Dynamic impedance matching charger and method based on battery aging, electronic equipment and medium

Abstract

The invention provides a dynamic impedance matching charger based on battery aging, a method, electronic equipment and a medium. The charger comprises a master control unit, and a battery state monitoring module, an aging grade evaluation module, a dynamic charging control module, a battery cell balancing module, an optimization control module and an anti-interference module which are connected with the master control unit to cooperatively realize charging closed-loop control. The battery state monitoring module injects a preset frequency alternating current signal and outputs an impedance parameter, a charge state and a temperature parameter; the aging grade evaluation module outputs a battery aging grade in combination with a preset model; the dynamic charging control module is matched with an adaptive charging mode instruction; the battery cell balancing module calculates the battery cell impedance difference, and balancing operation is executed if the battery cell impedance difference reaches the standard; the optimization control module generates an optimized current instruction through a preset algorithm according to the related parameters, and superposes the optimized current instruction to the reference current; the anti-interference module improves the signal-to-noise ratio of collected signals to a preset standard through isolation and shielding design.

Description

Technical Field

[0001] This invention relates to the field of electromagnetic shock technology, specifically to a dynamic impedance matching charger, method, electronic device, and medium based on battery aging. Background Technology

[0002] In high-power battery applications such as electric vehicles and energy storage devices, battery charging technology faces multiple industry pain points and technical bottlenecks, which restrict the improvement of charging safety, battery life, and operation and maintenance economy. Specific problems are as follows:

[0003] 1. Fast charging accelerates battery aging and poses significant safety risks: The current mainstream constant current-constant voltage (CC-CV) charging strategy uses fixed parameters to execute the charging process, without considering the impedance increase characteristics of the battery due to aging during its service life. This strategy is prone to abnormal fluctuations in battery terminal voltage during charging, leading to overvoltage risks. At the same time, the Joule heat accumulation accompanying the impedance increase is difficult to control effectively, significantly increasing the risk of thermal runaway and shortening battery cycle life.

[0004] 2. High maintenance costs for battery aging detection: In existing technologies, battery aging detection mostly relies on manual disassembly of battery packs for offline testing. This is not only cumbersome and inefficient, but also carries the risk of battery damage during the disassembly process. In addition, premature replacement of battery packs due to untimely aging detection is common. The cost of replacing aged battery packs accounts for too high a proportion of the total maintenance cost of the equipment throughout its life cycle, which increases the economic burden on users and enterprises.

[0005] 3. Existing charging control technologies have significant shortcomings: On the one hand, some existing solutions rely on cloud platforms to analyze battery status and issue charging commands. Cloud data transmission and processing have inherent delays, which cannot meet the millisecond-level real-time control requirements in fast charging scenarios, and easily lead to charging strategy adjustments lagging behind changes in battery status. On the other hand, neither traditional fixed-parameter charging strategies nor some simple adaptive strategies based solely on voltage and temperature can dynamically adapt to the dynamic aging state of the battery over time, making it difficult to balance charging efficiency and battery protection.

[0006] Therefore, how to provide a charging solution that balances real-time performance, adaptability, and safety is a pressing technical problem that needs to be solved. Summary of the Invention

[0007] In view of this, the present disclosure provides a dynamic impedance matching charger, method, electronic device and medium based on battery aging, which at least partially solves the problems existing in the prior art.

[0008] In a first aspect, embodiments of this disclosure provide a dynamic impedance matching charger based on battery aging, comprising:

[0009] The dynamic impedance matching charger includes: a battery status monitoring module, an aging level assessment module, a dynamic charging control module, a cell balancing module, an optimization control module, an anti-interference module, and a main control unit; the main control unit is electrically connected to the battery status monitoring module, the aging level assessment module, the dynamic charging control module, the cell balancing module, the optimization control module, and the anti-interference module respectively, and works together to realize closed-loop control of battery charging;

[0010] The battery status monitoring module is used to inject an AC signal of a preset frequency into the battery pack and output the battery's impedance parameters, state of charge, and temperature parameters.

[0011] The aging level assessment module is used to receive the impedance parameter, combine it with a preset aging assessment model, and output the battery aging level.

[0012] The dynamic charging control module is used to match and output an appropriate charging mode command according to the battery aging level.

[0013] The cell balancing module is used to receive the impedance parameters, calculate the impedance difference between cells, and generate and execute a balancing operation command when the impedance difference reaches a preset threshold.

[0014] The optimization control module constructs a multi-dimensional state vector including battery impedance increment, temperature change rate, current margin and basic state parameters, and normalizes the multi-dimensional state vector. It then uses a delayed update deep reinforcement learning algorithm that integrates historical monitoring data feature extraction mechanism to optimize the charging command in combination with a preset multi-dimensional reward function, and sends the optimized charging command to the dynamic charging control module.

[0015] The anti-interference module is used to improve the signal-to-noise ratio of the signal collected by the battery status monitoring module to a preset standard value through a preset isolation and shielding design.

[0016] According to a specific implementation of this disclosure, the frequency range of the AC signal is 0.1–1000Hz, and the impedance parameters include: solid electrolyte interface film resistance and charge transfer resistance.

[0017] According to a specific implementation of this disclosure, the impedance difference is the difference between the aging impedance of the target cell and the impedance of the minimum aging cell, and the equalization operation command is calculated and generated based on the impedance difference and the current maximum single-cell state of charge of the battery.

