Intelligent charging device and method for central emergency power supply system based on model prediction

Abstract

The invention relates to an intelligent charging device and method for a central emergency power supply system based on model prediction. The device comprises a charging module and a main control module. The main control module collects the real-time state of the battery, and predicts the state trend in a future time window by using an equivalent circuit model; and based on the objective function of efficiency maximization and temperature rise minimization and the security constraint, solving the optimal charging parameter sequence in a rolling manner, and outputting the first cycle parameter to the charging module to adjust the charging output. The method has the advantages that the strategy can be adaptively adjusted according to the dynamic characteristics of the battery, multi-target rolling optimization is achieved, temperature rise is effectively restrained while the charging efficiency is guaranteed, and the reliability of an emergency system is guaranteed.

Description

Technical Field

[0001] This application relates to the field of intelligent control, and in particular to an intelligent charging device and method for a central emergency power supply system based on model prediction. Background Technology

[0002] As a key component ensuring the continuity of power supply for buildings and industrial facilities, the central emergency power supply system's core energy storage battery pack's charging and maintenance level directly affects the system's standby reliability and lifespan. Existing charging devices generally employ traditional constant-current, constant-voltage charging strategies or proportional-integral-derivative (PID) logic control methods when charging energy storage batteries. These technical solutions primarily rely on real-time comparisons between the battery's current terminal voltage, current, or temperature and set threshold values ​​to adjust the charging module's output power. Essentially, their control logic is a feedback adjustment mechanism based on the current state deviation.

[0003] In practical applications, the electrochemical reaction process of energy storage batteries exhibits significant nonlinear and time-varying characteristics. Complex coupling relationships exist between battery terminal voltage fluctuations, internal polarization states, and temperature rise changes, with state changes often accompanied by time lags. Traditional feedback control methods, lacking the ability to predict battery state change trends over a preset time period, often only passively adjust after actual abrupt changes in state parameters or reaching safety boundaries. This lag in adjustment can easily lead to control overshoot or insufficient response when dealing with high-rate fast charging or scenarios with large ambient temperature fluctuations. For example, if the control system maintains a high current input based solely on the current temperature before the internal heat accumulation is fully reflected on the surface temperature sensor, thermal inertia may subsequently cause the battery temperature to exceed the safe range. Furthermore, existing technologies have limitations in handling multi-objective optimization problems, struggling to simultaneously consider multiple interdependent performance indicators such as charging rate, energy conversion efficiency, and battery health. Due to the lack of global solution capabilities based on system models, traditional devices typically operate according to factory-preset fixed curves, unable to dynamically find the optimal charging parameter sequence based on the actual aging of the battery or specific emergency conditions. This makes it difficult to achieve an ideal balance between shortening charging time and suppressing temperature rise during the charging process, which may accelerate battery wear over long-term operation. Summary of the Invention

[0004] To overcome the response lag problem caused by traditional feedback control adjusting only based on the current state deviation, this application provides a model-predictive intelligent charging device and method for a central emergency power supply system.

[0005] Firstly, this application provides a model-based intelligent charging device for a central emergency power supply system, employing the following technical solution: A model-based intelligent charging device for a central emergency power supply system includes: The charging module is used to connect to the energy storage battery pack and adjust the charging output according to the control signal; The main control module, connected to the charging module, is configured as follows: Real-time acquisition of battery status parameters of the energy storage battery pack; Based on the stored battery equivalent circuit model and the battery state parameters, the trend of battery state change within a future preset time window is predicted. Based on the objective function and constraints, the optimal charging parameter sequence is solved based on the battery state change trend. The first control cycle parameter in the optimal charging parameter sequence is output as the control signal to the charging module to execute the charging strategy.

[0006] Optionally, the charging module includes several independent charging branch units; the charging branch units are used to adjust the output voltage and output current in response to the control signals of the main control module.

[0007] The intelligent charging device also includes: An expansion interface module is provided, through which the several independent charging branch units are connected to the main control module.

[0008] Optionally, the main control module further includes: The fault detection unit is used to collect the operating parameters of the charging module, identify overcurrent, overvoltage, undervoltage, short circuit and insulation degradation fault types, and transmit a signal to the main control module to stop charging when a fault is detected. The communication unit is used to establish two-way communication with the external scheduling platform, receive external charging instructions containing the target remaining power and charging time limit, and upload charging status information.

