rate t, and current change rate.

[0078] GaN device characteristic parameters, including GaN switching frequency f GaN GaN junction temperature T j GaN switching loss P loss GaN driving voltage V GaN .

[0079] The two types of data, totaling 10 parameters, serve as the input layer data for the BP neural network. For the 10 acquired input parameters, a three-step targeted processing method is used, taking into account the characteristics of GaN, to complete the parameter coupling correction: Model structure adaptation: The 3-layer BP neural network structure was optimized by connecting 10 input parameters to the input layer, setting 12 neurons in the hidden layer, using the Sigmoid activation function, and setting the output layer to have 3 core parameters V. cell T j 、I.

[0080] The model was trained using measured data from GaN devices under all operating conditions.

[0081] GaN characteristic constraints are added to the model output layer to perform targeted optimization on the corrected parameters: When the GaN junction temperature T j When the temperature is above 120℃, the correction range of the charging current I is forcibly increased (by 30%) to avoid current acquisition deviation caused by excessive junction temperature.

[0082] When the GaN switching frequency f GaN When the frequency is >1.5MHz, the single-unit voltage V cell The correction error is controlled within ±3mV, suppressing voltage fluctuation deviations caused by high-frequency switching.

[0083] After model training and constraint correction, the precise correction values ​​of three core parameters are output, namely the corrected unit voltage V. cell GaN junction temperature T j The charging current I completely eliminates the acquisition deviation caused by multi-parameter coupling (especially the coupling between GaN characteristics and charging parameters).

[0084] A closed-loop preprocessing workflow deeply coupled with GaN characteristics. GaN device status monitoring is integrated into every step of the process, achieving full coupling between "data acquisition - noise reduction - correction - anomaly removal" and GaN characteristics. This ensures that the preprocessed data can directly support the MCU's precise control of the GaN device. The specific workflow is as follows: Raw data acquisition (including GaN-specific parameters), GaN state determination (real-time determination of the switching state of GaN devices, junction temperature T) j Switching frequency f GaN ), Dynamic adjustment of Kalman filter parameters (adjusting Q, R values ​​and filter bandwidth according to GaN state), Adaptive Kalman filter denoising (focusing on suppressing high-frequency noise in GaN while retaining overcharge precursor characteristics), BP neural network parameter correction (incorporating GaN characteristic constraints to achieve precise parameter correction), Outlier removal (adding a new GaN-specific outlier judgment standard: T) j >150℃, f GaN If the frequency deviates from the 1MHz-2MHz range and triggers an anomaly, the previous correction value will be used as the replacement value, and a warning will be sent to the MCU.

[0085] The sampling period of this process is maintained at 100ms, which is precisely matched with the response speed of GaN devices of <5ms. This ensures data real-time performance and improves data accuracy through full GaN characteristic coupling, avoiding overcharge protection failure due to data deviation. At the same time, it is adapted to the high-frequency, high-temperature, and high-power-density operating characteristics of GaN devices.

[0086] S2. Construct a multi-stage charging strategy and monitor and control the charger based on available data.

[0087] Furthermore, the specific details of the multi-stage charging strategy are as follows: Constant current fast charging stage: Relying on the high efficiency of gallium nitride devices, charging is performed at a constant current of 0.5C~1C until the single-cell voltage reaches 4.1V. The voltage rise rate is monitored. If dV / dt>0.1V / min, an abnormality is determined. The MCU immediately controls the gallium nitride device to reduce the current to 0.2C.

[0088] Constant voltage and current limiting phase: Maintain a constant voltage of 4.2V, with gallium nitride devices precisely regulating the output current to ensure a gradual decrease in current; Triggering full charge condition: A full charge is determined when the current drops to 0.03C and remains there for 10 minutes.

[0089] If the voltage remains stable at 4.2V±10mV for more than 30 minutes, it is considered fully charged.

[0090] Gallium nitride device junction temperature If the temperature exceeds 120℃, the power will be reduced simultaneously. If the temperature continues to rise, the system will be shut down when fully charged.

[0091] Trickle maintenance or shutdown phase: After full charging, the main charging circuit is cut off to control the gallium nitride device to shut down, leaving only the <0.01C micro-current charging option available. The gallium nitride device provides a stable micro-current to reduce power consumption.

