Control system for sharing charging voltage and current of energy storage power supply

Abstract

The invention relates to the technical field of power supply control, in particular to a shared energy storage power supply charging voltage and current magnitude control system, which comprises an information acquisition unit, a main control unit, a communication protocol processing unit, an auxiliary monitoring unit, a state monitoring unit and a parameter adjusting unit, the system combines a long and short term memory network (LSTM) in deep learning and a deep Q network (DQN) in reinforcement learning, analyzes historical battery state data through an LSTM module, accurately predicts a future state of charge (SOC), and dynamically selects an optimal charging strategy in combination with a DQN module, thereby ensuring that the charging process is efficient and safe, and improving the charging efficiency. And the communication protocol processing unit can automatically identify and adapt to various quick charging protocols.

Description

Technical Field

[0001] The present invention relates to the technical field of power supply control, and in particular to a control system for charging voltage and current of a shared energy storage power supply. Background Art

[0002] With the popularity of mobile devices and the growing demand for portable power, shared energy storage power sources (such as shared power banks, electric vehicle charging piles, etc.) have become an indispensable part of modern life. These devices not only provide users with convenient charging services, but also improve energy efficiency and user experience through intelligent management and optimization. However, with the advancement of battery technology and the increasing requirements of users for charging speed, traditional charging control systems have gradually exposed some limitations: low charging efficiency, traditional charging methods mostly use constant current or constant voltage modes with fixed parameters, and cannot dynamically adjust the charging strategy according to the battery status, resulting in long charging time and low efficiency. Insufficient safety. During the fast charging process, if the voltage and current are not accurately controlled, it may cause problems such as overheating and overcharging, and even cause battery damage or safety accidents. Poor compatibility. Devices of different brands and models support different charging protocols. The existing system is difficult to flexibly adapt to multiple fast charging standards, which limits the scope of use of users.

[0003] Lacking intelligent management, traditional systems often do not have the ability to learn and cannot predict battery behavior or optimize future charging processes based on historical data, affecting long-term performance.

[0004] Most charging control systems on the market currently rely mainly on preset charging curves to guide the charging process. Although this approach is simple and direct, it is inadequate when faced with complex and changeable actual application scenarios. For example, in the early stages of charging, due to the low battery voltage, a higher charging current is usually required to speed up the charging process; and when the battery is close to full charge, the current should be reduced to prevent overcharging. In addition, factors such as temperature changes and aging will also affect battery characteristics, and a single fixed charging solution is difficult to take into account all situations. Summary of the invention

[0005] In order to solve the problems mentioned in the above background technology, the present invention provides a control system for charging voltage and current of a shared energy storage power supply.

[0006] The present invention provides a control system for charging voltage and current of a shared energy storage power supply, which adopts the following technical solution: A control system for charging voltage and current of a shared energy storage power supply, characterized in that it includes an information acquisition unit, a main control unit, a communication protocol processing unit, an auxiliary monitoring unit, a state monitoring unit, and a parameter adjustment unit. The information acquisition unit is used to read the basic information of the inserted device, and after being processed by a signal conditioning circuit, the information is converted into a digital format and sent to the main control unit. The main control unit receives data from the information acquisition unit and calculates the most suitable charging curve based on the data. The main control unit includes an LSTM prediction module and a DQN optimization module. The LSTM prediction module is used to predict future SOC changes based on historical battery status data. The DQN optimization module is used to optimize the charging strategy through reinforcement learning based on the current battery status and the prediction results. The communication protocol processing unit is used to identify and process device requests that support fast charging protocols. The auxiliary monitoring unit is used to continuously monitor the working status of the main control unit and its output results. When an abnormality is detected, the auxiliary monitoring unit is switched to take over the task. The state monitoring unit is used to regularly check the status of internal components of the system. When a potential problem is found, the corresponding protection mechanism is triggered. The parameter adjustment unit is used to receive instructions from the main control unit and adjust the actual charging parameters accordingly to ensure that the changes in voltage and current during the charging process meet the preset optimal charging curve.

[0007] Furthermore, the information acquisition unit includes a sensing circuit, a signal conditioning circuit, and an interactive circuit. The sensing circuit includes a voltage sensor, a current sensor, and a temperature sensor. The signal conditioning circuit includes an electrically connected filter and an amplifier. The interactive circuit includes an electrically connected analog-to-digital converter and a microcontroller. The voltage sensor, the current sensor, and the temperature sensor are respectively electrically connected to a signal conditioning circuit and then connected to the interactive circuit.

[0008] Further, the main control unit is configured to execute the following steps: S1. Initialization phase: set the initial charging voltage V0 and current I0, define the default safety thresholds, including the maximum temperature, maximum charging current, and termination voltage; set up the reinforcement learning environment, define the state space, action space, and reward function, and initialize the Q value network parameters θ and target network parameters ; S2. Collect data from the information acquisition unit, including voltage V t 、Current I t , Temperature T t 、SOCS t , clean and smooth the data, remove outliers and noise, and convert the current and past battery status data [V t-k:t ,I t-k:t ,T t-k:t ,S t-k:t ] is input into the LSTM model and the following LSTM internal operations are performed: a. Forget Gate: ; In the formula, f t is the output of the forget gate at time step t, W f is the weight matrix, b f is the bias term, h t-1 is the hidden state at the previous moment, x t is the input vector at the current moment, including voltage, current, and temperature, and σ is the sigmoid activation function; b. Input gate: Update candidate values: ; In the formula, is the new candidate cell state at time step t, tanh is the hyperbolic tangent activation function, and W c is the weight matrix of the candidate cell state, b c is the bias term of the candidate cell state; Input gate signal: ; In the formula, i t is the output vector of the input gate at time step t, W i is the weight matrix of the input gate, b i is the bias term of the input gate; c. Cell status update: ; In the formula, c t is the updated cell state at time step t, f t is the output of the forget gate, C t-1 is the cell state at the previous moment, i t is the output of the input gate, is the new candidate cell state, ⊙ is the element-wise multiplication; d. Output gate: Output gate signal: ; In the formula, o t is the output vector of the output gate at time step t, W o is the weight matrix of the output gate, b o is the bias term of the output gate; Current hidden status: ; In the formula, h t is the hidden state at time step t, c t is the current cell state, Tanh (c t ) is the cell state processed by the activation function to ensure that the value range is between (1,1).