[0018] According to one specific implementation of the embodiments of this disclosure,

[0019] The delayed update deep reinforcement learning algorithm that utilizes the feature extraction mechanism of fused historical monitoring data, combined with a preset multi-dimensional reward function to optimize charging instructions, includes: using a dual-delay deep deterministic policy gradient algorithm that extracts subnets based on temporal features to construct a reinforcement learning model;

[0020] The reinforcement learning model is trained based on a multi-dimensional reward function that considers charging efficiency, battery aging suppression, temperature safety, parameter stability, and consistency of lifetime prediction.

[0021] A quantization strategy is adopted for the trained reinforcement learning model, and online inference is achieved through a cache alternation mechanism;

[0022] If the battery temperature exceeds the preset threshold or the single cell voltage exceeds the safety threshold, the hardware comparator will directly cut off the main relay within microseconds to achieve an emergency interruption of the charging main circuit.

[0023] According to a specific implementation of this disclosure, the anti-interference module includes: adding an isolation power supply at the power supply end of the current sampling circuit, and wrapping a μ-level metal shielding layer around the outer layer of the current sampling cable and the outer shell of the sampling module.

[0024] Secondly, embodiments of this disclosure provide a dynamic impedance matching charging method based on battery aging, which includes:

[0025] S1: Receives a charging request and initiates the charging process;

[0026] S2: Inject an AC signal of a preset frequency into the battery pack and collect the battery's impedance parameters, state of charge, and temperature parameters.

[0027] S3: Determine the battery aging level based on the impedance parameters;

[0028] S4: Match the appropriate charging mode according to the battery aging level;

[0029] S5: A delayed update deep reinforcement learning algorithm that integrates historical monitoring data feature extraction mechanism optimizes the charging command by combining a preset multi-dimensional reward function, and sends the optimized charging command to the dynamic charging control module.

[0030] S6: Calculate the impedance difference between cells in the battery pack. When the impedance difference reaches a preset threshold, start the impedance equalization operation.

[0031] S7: Perform temperature control monitoring and execute corresponding charging adjustment or protection operations based on the temperature parameters;

[0032] S8: Cycle through S2-S7 until charging is complete.

[0033] According to a specific implementation of this disclosure, step S7 further includes an anomaly diagnosis operation, which triggers a current reduction or charging stop operation when the rate of change of the impedance parameter per unit time exceeds a preset rate of change threshold, or the voltage of a single cell exceeds a preset voltage range.

[0034] Thirdly, embodiments of this disclosure provide an electronic device, the electronic device comprising:

[0035] At least one processor; and,

[0036] The memory is communicatively connected to the at least one processor; wherein,

[0037] The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the above-described dynamic impedance matching charging method based on battery aging.

[0038] Fourthly, embodiments of this disclosure provide a non-transitory computer-readable storage medium storing computer instructions for causing the computer to execute the aforementioned dynamic impedance matching charging method based on battery aging.

[0039] In summary, compared with the prior art, this embodiment has the following advantages:

[0040] 1. Significantly extends battery cycle life: Specifically designed for end-of-life batteries (e.g., R...). ct (With an increase of >50%), the current-limited constant-voltage charging strategy can effectively suppress lithium dendrite growth, reduce irreversible chemical reactions inside the battery, significantly increase the number of battery cycles in electric vehicles, energy storage devices and other scenarios, and extend the overall service life of the battery.

[0041] 2. Comprehensive improvement of fast charging safety performance: By monitoring the battery aging status in real time, the system automatically triggers current limiting and temperature control protection for high-risk aging batteries, and controls the battery temperature rise during charging (e.g., below 45℃). Compared with the temperature rise peak of the traditional solution (75℃), the safety is significantly improved. At the same time, based on multi-cell impedance balancing technology, it reduces the problem of local overheating caused by inconsistent cell aging, greatly reduces the risk of thermal runaway, and ensures safety throughout the charging process.

[0042] 3. Dynamically optimize overall charging efficiency: Adapt differentiated charging strategies for batteries with different aging levels. New batteries (early aging) adopt a stepped current ramp mode (0.5C→2C) to minimize fast charging time. Reinforcement learning optimization dynamically balances the three goals of charging speed, battery life and temperature rise control to avoid ineffective energy consumption and improve overall charging efficiency.

[0043] 4. Effectively reduce operation and maintenance costs: By using real-time EIS online diagnostic technology during charging, the traditional method of manually disassembling and inspecting battery aging status is replaced, significantly improving the efficiency of operation and maintenance inspection; the extension of battery life and the reduction of failure risk not only reduce the frequency of electric vehicle battery replacement, but also reduce the equipment maintenance costs in scenarios such as energy storage power stations, optimizing the economics of the entire life cycle.

[0044] 5. Possesses broad hardware compatibility: It can directly reuse mature power monitoring modules such as TIBQ27411-G1 without additional hardware architecture reconstruction, reducing adaptation costs; at the same time, it supports impedance model migration for different types of batteries such as lithium-ion batteries and solid-state batteries, adapting to various high-power battery application scenarios, and has strong technical scalability and feasibility for implementation.

[0045] In summary, this invention, through its core logic of real-time aging perception, dynamic strategy matching, multi-objective optimization, and end-to-end safety protection, not only solves the pain points of traditional charging solutions that ignore battery aging, such as shortened lifespan and safety hazards, but also achieves comprehensive optimization of charging efficiency, maintenance costs, and hardware compatibility, providing a more efficient, safe, and economical charging solution for high-power battery application scenarios. Attached Figure Description

[0046] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments and accompanying drawings. Here, the illustrative embodiments and descriptions of this invention are used to explain the invention, but are not intended to limit the invention.