[0092] Overtime protection: If the total charging time exceeds 8 hours, the gallium nitride device will be forcibly shut down to avoid overcharging caused by prolonged trickle charging.

[0093] S3. A composite risk factor R is constructed by incorporating the junction temperature parameters of gallium nitride devices to conduct overcharge risk warning and graded control.

[0094] Furthermore, the expression for the composite risk factor R in S3 is: .

[0095] in, This is the voltage risk weighting coefficient. This is the current risk weighting coefficient. This is the cell temperature risk weighting coefficient. This is the risk weighting coefficient for GaN junction temperature. This represents the real-time voltage of a single battery cell. The full charge voltage threshold for lithium batteries. For real-time charging circuit current, The threshold for trickle charging current. The safe temperature threshold for the battery cell. For real-time cell surface temperature, Real-time junction temperature for GaN devices The safe junction temperature threshold for GaN devices, This is an overfilled composite risk factor.

[0096] The specific content of overcharge risk warning and graded control includes: When R>a, an early warning is issued and the power of the gallium nitride device is reduced to 50%, the APP alarm is triggered, and the working status of the gallium nitride is displayed simultaneously.

[0097] When R>b and b>a, protection measures are implemented to shut down gallium nitride devices, initiate heat dissipation, and cut off the charging circuit.

[0098] When any of the total voltage, charger temperature, or gallium nitride device junction temperature exceeds the corresponding preset value, a hard cut-off is triggered. The protection IC independently shuts down the gallium nitride MOS, which cannot be automatically restored and requires power-off reset to completely prevent overcharging.

[0099] Furthermore, the monitoring method is also equipped with hardware-level redundancy protection, specifically including: Independent protection IC: Independent of software, the charging circuit is immediately cut off when the voltage is >4.35V or the junction temperature of the gallium nitride device is >150℃, forming dual redundancy with MCU control.

[0100] Physical protection: Cell expansion triggers a mechanical switch, forcibly disconnecting the charging interface and simultaneously shutting off the power supply to the gallium nitride device to prevent the device from being damaged under no-load conditions.

[0101] Finally, it should be noted that the above embodiments are only used to illustrate the technical methods of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical methods of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical methods to deviate from the spirit and scope of the technical methods of the present invention.

Claims

1. A charger that avoids overcharging, characterized in that, include: The main control unit includes a 32-bit MCU and an independent overcharge protection IC. The MCU is used to receive data collected by the multi-dimensional monitoring module, execute the overcharge protection algorithm, and output control commands to the gallium nitride power module and the shutdown circuit. An independent overcharge protection IC independently monitors specific parameters and directly performs a hard cutoff when the threshold is triggered. A gallium nitride power module, wherein the gallium nitride power module is integrated into the main charging power circuit; A multi-dimensional monitoring module works in conjunction with the gallium nitride power module to detect voltage parameters, current parameters, temperature parameters, and cell balancing parameters and transmit them to the MCU. At the same time, it synchronously transmits specific data to an independent protection IC. Power and protection circuits, including power circuits, protection circuits, and heat dissipation circuits; The communication and interaction module includes a communication module that receives charging parameters, device status, warning and fault data transmitted by the MCU and uploads them to the host computer, and receives parameter configuration instructions from the host computer and transmits them to the MCU, which then performs the corresponding parameter adjustments. The interactive module receives status commands from the MCU, controls the indicator lights to turn on and off and the buzzer to sound, providing feedback on the charger's working status.

2. A charger for preventing overcharging according to claim 1, characterized in that, The gallium nitride power module includes a GaN device, a driving circuit, an overcurrent detection resistor, and a junction temperature acquisition sensor integrated within a ceramic package structure. The GaN device is integrated on a ceramic packaged substrate, including a high thermal conductivity SiC substrate, an AlN buffer layer, a GaN epitaxial layer, a gate structure, and source / drain electrodes. The gate adopts a MIS-HEMT gate structure, and the gate length is controlled to be 0.15μm; Gate pitch is reduced by combining ALD ultrathin gate dielectric with LP-SiN passivation layer; The source and drain electrodes use ohmic contact electrodes; The substrate has reserved connection interfaces with the MCU, protection IC and heat dissipation module, and the ceramic package structure is directly embedded in the charger power circuit.