[0009] e. Final output: ; In the formula, is the predicted SOC at time step t+1, It is a fully connected layer.

[0010] S3. Prediction results based on LSTM , initially set the charging voltage V t+1 and current I t+1 , if the device supports the fast charging protocol, it cooperates with the communication protocol processing unit to determine the optimal fast charging parameters; S4: In the current state, use the optimization module to select an action a t , after executing the action, observe the new state s t+1 and instant rewards t , update the Q value network parameters θ and the target network parameters ; S5: The final determined charging parameter V t+1 and I t+1 The data is sent to the parameter adjustment unit for execution, and the feedback data is monitored in real time to further fine-tune the charging parameters according to the actual situation.

[0011] Further, the communication protocol processing unit is configured to execute the following steps: When the user inserts a device that supports the fast charging protocol, the communication protocol processing unit immediately starts a self-test program to confirm that all hardware components are working properly. Through the preset handshake protocol, it automatically detects and identifies the specific fast charging protocol type supported by the connected device. Once the fast charging protocol supported by the device is determined, the communication protocol processing unit communicates bidirectionally with the device to negotiate the optimal charging voltage and current parameters. Based on the negotiation results, the output parameters of the charging system are adjusted to ensure that they meet the safety charging specifications recommended by the device manufacturer.

[0012] Further, the auxiliary monitoring unit is configured to execute the following steps: Continuously listen to the status report of the main control unit, monitor the control instructions issued by the main control unit and their execution results. If the main control unit is found to have abnormal behavior, including instruction timeout, illegal operation code, and output beyond the safe range, the alarm mechanism will be triggered immediately. When it is confirmed that the main control unit has indeed failed or there is a safety hazard, the auxiliary monitoring unit automatically takes over the task and continues the unfinished charging process. It recalculates the optimal charging curve according to the latest battery status information and sends the adjusted parameters to the parameter adjustment unit. When the main control unit resumes normal operation, the auxiliary monitoring unit smoothly transfers the task to the main control unit.

[0013] Furthermore, the parameter adjustment unit includes a signal processing circuit and a control circuit. The signal processing circuit includes a digital-to-analog converter and an operational amplifier. The control circuit includes a DC-DC converter and a MOSFET. The digital-to-analog converter is connected to the main control unit, the output of the digital-to-analog converter is connected to the input of the operational amplifier in the power stage circuit, the DC-DC converter output is connected to the positive electrode of the battery, the gate of the MOSFET is connected to the output of the operational amplifier, and the drain of the MOSFET is connected to the node between the positive electrode of the battery and the output of the DC-DC converter.

[0014] Furthermore, the VIN pin on the input side of the DC-DC converter is connected to the system power input, the GND pin is connected to the common ground, an electrolytic capacitor is connected between the VIN pin and the GND pin, the VOUT pin on the output side of the DC-DC converter is connected to the positive electrode of the battery, a feedback voltage divider network composed of two resistors is connected between the VOUT pin and the GND pin, and the middle node of the two resistors is connected to the DC-DC converter feedback pin.

[0015] Further, the supply pin of the operational amplifier is connected to the power rail, the non-inverting input of the operational amplifier is connected to the analog output of the digital-to-analog converter, the inverting input of the operational amplifier is connected to the load current path through a current sensing resistor or to the battery terminal voltage through a voltage divider network, and the output of the operational amplifier is connected to the feedback input of the DC-DC converter.

[0016] Beneficial technical effects of the present invention: This application proposes an intelligent control system for controlling the charging voltage and current of a shared energy storage power supply. The system combines the long short-term memory network (LSTM) in deep learning and the deep Q network (DQN) in reinforcement learning. The technical effects are reflected in: Accurate prediction and optimization: The LSTM module analyzes historical battery status data to accurately predict the future state of charge (SOC), and combines the DQN module to dynamically select the optimal charging strategy to ensure that the charging process is both efficient and safe.

[0017] Wide compatibility: The communication protocol processing unit can automatically identify and adapt to a variety of fast charging protocols to meet the needs of different types of devices and provide a personalized charging experience.

[0018] Redundant design and safety assurance: The auxiliary monitoring unit can take over the task when the main control unit fails to ensure uninterrupted charging process; at the same time, a complete built-in protection mechanism can effectively prevent various potential risks.

[0019] Continuous learning and adaptation: The system has the ability to self-learn and can continuously optimize the charging strategy based on accumulated experience, adapt to different types of batteries, and improve overall performance over time.