[0009] Optionally, the main control module integrates a battery status acquisition unit, which includes a voltage sensor, a temperature sensor, a current sensor, and a SOC estimation unit. The battery status parameters include at least the individual cell terminal voltage, the battery pack terminal voltage, the battery temperature, the charging current, and the remaining charge (SOC). The SOC estimation unit is configured to perform calculations based on the extended Kalman filter algorithm, combined with the collected battery cell terminal voltage, battery temperature and charging current, to obtain the remaining power SOC.

[0010] Optionally, it also includes a fan cooling module, which is connected to the main control module; the main control module is further configured to perform forward-looking heat dissipation control: based on the predicted battery temperature change trend in the battery state change trend, it determines whether the future temperature will reach a set threshold; if the determination result is yes, it adjusts the fan speed of the fan cooling module in advance before the battery temperature actually rises.

[0011] Optionally, the main control module is further configured to identify the type of the connected energy storage battery pack and adaptively adjust the parameters of the battery equivalent circuit model and the weight coefficients of the optimization objective function according to the type.

[0012] Optionally, the battery equivalent circuit model adopts a second-order RC equivalent circuit model; the optimization objective function is configured to maximize charging efficiency and minimize battery temperature rise. The optimization objective function is expressed as: Where J is the objective function value, , These are the weighting coefficients for charging efficiency and temperature rise, respectively. , , η(k) is the charging efficiency in the k-th control cycle. This represents the battery temperature rise in the (k+1)th control cycle relative to the kth control cycle. The method for solving the optimal charging parameter sequence based on the battery state change trend, based on the optimization objective function and constraints, includes the following steps: (1) State initialization: Based on the real-time parameters of the battery state acquisition unit, initialize the state variables of the battery equivalent circuit model; (2) Rolling prediction: Based on the battery equivalent circuit model, predict the battery SOC, temperature and terminal voltage change trends within the preset time window, wherein the preset time window includes N future control cycles; the value of N ranges from 5 to 10. (3) Optimization solution: With the optimization objectives of maximizing charging efficiency and minimizing battery temperature rise, an optimization objective function is constructed. Under the conditions of satisfying battery voltage constraints, current constraints, temperature constraints and charging time constraints, the optimal charging parameter sequence of the preset time window is obtained; wherein, the optimal charging parameter sequence is the optimal charging current / voltage sequence; (4) Control execution: Output the first control cycle parameter of the optimal charging parameter sequence to the charging module to control the charging process; (5) Feedback correction: In the next control cycle, new battery state parameters are collected, the state variables of the battery equivalent circuit model are corrected, and steps (2)-(4) are repeated to achieve rolling optimization and feedback correction.

[0013] Optionally, the constraints include: Current constraint: Where t is the prediction step size, t∈[1,N], The minimum allowable charging current for the battery. The maximum allowable charging current for the battery. , C is a current value reference equal to the rated capacity of the energy storage battery pack; Voltage constraint: ,in This is the minimum permissible terminal voltage of the battery. This is the maximum permissible terminal voltage of the battery; Temperature constraints: ,in, The battery safety temperature threshold, , t∈[1,N]; Charging time constraint: The total time to complete charging is less than the charging time limit specified in the received external charging command.

[0014] Secondly, this application provides a model-based intelligent charging method for a central emergency power supply system, employing the following technical solution: A model-based intelligent charging method for a central emergency power supply system includes the following steps: Step 1: Real-time acquisition of battery status parameters of the energy storage battery pack; Step 2: Based on the pre-stored battery equivalent circuit model and the collected battery state parameters, predict the battery state change trend within a preset time window in the future. Step 3: Combining the objective function and constraints, solve for the optimal charging parameter sequence based on the battery state change trend; Step 4: Output the first control cycle parameter in the optimal charging parameter sequence to the charging module to control the charging module to perform the charging operation; Step 5: Enter the next control cycle and repeat steps one through four to achieve rolling optimization control.

[0015] In summary, this application includes at least one of the following beneficial technical effects: 1. This application realizes the transformation from passive response to active predictive control by establishing an equivalent circuit model of the battery and predicting the state change trend within a preset time window. It effectively solves the regulation lag problem of traditional constant current constant voltage or proportional integral derivative control when facing nonlinear battery characteristics, and improves the system's response speed to changes in operating conditions.