[0047] The term "comprising" and its variations as used herein signify open inclusion, i.e., "including but not limited to". Unless otherwise stated, the term "or" means "and / or". The term "based on" means "at least partially based on". The terms "one example embodiment" and "one embodiment" mean "at least one example embodiment". The term "another embodiment" means "at least one additional embodiment". The terms "first", "second", etc., may refer to different or the same objects. Other explicit and implicit definitions may also be included below.

[0048] Figure 1 This is an intentional description of a dynamic impedance matching charger based on battery aging, as provided in the first embodiment of the present invention.

[0049] Figure 2 A schematic flowchart of the dynamic impedance matching charging method based on battery aging provided in the second embodiment of the present invention;

[0050] Figure 3 An exemplary structural diagram of a device capable of implementing the method according to an embodiment of the present invention is shown. Detailed Implementation

[0051] The embodiments of this disclosure will now be described in detail with reference to the accompanying drawings.

[0052] The following specific examples illustrate the implementation of this disclosure. Those skilled in the art can easily understand other advantages and effects of this disclosure from the content disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of this disclosure, and not all of them. This disclosure can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of this disclosure. It should be noted that, in the absence of conflict, the following embodiments and features in the embodiments can be combined with each other. Based on the embodiments in this disclosure, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this disclosure.

[0053] It should be noted that various aspects of embodiments within the scope of the appended claims are described below. It will be apparent that the aspects described herein can be embodied in a wide variety of forms, and any particular structure and / or function described herein is merely illustrative. Based on this disclosure, those skilled in the art will understand that one aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. For example, any number of aspects set forth herein can be used to implement the device and / or practice the method. Additionally, this device and / or method can be implemented using structures and / or functionalities other than one or more of the aspects set forth herein.

[0054] It should also be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of this disclosure. The drawings only show the components related to this disclosure and are not drawn according to the number, shape and size of the components in actual implementation. In actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0055] Furthermore, specific details are provided in the following description to facilitate a thorough understanding of the examples. However, those skilled in the art will understand that the described aspects can be practiced without these specific details.

[0056] Please see Figure 1 This application provides a schematic diagram of a prior art dynamic impedance matching charger based on battery aging, as shown in the embodiment. Figure 1The dynamic impedance matching charger based on battery aging shown includes: a battery status monitoring module, an aging level assessment module, a dynamic charging control module, a cell balancing module, an optimization control module, an anti-interference module (not shown), and a main control unit (not shown).

[0057] exist Figure 1 In the embodiment shown, the battery status monitoring module can be an EIS monitoring module. During the charging process, the charging motor injects an AC signal of a preset frequency into the vehicle battery through the charging gun wire, and outputs the battery's impedance parameters, state of charge, and temperature parameters.

[0058] Preferably, the frequency range of the AC signal is 0.1-1000Hz, and the impedance parameters include: solid electrolyte interface film resistance and charge transfer resistance.

[0059] The battery status monitoring module analyzes the impedance spectrum characteristics of the battery in real time (R0). Ω R ct ), where R Ω R is the interfacial film resistance of the solid electrolyte. ct R is the charge transfer resistor. Ω Reflecting the state of the solid electrolyte interface film in the battery, accelerated battery aging will lead to R Ω Increase, R ct The efficiency of charge transfer within a battery is a core parameter for assessing its aging level; battery aging grades are directly based on R. ct Increase determination.

[0060] Preferably, the aging level assessment module receives impedance parameters, combines them with a preset aging assessment model, and outputs the battery aging level. The aging assessment model is a multi-stage model, dividing aging levels according to the increase in charge transfer resistance. The dynamic charging control module dynamically switches the charging mode according to the aging level, matching and outputting an appropriate charging mode command. For example, the aging level assessment module can output the aging level based on a three-stage aging model, where the aging level can be determined according to the initial stage R... ct The increase is less than 20%; mid-term R ct The increase is greater than 20% but less than 50%; R in the final stage ct We categorize companies based on increases greater than 50%.

[0061] In other words, R ct The increase is less than 20% compared to the battery's factory rated value. At this point, the battery has high internal electrochemical activity and low impedance, enabling it to withstand higher charging currents; R ct When the increase is in the 20%–50% range, phenomena such as thickening of the solid electrolyte interface film and degradation of active materials begin to appear inside the battery, leading to increased impedance. A balance needs to be struck between charging speed and the risk of battery damage. ctWhen the increase exceeds 50%, the battery's electrochemical performance deteriorates significantly, impedance increases substantially, and safety hazards such as overheating and overvoltage during charging are likely to occur. Low-current, low-risk charging should be the primary method.

[0062] The charger's charging control module will switch charging modes in real time based on the above aging level determination results, and generate a charging curve adapted to the current aging state. The specific strategy is as follows:

[0063] In the early stages of aging, a stepped current ramp-up strategy (e.g., 0.5C → 2C) is employed. For example, for new batteries or slightly aged batteries, a gradual current ramp-up logic is used to avoid the impact of an initial large current on the battery. During charging startup, the battery is preheated with a low current of 0.5C (C is a multiple of the battery capacity, such as 50A for a 100Ah battery). Then, the charging current is gradually increased at preset intervals (e.g., 5 minutes, which can be adjusted according to battery specifications) until the rated maximum charging current of 2C is reached. This maximizes charging efficiency and shortens charging time while ensuring battery safety.

