Charging control method and device, battery management system and nonvolatile storage medium
By intelligently adjusting the charging current and optimizing the charging strategy according to the battery status, the problems of low energy utilization and large battery loss caused by the volatility of traditional charging strategies are solved, extending the service life of the battery pack and improving the stability of the charging process.
Patent Information
- Application Number
- CN202411025021.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-29
- Publication Date
- 2024-11-12
AI Technical Summary
Traditional battery charging strategies have high volatility, resulting in low energy utilization and large battery losses, affecting battery performance and service life.
By determining the battery terminal voltage, temperature and health status of the battery pack, the charging action with the largest cumulative reward value in the action space is performed as the target charging action. The intelligent body is trained based on the empirical sample data set, optimize the charging process, and intelligently adjust the charging current to Optimize energy utilization efficiency and reduce losses.
Effectively control the battery charging process, improve energy utilization efficiency, reduce loss during the system charging process, extend the service life of the battery pack, and improve the stability and reliability of the charging process.
Smart Images

Figure CN118944241A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of battery management technology, and in particular to a charging control method, device, battery management system and non-volatile storage medium. Background Art
[0002] Charging control technology is crucial to meeting the demands for charger functionality and safety in modern life. However, traditional battery charging strategies in related technologies are highly volatile, resulting in low energy utilization during battery charging and high battery losses, which may affect the performance and service life of backup energy equipment.
[0003] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention
[0004] The embodiments of the present application provide a charging control method, device, battery management system and non-volatile storage medium to at least solve the technical problems in the related art that the traditional charging strategy has large volatility, resulting in low energy utilization and large battery loss during the battery charging process.
[0005] According to one aspect of an embodiment of the present application, a charging control method is provided, including: determining a battery state of a battery pack at a first time step, wherein the battery state includes: battery terminal voltage, temperature and health status of the battery pack; determining a cumulative reward value corresponding to executing each charging action in an action space under the battery state at the first time step, wherein the cumulative reward value is used to characterize the cumulative value of the expected return that can be obtained by taking a charging action under a given battery state; determining the charging action with the largest cumulative reward value in the action space as the target charging action corresponding to the first time step; after executing the target charging action, determining the battery state of the battery pack at a second time step, repeating the steps of determining and executing the target charging action until the battery pack is fully charged, wherein the second time step is the next time step immediately adjacent to the first time step.
[0006] Optionally, the charging action includes: adjusting the charging voltage or current; the method also includes: obtaining an experience sample data set, wherein the experience sample data set includes multiple experience samples obtained from the actual charging process of the vehicle battery, each experience sample corresponds to a time step, and the experience sample includes: the battery state corresponding to the first time step, the charging action performed in the first time step, the battery state corresponding to the second time step, and the immediate reward obtained after the charging action is performed in the first time step; based on the experience sample data set, the intelligent agent is trained, wherein the trained intelligent agent is used to determine the cumulative reward value corresponding to performing the charging action under a given battery state.
[0007] Optionally, training the intelligent agent includes: determining the battery state corresponding to the first time step and the feature vector corresponding to the charging action, as well as the parameter vector corresponding to the intelligent agent at the first time step; during the training process, using the intelligent agent, based on the feature vector and parameter vector corresponding to the first time step, determining the predicted value of the cumulative reward after performing the charging action; based on the empirical sample data set, calculating the true value of the cumulative reward after performing the charging action in the first time step, and determining the mean square error between the predicted value of the cumulative reward and the true value of the cumulative reward; using the gradient descent algorithm, based on the mean square error, updating the parameter vector corresponding to the intelligent agent, and using the intelligent agent after the parameter vector is updated, repeating the above training process until the mean square error is no greater than a preset error threshold.
[0008] Optionally, the target charging action includes: a random action and a greedy action, wherein the random action is any random charging action in the action space, and the greedy action is the charging action with the largest cumulative reward value in the action space; the method also includes: at a first time step, according to a preset probability, determining the random action or the greedy action as the target charging action and executing it, wherein the preset probability includes: a first probability and a second probability, the random action is selected as the target charging action with the first probability, and the greedy action is selected as the target charging action with the second probability, and the sum of the first probability and the second probability is one.
[0009] Optionally, the method also includes: determining the battery state of the battery pack at the second time step after the target charging action is performed under the battery state at the first time step, and the value of the immediate reward obtained by performing the target charging action; adding the battery state and target charging action corresponding to the first time step, the battery state corresponding to the second time step, and the value of the immediate reward obtained by performing the target charging action as a new experience sample to the experience sample data set, and determining the time step corresponding to the experience sample; when the time step corresponding to the experience sample is an integer multiple of the preset time step, updating the parameter vector of the intelligent agent according to the updated experience sample data set.
[0010] Optionally, the battery pack is composed of multiple batteries interconnected in series and parallel, and each battery is connected by a nickel-plated copper strip. The battery pack also includes a battery management system, wherein the battery management system enters a sleep mode when receiving a sleep command. In the sleep mode, the power consumption of the battery management system is less than a preset power consumption threshold.
[0011] Optionally, during the charging process of the battery pack, the method also includes: when the voltage difference between the highest voltage and the lowest voltage in each battery of the battery pack exceeds a preset voltage difference, inputting a balancing current into the battery pack, wherein the balancing current is used to balance the voltage between each battery in the battery pack; during the charging process, after each preset time period, terminating the input of the balancing current and the charging current, and measuring the voltage of each battery in the battery pack; after the voltage measurement is completed, resuming the charging current, and re-determining whether it is necessary to continue to input the balancing current based on the measured voltage.
[0012] According to another aspect of an embodiment of the present application, a charging control device is also provided, including: a battery status determination module, used to determine the battery status of a battery pack at a first time step, wherein the battery status includes: battery terminal voltage, temperature and health status of the battery pack; a cumulative reward determination module, used to determine the cumulative reward value corresponding to executing each charging action in the action space under the battery status at the first time step, wherein the cumulative reward value is used to characterize the cumulative value of the expected return that can be obtained by taking a charging action under a given battery status; a charging action determination module, used to determine the charging action with the largest cumulative reward value in the action space as the target charging action corresponding to the first time step; a loop control module, used to determine the battery status of the battery pack at a second time step after executing the target charging action, and repeat the steps of determining and executing the target charging action until the battery pack is fully charged, wherein the second time step is the next time step adjacent to the first time step.