[0016] 2. This application employs a rolling optimization mechanism, re-acquiring real-time data and solving for the optimal sequence within each control cycle, and only executing the parameters for the first control cycle. This closed-loop feedback correction mechanism can promptly compensate for errors caused by model mismatch or external disturbances, ensuring the control accuracy and system stability of the charging process throughout the entire time axis. 3. This application combines an optimization objective function with multiple constraints when solving for the optimal charging parameter sequence, automatically finding a balance between maximizing charging efficiency and minimizing temperature rise. This effectively improves charging performance and helps extend the battery pack's cycle life while ensuring safe operation, preventing overcharging and overheating. Attached Figure Description

[0017] Figure 1 A schematic diagram of the module structure of an intelligent charging device for a central emergency power supply system based on model prediction is shown in one embodiment of the present invention; Figure 2 A schematic diagram illustrating the workflow of the main control module in one embodiment of the present invention is shown. Figure 3 A flowchart illustrating a model-predictive intelligent charging method for a central emergency power supply system is shown in one embodiment of the present invention. Detailed Implementation

[0018] The present application will be further described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are merely illustrative of the application and are not intended to limit the scope of the application.

[0019] In the following description, numerous specific details are set forth for purposes of explanation in order to provide a thorough understanding of the inventive concept. As part of this specification, some of the accompanying drawings of this disclosure are block diagrams illustrating structures and devices to avoid complicating the disclosed principles. For clarity, not all features of the actual embodiment need to be described. Furthermore, the language used in this disclosure has been primarily chosen for readability and instructional purposes and may not have been chosen to define or limit the subject matter of the invention, thus requiring the necessary claims to determine such inventive subject matter. References to “an embodiment” or “an embodiment” in this disclosure mean that a particular feature, structure, or characteristic described in connection with that embodiment is included in at least one embodiment, and multiple references to “an embodiment” or “an embodiment” should not be construed as necessarily referring to the same embodiment.

[0020] Unless explicitly defined, the terms “a,” “an,” and “the” are not intended to refer to a singular entity, but rather to include a general category whose specific examples can be used for illustration. Therefore, the use of the terms “a” or “an” can mean any number of at least one, including “a,” “one or more,” “at least one,” and “one or more.” The term “or” means any of the options and any combination of the options, including all options unless explicitly indicated that the options are mutually exclusive. The phrase “at least one of” when combined with a list of items refers to a single item in the list or any combination of items in the list. The phrase does not require all items listed unless explicitly defined as such.

[0021] This application discloses an intelligent charging device for a central emergency power supply system based on model prediction, referring to... Figure 1 It consists of a main control module, a charging module, an expansion interface module, and supporting power supply module and fan cooling module.

[0022] The charging module connects to the energy storage battery pack and employs a parallel redundant architecture, comprising several independent charging branch units. Each charging branch unit integrates a high-frequency switching power supply circuit (such as an LLC resonant converter) to respond to control signals from the main control module and adjust the output voltage and current. The expansion interface module serves as a connection hub, featuring a hot-swappable backplane interface design that allows several independent charging branch units to be easily connected to the system bus and the main control module. A significant advantage of this design is the system's scalability and ease of maintenance: when the battery capacity of the central emergency power supply system increases, users can simply insert more charging branch units through the expansion interface to expand the capacity without replacing the entire unit; simultaneously, if a charging branch fails, it can be directly removed and replaced without affecting the normal operation of other branches, greatly improving the availability of the emergency power supply system.

[0023] The main control module is the core of the entire intelligent charging device, and it is electrically connected to the charging module and the fan cooling module. The main control module integrates a battery status acquisition unit, a fault detection unit, a model storage unit, a communication unit, and a model prediction control unit.

[0024] The battery state acquisition unit is responsible for providing data input to the control algorithm. It includes voltage sensors, distributed temperature sensors, current sensors, and a State of Charge (SOC) estimation unit. The acquired battery state parameters include at least the individual cell voltage, battery pack voltage, battery temperature, charging current, and remaining SOC. For example, the temperature sensor can be an NTC thermistor or a fiber optic temperature probe attached to the battery surface, and the current sensor can be a Hall effect current sensor. Due to the nonlinear characteristics of the internal chemical reactions in energy storage batteries, especially lithium-ion batteries, a simple ampere-hour integration method is prone to cumulative errors. Therefore, the SOC estimation unit in this embodiment is configured to operate based on the Extended Kalman Filter (EKF) algorithm. The EKF algorithm uses the individual cell voltage, temperature, and charging current as observations and iteratively refines the SOC estimate, thereby obtaining high-precision remaining SOC data during dynamic charging.