[0064] During the mid-aging stage, a pulse charging strategy (e.g., 10s on / 5s off) is used, while the battery impedance (R) during mid-aging... Ω R ct When the temperature rises, continuous high-current charging can easily lead to heat accumulation. Therefore, a pulse mode of alternating charging and rest is adopted. For example, a cycle of 10 seconds charging (on) and 5 seconds rest (off): during the charging phase, a moderate current (e.g., 1C–1.5C) is used to replenish the charge, while during the rest phase, the concentration polarization and ohmic polarization are alleviated by redistributing ions inside the battery, thereby reducing the battery temperature rise. The 10-second on / 5-second off time parameter is used as an example. In practice, the charging-rest duty cycle can be dynamically fine-tuned (e.g., ±10%) using a reinforcement learning optimization engine to balance the charging speed and polarization suppression effect. This application does not impose specific limitations on the charging-rest duty cycle in its embodiments.

[0065] In the late aging stage (severe aging), a current-limiting and constant-voltage strategy (e.g., ≤0.3C) is adopted. For severely aged batteries, charging safety is prioritized, and a conservative strategy of low current and constant voltage is used. For example, during the charging process, the charging current is strictly limited to a low rate range of ≤0.3C to avoid lithium dendrite growth or thermal runaway caused by high current. At the same time, the rated voltage of the battery (e.g., 4.2V for lithium-ion batteries) is used as the constant voltage target. When the battery voltage approaches the rated value, the voltage is maintained by slowly decreasing the current until the battery SOC (state of charge) reaches more than 95%, minimizing damage to the aged battery during the charging process and extending its remaining service life.

[0066] It should be noted that the values ​​mentioned above, such as 0.5C, 2C, 10s on / 5s off, and ≤0.3C, are all basic examples. In actual applications, they need to be adjusted according to the specific specifications of the battery (such as capacity, material system, and rated voltage), application scenarios (such as electric vehicles and energy storage devices), and charger power configuration. For example, the current ramp-up rate and constant voltage value of lithium iron phosphate batteries and ternary lithium batteries need to be set differently. The pulse interval of high-power energy storage batteries can also be optimized according to heat dissipation conditions to ensure the universality and adaptability of the dynamic switching strategy.

[0067] The following is an introduction to the cell balancing module.

[0068] Optionally, the impedance difference between the cells is the difference between the aging impedance of the target cell and the impedance of the minimum aging cell, and the equalization operation command is calculated and generated based on the impedance difference and the current maximum single-cell state of charge of the battery.

[0069] The charger uses an online monitoring module to collect the aging impedance (denoted as R_{aging-target}, i.e., the target cell aging impedance of each cell to be evaluated) of all individual cells in the battery pack in real time through the battery status monitoring module, and selects the minimum aging impedance (denoted as R_{aging-min}). The difference between the target cell aging impedance and the minimum aging cell impedance of each cell is calculated using the formula: ΔR=R_{aging-target}-R_{aging-min}. This quantifies the degree of aging inconsistency between cells. That is, the larger ΔR is, the higher the degree of aging of the cell is compared with the healthy cells in the battery pack, and the more targeted balancing is needed.

[0070] Based on the above ΔR value, using formula Q bal =k·ΔR·SOC max Calculate the required equalization capacity for each cell. Where Q... bal To balance capacity, k is the cell consistency coefficient, which needs to be preset based on the cell type of the battery pack (such as ternary lithium, lithium iron phosphate), the number of cells connected in series, and heat dissipation conditions (for example, the k value of a 12-cell ternary lithium battery pack can be set to 0.02-0.05, but specific values ​​need to be calibrated experimentally). SOC max This is the maximum real-time state of charge (SOC) value among all individual cells in the current battery pack, ensuring that the balancing process does not exceed the battery's safe charging limit and avoiding the risk of overcharging.

[0071] To avoid unnecessary balancing operations that could increase energy consumption, a trigger threshold can be set. When the impedance difference ΔR between any two cells is greater than 5%, the charger's balancing control module will immediately activate the active balancing mechanism. When ΔR ≤ 5%, the aging differences between cells have little impact on charging safety and consistency, and balancing is not required. When ΔR > 5%, cells with higher aging levels are prone to problems such as local current concentration and excessive temperature rise during charging, which must be alleviated through balancing adjustment.

[0072] The optimization control module will be described below.

[0073] Optionally, a multi-dimensional state vector is constructed, including battery impedance increment, temperature change rate, current margin, and basic state parameters. The multi-dimensional state vector is then normalized. A delayed update deep reinforcement learning algorithm that integrates historical monitoring data feature extraction mechanism is used to optimize the charging command in combination with a preset multi-dimensional reward function. The optimized charging command is then sent to the dynamic charging control module.

[0074] The fundamental state parameters include: charge transfer resistance (R0). ct ), SOC (State of Charge) and cell temperature (T).

[0075] For example, to eliminate the dimensional differences of various state parameters under high battery impedance scenarios and ensure the consistency and stability of the input data of the reinforcement learning model, this application extends the basic battery state parameters to construct a six-dimensional state vector, specifically including:

[0076] Normalized impedance increment ΔR ct =(R ct_now -R ct_new ) / R ct_new , where R ct_now R is the measured charge transfer resistance of the battery at the current moment. ct_new The initial charge transfer resistance, ΔR, is the value specified by the battery at the factory. ct The value range is [0,1], used to quantify the dynamic changes in battery aging; the temperature change rate dT / dt is expressed by the formula dT / dt=(T t -T t-1 The value is obtained by calculating ) / Δt, where T t T represents the current cell temperature. t-1 The cell temperature is the value at the previous sampling time, and Δt is the sampling period (e.g., Δt = 1s). This is used to reflect the real-time trend of battery temperature changes and to predict the risk of thermal runaway in advance; current margin I margin =I max (SOC, T)-I refThe current safety boundary is characterized by the 6-dimensional state vector, which is then normalized and input into the policy network to solve the gradient vanishing problem.