[0013] According to another aspect of the embodiment of the present application, a battery management system is provided, including: a memory and a processor, the processor being used to run a program stored in the memory, wherein the charging control method is executed when the program is run.
[0014] According to another aspect of the embodiments of the present application, a non-volatile storage medium is provided, the non-volatile storage medium includes a stored computer program, wherein the device where the non-volatile storage medium is located executes the charging control method by running the computer program.
[0015] In an embodiment of the present application, a battery state of a battery pack at a first time step is determined, wherein the battery state includes: battery terminal voltage, temperature and health state of the battery pack; a cumulative reward value corresponding to executing each charging action in the action space under the battery state at the first time step is determined, wherein the cumulative reward value is used to characterize the cumulative value of the expected return that can be obtained by taking a charging action under a given battery state; the charging action with the largest cumulative reward value in the action space is determined as the target charging action corresponding to the first time step; after executing the target charging action, the battery state of the battery pack at a second time step is determined, and the steps of determining and executing the target charging action are repeated until the battery pack is charged, wherein the second time step is the next time step immediately adjacent to the first time step, and the charging current is intelligently adjusted according to parameters such as battery temperature, health state and terminal voltage, so as to achieve the purpose of effectively controlling the charging process of the battery, optimizing energy utilization efficiency, and reducing the loss of the system during the charging process, thereby solving the technical problems of low energy utilization and large battery loss caused by the large volatility of traditional charging strategies in related technologies. BRIEF DESCRIPTION OF THE DRAWINGS
[0016] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:
[0017] Figure 1 It is a hardware structure block diagram of a computer terminal (or electronic device) for implementing a charging control method provided in an embodiment of the present application;
[0018] Figure 2 is a schematic diagram of a charging control method flow according to an embodiment of the present application;
[0019] Figure 3 It is a schematic diagram of a design process of a backup energy adaptive slow charging balancing strategy provided in an embodiment of the present application;
[0020] Figure 4 is a schematic diagram of a battery testing process provided according to an embodiment of the present application;
[0021] Figure 5 It is a structural schematic diagram of a charging control device provided according to an embodiment of the present application. DETAILED DESCRIPTION
[0022] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.
[0023] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.
[0024] In order to improve the performance and life of batteries, the performance and life of batteries can be improved by developing higher-performance battery materials, optimizing battery design and control systems, and implementing intelligent charging management. At the same time, for the real-time challenges of battery packs under different discharge rates, operating temperatures, and road conditions, it is possible to consider introducing an intelligent battery management system to monitor the battery status in real time and adjust the operating parameters to maximize the protection of the battery and extend its life. This application proposes an adaptive slow charging equalization strategy technology that can dynamically adjust the charging rate by real-time monitoring of factors such as system energy status, environmental conditions, and user needs, so that the system energy is in a balanced state. This strategy not only helps to optimize energy utilization efficiency, but also effectively reduces the loss during system charging and extends the service life of the backup energy system. At the same time, the adaptive slow charging equalization strategy can also improve the stability of the system during charging, reduce system fluctuations, and enhance the reliability of the system. In general, the design of intelligent charging control technology is crucial to meet the needs of charger functionality and safety in modern life, and is expected to provide people with a more convenient and safer charging experience, while extending the service life of the device and improving the overall user experience.
[0025] The following is a detailed description of the relevant solutions in the embodiments of the present application for solving technical problems such as low energy utilization and large battery loss during battery charging due to the large volatility of traditional charging strategies in related technologies.
[0026] According to an embodiment of the present application, a method embodiment of charging control is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.
[0027] The method embodiments provided in the embodiments of the present application can be executed in a mobile terminal, a computer terminal or a similar computing device. Figure 1 FIG. 1 shows a hardware structure block diagram of a computer terminal (or electronic device) for implementing a charging control method. Figure 1 As shown, the computer terminal 10 (or electronic device) may include one or more (102a, 102b, ..., 102n are used to illustrate) processors 102 (the processor 102 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA), a memory 104 for storing data, and a transmission device 106 for communication functions. In addition, it may also include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of the BUS bus), a network interface, a power supply and / or a camera. It can be understood by those skilled in the art that Figure 1 The structure shown is only for illustration and does not limit the structure of the above electronic device. Figure 1 More or fewer components as shown, or with Figure 1 Different configurations shown.
[0028] It should be noted that the one or more processors 102 and / or other data processing circuits described above may generally be referred to herein as "data processing circuits". The data processing circuits may be embodied in whole or in part as software, hardware, firmware, or any other combination thereof. In addition, the data processing circuit may be a single independent processing module, or may be incorporated in whole or in part into any of the other components in the computer terminal 10 (or electronic device). As described in the embodiments of the present application, the data processing circuit acts as a processor control (e.g., selection of a variable resistor terminal path connected to an interface).
[0029] The memory 104 can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the charging control method in the embodiment of the present application. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, that is, the above-mentioned charging control method is realized. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include a memory remotely arranged relative to the processor 102, and these remote memories may be connected to the computer terminal 10 via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.
[0030] The transmission device 106 is used to receive or send data via a network. The specific example of the above network may include a wireless network provided by a communication provider of the computer terminal 10. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 can be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.
[0031] The display may be, for example, a touch screen liquid crystal display (LCD) that enables a user to interact with a user interface of the computer terminal 10 (or electronic device).