[0025] The fault detection unit is used to collect the battery status parameters of the charging module in real time. Once a fault type such as overcurrent, overvoltage, undervoltage, short circuit, or decreased insulation resistance (insulation degradation) is detected, a hardware interrupt will be immediately triggered, transmitting a signal to the main control module to cut off the charging output at a speed of milliseconds to prevent the accident from escalating. The communication unit is used to establish two-way communication between the device and the external scheduling platform, receiving external charging instructions containing the target remaining power (e.g., setting to charge to 95%) and charging time limit (e.g., requiring completion within 2 hours), and uploading charging status information.

[0026] Currently, traditional central emergency power supply charging devices mostly employ a constant current-constant voltage strategy. While this strategy is simple, it suffers from significant lag. For example, when an excessively high battery temperature is detected, heat accumulation inside the battery has often already occurred. At this point, reducing the current is often insufficient to quickly suppress the temperature rise, potentially leading to shortened battery life or even thermal runaway. To address this issue, this application introduces a model predictive control unit.

[0027] The main control module stores a battery equivalent circuit model, and in this embodiment, a second-order RC equivalent circuit model is preferred. This model consists of an ohmic internal resistance, two parallel RC circuits (representing electrochemical polarization and concentration polarization respectively), and an open-circuit voltage source connected in series. It can accurately characterize the dynamic response characteristics of the battery's terminal voltage, SOC, and temperature under different current excitations.

[0028] Specifically, refer to Figure 2 The working logic of the model prediction control unit is as follows: step ad.

[0029] a. Real-time acquisition of battery status parameters of the energy storage battery pack.

[0030] b. Based on the stored battery equivalent circuit model and the battery state parameters, predict the battery state change trend within a preset time window in the future.

[0031] c. Based on the objective function and constraints, the optimal charging parameter sequence is solved based on the battery state change trend.

[0032] d. The first control cycle parameter in the optimal charging parameter sequence is output as the control signal to the charging module to execute the charging strategy.

[0033] First, based on real-time collected battery state parameters, it predicts the battery state change trend within a preset time window (e.g., the next 5-10 control cycles, each 100ms). Second, based on an optimization objective function and constraints, it solves an optimization problem mathematically to obtain an optimal charging parameter sequence. The optimization objective function is configured to maximize charging efficiency and minimize battery temperature rise. Its mathematical expression is: .

[0034] Where J is the objective function value, , These are the weighting coefficients for charging efficiency and temperature rise, respectively. , , η(k) is the charging efficiency in the k-th control cycle. This represents the battery temperature rise in the (k+1)th control cycle relative to the kth control cycle. It's worth noting that the weighting coefficients can convert and unify the dimensions.

[0035] The design of this function essentially addresses the contradiction between fast charging and battery safety. At a physical level, improving charging efficiency and shortening charging time typically requires increasing the charging current (i.e., increasing η(k)); however, a large current inevitably leads to high internal resistance losses, causing a sharp rise in battery temperature (i.e., increasing...). Traditional charging methods often suffer from one drawback after another, while the function in this application introduces a weighting coefficient. and This transforms the multi-objective optimization problem into a single-objective scalar problem. The first term (1−η(k)) aims to minimize energy loss, driving the system to output a large current to improve efficiency; the second term… The aim is to punish temperature rise and suppress heat accumulation caused by excessive current. This is achieved through regulation. and By adjusting the ratio, the system can find the optimal balance between aggressive charging and conservative protection. For example, when a rapid power supply is needed in an emergency, the ratio can be appropriately increased. When heat dissipation conditions are poor, increase This can force the algorithm to choose a gentler current sequence.