[0077] Optionally, a reinforcement learning model can be constructed using a dual-delay deep deterministic policy gradient algorithm for extracting subnets based on temporal features.

[0078] The reinforcement learning model is trained based on a multi-dimensional reward function that considers charging efficiency, battery aging suppression, temperature safety, parameter stability, and consistency of lifetime prediction.

[0079] A quantization strategy is adopted for the trained reinforcement learning model, and online inference is achieved through a cache alternation mechanism;

[0080] If the battery temperature exceeds the preset threshold or the single cell voltage exceeds the safety threshold, the hardware comparator will directly cut off the main relay within microseconds to achieve an emergency interruption of the charging main circuit.

[0081] Specifically, the delayed-update deep reinforcement learning algorithm that integrates the historical monitoring data feature extraction mechanism can be the dual-delay deep deterministic policy gradient (TD3) algorithm. Simultaneously, an impedance feature extraction subnet is introduced, forming a network architecture that balances temporal feature mining and policy optimization (hereinafter referred to as the TD3-IMP architecture). The specific structural design is as follows:

[0082] The backbone network may include one command I for outputting charging current. _cmd The model consists of an Actor policy network and two Critic value networks (Q1, Q2) for evaluating action value and suppressing Q-value overestimation. The update frequency of the Actor network and the corresponding target network is set to half the update frequency of the Critic network. This delayed update mechanism avoids the superposition of policy update and value estimation bias, thereby improving the stability of the model output.

[0083] In addition, an independent one-dimensional convolutional CNN subnetwork is embedded in the front end of the policy network. The kernel size of this subnetwork can be set to 3×1, the stride to 1, and the number of output channels to 16. It is used to extract features from the electrochemical impedance spectroscopy (EIS) sequence data collected continuously over a preset number of steps (e.g., 20 steps), mine the temporal correlation information of impedance changes, and finally output a 32-dimensional hidden feature vector.

[0084] Then, the 32-dimensional latent feature vector output by the one-dimensional CNN subnet is concatenated with the six-dimensional state vector after layer normalization to form a 38-dimensional fused feature vector. This fused feature vector is then fed into the subsequent fully connected layer for feature mapping and policy calculation, ensuring that the model simultaneously utilizes the real-time battery status and historical impedance timing information to optimize charging commands.

[0085] Furthermore, to balance the model's exploration capability and convergence speed, exploration noise generated by the Ornstein-Uhlenbeck process is introduced at the output of the Actor network, with noise intensity σ. _t According to formula σ _t =σ o • exp(-t / τ) dynamic decay, where the initial noise intensity σ0=0.3 and the decay coefficient τ=30min, automatically reduces noise interference as the charging process progresses, ensuring the stability of the strategy in the later stages of charging.

[0086] Furthermore, to achieve multi-objective optimization of improving charging efficiency, suppressing battery aging, and preventing safety risks, a multi-dimensional weighted reward function can be constructed by extending the traditional reward function, as follows:

[0087] In the traditional reward function R = α·I charge -β·ΔR ct Based on -γ·max(T-45,0), add current fluctuation penalty ε·|I t -I (t-1) |, ε=0.02, safety boundary reward ζ·min(I margin ,0),ζ=0.5, and lifetime prediction consistency reward η·exp(-|SOH pred -SOH label |, where SOH pred By R ct Online estimation of SOH label For the BMS (Battery Management System), the factory calibration is η = 1.0 to suppress frequent relay jitter. Wherein, α·I charge For charging efficiency rewards, α is the efficiency weighting coefficient, and I _charge The current charging current is used to encourage the model to increase the charging current within a safe range, thereby shortening the charging time; -β·ΔR ct The term represents the aging inhibition penalty, β is the aging weighting coefficient, and ΔR is the aging inhibition penalty term. ct The difference between the charge transfer resistance at the current moment and the initial moment is used to penalize charging behavior that leads to increased battery impedance and delay battery aging; -γ·max(T-45,0) is a temperature over-limit penalty term, where γ is a temperature weighting coefficient. When the battery temperature T>45℃, this term generates a penalty value positively correlated with the over-temperature amplitude; when T≤45℃, this term is 0, used to suppress excessively high battery temperatures; ε·|It-I (t-1) | represents the current fluctuation penalty term, ε is the fluctuation weighting coefficient, and I t For the current charging current, I (t-1)The charging current from the previous moment is used to suppress drastic fluctuations in the charging current and reduce losses caused by frequent relay operation; ζ·min(I margin ,0) represents the safety boundary reward term, and ζ represents the safety weight coefficient. When the current margin I margin When I < 0 (i.e., the current current exceeds the safe limit), this item generates a negative reward; when I margin When ≥0, this term is 0, used to guide the model to fully utilize the safety margin without exceeding the safety boundary; η·exp(-|SOH pred -SOH label | represents the consistency reward for lifetime prediction, η is the consistency weight coefficient, and SOH pred For R-based ct Online estimated battery health status, SOH label This is the baseline health status value set at the battery's factory, and this item rewards SOH through an exponential function. pred With SOH label Consistency in this regard guides the model to output a charging strategy that aligns with the battery's electrochemical characteristics.