[0020] In summary, the intelligent control system provided in this application not only overcomes the limitations of traditional methods, but also achieves technological innovation at multiple levels, bringing more advanced and reliable solutions to the field of shared energy storage power supplies. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 is a system connection block diagram of the present invention; Figure 2 It is a circuit schematic diagram of the information acquisition unit of the present invention. DETAILED DESCRIPTION

[0022] The embodiment of the present invention discloses a control system for charging voltage and current of a shared energy storage power supply. A control system for charging voltage and current of a shared energy storage power supply comprises an information acquisition unit, a main control unit, a communication protocol processing unit, an auxiliary monitoring unit, a state monitoring unit and a parameter adjustment unit. The information acquisition unit is used to read basic information of an inserted device, and after being processed by a signal conditioning circuit, the information is converted into a digital format and sent to the main control unit. The main control unit receives data from the information acquisition unit and calculates the most suitable charging curve based on the data. The main control unit comprises an LSTM prediction module and a DQN optimization module. The LSTM prediction module is used to predict future SOC changes based on historical battery state data. The DQN optimization module is used to optimize the charging strategy through reinforcement learning based on the current battery state and the prediction result. The communication protocol processing unit is used to identify and process device requests supporting a fast charging protocol. The auxiliary monitoring unit is used to continuously monitor the working state of the main control unit and its output results. When an abnormality is detected, the auxiliary monitoring unit is switched to take over the task. The state monitoring unit is used to regularly check the state of components inside the system. When a potential problem is found, the corresponding protection mechanism is triggered. The parameter adjustment unit is used to receive instructions from the main control unit and adjust the actual charging parameters accordingly to ensure that the changes in voltage and current during the charging process meet the preset optimal charging curve.

[0023] refer to Figure 1As shown in the figure, the information acquisition unit is connected to the main control unit. The information acquisition unit sends the processed digital signal to the main control unit through the communication interface (such as I2C, SPI). It is responsible for reading the basic information of the inserted device (such as voltage, current, temperature, etc.), and converting it into a format suitable for ADC processing through the signal conditioning circuit, and then converting it into a digital signal by the analog-to-digital converter, and finally sending it to the main control unit after preliminary processing by the microcontroller (MCU). The main control unit is connected to the information acquisition unit, receives data from the information acquisition unit, and calculates the most suitable charging curve based on the data; the main control unit is connected to the communication protocol processing unit, and when it needs to support the fast charging protocol, the main control unit cooperates with the communication protocol processing unit to determine the optimal fast charging parameters, the main control unit is connected to the auxiliary monitoring unit, and the main control unit reports its working status and output results to the auxiliary monitoring unit so that the auxiliary monitoring unit can monitor in real time and prepare to take over the task; the main control unit is connected to the status monitoring unit, and the main control unit regularly checks the status of the internal components of the system, and triggers the protection mechanism according to the information provided by the status monitoring unit; the main control unit is connected to the parameter adjustment unit, and the main control unit generates the final charging parameters according to the optimization strategy, and sends instructions to the parameter adjustment unit for execution through the SPI / I2C communication interface. The connection between the communication protocol processing unit and the main control unit. After determining the fast charging protocol supported by the device, the communication protocol processing unit collaborates with the main control unit to negotiate the optimal charging voltage and current parameters, and adjusts the output parameters of the charging system to meet the safe charging specifications recommended by the device manufacturer; the connection between the communication protocol processing unit and the external device automatically detects and identifies the specific fast charging protocol type supported by the connected device through the handshake protocol to ensure the compatibility and interoperability between the system and other devices. The connection between the auxiliary monitoring unit and the main control unit continuously monitors the status report of the main control unit, takes over the task when abnormal behavior is detected, and continues the unfinished charging process. Once the main control unit returns to normal, the auxiliary monitoring unit smoothly transfers to the main control unit. The connection between the auxiliary monitoring unit and the parameter adjustment unit. If the main control unit fails, the auxiliary monitoring unit can directly send the adjusted parameters to the parameter adjustment unit to ensure the continuous operation of the charging process. The connection between the status monitoring unit and the main control unit regularly checks the status of the internal components of the system and provides relevant information to the main control unit to help the main control unit evaluate whether the protection mechanism needs to be triggered. The parameter adjustment unit is connected to the main control unit, receives instructions from the main control unit, and adjusts the actual charging parameters accordingly to ensure that the changes in voltage and current during the charging process meet the preset optimal charging curve. The parameter adjustment unit is connected to the auxiliary monitoring unit, and in the event of a failure of the main control unit, it receives instructions from the auxiliary monitoring unit to ensure the continuity of the charging process.

[0024] refer to Figure 2As shown, the specific structure of the information acquisition unit includes a sensing circuit, a signal conditioning circuit, and an interactive circuit. The sensing circuit includes a voltage sensor, a current sensor, a temperature sensor, and a communication interface. The signal conditioning circuit includes an electrically connected filter and an amplifier. The interactive circuit includes an electrically connected analog-to-digital converter and a microcontroller. The voltage sensor, current sensor, and temperature sensor are respectively electrically connected to a corresponding signal conditioning circuit and then connected to the interactive circuit. The voltage sensor of the sensing circuit is used to measure the battery terminal voltage. The current sensor is used to monitor the charging current in real time, the temperature sensor is used to monitor the battery surface or ambient temperature, and the communication interface is I2C / SPI, which is used to read the information in the battery management chip. The voltage sensor and current sensor are installed close to the battery connection point to reduce the lead length and electromagnetic interference. The temperature sensor is installed near the charging interface or in the key area inside the charging station to ensure that the temperature change of the battery surface can be sensed. Each sensor corresponds to an independent signal conditioning circuit, including: The filter is used to remove high-frequency noise to ensure signal purity, and the amplifier is used to amplify the weak signal output by the sensor to a level suitable for ADC processing. The analog-to-digital converter of the interactive circuit is a multi-channel analog-to-digital converter (ADC), which is used to receive analog signals from various signal conditioning circuits and convert them into digital format. Microcontroller (MCU): Responsible for coordinating the work of each sensor and exchanging information with the main control DSP through the selected communication protocol. The voltage sensor, current sensor and temperature sensor are respectively connected to their respective signal conditioning circuits, each of which includes a filter and an amplifier.

[0025] The filtered and amplified signals are output from the signal conditioning circuit of each sensor and then connected to a multi-channel ADC. The ADC converts the analog signals of all sensors into digital signals and sends them to the MCU for further processing. The MCU exchanges data with the main control DSP or other system components through communication interfaces (such as UART, I2C, SPI).