[0036] At the same time, the solution process needs to satisfy constraints, including: Current constraint: Where t is the prediction step size, , The minimum allowable charging current for the battery. The maximum allowable charging current for the battery. , C is a current value reference equal to the rated capacity of the energy storage battery pack; Voltage constraint: ,in This is the minimum permissible terminal voltage of the battery. This is the maximum permissible terminal voltage of the battery; Temperature constraints: ,in, The battery safety temperature threshold, , t∈[1,N]; Charging time constraint: The total time to complete charging is less than the charging time limit specified in the received external charging command.

[0037] The specific analysis of the above constraints is as follows: Current constraints not only limit the upper limit To prevent excessive current from causing lithium plating inside the battery or overloading electronic components, a lower limit is also set. This ensures that the charging process remains within the effective operating range, avoiding problems such as excessively long charging time or low inverter efficiency due to insufficient current.

[0038] Voltage constraints specify the permissible range of battery terminal voltage. To prevent battery overcharging (which could lead to fire and explosion) and over-discharging (which could cause irreversible capacity decay), the model eliminates any control sequences that could cause voltage out-of-bounds errors during the prediction process.

[0039] Temperature constraints set an absolute safety threshold. (e.g., 55℃) This differs from minimizing temperature rise in the objective function, which aims for "as low as possible," while this constraint absolutely cannot exceed it. Once the predicted temperature reaches this red line, the system must forcibly reduce the current, regardless of how low the efficiency is.

[0040] The charging time constraint is to meet the business needs of the central emergency power supply system and ensure that the battery can be fully charged within the specified emergency response time (such as 2 hours required by the dispatch instruction). This requires that when the algorithm solves the optimal solution, it cannot extend the charging time indefinitely in order to excessively pursue low temperature. Instead, it must find the maximum allowable current under the pressure of the time limit.

[0041] Suppose that in an emergency charging task, the scheduling platform requires the battery to be fully charged within one hour (a tight charging time constraint), while the current battery temperature is low (25°C). In the initial stage of charging, the model predicts that the battery temperature will move away from its normal range. (55℃), and the voltage has not reached its upper limit. At this point, in order to meet the time constraint and maximize efficiency (in the objective function...) (The term plays a dominant role), and the optimization algorithm will calculate a large current sequence (close to...) ), because at this time although There will be an increase, but its impact on the total objective function value J is less than the gains from efficiency improvements, and it does not violate any hard constraints.

[0042] As charging progresses, the battery temperature gradually rises to 50°C, approaching its maximum temperature. At this point, in the objective function This value increases rapidly, and to minimize J, the algorithm automatically reduces the predicted current value. More importantly, if the high current is maintained, the model's predicted future temperature will... The temperature will exceed 55°C (violating temperature constraints). Therefore, the algorithm will be forced to output a reduced current sequence, sacrificing some charging speed in exchange for a temperature drop.

[0043] Ultimately, through this dynamic game, the system automatically generated an optimal charging curve that was fast at first and then slow, ensuring that charging was completed within 1 hour (meeting the time limit), that the temperature never exceeded 55℃ (meeting safety requirements), and that efficiency was maximized throughout the process.

[0044] Although the main control module calculates the optimal sequence for the next N cycles, it only outputs the parameters for the first control cycle to the charging module during actual execution. In the next cycle, the system re-collects the latest battery state (feedback) and predicts and solves again based on the new state. This mechanism effectively overcomes errors caused by inaccurate model parameters or environmental interference, ensuring that the charging process always stays on the optimal trajectory.

[0045] Based on the constructed objective function model and the established physical constraints, the model prediction control unit does not calculate all parameters for the entire charging cycle at once, but rather generates control commands in real time through a high-frequency iterative closed-loop process. Specifically, the complete execution process for solving the optimal charging parameter sequence based on the objective function and constraints, and based on the battery state change trend, includes the following steps: (1) State initialization: Based on the real-time parameters of the battery state acquisition unit, initialize the state variables of the battery equivalent circuit model; (2) Rolling prediction: Based on the battery equivalent circuit model, predict the battery SOC, temperature and terminal voltage change trends within the preset time window, wherein the preset time window includes N future control cycles; the value of N ranges from 5 to 10. (3) Optimization solution: With the optimization objectives of maximizing charging efficiency and minimizing battery temperature rise, an optimization objective function is constructed. Under the conditions of satisfying battery voltage constraints, current constraints, temperature constraints and charging time constraints, the optimal charging parameter sequence of the preset time window is obtained; wherein, the optimal charging parameter sequence is the optimal charging current / voltage sequence; (4) Control execution: Output the first control cycle parameter of the optimal charging parameter sequence to the charging module to control the charging process; (5) Feedback correction: In the next control cycle, new battery state parameters are collected, the state variables of the battery equivalent circuit model are corrected, and steps (2)-(4) are repeated to achieve rolling optimization and feedback correction.