[0032] Under the above operating environment, the embodiment of the present application provides a charging control method. Figure 2 is a schematic diagram of a charging control method flow according to an embodiment of the present application, such as Figure 2 As shown, the method comprises the following steps:
[0033] Step S202, determining a battery state of the battery pack at a first time step, wherein the battery state includes: battery terminal voltage, temperature and health state of the battery pack;
[0034] Step S204, determining the cumulative reward value corresponding to executing each charging action in the action space under the battery state at the first time step, wherein the cumulative reward value is used to represent the cumulative value of the expected return that can be obtained by taking the charging action under the given battery state;
[0035] Step S206, determining the charging action with the largest cumulative reward value in the action space as the target charging action corresponding to the first time step;
[0036] Step S208, after executing the target charging action, determine the battery state of the battery pack at the second time step, and repeat the steps of determining and executing the target charging action until the battery pack is fully charged, wherein the second time step is the next time step immediately adjacent to the first time step.
[0037] Through the above steps, by intelligently adjusting the charging current according to parameters such as battery temperature, health status and terminal voltage, the purpose of effectively controlling the battery charging process, optimizing energy utilization efficiency and reducing system losses during charging is achieved, thereby solving the technical problem of low energy utilization and large battery losses during battery charging caused by large fluctuations in traditional charging strategies in related technologies.
[0038] The following further introduces the charging control method in steps S202 to S208 of the embodiment of the present application.
[0039] The embodiment of the present application can be used for a power station backup power battery pack in an environment of a real station scene, such as Figure 3 As shown, in the face of problems that may affect battery performance and life, the embodiment of the present application can optimize performance and parameters in the early stage of the battery design phase, and at the same time propose an adaptive slow charging strategy that intelligently adjusts the charging current according to parameters such as battery temperature, health status and terminal voltage to effectively control the battery charging process. In addition, a new balancing technology is introduced, which involves temporarily suspending balancing and charging current during voltage measurement, thereby reducing errors related to ohmic drop and improving the performance and life of the battery pack. Each link is described in detail below.
[0040] In the embodiment of the present application, the battery pack is described by taking the 2.86kWh battery pack of the station backup power supply as an example. The battery pack is constructed using prismatic lithium-ion batteries composed of NMC (Nickel Manganese Cobalt Oxide) + LMO (Lithium Manganese Oxide) chemical composition, with a rated nominal voltage of 3.63V and a capacity of 14Ah. In some embodiments of the present application, the battery pack is composed of a plurality of batteries interconnected in series and parallel, and each battery is connected by a nickel-plated copper strip. The battery pack also includes a battery management system, wherein the battery management system enters a sleep mode when receiving a sleep command, and in the sleep mode, the power consumption of the battery management system is less than a preset power consumption threshold.
[0041] Specifically, four cells can be connected in parallel by electric welding to form a 4P (Parallel) module, and then two such 4P modules can be connected using aluminum bars, the modules are tightened, and then the battery pack is assembled from 14 such module batteries, and the modules are connected in series to obtain the required voltage. These series and parallel interconnections between batteries can be established by various technologies such as spot welding and laser welding. The Battery Management System (BMS) is installed on the battery pack to perform all necessary functions such as protection, charge control, data logging and estimation. The test process of this battery pack is as follows Figure 4 As shown, the battery pack can be charged at the replacement station. The battery pack can also be taken to the laboratory for regular measurement of SOH (State of Health).
[0042] The embodiment of the present application performs an initial optimization of the battery pack during the battery pack design stage, which is described in detail below.
[0043] As for the resistance inside the battery pack, since the battery pack is composed of multiple batteries connected in series and parallel. Taking the 2.86kwh battery pack as an example, the device has a 4P (Parallel) 14S (Series) configuration. The series and parallel interconnections between batteries can be established by various techniques such as spot welding and laser welding, but due to the resistance of the aluminum strip and the contact resistance between the aluminum strip and the busbar, the voltage difference between the modules will be large under high current. In this embodiment, the problem caused by the aluminum strip can be solved by replacing the aluminum strip with a continuous busbar welded on the battery. However, since some modules are still connected by wires, additional resistance is added, which may cause further problems. Therefore, in this embodiment, nickel-plated copper strips are finally used to replace these wires to achieve series and parallel connection between batteries, which helps to balance and reduce the resistance at the module interconnection.
[0044] Regarding the current consumption of the BMS, the BMS is responsible for performing various functions such as safety, protection, and evaluation to provide power for the battery pack. For the battery pack, the power consumption of the BMS during normal use is about 500-800 mW. Although this part of the consumption seems to be only a small part of the total capacity of the battery, it becomes critical when the battery is close to fully discharged. When the discharged battery is in the low SOC (State of Charge) area, the energy is minimal. When the battery pack in this state remains idle for more than a few days with an active BMS, the small power consumption of the BMS is enough to exhaust the battery and put it into a deep discharge state, resulting in rapid degradation of the battery. To solve this problem, this embodiment implements a sleep mode in the BMS, in which the BMS only activates basic functions, reducing the power consumption of the BMS to tens of milliwatts. In order to further reduce consumption, an external manual switch can also be introduced, which can be used to completely disconnect the battery from the BMS when the vehicle is parked for a long time. This ensures the proper open circuit condition of the battery and eliminates the problem of BMS power consumption.
[0045] In addition, the embodiment of the present application also improves the connector of the battery pack. Specifically, a battery pack is composed of different connectors for different purposes. Due to the inevitable bumps and vibrations during transportation, these connectors sometimes become loose due to quality or curling problems, resulting in poor contact or incorrect readings. To solve this problem, the connector used in this embodiment is replaced with a higher quality alternative. In addition, the connection between the connector and the wire is filled with silicone to strengthen the connection.
[0046] For the above-mentioned improved battery pack, in the related art, for lithium-ion batteries, the traditional charging method is CC-CV (Constant Current-Constant Voltage) charging, which includes two steps: CC step and CV step. In CC mode, the battery is charged with a constant current until a predefined voltage upper limit. Once the battery voltage reaches this level, the charger switches to CV mode, in which the voltage remains constant and the current is allowed to drop. When this current drops below a certain cutoff current, the charging process is called complete. The lower the cutoff current, the longer it takes to fully charge. The battery charging steps of this method are as follows.