[0026] When the user plugs in the shared energy storage power supply, the information acquisition unit immediately starts the self-test program to confirm that all hardware components are working properly.

[0027] MCU controls each sensor to start working, reads the battery voltage, current, and temperature parameters in turn, and obtains the battery model, SOC, and SOH information from the BMS. The collected data is first processed by the signal conditioning circuit and then converted into digital format by the ADC. The MCU further processes these digital signals, calculates the average value, removes abnormal points, and ensures data quality. The processed data is sent to the main control DSP through the selected communication protocol. In order to ensure the stability and real-time performance of data transmission, a double buffering mechanism is adopted, that is, when one buffer is being read, the other buffer continues to receive new data. During the entire charging process, the information acquisition unit remains activated and updates the battery status information regularly (such as once a second) to provide the main control unit with the latest reference basis.

[0028] The communication protocol processing unit is a key component in the charging control system responsible for identifying and processing device requests that support fast charging protocols (such as USB-PD, QC). It can automatically detect the fast charging protocol supported by the connected device and exchange data with it, providing users with safe and efficient fast charging services. In addition, the unit also ensures compatibility and interoperability between the system and other devices.

[0029] Instructions for configuring the communication protocol processing unit to perform the following steps: Protocol identification: When the user plugs in a device that supports the fast charging protocol, the communication protocol processing unit immediately starts a self-test program to confirm that all hardware components are working properly. Through the preset handshake protocol, it automatically detects and identifies the specific fast charging protocol type (such as USB-PD, QC, SCP) supported by the connected device.

[0030] Parameter negotiation: Once the fast charging protocol supported by the device is determined, the communication protocol processing unit conducts two-way communication with the device to negotiate the optimal charging voltage and current parameters. Based on the negotiation results, the output parameters of the charging system are adjusted to ensure that they meet the safe charging specifications recommended by the device manufacturer.

[0031] Data exchange: During the entire charging process, the battery status and other related parameters are continuously monitored, and updated information is sent to the device in real time. If there are any changes on the device side (such as temperature rise, power reaching a threshold), the charging strategy is adjusted in time to ensure the safety and efficiency of the charging process.

[0032] Abnormal handling: Monitor the status of the entire communication link. If any abnormal situation is found (such as communication interruption, illegal command), take protective measures immediately, such as suspending charging or disconnecting.

[0033] Record detailed log information for subsequent analysis and troubleshooting.

[0034] The auxiliary monitoring unit is a key redundant component in the charging control system and works independently of the main control unit. Its main responsibility is to continuously monitor the working status of the main control unit and its output results to ensure the stability and reliability of the system. When any abnormal situation is detected, the auxiliary monitoring unit can quickly take over the task to ensure the continuous operation of the charging process and take necessary protective measures to avoid potential safety risks. The built-in self-diagnosis function regularly checks its own operating status, including memory consistency and program flow correctness, to ensure its healthy operation. Dual power input or backup battery is used to ensure that basic functions can be maintained even in the event of a main power failure.

[0035] Configure the auxiliary monitoring unit to perform the following steps: Initialization and self-test: When the system starts, the auxiliary monitoring unit first performs a self-test procedure to confirm that all hardware components are working properly and initialize internal variables and parameters. It establishes a connection with the main control unit and is ready to receive status information from the main control unit.

[0036] Real-time monitoring: Continuously monitor the status report of the main control unit, including but not limited to CPU load, memory usage, and program execution progress. Monitor the control instructions issued by the main control unit and their execution results to ensure that these instructions meet the preset safety specifications and logical requirements. Simultaneously record the key operation logs of the main control unit to facilitate subsequent analysis and troubleshooting.

[0037] Anomaly detection: If abnormal behavior is detected in the main control unit, including instruction timeout, illegal opcode, and output beyond the safe range, the alarm mechanism is triggered immediately. Detailed abnormality logs are recorded for subsequent analysis and troubleshooting. Users are notified of information about abnormal situations to maintain transparency.

[0038] Take over the task: When it is confirmed that the main control unit has indeed failed or there is a safety hazard, the auxiliary monitoring unit automatically takes over the task and continues the unfinished charging process. The optimal charging curve is recalculated based on the latest battery status information, and the adjusted parameters are sent to the parameter adjustment unit. Once any situation that may cause danger (such as overheating, overcharging) is detected, the protection mechanism is immediately activated, such as reducing power, suspending charging, or disconnecting the power supply.

[0039] When the main control unit resumes normal operation, the auxiliary monitoring unit smoothly transfers the task to the main control unit and generates a detailed event report. The report includes the time point of the failure, cause analysis, measures taken and the final result, which helps system maintenance personnel understand the problem and make improvements.

[0040] Under normal circumstances, the auxiliary monitoring unit works closely with the main control unit to ensure the efficient operation of the system. The status information between the two is synchronized regularly to ensure consistency and coordination.

[0041] The parameter adjustment unit is a key component in the charging control system that is responsible for executing instructions from the main control unit, ensuring that the changes in voltage and current during the charging process strictly follow the preset optimal charging curve. It can not only accurately control the charging parameters, but also respond to the adjustment requests of the main control unit in real time to ensure the safety and efficiency of the charging process.