[0088] It should be noted that α, β, γ, δ, ε, ζ, and η are all configurable weighting coefficients that can be adaptively adjusted according to different battery types (such as lithium-ion batteries and solid-state batteries) and application scenarios (such as electric vehicles and energy storage devices) to meet actual usage requirements.

[0089] In addition, to ensure the generalization ability and stability of the reinforcement learning model under different battery aging states and different ambient temperatures, a training process combining offline pre-training and online fine-tuning can be used to train the reinforcement learning model, as detailed below:

[0090] First, a training environment can be built based on a digital twin battery model, whose parameters cover the aging states of the battery throughout its entire life cycle (e.g., R...). ct In scenarios with a power increase of 0-80% and wide temperature range (e.g., -10℃ to 60℃), 200k training samples are randomly sampled in this environment to pre-train the reinforcement learning model offline. Through extensive simulation data, the reinforcement learning model learns the basic charging strategy, obtains initial network weights, and ensures that the model possesses basic charging optimization capabilities.

[0091] Then, in actual charging scenarios, after each complete charging process, the real-time state data (which may include a six-dimensional state vector, actual charging current, reward value, etc.) of the final time period (e.g., 10 minutes) at the end of the charging phase is stored in an experience replay pool. For example, the maximum storage capacity of the experience replay pool is set to 50k data entries. When the amount of data stored in the experience replay pool exceeds 5k entries, the online model fine-tuning process is triggered, for example, with a batch size of 256 and a 1×10... -4The learning rate is used to fine-tune the model for one training cycle, enabling the model to adapt to dynamic factors such as battery batch differences and seasonal temperature changes in real-world applications, thus avoiding model performance drift.

[0092] In addition, to prevent model anomalies during online fine-tuning, a Q-value error monitoring step can be set up to calculate the error between the Q-value predicted by the model online and the actual return value generated during the actual charging process in real time. If the absolute value of the error exceeds the preset upper limit (e.g., 15%), the model is determined to be in an abnormal state, the network parameters are immediately frozen, and a conservative PID (Proportional-Integral-Derivative Controller) baseline charging strategy is switched until the model is readjusted and returns to normal, ensuring the safety and continuity of the charging process.

[0093] Furthermore, to meet the real-time requirements of the charger for charging commands and to construct a hardware-level safety fallback mechanism, the embodiments of this application are designed from two aspects: model inference acceleration and emergency protection, as detailed below:

[0094] First, the trained reinforcement learning model is quantized using the INT8 quantization strategy, converting the model weights from 32-bit floating-point numbers to 8-bit integers. This significantly reduces the computational load while ensuring that the model accuracy loss is ≤5%, keeping the inference latency within 0.8ms. Simultaneously, two independent random access memories (RAMs) are allocated on the microcontroller unit (MCU) as dual buffers, alternating between state data writing and model inference output operations. This ensures that the output control cycle of charging commands remains stable within 2ms, meeting the real-time control requirements of fast charging scenarios.

[0095] Furthermore, to address software malfunctions or extreme emergencies, a hardware comparator independent of the software system can be configured: This comparator monitors battery temperature and individual cell voltage in real time. When the battery temperature T > 52℃ (or reaches other preset temperature safety thresholds), or the individual cell voltage > 4.25V (or reaches other preset overvoltage safety thresholds), the hardware comparator directly pulls down I without going through the software control process. GBT The drive signal cuts off the main relay, achieving an emergency interruption of the charging main circuit with a microsecond-level response, forcibly terminating the charging process, and ensuring that the battery and charger system are out of danger.

[0096] The following section introduces the anti-jamming module.

[0097] Optional, the anti-interference module includes: adding an isolation power supply at the power supply end of the current sampling circuit, and wrapping the outer layer of the current sampling cable and the outer shell of the sampling module with a μ-level metal shielding layer.

[0098] Specifically, an isolation power supply (such as an optocoupler or magnetic coupling module) is added to the power supply end of the current sampling circuit to achieve electrical isolation between the sampling circuit and the main charging circuit. In high-power charging scenarios, the main charging circuit (typically hundreds of volts high voltage and tens of amps high current) generates strong electromagnetic radiation and voltage fluctuations. If the sampling circuit and the main circuit share the same power supply, these interferences will enter the sampling system through power coupling, causing EIS signal distortion. Power isolation can cut off this common ground interference path to ensure the stability of the sampling circuit power supply and avoid the impact of main circuit interference on the acquisition of micro-current signals.

[0099] Alternatively, a highly conductive metal material (such as copper or aluminum foil) at the μm level (micrometer level, typically 5-20 μm thick) can be wrapped around the outer layer of the current sampling cable and the outer shell of the sampling module, and reliably grounded. This physical shielding structure effectively blocks external electromagnetic noise (such as radiated noise generated by the high-frequency switching of the charging switch tube and electromagnetic interference from the external environment) from entering the sampling circuit. When the EIS signal is a micro-current signal <10mA, it is extremely sensitive to external noise. The μm-level metal shielding layer can absorb external radiated noise through the principle of electromagnetic induction cancellation, preventing it from superimposing on the EIS signal and ensuring the purity of the acquired signal.