[0047] 1) Soak the battery pack at 25℃ for 2 hours and discharge the battery pack to the minimum voltage of 42V for testing. 2) Charge the battery pack using the CC-CV profile. In constant current mode, the current of C / 3 can reach up to 58.8V. When the current reaches C / 10, the constant voltage mode is terminated. 3) Rest for 30 minutes. 4) Discharge the battery with a constant current of C / 3 (i.e., a current of one-third of the rated capacity C of the battery) until the low voltage cutoff of 42V. 5) Rest for 30 minutes. 6) Repeat steps 2)-5) twice.
[0048] Based on the traditional charging algorithm, the embodiment of the present application takes voltage, temperature and health status as key parameters to be considered in implementing the adaptive charging strategy, which is described in detail below.
[0049] Lithium-ion batteries undergo complex chemical reactions during the charging process. These reactions are voltage-dependent, and controlling the voltage allows the charging algorithm to regulate the progress of these reactions, thereby optimizing performance. High voltage during the charging process causes metallic lithium to form on the anode (a process called lithium plating), which reduces the capacity of the battery. This is because under high voltage conditions, lithium ions are irreversibly reduced to metallic lithium, which deposits and forms a solid on the anode and can no longer participate in the normal charging and discharging process of the battery, thereby reducing the available capacity of the battery. The embodiments of the present application can reduce energy waste and minimize heat generation by controlling the voltage and optimizing the energy transfer of the charging algorithm. Proper voltage regulation helps to reduce these harmful effects and keep the battery healthy.
[0050] Temperature is another key parameter that affects the long-term health and performance of lithium-ion batteries. Temperature affects the rate at which lithium-ion batteries can be safely charged. At lower temperatures, the internal resistance of the battery increases, limiting its ability to absorb energy quickly. Low temperature charging can also cause lithium plating on the anode, causing permanent damage to the battery. Conversely, at higher temperatures, the battery can accept charging more quickly, but there is a risk of overheating and thermal runaway. The charging strategy developed in the embodiments of the present application helps to minimize the stress on the electrodes and electrolyte by keeping the battery within a specific temperature range during the charging process, thereby maintaining the service life of the battery.
[0051] The next key factor is the battery's SOH. As a battery approaches the end of its useful life, its capacity declines and its ability to deliver power decreases due to increased internal resistance. Therefore, in order to reduce ohmic heating, a battery near the end of its life will require a low current charge to avoid accelerated degradation, while newer batteries can withstand more aggressive charging without compromising their health. Additionally, degraded batteries may exhibit unpredictable behavior during charging, including increased risk of thermal runaway or overcharging. Battery capacity as a parameter to estimate degradation: Among them, the capacity Q p and Qo The capacity is obtained by the charge and discharge rate of C / 3 and 25℃. o Indicates the original capacity of the battery, that is, the amount of power that the battery can provide in its new state. p Indicates the current capacity of the battery, that is, the amount of electricity the battery can provide under specific test conditions.
[0052] In the embodiment of the present application, the charging current or voltage can be intelligently adjusted according to parameters such as battery temperature, health status and terminal voltage to control the charging rate Charge rate = f(V, T, SOH), where V is the terminal voltage of the highest voltage battery, T is the highest battery temperature recorded by the thermistor, SOH is the health status of the battery, and f() represents the charging control strategy in the present application. The charging strategy proposed in the present application can adapt to the SOH of the battery and ensure safe and reliable charging.
[0053] The charging strategy in this application is further introduced below.
[0054] In the embodiment of the present application, the charging control strategy is initialized by using a Q learning framework based on experience playback, and the prediction result of the LSTM (Long Short-Term Memory)-attention-A model is used as one of the input features. In addition, the Markov (MDP) decision process is used to initialize the parameters of the charging control strategy MDP=<S,A,γ,P,R> , where S represents the state space, A represents the action space, γ represents the discount factor, P represents the state transition probability, and R represents the deterministic reward for each state-action pair.
[0055] First, the LSTM (Long Short-Term Memory)-attention-A model is used as the intelligent agent, and the experience sample database is used to train the intelligent agent. The specific steps are as follows.
[0056] In some embodiments of the present application, the charging action includes: adjusting the charging voltage or current; the method also includes the following steps: obtaining an experience sample data set, wherein the experience sample data set includes multiple experience samples obtained from the actual charging process of the vehicle battery, each experience sample corresponds to a time step, and the experience sample includes: the battery state corresponding to the first time step, the charging action performed in the first time step, the battery state corresponding to the second time step, and the immediate reward obtained after the charging action is performed in the first time step; based on the experience sample data set, the intelligent agent is trained, wherein the trained intelligent agent is used to determine the cumulative reward value corresponding to performing the charging action under a given battery state.
[0057] Specifically, the agent learns from the actual vehicle charging state and uses it as an experience sample (s t ,a t ,s t+1 ,r t+1 ) is stored in the empirical sample data set D and is expressed as:
[0058] D={l,t,seq,s t ,a t ,s t+1 ,r t+1}
[0059] Among them, l represents the round number, t represents the time step, seq represents the sequence number of the experience learning round, and s t represents the state at time t (i.e., the first time step above), a t represents the charging action performed at time t, s t+1 Indicates that charging action a is taken t Then it is transferred to the state at time t+1 (i.e. the second time step mentioned above), r t+1 Indicates that charging action a is taken t Instant rewards after.
[0060] The training process of the intelligent agent is as follows.
[0061] In some embodiments of the present application, training the intelligent agent includes the following steps: determining the battery state corresponding to the first time step and the feature vector corresponding to the charging action, as well as the parameter vector corresponding to the intelligent agent at the first time step; during the training process, using the intelligent agent, based on the feature vector and parameter vector corresponding to the first time step, determining the predicted value of the cumulative reward after performing the charging action; based on the empirical sample data set, calculating the true value of the cumulative reward after performing the charging action at the first time step, and determining the mean square error between the predicted value of the cumulative reward and the true value of the cumulative reward; using the gradient descent algorithm, based on the mean square error, updating the parameter vector corresponding to the intelligent agent, and using the intelligent agent after the updated parameter vector, repeating the above training process until the mean square error is no greater than a preset error threshold.