[0042] The parameter adjustment unit includes a digital-to-analog converter, a power stage circuit, a MOSFET drive circuit, a feedback loop, and a protection circuit. The digital control signal of the main control unit is converted into an analog voltage through the digital-to-analog converter. The analog voltage is used as the input of the operational amplifier in the power stage circuit to accurately adjust the output voltage and current of the DC-DC converter. The input of the DC-DC converter is connected to the system power supply and the voltage is stabilized by a capacitor; the output of the DC-DC converter is connected to the positive electrode of the battery and the output is ensured to be stable through a feedback voltage divider network and a low ESR ceramic capacitor. The MOSFET drive circuit switches quickly according to the PWM signal to control the charging current. The feedback loop uses an operational amplifier to monitor and adjust the output voltage and current in real time to form a closed-loop control system. The protection circuit includes overvoltage, overcurrent and short-circuit protection mechanisms. The safety and reliability of the system are ensured through TVS diodes, fuses and intelligent power switching components. Specifically, the power pin of the digital-to-analog converter is connected to the power rail of the system, and the SPI / I2C communication interface of the digital-to-analog converter is connected to the main control unit for receiving digital control signals; the output end of the digital-to-analog converter is connected to the input end of the operational amplifier in the power stage circuit, and the power stage circuit includes a DC-DC converter, the VIN pin on the input side of the DC-DC converter is connected to the system power input, the GND pin is connected to the common ground, a 10µF to 100µF electrolytic capacitor is connected between the VIN pin and the GND pin, the VOUT pin on the output side of the DC-DC converter is connected to the positive electrode of the battery, a low ESR ceramic capacitor is connected between the VOUT pin and the GND pin, a feedback voltage divider network composed of two resistors is connected between VOUT and GND, and the middle node of the two resistors is connected to the feedback pin, the MOSFET drive circuit adopts the IRFZ44NN channel MOSFET, the gate of the MOSFET is connected to the output of the operational amplifier to receive the switching signal; the source of the MOSFET is connected to the system ground, the MOSFET drain is connected to the node between the positive electrode of the battery and the output of the DC-DC converter, the feedback loop includes an operational amplifier, the power supply pin of the operational amplifier is connected to the power rail, and a decoupling capacitor is added, the non-inverting input of the operational amplifier is connected to the analog output of the digital-to-analog converter, the inverting input of the operational amplifier is connected to the load current path through a current detection resistor or to the battery terminal voltage through a voltage divider network, the output of the operational amplifier is connected to the feedback input of the DC-DC converter to form a closed-loop control system, the protection circuit includes overvoltage protection (OVP), overcurrent protection (OCP) and short circuit protection (SCP), a TVS diode or a dedicated OVP chip is connected between the positive electrode of the battery and the ground to prevent excessive voltage from damaging the system, a fuse or electronic fuse is connected in series in the power input or the positive electrode line of the battery, and the circuit is immediately disconnected once the current exceeds the safety threshold, a current limiting resistor or an intelligent power switch is connected in series in the positive electrode line of the battery to monitor the current in real time, and quickly cut off the power supply when a short circuit occurs to ensure system safety.

[0043] The working principle of the parameter adjustment unit is: establish a communication connection with the main control unit, continuously listen to the instructions from the main control unit, including the latest charging parameter settings, and use a digital-to-analog converter to convert the digital charging parameters into analog control signals according to the received instructions. Monitor the output voltage and current in real time, and feed back the actual measured values ​​to the main control unit through a high-speed feedback loop. If deviations are found, fine-tune them immediately to ensure that the output is always close to the preset optimal charging curve. When any situation that may endanger the safety of the system or battery is detected, the corresponding protection mechanism is immediately activated, and the fault information is reported to the main control unit for further analysis and processing. The main control unit is regularly reported its own working status, including but not limited to the current output parameters and health status.

[0044] The main control unit is the brain of the charging control system. It is responsible for receiving data from the information acquisition unit and calculating the most suitable charging curve based on this data. It dynamically adjusts the charging parameters (such as voltage and current) according to the current battery status and sends instructions to the execution unit to ensure the safety and efficiency of the charging process. In addition, the main control unit has a built-in intelligent charging algorithm module, which enables the system to better adapt to different types of batteries by continuously learning and optimizing charging strategies.

[0045] The main control unit includes an LSTM prediction module and a DQN optimization module. The LSTM prediction module predicts future SOC changes based on historical battery status data, and the DQN optimization module optimizes the charging strategy through reinforcement learning based on the current battery status and prediction results.

[0046] The model structure of the LSTM prediction module includes: Input layer: receives battery status data (voltage V t 、Current I t , Temperature T t 、SOCS t ).

[0047] Hidden layers: Use multiple layers of LSTM units to capture temporal dependencies and long-term memory.

[0048] Output layer: predict the SOC change trend in the future.

[0049] LSTM internal operations include: 1. ForgetGate ; f t : The output vector of the forget gate at time step t determines the cell state c at the previous moment t-1 What information should be kept or forgotten.

[0050] σ: sigmoid activation function, which compresses the input to the (0,1) interval, indicating the probability of retention or forgetting.

[0051] W f : The weight matrix of the forget gate, used to connect the hidden state h of the previous moment t-1 and the current input x t .

[0052] [h t-1 ,x t ]: The hidden state h at the previous moment t-1 and the current input x t splicing.

[0053] b f : Bias term of forget gate.

[0054] 2. Input Gate Update candidate values: ; is the new candidate cell state at time step t, which is new information that may be added to the current cell state.

[0055] tanh: Hyperbolic tangent activation function, which compresses the input to the (1,1) interval to ensure numerical stability.

[0056] W c : Weight matrix of candidate cell states.

[0057] b c : Bias term for candidate cell states.

[0058] Input gating signal: ; i t : The output vector of the input gate at time step t, which determines how much of the new information should be added to the cell state.

[0059] W i : Weight matrix of input gate.

[0060] b i : Bias term for the input gate.

[0061] 3. Cell State Update (CellStateUpdate) ; c t : The updated cell state at time step t.

[0062] f t: The output of the forget gate determines how much old information to retain.

[0063] C t-1 : The cell state at the previous moment.

[0064] i t : The output of the input gate, which determines how much new information to add.

[0065] : New candidate cell state.

[0066] ⊙: Element-wise multiplication.