[0100] Signal-to-noise ratio (SNR) is a core metric for measuring signal quality. For example, a pre-set SNR standard of greater than 60dB means that the effective signal strength of the EIS is more than 1000 times the noise signal strength (for every 20dB increase in SNR, the signal strength is 10 times the noise; 60dB is 10...). 3 =1000 times). Achieving this specification ensures the charger's accuracy in resolving EIS signals, even under strong interference conditions such as high-power charging (e.g., 2C fast charging), accurately separating the ohmic resistance (R) reflecting the battery's aging state. Ω ), charge transfer resistance (R) ct Key parameters such as noise interference are used to avoid misjudgment of aging levels and provide a reliable data foundation for subsequent dynamic charging strategy switching.

[0101] Secondly, this application also provides a dynamic impedance matching charging method based on battery aging, as detailed in the embodiments below. Figure 2 This can be executed by an electronic device, which can act as a host computer, specifically by one or more processors within the electronic device, to achieve the following steps:

[0102] S201: Receive charging request and start charging process.

[0103] Specifically, after the charging gun is plugged in and the battery management system successfully hands over, the main control unit receives the charging demand command, and the entire charging control process is then officially started. This is the pre-trigger step for the charger to enter the subsequent battery status monitoring module for monitoring, aging determination and execution of dynamic charging strategies.

[0104] S202. Inject an AC signal of a preset frequency into the battery pack and collect the battery's impedance parameters, state of charge, and temperature parameters.

[0105] Specifically, the main control unit synchronously starts the EIS module, injecting a 0.1-1000Hz microsecond-level AC signal into the battery via the gun line; a high-precision current sampling circuit with isolation and μ-level shielding captures the echo in real time, and then analyzes it through FFT to finally output the SEI film resistance R. Ω Charge transfer resistance R ct Battery SOC and cell temperature T.

[0106] S203. Determine the battery aging level based on impedance parameters.

[0107] Specifically, aging levels are determined based on the increase in charge transfer resistance. The increase in charge transfer resistance is calculated as: (Current charge transfer resistance – Battery factory-rated charge transfer resistance) / Battery factory-rated charge transfer resistance × 100%. The aging level can be determined according to the initial stage R... ct The increase is less than 20%; mid-term R ct The increase is greater than 20% but less than 50%; R in the final stage ct If the increase is greater than 50%, a division is performed. The division result can be written to a register in the form of a numerical tag for subsequent strategy selection.

[0108] It should be noted that the battery's factory-calibrated charge transfer resistance is pre-stored in the non-volatile memory of the battery management system.

[0109] S204. Match the appropriate charging mode according to the battery aging level.

[0110] Specifically, the charging mode is selected based on the aging level label. In the initial aging stage, a stepped current ramp-up strategy (e.g., 0.5C → 2C) is employed. For example, for new or slightly aged batteries, a gradual current ramp-up logic is used to avoid impacting the battery with an initial large current. Charging begins with a low current of 0.5C (C is a multiple of the battery capacity, e.g., 0.5C for a 100Ah battery is 50A) to preheat the battery. Then, the charging current is gradually increased at preset intervals (e.g., 5 minutes, which can be adjusted according to battery specifications) until the rated maximum charging current of 2C is reached. This maximizes charging efficiency and shortens charging time while ensuring battery safety.

[0111] During the mid-aging stage, a pulse charging strategy (e.g., 10s on / 5s off) is used, while the battery impedance (R) during mid-aging... Ω R ct When the temperature rises, continuous high-current charging can easily lead to heat accumulation. Therefore, a pulse mode of alternating charging and rest is adopted. For example, a cycle of 10 seconds charging (on) and 5 seconds rest (off): during the charging phase, a moderate current (e.g., 1C–1.5C) is used to replenish the charge, while during the rest phase, the concentration polarization and ohmic polarization are alleviated by redistributing ions inside the battery, thereby reducing the battery temperature rise. The 10-second on / 5-second off time parameter is used as an example. In practice, the charging-rest duty cycle can be dynamically fine-tuned (e.g., ±10%) using a reinforcement learning optimization engine to balance the charging speed and polarization suppression effect. This application does not impose specific limitations on the charging-rest duty cycle in its embodiments.

[0112] In the late aging stage (severe aging), a current-limiting and constant-voltage strategy (e.g., ≤0.3C) is adopted. For severely aged batteries, charging safety is prioritized, and a conservative strategy of low current and constant voltage is used. For example, during the charging process, the charging current is strictly limited to a low rate range of ≤0.3C to avoid lithium dendrite growth or thermal runaway caused by high current. At the same time, the rated voltage of the battery (e.g., 4.2V for lithium-ion batteries) is used as the constant voltage target. When the battery voltage approaches the rated value, the voltage is maintained by slowly decreasing the current until the battery SOC (state of charge) reaches more than 95%, minimizing damage to the aged battery during the charging process and extending its remaining service life.

[0113] It should be noted that the values ​​mentioned above, such as 0.5C, 2C, 10s on / 5s off, and ≤0.3C, are all basic examples. In actual applications, they need to be adjusted according to the specific specifications of the battery (such as capacity, material system, and rated voltage), application scenarios (such as electric vehicles and energy storage devices), and charger power configuration. For example, the current ramp-up rate and constant voltage value of lithium iron phosphate batteries and ternary lithium batteries need to be set differently. The pulse interval of high-power energy storage batteries can also be optimized according to heat dissipation conditions to ensure the universality and adaptability of the dynamic switching strategy.