[0062] Specifically, a linear function approximator is used to approximate the Q-value function and evaluate the effect of taking a charging action a under a specific state. t The value of the cumulative reward is as follows:
[0063]
[0064] in, is a state-action pair (s t , a t ), that is, the characteristic vector corresponding to the battery state and charging action corresponding to the first time step, θt is the parameter vector obtained from learning experience at time t, that is, the parameter vector corresponding to the agent in the first time step, It means that at time t, the parameter vector θ t and the eigenvector The calculated value estimate is the cumulative reward prediction value mentioned above, γ n represents the cumulative effect of the discount factor on future rewards, r t+n+1 Represents the immediate reward at time t+n+1.
[0065] Construct the parameter vector θ at time t (the first time step) t The target equation is to perform charging action a t The expected cumulative reward is expressed as Q value. By minimizing the mean square error between the predicted Q value and the true Q value, the estimated Q value is closer to the true value. The formula used is as follows:
[0066] J=∑p(s,a)(Q h (s,a)-Q t (s,a)) 2
[0067] Among them, p(s,a) is the proportion of state-action pairs (s,a) in all state-action pairs, Q h (s, a) represents the true value of any state-action pair (s, a) under strategy h, that is, the true value of the cumulative reward, Q t (s,a) represents the estimated Q value of any state-action pair (s,a) at time step t, that is, the cumulative reward prediction value. In the continuous state space, the gradient descent algorithm is used to solve the parameter vector, that is, the parameter vector corresponding to the agent is updated according to the mean square error. The specific formula is as follows:
[0068]
[0069] Among them, α represents the learning rate, represents the parameter vector θ t is the transpose of , r is the immediate reward value, γ is the discount factor, and φ is the feature vector of the state-action pair.
[0070] After the mean square error is optimal, the optimal strategy can be obtained through one-step mapping: That is, among all possible charging actions a in the action space, find the charging action that maximizes the cumulative reward value under a given battery state s, where θ * is the optimal parameter vector obtained through training.
[0071] After the agent determines the charging action with the largest cumulative reward value corresponding to the first time step, the ε-greedy strategy can be used to execute the target charging action. The specific steps are as follows.
[0072] In some embodiments of the present application, the target charging action includes: a random action and a greedy action, wherein the random action is any random charging action in the action space, and the greedy action is the charging action with the largest cumulative reward value in the action space; the method also includes the following steps: at a first time step, according to a preset probability, the random action or the greedy action is determined as the target charging action and executed, wherein the preset probability includes: a first probability and a second probability, the random action is selected as the target charging action with the first probability, and the greedy action is selected as the target charging action with the second probability, and the sum of the first probability and the second probability is one.
[0073] Specifically, for each time step t, according to the charging demand predicted by the LSTM-attention-A model in the above steps, the appropriate charging action a is selected in combination with the ε-greedy strategy. t , that is, the agent selects a random action in the action space A with a probability of ε (i.e., the first probability mentioned above), and selects the greedy action a with the largest Q value with a probability of 1-ε (i.e., the second probability mentioned above). * To maximize the cumulative reward, the formula used is as follows:
[0074]
[0075] Among them, ε is the exploration factor.
[0076] In addition, after executing the target charging action corresponding to each time step, the data corresponding to that time will be added to the data set as a new experience sample. When the number of new samples in the experience sample data set reaches a certain level, the parameter vector of the agent can be updated. The specific steps are as follows.
[0077] In some embodiments of the present application, the method also includes the following steps: determining the battery state of the battery pack at the second time step after performing the target charging action under the battery state at the first time step, and the value of the immediate reward obtained by performing the target charging action; adding the battery state and target charging action corresponding to the first time step, the battery state corresponding to the second time step, and the value of the immediate reward obtained by performing the target charging action as a new experience sample to the experience sample data set, and determining the time step corresponding to the experience sample; when the time step corresponding to the experience sample is an integer multiple of the preset time step, updating the parameter vector of the intelligent agent based on the updated experience sample data set.
[0078] Specifically, the action performs the charging action a at the current time step t, get the next state s t+1 and instant rewards t+1 , calculate the sample sequence number seq, the formula used is as follows:
[0079] seq = nl + t
[0080] n is a constant used to distinguish different rounds, l represents the round number, and t is the time step.
[0081] Update the data set D and get D = D∪{l,t,seq,s t ,a t ,s t+1 ,r t+1}, learning parameter vector θ t , if the current time step t is an integer multiple of the maximum time step T, then the parameter vector θ is adjusted according to the online learning algorithm t To update, follow these steps:
[0082] Read samples {l, t, seq from the dataset D using uniform distribution ′ ,s t ,a t ,s t+1 ,r t+1}, with the same state s t and charging action a t All samples of s are taken out to make the next state s t+1 The greedy action a with the maximum Q value * , the formula used is as follows:
[0083] a * = argmax a Q(s t+1 ,a)
[0084] Generate state-action pairs (s t+1 ,a) The corresponding eigenvector Use the formula Update the parameter θ t , repeat the steps to output the value function parameter vector θ t , update the current state s t+1 Optimal strategy generation, when the parameter vector θ t When the value no longer changes, that is, the value function Q(s, a) reaches a balance in weight for a specific state-action pair, the value function converges to the local optimal solution, and the optimal control strategy for wireless charging is generated.
[0085] Through tests under actual conditions, it is proved that the charging control strategy of the embodiment of the present application is effective in extending the life of the battery pack and improving the performance, solves the real-time challenges faced by using the battery pack under external conditions, and can maximize the performance and life of the power station backup power battery pack. The following is an example of a case in which the charging control strategy in the embodiment of the present application is used for adjustment during the charging process.
[0086] The potential difference of the battery electrodes plays a very important role in determining the battery degradation. When the voltage drops below 2.5V, the battery used to make the battery pack presents a deep discharge state. In addition, the voltage below 3V is called the very low voltage area. Charging at a large current below 3V will cause permanent capacity loss. To alleviate this situation, the charging current is kept at a very low level when the battery voltage drops below 3V. When the battery voltage is below 3V and 2.5V, respectively, currents of 0.04C and 0.01C are used. With these low currents, the battery is first recovered from the low voltage area, and then a charging current corresponding to CC is applied to the voltage above 3V. The energy content of the battery is minimal below 3V; therefore, a low current is sufficient to increase the battery voltage.