[0046] In layman's terms, this process is equivalent to taking one step, observing N steps, and correcting one step at a time. Traditional charging control is often based on looking up tables (such as the SOC-current curve) in the current state, which is static and lagging. In the process of this application, the rolling prediction in step (2) can predict whether the current high current will cause the temperature to exceed the limit in the next few seconds; step (4) only executes the first value in the sequence, which is for the prudent handling of model error; and the feedback correction in step (5) uses the measured value at the next moment to eliminate the deviation between the model prediction and the actual battery state (such as the change in internal resistance caused by battery aging). This rolling optimization + feedback correction mechanism enables the system to dynamically adapt to the time-varying characteristics of the battery's internal parameters and the disturbances of the external environment, so as to always maintain the optimal control performance in complex emergency power supply scenarios.

[0047] The following example illustrates how this process works: Assuming in At a given moment, the battery temperature is 30°C, and the system predicts the optimal current sequence for the next 5 cycles to be {100A, 98A, 96A, 94A, 92A}. In step (4), the system only outputs the first value, 100A, to the charging module. After one control cycle, the current reaches... At that moment, due to a sudden increase in ambient temperature or obstruction of the cooling fan's air intake, the battery temperature measured by the sensor unexpectedly rose to 32℃ (higher than normal). The predicted temperature is 31℃. Entering the feedback correction phase of step (5), the main control module immediately detects this deviation and corrects the thermal parameters of the battery model. Then, steps (2) and (3) are re-executed, and based on the corrected model and the current temperature of 32℃, the sequence for the next 5 cycles is recalculated. At this point, to prevent further temperature runaway, the new optimal sequence may become 90A (lower than the previous planned 98A), 88A, 86A... The system then executes the new first value of 90A. In this way, the system automatically corrects the original plan and achieves adaptive adjustment to sudden thermal disturbances.

[0048] This application also introduces a forward-looking heat dissipation control mechanism. This is fundamentally different from traditional temperature control logic (i.e., the fan only turns on when the temperature exceeds a set value). The main control module uses the predictive capability of MPC to analyze the battery temperature change trend. If the prediction algorithm finds that, based on the current high charging current, the battery temperature will reach a set threshold (e.g., 45°C) within the next 5 minutes, even if the current temperature may only be 35°C, the main control module will still make a judgment and adjust the fan speed of the cooling module in advance (e.g., from 20% to 80%). This control method utilizes the thermal inertia of fan cooling, suppressing the temperature before it actually spikes by enhancing convection heat transfer, thereby achieving more stable temperature control. This allows the system to use a larger current for charging without triggering overheat protection, indirectly shortening the charging time.

[0049] Furthermore, considering that the central emergency power supply system may connect to different types of energy storage batteries (such as lithium iron phosphate batteries, ternary lithium batteries, or lead-carbon batteries), the main control module also has an adaptive identification function. When different types of battery packs are connected, the main control module will adaptively adjust the internal parameters (such as R and C values) of the battery equivalent circuit model and the weighting coefficients of the objective function based on the identified type. For example, for ternary lithium batteries with poor thermal stability, the system will automatically increase the temperature rise weight. This involves sacrificing some charging speed for higher safety; while for lithium iron phosphate batteries with better temperature resistance, the efficiency weight can be appropriately increased. .

[0050] Accordingly, this application also discloses a smart charging method for a central emergency power supply system based on model prediction, referring to... Figure 3 This includes the following steps: S1: System Initialization. After the main control module starts, it first performs a self-test on the electrical connections and communication functions of each hardware module in the system (including the charging module, battery status acquisition unit, fault detection unit, etc.). After the self-test passes, the main control module loads the battery equivalent circuit model (such as a second-order RC model) and the weight coefficients of the optimization objective function from the integrated model storage unit. and ) and constraint parameters of each physical quantity ( , (etc.). If a module failure or model loading failure is detected during the self-test, the system will display the specific fault code through the human-computer interaction module and send an alarm message to the remote dispatch platform through the integrated communication unit. At this time, the system will be locked and the charging program will not be started.