3. A charger for preventing overcharging according to claim 2, characterized in that, The power circuit adopts a dual gallium nitride MOSFET series structure to form a dual cutoff path. One path is controlled by the MCU and is used for power regulation during normal charging and cutoff after full charge. Another route is independently protected by an IC control for hard cutoff in emergency situations; the dual cutoff paths are independent of each other. The protection circuit monitors the input / output voltage and current in real time. When an abnormality is detected, it immediately outputs a signal to the MCU and the independent protection IC. The MCU and the independent protection IC output control commands to the dual GaN MOS shutdown circuit to execute the cut-off action. The heat dissipation circuit receives temperature data from the temperature monitoring module, triggers the fan to start and stop, and simultaneously feeds back the heat dissipation status to the MCU. It includes a heat conduction plate and a temperature-controlled fan. The heat conduction plate is attached to the gallium nitride power module to quickly conduct the heat generated by the device during operation to the heat sink. The temperature-controlled fan is linked with the temperature monitoring module. When the temperature of the gallium nitride device is detected to be higher than the first preset temperature or the internal temperature of the charger is higher than the first preset temperature, the fan is activated to dissipate heat. When the temperature drops below the second preset temperature, the fan is turned off.

4. A monitoring method for a charger that avoids overcharging, applied to a charger that avoids overcharging as described in any one of claims 1-3, characterized in that, Includes the following steps: S1. Collect real-time parameters covering charging state parameters and GaN device state parameters, and preprocess the real-time parameters using an improved Kalman filter algorithm and a BP neural network fitting model to obtain usable data; S2. Construct a multi-stage charging strategy and monitor and control the charger based on available data; S3. A composite risk factor R is constructed by incorporating the junction temperature parameters of gallium nitride devices to conduct overcharge risk warning and graded control.

5. The method for monitoring a charger to avoid overcharging according to claim 4, characterized in that, Charging status parameters include individual cell voltage V cell Total voltage of the entire battery pack V pack Charging circuit current I, cell surface temperature, gallium nitride device surface temperature, voltage change rate, current change rate, gallium nitride device junction temperature T j .

6. The monitoring method for a charger to avoid overcharging according to claim 5, characterized in that, The specific content of preprocessing real-time parameters using the improved Kalman filter algorithm includes: Obtain raw data, including: real-time parameters and GaN device junction temperature T. j Charging current I and GaN device-specific state data, i.e., real-time switching frequency f GaN GaN switching state; To dynamically optimize the filter bandwidth for high-frequency current noise generated by GaN device switching, the filter cutoff frequency is adjusted to the real-time acquired GaN switching frequency f. GaN 1 / 5; The acquired GaN switch state and real-time GaN device junction temperature T j Incorporate Kalman filter parameter adjustment logic; Based on the GaN state, the process noise covariance Q and observation noise covariance R are dynamically adjusted, and then iterative calculations are performed. The specific steps are as follows: State prediction: Substituting the GaN switch state weighting coefficient B, where B=0.8 during switching and B=0.2 during turning off, the formula is used... By combining the filtered data from the previous moment, the estimated parameter values ​​for the current moment are predicted. in, Here is the state transition matrix. These are the weighting coefficients for the GaN switching states. The data after filtering at time k-1 The filtered parameter estimate at time k refers to the GaN junction temperature T. j Charging current I, This is the control input at time k-1; Covariance prediction: using the formula The covariance is calculated to determine the degree of error in the predicted data. in, The covariance of the state at time k is estimated. Here is the state transition matrix. For transpose, Estimate the covariance of the state at time k-1; Adaptive gain adjustment: via formula Calculate the filter gain K k R and Q are synchronized and dynamically adjusted according to the GaN state; in, This is the observation matrix, which has no unit and takes a value of 1.

0. Status update: via formula Combined with the original collected data Correct the predicted value, and at the same time when T j At >120℃, increase K k Weighting accelerates the update speed of junction temperature data; After the above processing, the filtered, accurate data at time k is output. That is, the optimized GaN device junction temperature T j Charging current I.