[0067] 4. Output Gate Output gate signal: ; o t : The output vector of the output gate at time step t, which determines how much information in the cell state should be output as the hidden state.

[0068] W o : Weight matrix of the output gate.

[0069] b o : Bias term for the output gate.

[0070] Current hidden status: ; h t : The hidden state at time step t is the information that is finally output to the next layer.

[0071] c t : The current cell state.

[0072] Tanh (c t ): The cell state processed by the activation function ensures that the value range is between (1,1).

[0073] 5. Final output (PredictedSOC) ; : Predicted SOC (State of Charge) at time step t+1.

[0074] : A fully connected layer used to transform the hidden state into the predicted future SOC.

[0075] The model structure of the DQN optimization module includes: State space, including the current state of the battery (such as SOC, SOH, temperature), charging parameters (voltage, current), and other relevant environmental information.

[0076] Action space, the actions that can be taken a t The collection includes, but is not limited to, adjusting the charging current, increasing the charging current, and increasing the charging speed; reducing the charging current, reducing the charging speed, and reducing heat; adjusting the charging voltage, increasing the charging voltage, and increasing the voltage within the allowable range can charge the power faster; reducing the charging voltage, slowing down the charging speed, and avoiding the risk of overvoltage; switching the charging mode, constant current (CC) mode, keeping the charging current constant, suitable for the initial charging stage of the battery; constant voltage (CV) mode, keeping the charging voltage constant, suitable for preventing overcharging when the battery is close to full charge; fast charging mode, if the device supports the fast charging protocol, switch to the fast charging mode; standard charging mode, restore to the default charging parameters, suitable for not supporting fast charging or safety considerations; pause / resume charging, suspend charging, temporarily stop charging when an abnormal situation is detected (such as excessive temperature, large voltage fluctuations); resume charging, after confirming that the abnormal situation has been resolved, restart charging; preheating / cooling control, start the preheating function, start in a low temperature environment Automatic heating elements ensure that the battery is at the optimal operating temperature; start the cooling function to reduce the temperature through heat dissipation devices (such as fans and heat sinks) to prevent overheating; adjust the charging rate to accelerate charging, and speed up the charging rate when the battery status and environmental conditions permit; slow down charging, and slow down the charging rate to protect the battery life when it is close to full charge or potential risks are detected; perform maintenance operations, perform balanced charging, and balance the voltage between cells for multi-cell battery packs; perform discharge cycles, and regularly perform shallow discharge and recharge on the battery to maintain battery health; adjust the charge termination conditions, adjust the termination voltage, and set a stricter termination voltage threshold to prevent overcharging; set the remaining capacity threshold, terminate charging in advance as needed, and retain a certain proportion of uncharged capacity; update the protection mechanism, enable / disable overvoltage protection (OVP), overcurrent protection (OCP), short circuit protection (SCP), etc.; adjust the protection threshold, and dynamically adjust the trigger point of the protection mechanism according to real-time monitoring data.

[0077] Reward function, defines the goal, such as maximizing charging efficiency, minimizing charging time, and keeping temperature within a safe range.

[0078] The Q-value network uses a neural network to evaluate the value (Q-value) of taking each action in each state.

[0079] The internal operations of the DQN optimization module include: 1. Q-value function approximation: [Qs,a;θ]; s is the current state, a is the action taken, and θ is the parameter of the Q-value network.

[0080] 2. Loss Function: ; r is the immediate reward, γ is the discount factor, θ - are fixed target network parameters and D is the experience replay buffer.

[0081] 3. Gradient descent update rule: ; α is the learning rate.

[0082] 4. Experience Replay: Store past experiences s, a, r, s' into buffer D and randomly extract a batch of samples for update during each training.

[0083] 5. Fixed Q targets (FixedQtargets): Periodically copy the current network parameters to the target network θ - , to stabilize the training process.

[0084] The calculation process of the main control unit: Initialization phase 1. Initialization parameters: Set the initial charging voltage V0 and current I0. Define default safety thresholds (such as maximum temperature, maximum charging current, and termination voltage).

[0085] 2. Set up the reinforcement learning environment: Define the state space, action space and reward function. Initialize the Q value network parameters θ and the target network parameters θ - .

[0086] Real-time control and learning loop Step 1: Data collection and preprocessing Collect data from the information acquisition unit, including the battery voltage V t 、Current I t , Temperature T t 、SOCS t .

[0087] Clean and smooth the data to remove outliers and noise.

[0088] Step 2: Use LSTM to predict future states The current and past battery status data [V t-k:t , I t-k:t , T t-k:t , S t-k:t] is input into the LSTM model. LSTM outputs the prediction of future SOC change trends .

[0089] Step 3: Dynamically adjust charging parameters According to the prediction results of LSTM , initially set the charging voltage V t+1 and current I t+1 ,If the device supports the fast charging protocol, it collaborates with the communication protocol processing unit to determine the optimal fast charging parameters.

[0090] Step 4: Apply DQN optimization strategy In the current state, use DQN to select an action a t (Such as adjusting the charging current).

[0091] After executing the action, observe the new state s t+1 (such as new SOC, temperature changes) and instant rewards t (Such as charging rate increases and temperature does not exceed safe range).

[0092] Update the Q value network parameter θ and the target network parameter θ - , to reflect the newly learned experience.

[0093] Step 5: Implementing Charging Control The final charging parameter V t+1 and I t+1 Sent to the parameter adjustment unit for execution.

[0094] Monitor feedback data (such as actual voltage, current, and temperature) in real time, and further fine-tune charging parameters based on actual conditions.

[0095] Through the above configuration, the main control unit can realize intelligent control of the charging voltage and current of the shared energy storage power supply. It can not only dynamically adjust the charging parameters according to real-time data to ensure the safety and efficiency of the charging process, but also enhance the system's adaptability to different types of batteries by continuously learning and optimizing charging strategies. This combined method fully utilizes the advantages of LSTM and DQN to provide a flexible, adaptive and efficient solution.