[0051] S2: Parameter Acquisition and Command Reception. The main control module collects real-time status parameters of the energy storage battery pack through its integrated battery status acquisition unit, including individual cell terminal voltage, total battery pack voltage, battery temperature, and charging circuit current. The collected raw data is input to the SOC estimation unit, which calculates the current remaining SOC using an extended Kalman filter algorithm and transmits all data to the core processing unit of the main control module. Simultaneously, the main control module collects operating parameters of the charging circuit (such as insulation resistance and contact resistance) through its integrated fault detection unit to ensure hardware safety. Furthermore, the main control module receives charging commands from the scheduling platform via the communication unit, which specify the target SOC and allowed charging time limit for this charging task.

[0052] S3: Model Prediction and Optimization Solution. The model prediction control unit within the main control module initializes the state variables of the battery equivalent circuit model based on the real-time battery state parameters acquired in step S2. Subsequently, it predicts the battery's SOC growth curve, temperature change curve, and terminal voltage change curve over the next N control cycles based on this model. Based on these prediction data, an optimization function is constructed with the objectives of maximizing charging efficiency and minimizing battery temperature rise. Under the constraints of voltage, current, temperature, and total time limit, a numerical optimization algorithm is used to solve the function, thereby obtaining the optimal charging parameter sequence (i.e., the optimal current sequence) covering the next N control cycles. In this step, the model prediction control unit also possesses adaptive adjustment capabilities. It can automatically adjust the internal parameters (R and C values) of the battery equivalent circuit model and the weighting coefficients of the optimization objective function according to the characteristics of different types of energy storage battery packs (such as lithium batteries or lead-acid batteries) to adapt to the charging requirements of various energy storage batteries.

[0053] S4: Charging Execution and Proactive Heat Dissipation Control. The main control module extracts the first control cycle parameter from the optimal charging parameter sequence and outputs it to the charging module to control the DC / DC conversion unit to adjust the output current and execute the charging operation. Simultaneously, based on predicted battery temperature trends, the main control module implements a proactive heat dissipation strategy: if the predicted data indicates that the temperature will significantly increase and approach the warning value in the near future, the main control module will proactively control the fan cooling module to increase the fan speed, using thermal inertia to intervene in advance and prevent subsequent temperature overshoot.

[0054] S5: Feedback Correction and Rolling Optimization. The system enters the next control cycle (e.g., after 100ms). At this time, step S2 is repeated to collect the latest battery state parameters. Using these measured new parameters, the state variables in the battery equivalent circuit model are corrected to eliminate the deviation between the model prediction and the actual system. Subsequently, based on the corrected state, steps S3 and S4 are re-executed. This process is continuously cyclical, realizing rolling optimization and ensuring that the control strategy is always adjusted based on the latest system state.

[0055] S6: Fault Diagnosis and Handling. During the entire charging process, if the main control module determines a hardware fault (such as overcurrent or insulation failure) based on real-time parameters from the fault monitoring module, or if it predicts that the future battery state will not meet safety constraints (e.g., regardless of current adjustments, the future temperature will exceed the limit), it immediately triggers the fuse mechanism, controlling the charging module to stop charging and issuing a fault alarm. After the fault is resolved, the process must restart from step S1.

[0056] S7: Charging Termination Judgment. In the predictive calculation of each control cycle, if it is predicted that the battery SOC will reach the target value set by the scheduling command within the next N control cycles, or the used charging time has reached the set time limit, the main control module will control the charging module to smoothly stop charging. Subsequently, the charging completion information and the final battery status data are uploaded to the scheduling platform through the communication unit; otherwise, the system will return to step S5 and continue the rolling optimization charging for the next cycle.

[0057] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0058] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A smart charging device for a central emergency power supply system based on model prediction, characterized in that, include: The charging module is used to connect to the energy storage battery pack and adjust the charging output according to the control signal; The main control module, connected to the charging module, is configured as follows: Real-time acquisition of battery status parameters of the energy storage battery pack; Based on the stored battery equivalent circuit model and the battery state parameters, the trend of battery state change within a future preset time window is predicted. Based on the objective function and constraints, the optimal charging parameter sequence is solved based on the battery state change trend. The first control cycle parameter in the optimal charging parameter sequence is output as the control signal to the charging module to execute the charging strategy.