7. The monitoring method for a charger to avoid overcharging according to claim 6, characterized in that, The specific preprocessing of real-time parameters by the BP neural network fitting model includes: The input data includes real-time data and denoised data after Kalman filtering, including the individual unit voltage V. cell Total voltage V pack Optimized charging current I, cell temperature T, voltage change rate t, and current change rate; GaN device characteristic parameters, including GaN switching frequency f GaN GaN junction temperature T j GaN switching loss P loss GaN driving voltage V GaN ; The two types of data, totaling 10 parameters, serve as the input layer data for the BP neural network. For the 10 acquired input parameters, a three-step targeted processing method is used, taking into account the characteristics of GaN, to complete the parameter coupling correction: Model structure adaptation: The 3-layer BP neural network structure was optimized by connecting 10 input parameters to the input layer, setting 12 neurons in the hidden layer, using the Sigmoid activation function, and setting the output layer to have 3 core parameters V. cell T j 、I; The model was trained using measured data from GaN devices under all operating conditions. GaN characteristic constraints are added to the model output layer to perform targeted optimization on the corrected parameters: When the GaN junction temperature T j When the temperature is above 120℃, the correction amplitude of the charging current I is forcibly increased to avoid current acquisition deviation caused by excessive junction temperature. When the GaN switching frequency f GaN When the frequency is >1.5MHz, the single-unit voltage V cell The correction error is controlled within ±3mV to suppress voltage fluctuation deviation caused by high-frequency switching; After model training and constraint correction, the precise correction values ​​of three core parameters are output, namely the corrected unit voltage V. cell GaN junction temperature T j Charging current I.

8. The monitoring method for a charger to avoid overcharging according to claim 7, characterized in that, The specific details of the multi-stage charging strategy are as follows: Constant current fast charging stage: Relying on the high efficiency of gallium nitride devices, charging is performed at a constant current of 0.5C~1C until the single-cell voltage reaches 4.1V. The voltage rise rate is monitored. If dV / dt>0.1V / min, an abnormality is determined. The MCU immediately controls the gallium nitride device to reduce the current to 0.2C. Constant voltage and current limiting phase: Maintain a constant voltage of 4.2V, with gallium nitride devices precisely regulating the output current to ensure a gradual decrease in current; Triggering full charge condition: A full charge is determined when the current drops to 0.03C and remains there for 10 minutes. A full charge is determined when the voltage remains stable at 4.2V±10mV for more than 30 minutes. Gallium nitride device junction temperature If the temperature exceeds 120℃, the power will be reduced simultaneously. If the temperature continues to rise, the full charge shutdown will be triggered. Trickle maintenance or shutdown phase: After full charge, the main charging circuit is cut off to control the gallium nitride device to turn off, and only the <0.01C micro-current charging option is retained. The gallium nitride device provides a stable micro-current to reduce power consumption. Overtime protection: If the total charging time exceeds 8 hours, the gallium nitride device will be forcibly shut down to avoid overcharging caused by prolonged trickle charging.

9. A method for monitoring a charger to avoid overcharging according to claim 8, characterized in that, The expression for the composite risk factor R in S3 is: ； in, This is the voltage risk weighting coefficient. This is the current risk weighting coefficient. This is the cell temperature risk weighting coefficient. This is the risk weighting coefficient for GaN junction temperature. This represents the real-time voltage of a single battery cell. The full charge voltage threshold for lithium batteries. For real-time charging circuit current, The threshold for trickle charging current. The safe temperature threshold for the battery cell. For real-time cell surface temperature, Real-time junction temperature for GaN devices The safe junction temperature threshold for GaN devices, For overfilling composite risk factors; The specific content of overcharge risk warning and graded control includes: When R>a, an early warning is issued and the power of the gallium nitride device is reduced to 50%, the APP alarm is triggered, and the working status of the gallium nitride is simultaneously indicated. When R>b and b>a, protection measures are implemented to control the gallium nitride device to shut down, initiate heat dissipation, and cut off the charging circuit. When any of the total voltage, charger temperature, or gallium nitride device junction temperature exceeds the corresponding preset value, a hard cutoff is performed, and the protection IC independently shuts down the gallium nitride MOS. It cannot be automatically restored and requires power-off reset.

10. A method for monitoring a charger to avoid overcharging according to claim 9, characterized in that, The monitoring method is also equipped with hardware-level redundancy protection, specifically including: Independent protection IC: Independent of software, the charging circuit is immediately cut off when the voltage is >4.35V or the junction temperature of the gallium nitride device is >150℃, forming dual redundancy with MCU control; Physical protection: Cell expansion triggers a mechanical switch, forcibly disconnecting the charging interface and simultaneously shutting off the power supply to the gallium nitride device to prevent the device from being damaged under no-load conditions.

Images