[0096] In terms of specific circuits, hardware structures or physical entities, the main control unit can use a microcontroller (MCU) or a digital signal processor (DSP) as the core processing component. For example, MCUs like the STM32 series have powerful computing capabilities, rich peripheral interfaces and flexible programming features, and can execute complex control algorithms and data processing tasks. It can receive data from the information acquisition unit, coordinate the collaborative interaction between modules according to the established program logic, and is the core computing and control basis for realizing the entire charging control strategy. The main control unit must complete the calculation based on the LSTM prediction module and the DQN optimization module. The MCU or DSP can efficiently handle tasks such as battery status data analysis, prediction, and optimal selection of charging strategies, so as to dynamically adjust the charging parameters based on real-time data, ensure the charging process is safe and efficient, and adapt to the functions of different battery types through continuous learning.

[0097] A certain capacity of built-in flash memory is used to store historical battery status data and related program codes, etc. Because the LSTM prediction module relies on historical data to predict future SOC changes, such a storage unit is needed to preserve data for a long time and facilitate analysis at any time.

[0098] Establish connections with other units through communication interface circuits such as I2C and SPI. For example, use the I2C interface to communicate with the information acquisition unit and receive the digital signal containing basic information such as battery voltage, current, and temperature; connect to the parameter adjustment unit through the SPI interface, send the finalized charging voltage, current and other parameters in the form of digital instructions, and command the parameter adjustment unit to perform the corresponding operation; and the communication protocol processing unit also relies on the corresponding communication line to exchange data, and when the device supports the fast charging protocol, jointly complete the negotiation and determination of charging parameters. In addition, through a specific communication link, the auxiliary monitoring unit reports its own working status and other information to it, so that the auxiliary monitoring unit can monitor the operation of the main control unit in real time.

[0099] In the specific circuit, hardware structure or physical entity, the auxiliary monitoring unit uses a monitoring chip to monitor the working status of the main control unit and its output results in real time. For example, some circuit modules with comparator and logic judgment functions can continuously compare the control instructions issued by the main control unit and their execution results to see if they meet the preset safety specifications and logic requirements, and promptly discover abnormal behaviors such as instruction timeout, illegal operation code, and output beyond the safety range. In addition, a corresponding timer circuit is built in to assist in judging whether the instruction has timed out, etc., to ensure the accuracy and timeliness of the main control unit status monitoring.

[0100] In specific circuits, hardware structures or physical entities, the state monitoring unit can use a variety of sensors to obtain the status information of the internal components of the system. For example, temperature sensors are used to monitor the temperature changes of key circuit parts to prevent abnormalities such as overheating; voltage sensors are distributed at various key power supply nodes to detect whether the power supply voltage of each main module is normal; current sensors are installed in appropriate locations to monitor the current size and stability in the system. The signals collected by these sensors are often weak or have noise interference, so they are matched with corresponding signal conditioning circuits, such as circuits composed of amplifiers and filters, to amplify and filter the sensor signals and convert them into standard signal forms suitable for subsequent processing, so as to accurately judge the actual status of the internal components of the system. Equipped with a microcontroller (MCU) to coordinate the work of each sensor and collect, analyze and process the conditioned sensor signals.

[0101] In the specific circuit, hardware structure or physical entity, the communication protocol processing unit uses a protocol identification and processing chip, such as an application-specific integrated circuit (ASIC) that supports multiple fast charging protocol identification. When the user inserts a device that supports the fast charging protocol, it can quickly start the self-test program, confirm that all hardware components are working properly, and then rely on the built-in protocol parsing module to automatically detect and identify the specific fast charging protocol type supported by the connected device (such as USB-PD, QC, SCP, etc.) through the preset handshake protocol. These chips have solidified the standard communication processes and parameter parsing rules of different protocols, and can complete the protocol identification work efficiently and accurately.

[0102] The above are all preferred embodiments of the present invention, and are not intended to limit the protection scope of the present invention. Therefore, any equivalent changes made based on the structure, shape, and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A control system for charging voltage and current of a shared energy storage power source, characterized in that: It includes an information acquisition unit, a main control unit, a communication protocol processing unit, an auxiliary monitoring unit, a state monitoring unit, and a parameter adjustment unit. The information acquisition unit is used to read the basic information of the inserted device, and after being processed by the signal conditioning circuit, it is converted into a digital format and sent to the main control unit. The main control unit is used to receive data from the information acquisition unit and calculate the most suitable charging curve based on these data. The main control unit includes an LSTM prediction module and a DQN optimization module. The LSTM prediction module is used to predict future SOC changes based on historical battery status data. The DQN optimization module is used to optimize the charging strategy through reinforcement learning based on the current battery status and prediction results. The communication protocol processing unit is used to identify and process device requests that support the fast charging protocol. The auxiliary monitoring unit is used to continuously monitor the working status of the main control unit and its output results. When an abnormality is detected, it is switched to the auxiliary monitoring unit to take over the task. The state monitoring unit is used to regularly check the status of the internal components of the system. When potential problems are found, the corresponding protection mechanism is triggered. The parameter adjustment unit is used to receive instructions from the main control unit and adjust the actual charging parameters accordingly to ensure that the changes in voltage and current during the charging process meet the preset optimal charging curve.

2. A control system for charging voltage and current of a shared energy storage power supply according to claim 1, characterized in that: The information acquisition unit includes a sensing circuit, a signal conditioning circuit, and an interactive circuit. The sensing circuit includes a voltage sensor, a current sensor, and a temperature sensor. The signal conditioning circuit includes an electrically connected filter and an amplifier. The interactive circuit includes an electrically connected analog-to-digital converter and a microcontroller. The voltage sensor, the current sensor, and the temperature sensor are respectively electrically connected to a corresponding signal conditioning circuit and then connected to the interactive circuit.