2. The intelligent charging device for the central emergency power supply system according to claim 1, characterized in that, The charging module includes several independent charging branch units; the charging branch units are used to respond to the control signals of the main control module and adjust the output voltage and output current. The intelligent charging device also includes: An expansion interface module is provided, through which the several independent charging branch units are connected to the main control module.

3. The intelligent charging device for the central emergency power supply system according to claim 1, characterized in that, The main control module also includes: The fault detection unit is used to collect the battery status parameters of the charging module, identify overcurrent, overvoltage, undervoltage, short circuit and insulation degradation fault types, and transmit a signal to the main control module to stop charging when a fault is detected. The communication unit is used to establish two-way communication with the external scheduling platform, receive external charging instructions containing the target remaining power and charging time limit, and upload charging status information.

4. The intelligent charging device for the central emergency power supply system according to claim 1, characterized in that, The main control module integrates a battery status acquisition unit, which includes a voltage sensor, a temperature sensor, a current sensor, and a SOC estimation unit. The battery status parameters include at least the individual cell terminal voltage, the battery pack terminal voltage, the battery temperature, the charging current, and the remaining charge (SOC). The SOC estimation unit is configured to perform calculations based on the extended Kalman filter algorithm, combined with the collected battery cell terminal voltage, battery temperature and charging current, to obtain the remaining power SOC.

5. The intelligent charging device for the central emergency power supply system according to claim 1, characterized in that, It also includes a fan cooling module, which is connected to the main control module; the main control module is further configured to perform forward-looking heat dissipation control: based on the predicted battery temperature change trend in the battery state change trend, it determines whether the future temperature will reach a set threshold; if the determination result is yes, then the fan speed of the fan cooling module is adjusted in advance before the battery temperature actually rises.

6. The intelligent charging device for the central emergency power supply system according to claim 1, characterized in that, The main control module is also configured to identify the type of the connected energy storage battery pack and adaptively adjust the parameters of the battery equivalent circuit model and the weight coefficients of the optimization objective function according to the type.

7. The intelligent charging device for the central emergency power supply system according to claim 1, characterized in that, The battery equivalent circuit model adopts a second-order RC equivalent circuit model; the optimization objective function is configured to maximize charging efficiency and minimize battery temperature rise. The method for solving the optimal charging parameter sequence based on the battery state change trend, based on the optimization objective function and constraints, includes the following steps: (1) State initialization: Based on the real-time parameters of the battery state acquisition unit, initialize the state variables of the battery equivalent circuit model; (2) Rolling prediction: Based on the battery equivalent circuit model, predict the battery SOC, temperature and terminal voltage change trends within the preset time window, wherein the preset time window includes N future control cycles; the value of N ranges from 5 to 10. (3) Optimization solution: With the optimization objectives of maximizing charging efficiency and minimizing battery temperature rise, an optimization objective function is constructed. Under the conditions of satisfying battery voltage constraints, current constraints, temperature constraints and charging time constraints, the optimal charging parameter sequence of the preset time window is obtained; wherein, the optimal charging parameter sequence is the optimal charging current / voltage sequence; (4) Control execution: Output the first control cycle parameter of the optimal charging parameter sequence to the charging module to control the charging process; (5) Feedback correction: In the next control cycle, new battery state parameters are collected, the state variables of the battery equivalent circuit model are corrected, and steps (2)-(4) are repeated to achieve rolling optimization and feedback correction.

8. A smart charging method for a central emergency power supply system based on model prediction, characterized in that, The intelligent charging device for a central emergency power supply system based on model prediction as described in any one of claims 1-7 includes the following steps: Step 1: Real-time acquisition of battery status parameters of the energy storage battery pack; Step 2: Based on the pre-stored battery equivalent circuit model and the collected battery state parameters, predict the battery state change trend within a preset time window in the future. Step 3: Combining the objective function and constraints, solve for the optimal charging parameter sequence based on the battery state change trend; Step 4: Output the first control cycle parameter in the optimal charging parameter sequence to the charging module to control the charging module to perform the charging operation; Step 5: Enter the next control cycle and repeat steps one to four to achieve rolling optimization control.

Images