[0087] Li-ion cells exhibit a considerable variation in internal resistance with respect to voltage, with resistance increasing for voltages below 3.4V or 5% SOC. This high resistance results in higher ohmic losses and thus higher cell temperatures. To mitigate the temperature increase, current is limited to 0.2C for voltages < 3.5V. Additionally, current varies with increasing cell / ambient temperature. For temperatures above 35C, current is reduced to 0.1C during the CC phase.
[0088] The voltage range of 3.5V to 4V is the area where the battery lasts the longest, and we specify the maximum current in this area. When the temperature is above 35 degrees Fahrenheit, the current gradually decreases. When the battery approaches the end of the CC phase, the battery temperature will be high. Therefore, when the voltage is above 4v, the charging current is reduced again until the CV phase voltage is reached. In CV mode, charging is terminated when the current gradually decreases to 0.1C.
[0089] On the other hand, a new balancing technology is also introduced in the embodiment of the present application, which can temporarily suspend the balancing and charging current during the voltage measurement, thereby reducing the error associated with the ohmic drop and improving the performance and life of the battery pack.
[0090] Cell balancing is the process of equalizing the charge levels of individual cells within a battery pack. Because the capacity and performance characteristics of batteries can vary slightly, some cells may charge or discharge more than others over time during charge and discharge cycles. This imbalance can reduce the overall capacity of the battery, efficiency, and safety issues. The 2.86kWh battery pack uses a passive cell balancing technique that dissipates excess energy through resistors. The BMS uses a top-level balancing method where balancing only occurs during charging, not during discharge. The commercial BMS chip used does not support simultaneous balancing of alternating cells within a register (an 8-bit register used to trigger balancing). The chip can be used for 16 cells in series and has two sets of balancing registers, A and B; each register can balance 8 cells, from C0 to C7 and C8 to C15.
[0091] During the charging process, when the balancing mechanism is enabled, the embodiments of the present application propose solutions for the following two balancing problems.
[0092] The first problem is the problem of over-balancing. The trigger voltage of balancing is initially set to 10mV, which means that the balancing process will start when the voltage difference between the battery with the highest voltage and the battery with the lowest voltage exceeds 10mV. This causes the battery pack to remain in a constant balance state. Therefore, this embodiment adopts the method of raising the threshold to 50mV (i.e., the above-mentioned preset voltage difference) to solve this problem.
[0093] The second problem involves small voltage fluctuations in the cell being balanced. The voltage measurement is obtained during the charging phase, when both the charging current and the balancing current are flowing. These currents have the potential to produce an ohmic drop that can lead to inaccurate voltage measurements. Therefore, the present embodiment proposes a balancing algorithm that aims to alleviate the voltage fluctuation problem by deactivating the balancing current and the charging current during the voltage acquisition process. In addition, the voltage reading obtained at zero current can be very close to the actual cell voltage. The algorithm is programmed to run in a 5-minute cycle. During the first 4 minutes, the balancing mechanism is enabled. At the beginning of the 5th minute, both the balancing current and the charging current are turned off. After a rest of 10 seconds, the voltage of all cells is read, the voltage loss is calculated, and then used as a reference to trigger balancing. Balancing is turned on again, and the cycle is repeated to significantly improve the performance. The specific steps are as follows.
[0094] In some embodiments of the present application, during the charging process of the battery pack, the method also includes the following steps: when the voltage difference between the highest voltage and the lowest voltage in each battery in the battery pack exceeds a preset voltage difference, inputting a balancing current into the battery pack, wherein the balancing current is used to balance the voltage between each battery in the battery pack; during the charging process, after each preset time period, terminating the input of the balancing current and the charging current, and measuring the voltage of each battery in the battery pack; after the voltage measurement is completed, resuming the charging current, and re-determining whether it is necessary to continue to input the balancing current based on the measured voltage.
[0095] Specifically, first, the importance of setting an appropriate balancing threshold voltage to ensure that the battery pack enters balancing mode only when needed. If the threshold is set too high, the battery pack may be constantly out of balance. Second, recognize the importance of following proper measurement procedures during the balancing process. Even a small resistance can cause a large voltage drop, resulting in incorrect measurements. Therefore, not only should additional resistance not be introduced, but measurements should also be taken without current flowing to obtain accurate results.
[0096] This application scheme optimizes various parts of the battery pack. The battery pack uses nickel-plated copper strips instead of wires, which helps to balance and reduce the resistance at the module interconnection; a sleep mode is implemented in the BMS, in which only basic functions are activated, reducing the power consumption of the BMS to tens of milliwatts. In order to further reduce consumption, an external manual switch is introduced, which can be used to completely disconnect the battery from the BMS when the vehicle is parked for a long time, which ensures the proper open circuit condition of the battery and eliminates the problem of BMS power consumption; in the face of bumps and shakes, the connectors used are replaced with higher quality alternatives, and silicone is filled at the connection to strengthen the connection.
[0097] In addition, innovative adaptive charging strategies and balancing mechanisms are added to extend the life of the battery. In conventional charging algorithms, voltage, temperature, and health status are proposed as key parameters that should be considered in implementing adaptive charging strategies. In addition, a new balancing mechanism is introduced, which involves disabling balancing and charging current during voltage measurement, effectively eliminating the voltage measurement error caused by ohmic drop for more accurate results.
[0098] According to an embodiment of the present application, an embodiment of a charging control device is also provided. Figure 5 is a schematic diagram of the structure of a charging control device provided according to an embodiment of the present application. Figure 5 As shown, the device comprises:
[0099] A battery state determination module 50 is used to determine a battery state of the battery pack at a first time step, wherein the battery state includes: a battery terminal voltage, a temperature and a health state of the battery pack;
[0100] The cumulative reward determination module 52 is used to determine the cumulative reward value corresponding to each charging action in the action space under the battery state at the first time step, wherein the cumulative reward value is used to represent the cumulative value of the expected return that can be obtained by taking the charging action under the given battery state;
[0101] A charging action determination module 54, used to determine the charging action with the largest cumulative reward value in the action space as the target charging action corresponding to the first time step;
[0102] The loop control module 56 is used to determine the battery state of the battery pack at a second time step after executing the target charging action, and repeat the steps of determining and executing the target charging action until the battery pack is fully charged, wherein the second time step is the next time step immediately adjacent to the first time step.