3. A control system for charging voltage and current of a shared energy storage power supply according to claim 1, characterized in that: Configure the main control unit to perform the following steps: S1. Initialization phase: set the initial charging voltage V0 and current I0, define the default safety thresholds, including the maximum temperature, maximum charging current, and termination voltage; set up the reinforcement learning environment, define the state space, action space, and reward function, and initialize the Q value network parameters θ and target network parameters ; S2. Collect data from the information acquisition unit, including voltage V t 、Current I t , Temperature T t 、SOCS t , clean and smooth the data, remove outliers and noise, and convert the current and past battery status data [V t-k:t ,I t-k:t ,T t-k:t ,S t-k:t ] is input into the LSTM model and the following LSTM internal operations are performed: a. Forget Gate: ; In the formula, f t is the output of the forget gate at time step t, W f is the weight matrix, b f is the bias term, h t-1 is the hidden state at the previous moment, x t is the input vector at the current moment, including voltage, current, and temperature, and σ is the sigmoid activation function; b. Input gate: Update candidate values: ; In the formula, is the new candidate cell state at time step t, tanh is the hyperbolic tangent activation function, and W c is the weight matrix of the candidate cell state, b c is the bias term of the candidate cell state; Input gating signal: ; In the formula, i t is the output vector of the input gate at time step t, W i is the weight matrix of the input gate, b i is the bias term of the input gate; c. Cell status update: ; In the formula, c t is the updated cell state at time step t, f t is the output of the forget gate, C t-1 is the cell state at the previous moment, i t is the output of the input gate, is the new candidate cell state, ⊙ is the element-wise multiplication; d. Output gate: Output gate signal: ; In the formula, o t is the output vector of the output gate at time step t, W o is the weight matrix of the output gate, b o is the bias term of the output gate; Current hidden status: ; In the formula, h t is the hidden state at time step t, c t is the current cell state, Tanh (c t ) is the cell state processed by the activation function, ensuring that the value range is between (1,1); e. Final output: ; In the formula, is the predicted SOC at time step t+1, is a fully connected layer; S3. Prediction results based on LSTM , initially set the charging voltage V t+1 and current I t+1 , if the device supports the fast charging protocol, it cooperates with the communication protocol processing unit to determine the optimal fast charging parameters; S4: In the current state, use the optimization module to select an action a t , after executing the action, observe the new state s t+1 and instant rewards t , update the Q value network parameters θ and the target network parameters ; S5: The final determined charging parameter V t+1 and I t+1 The data is sent to the parameter adjustment unit for execution, and the feedback data is monitored in real time to further fine-tune the charging parameters according to the actual situation.

4. A control system for charging voltage and current of a shared energy storage power supply according to claim 1, characterized in that: Instructions for configuring the communication protocol processing unit to perform the following steps: When the user inserts a device that supports the fast charging protocol, the communication protocol processing unit immediately starts a self-test program to confirm that all hardware components are working properly. Through the preset handshake protocol, it automatically detects and identifies the specific fast charging protocol type supported by the connected device. Once the fast charging protocol supported by the device is determined, the communication protocol processing unit communicates bidirectionally with the device to negotiate the optimal charging voltage and current parameters. Based on the negotiation results, the output parameters of the charging system are adjusted to ensure that they meet the safety charging specifications recommended by the device manufacturer.

5. A control system for charging voltage and current of a shared energy storage power supply according to claim 1, characterized in that: Configure the auxiliary monitoring unit to perform the following steps: Continuously listen to the status report of the main control unit, monitor the control instructions issued by the main control unit and their execution results. If the main control unit is found to have abnormal behavior, including instruction timeout, illegal operation code, and output beyond the safe range, the alarm mechanism will be triggered immediately. When it is confirmed that the main control unit has indeed failed or there is a safety hazard, the auxiliary monitoring unit automatically takes over the task and continues the unfinished charging process. It recalculates the optimal charging curve according to the latest battery status information and sends the adjusted parameters to the parameter adjustment unit. When the main control unit resumes normal operation, the auxiliary monitoring unit smoothly transfers the task to the main control unit.

6. A control system for charging voltage and current of a shared energy storage power supply according to claim 1, characterized in that: The parameter adjustment unit includes a signal processing circuit and a control circuit. The signal processing circuit includes a digital-to-analog converter and an operational amplifier. The control circuit includes a DC-DC converter and a MOSFET. The digital-to-analog converter is connected to the main control unit. The output of the digital-to-analog converter is connected to the input of the operational amplifier in the power stage circuit. The output of the DC-DC converter is connected to the positive electrode of the battery. The gate of the MOSFET is connected to the output of the operational amplifier. The drain of the MOSFET is connected to the node between the positive electrode of the battery and the output of the DC-DC converter.

7. A control system for charging voltage and current of a shared energy storage power supply according to claim 6, characterized in that: The VIN pin on the input side of the DC-DC converter is connected to the system power input, the GND pin is connected to the common ground, an electrolytic capacitor is connected between the VIN pin and the GND pin, the VOUT pin on the output side of the DC-DC converter is connected to the positive terminal of the battery, a feedback voltage divider network consisting of two resistors is connected between the VOUT pin and the GND pin, and the middle node of the two resistors is connected to the DC-DC converter feedback pin.

8. A control system for charging voltage and current of a shared energy storage power supply according to claim 6, characterized in that: The supply pin of the operational amplifier is connected to the power rail, the non-inverting input of the operational amplifier is connected to the analog output of the digital-to-analog converter, the inverting input of the operational amplifier is connected to the load current path through a current sense resistor or to the battery terminal voltage through a voltage divider network, and the output of the operational amplifier is connected to the feedback input of the DC-DC converter.

Images