[0103] Optionally, the charging action includes: adjusting the charging voltage or current; the charging control device is also used to: obtain an experience sample data set, wherein the experience sample data set includes multiple experience samples obtained from the actual charging process of the vehicle battery, each experience sample corresponds to a time step, and the experience sample includes: the battery state corresponding to the first time step, the charging action performed in the first time step, the battery state corresponding to the second time step, and the immediate reward obtained after the charging action is performed in the first time step; based on the experience sample data set, the intelligent agent is trained, wherein the trained intelligent agent is used to determine the cumulative reward value corresponding to performing the charging action under a given battery state.
[0104] Optionally, training the intelligent agent includes: determining the battery state corresponding to the first time step and the feature vector corresponding to the charging action, as well as the parameter vector corresponding to the intelligent agent at the first time step; during the training process, using the intelligent agent, based on the feature vector and parameter vector corresponding to the first time step, determining the predicted value of the cumulative reward after performing the charging action; based on the empirical sample data set, calculating the true value of the cumulative reward after performing the charging action in the first time step, and determining the mean square error between the predicted value of the cumulative reward and the true value of the cumulative reward; using the gradient descent algorithm, based on the mean square error, updating the parameter vector corresponding to the intelligent agent, and using the intelligent agent after the parameter vector is updated, repeating the above training process until the mean square error is no greater than a preset error threshold.
[0105] Optionally, the target charging action includes: a random action and a greedy action, wherein the random action is any random charging action in the action space, and the greedy action is the charging action with the largest cumulative reward value in the action space; the charging control device is also used to: at a first time step, according to a preset probability, determine the random action or the greedy action as the target charging action and execute it, wherein the preset probability includes: a first probability and a second probability, the random action is selected as the target charging action with the first probability, and the greedy action is selected as the target charging action with the second probability, and the sum of the first probability and the second probability is one.
[0106] Optionally, the charging control device is also used to: determine the battery state of the battery pack at the second time step after the target charging action is performed under the battery state at the first time step, and the value of the immediate reward obtained by performing the target charging action; add the battery state and target charging action corresponding to the first time step, the battery state corresponding to the second time step, and the value of the immediate reward obtained by performing the target charging action as a new experience sample to the experience sample data set, and determine the time step corresponding to the experience sample; when the time step corresponding to the experience sample is an integer multiple of the preset time step, update the parameter vector of the intelligent agent according to the updated experience sample data set.
[0107] Optionally, the battery pack is composed of multiple batteries interconnected in series and parallel, and each battery is connected by a nickel-plated copper strip. The battery pack also includes a battery management system, wherein the battery management system enters a sleep mode when receiving a sleep command. In the sleep mode, the power consumption of the battery management system is less than a preset power consumption threshold.
[0108] Optionally, during the charging process of the battery pack, the charging control device is also used to: input a balancing current to the battery pack when the voltage difference between the highest voltage and the lowest voltage in each battery of the battery pack exceeds a preset voltage difference, wherein the balancing current is used to balance the voltage between each battery in the battery pack; during the charging process, after each preset time period, terminate the input of the balancing current and the charging current, and measure the voltage of each battery in the battery pack; after the voltage measurement is completed, resume the charging current, and re-judge whether it is necessary to continue to input the balancing current based on the measured voltage.
[0109] It should be noted that the various modules in the above-mentioned charging control device can be program modules (for example, a set of program instructions that implement a certain specific function) or hardware modules. For the latter, it can be expressed in the following forms, but is not limited to this: the expression form of each of the above-mentioned modules is a processor, or the functions of each of the above-mentioned modules are implemented by a processor.
[0110] It should be noted that the charging control device provided in this embodiment can be used to execute Figure 2The charging control method shown, therefore, the relevant explanations and descriptions of the above charging control method are also applicable to the embodiments of the present application and will not be repeated here.
[0111] An embodiment of the present application also provides a battery management system, including: a memory and a processor, the processor is used to run a program stored in the memory, wherein the program executes a charging control method when it is running: determining a battery state of a battery pack at a first time step, wherein the battery state includes: battery terminal voltage, temperature and health status of the battery pack; determining a cumulative reward value corresponding to executing each charging action in an action space under the battery state at the first time step, wherein the cumulative reward value is used to characterize the cumulative value of the expected return that can be obtained by taking a charging action under a given battery state; determining the charging action with the largest cumulative reward value in the action space as the target charging action corresponding to the first time step; after executing the target charging action, determining the battery state of the battery pack at a second time step, and repeating the steps of determining and executing the target charging action until the battery pack is fully charged, wherein the second time step is the next time step adjacent to the first time step.
[0112] An embodiment of the present application also provides a non-volatile storage medium, the non-volatile storage medium includes a stored computer program, wherein the device where the non-volatile storage medium is located executes the following charging control method by running the computer program: determining a battery state of a battery pack at a first time step, wherein the battery state includes: battery terminal voltage, temperature and health status of the battery pack; determining a cumulative reward value corresponding to executing each charging action in an action space under the battery state at the first time step, wherein the cumulative reward value is used to characterize the cumulative value of the expected return that can be obtained by taking a charging action under a given battery state; determining the charging action with the largest cumulative reward value in the action space as the target charging action corresponding to the first time step; after executing the target charging action, determining the battery state of the battery pack at a second time step, and repeating the steps of determining and executing the target charging action until the battery pack is fully charged, wherein the second time step is the next time step immediately adjacent to the first time step.
[0113] An embodiment of the present application also provides a computer program product, including a computer program, which, when executed by a processor, implements the steps of the charging control method described in each embodiment of the present application: determining a battery state of a battery pack at a first time step, wherein the battery state includes: battery terminal voltage, temperature and health status of the battery pack; determining a cumulative reward value corresponding to executing each charging action in an action space under the battery state at the first time step, wherein the cumulative reward value is used to characterize the cumulative value of the expected return that can be obtained by taking a charging action under a given battery state; determining the charging action with the largest cumulative reward value in the action space as the target charging action corresponding to the first time step; after executing the target charging action, determining the battery state of the battery pack at a second time step, and repeating the steps of determining and executing the target charging action until the battery pack is fully charged, wherein the second time step is the next time step adjacent to the first time step.
[0114] The serial numbers of the above-mentioned embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.
[0115] In the above embodiments of the present application, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, please refer to the relevant description of other embodiments.
[0116] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units can be a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.
[0117] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.
[0118] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.
[0119] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, disk or optical disk and other media that can store program codes.
[0120] The above is only a preferred implementation of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.
Claims
1. A charging control method, characterized in that: include: Determine a battery state of a battery pack at a first time step, wherein the battery state includes: a battery terminal voltage, a temperature, and a health state of the battery pack; Determine a cumulative reward value corresponding to executing each charging action in the action space under the battery state at the first time step, wherein the cumulative reward value is used to represent the cumulative value of the expected return that can be obtained by taking the charging action under a given battery state; Determine the charging action with the largest cumulative reward value in the action space as the target charging action corresponding to the first time step; After executing the target charging action, determining the battery state of the battery pack at a second time step, repeating the steps of determining and executing the target charging action until the battery pack is fully charged, wherein the second time step is the next time step immediately adjacent to the first time step.
2. The charging control method according to claim 1, characterized in that: The charging action includes: adjusting the charging voltage or current; the method also includes: Acquire an experience sample data set, wherein the experience sample data set includes a plurality of experience samples acquired from an actual charging process of a vehicle battery, each of the experience samples corresponds to a time step, and the experience samples include: a battery state corresponding to the first time step, a charging action performed in the first time step, a battery state corresponding to the second time step, and an immediate reward obtained after performing the charging action in the first time step; The intelligent agent is trained based on the experience sample data set, wherein the intelligent agent after training is used to determine the cumulative reward value corresponding to performing a charging action under a given battery state.
3. The charging control method according to claim 2, characterized in that: Training an agent involves: Determine the battery state corresponding to the first time step and the feature vector corresponding to the charging action, and the parameter vector corresponding to the agent at the first time step; During the training process, using the agent, according to the feature vector and the parameter vector corresponding to the first time step, determining a predicted value of a cumulative reward after executing the charging action; Calculating, based on the empirical sample data set, a true value of a cumulative reward after executing the charging action at the first time step, and determining a mean square error between the predicted value of the cumulative reward and the true value of the cumulative reward; A gradient descent algorithm is used to update the parameter vector corresponding to the agent according to the mean square error, and the agent after the updated parameter vector is used to repeat the above training process until the mean square error is no greater than a preset error threshold.
4. The charging control method according to claim 3, characterized in that: The target charging action includes: a random action and a greedy action, wherein the random action is any random charging action in the action space, and the greedy action is the charging action with the largest cumulative reward value in the action space; The method further comprises: At the first time step, according to a preset probability, the random action or the greedy action is determined as the target charging action and executed, wherein the preset probability includes: a first probability and a second probability, the random action is selected as the target charging action with the first probability, and the greedy action is selected as the target charging action with the second probability, and the sum of the first probability and the second probability is one.
5. The charging control method according to claim 4, characterized in that: The method further comprises: Determine the battery state of the battery pack at the second time step after performing the target charging action under the battery state at the first time step, and the value of the immediate reward obtained by performing the target charging action; Adding the battery state and target charging action corresponding to the first time step, the battery state corresponding to the second time step, and the value of the instant reward obtained by executing the target charging action as a new experience sample to the experience sample data set, and determining the time step corresponding to the experience sample; When the time step corresponding to the experience sample is an integer multiple of the preset time step, the parameter vector of the agent is updated according to the updated experience sample data set.
6. The charging control method according to claim 1, characterized in that: The battery pack is composed of a plurality of batteries interconnected in series and parallel, and each of the batteries is connected by a nickel-plated copper strip. The battery pack also includes a battery management system, wherein the battery management system enters a sleep mode when receiving a sleep command. In the sleep mode, the power consumption of the battery management system is less than a preset power consumption threshold.
7. The charging control method according to claim 6, characterized in that: During the charging process of the battery pack, the method further includes: When a voltage difference between a highest voltage and a lowest voltage in each battery of the battery pack exceeds a preset voltage difference, inputting a balancing current into the battery pack, wherein the balancing current is used to balance the voltages between the batteries in the battery pack; During the charging process, after every preset time, the input of the balancing current and the charging current is stopped, and the voltage of each battery in the battery pack is measured; After the voltage measurement is completed, the charging current is restored, and based on the measured voltage, it is re-determined whether the balancing current needs to be continuously input.
8. A charging control device, characterized in that: include: A battery status determination module, used to determine a battery status of a battery pack at a first time step, wherein the battery status includes: a battery terminal voltage, a temperature, and a health status of the battery pack; a cumulative reward determination module, used to determine the cumulative reward value corresponding to executing each charging action in the action space under the battery state at the first time step, wherein the cumulative reward value is used to represent the cumulative value of the expected return that can be obtained by taking the charging action under a given battery state; a charging action determination module, configured to determine the charging action with the largest cumulative reward value in the action space as the target charging action corresponding to the first time step; A loop control module is used to determine the battery state of the battery pack at a second time step after executing the target charging action, and repeat the steps of determining and executing the target charging action until the battery pack is fully charged, wherein the second time step is the next time step immediately adjacent to the first time step.
9. A battery management system, characterized in that: include: A memory and a processor, wherein the processor is used to run a program stored in the memory, wherein the charging control method according to any one of claims 1 to 7 is executed when the program is run.
10. A non-volatile storage medium, characterized in that: The non-volatile storage medium includes a stored computer program, wherein the device where the non-volatile storage medium is located executes the charging control method according to any one of claims 1 to 7 by running the computer program.