[0114] S205. The charging command is optimized by using a delayed update deep reinforcement learning algorithm that integrates the feature extraction mechanism of historical monitoring data and a preset multi-dimensional reward function, and the optimized charging command is sent to the dynamic charging control module.

[0115] Specifically, a six-dimensional state vector is first constructed, which specifically includes the normalized impedance increment Δ. Rct Temperature change rate dT / dt, current margin I _margin and the original charge transfer resistance R ctThe six-dimensional state vector is then subjected to layer normalization to eliminate the dimensional differences between parameters.

[0116] Next, the dual-delay deep deterministic policy gradient (TD3) algorithm is adopted, in which the policy network μ is responsible for outputting the charging current command I_ cmd The dual Critic networks Q and Q' are used to suppress the overestimation of Q value; meanwhile, the Actor network and its corresponding target network adopt a delayed update method, and their update frequency is set to 1 / 2 of the update frequency of the Critic network.

[0117] A one-dimensional convolutional neural network (CNN) is embedded at the front end of the policy network. This subnetwork extracts features from the electrochemical impedance spectroscopy (EIS) sequence data collected in the past 20 steps and outputs a 32-dimensional hidden feature vector. This hidden feature vector is then concatenated with a six-dimensional state vector that has been normalized by the layer to form a fused feature vector, which is then fed into a fully connected layer for subsequent operations.

[0118] The designed reward function is R = α·I. charge -β·ΔR ct -γ·max(T-45,0)-δ·max(ΔR ct -ΔR ct_prev ,0)-ε·|I t -I t-1 |+ζ·min(I margin ,0)+η·exp(-|SOH pred -SOH label |), where α, β, γ, δ, ε, ζ, and η are all weight coefficients that can be configured according to the actual scenario.

[0119] The online inference stage adopts the INT8 quantization strategy and a dual-buffered alternation mechanism at the microcontroller unit (MCU) to ensure that the control cycle of the charging command is ≤2ms. When the battery temperature is detected to be >52℃ or the single cell voltage is detected to be >4.25V, the hardware comparator can directly cut off the main relay within microseconds to realize the emergency interruption of the charging main circuit.

[0120] S206. Calculate the impedance difference between cells in the battery pack. When the impedance difference reaches a preset threshold, start the impedance balancing operation.

[0121] Specifically, the difference between the target aging impedance and the minimum aging impedance of each battery cell is calculated using the formula ΔR = R_{aging-target} - R_{aging-min}. This quantifies the degree of aging inconsistency between cells; a larger ΔR indicates a higher degree of aging compared to the healthy cells in the battery pack, requiring more targeted balancing. When the impedance difference between any two cells reaches a preset threshold, for example, ΔR > 5%, the charger's balancing control module immediately activates the active balancing mechanism.

[0122] Based on the above ΔR value, using formula Q bal =k·ΔR·SOC max Calculate the required equalization capacity for each battery cell.

[0123] S207 performs temperature control monitoring and executes corresponding charging adjustment or protection operations based on temperature parameters.

[0124] Specifically, a multi-layered safety protection and anomaly diagnosis mechanism is set up. For example, the first-level temperature control protection reduces the current by 20% every 3 seconds when T>45℃ for 3 seconds; the second-level temperature control protection cuts off the main relay within 100ms when T>50℃; anomaly diagnosis refers to checking the current every minute R... ct When the rate of change exceeds 10% or the individual cell voltage is abnormal (>4.2V / <2.8V), current reduction or charging is triggered. Through these charging adjustment or protection actions, a closed-loop charging process of aging detection, strategy matching, real-time optimization, and safety protection can be achieved.

[0125] S208, S201-S207 are executed in a loop until charging is complete.

[0126] Specifically, the charging completion process can be initiated if any of the following conditions are met:

[0127] Normal full charge termination: When the battery SOC (State of Charge) is ≥95%, it means that the battery is close to full capacity and has reached the normal charging target, so there is no need to continue charging.

[0128] Fault recovery termination: If the current reduction / charging stop was triggered by a fault (such as over-temperature or abnormal voltage), the system can end the process once the fault is cleared and there is no need to maintain charging.

[0129] User-initiated termination: The user actively stops charging by operating the vehicle or charger (such as unplugging the charging gun or clicking the stop button), and the system terminates the charging in response to the user's command.

[0130] When any of the above conditions are met, the electronic device can issue a shutdown command through the charger's main control unit. Specific actions may include: cutting off the main charging circuit, stopping EIS monitoring and equalization control, etc., ultimately bringing the entire charging process to a complete end and ensuring the safety of the device and battery.

[0131] In summary, compared with the prior art, this embodiment has the following advantages:

[0132] 1. Significantly extends battery cycle life: Specifically designed for end-of-life batteries (e.g., R...). ct (With an increase of >50%), the current-limited constant-voltage charging strategy can effectively suppress lithium dendrite growth, reduce irreversible chemical reactions inside the battery, significantly increase the number of battery cycles in electric vehicles, energy storage devices and other scenarios, and extend the overall service life of the battery.

[0133] 2. Comprehensive improvement of fast charging safety performance: By monitoring the battery aging status in real time, the system automatically triggers current limiting and temperature control protection for high-risk aging batteries, and controls the battery temperature rise during charging (e.g., below 45℃). Compared with the temperature rise peak of the traditional solution (75℃), the safety is significantly improved. At the same time, based on multi-cell impedance balancing technology, it reduces the problem of local overheating caused by inconsistent cell aging, greatly reduces the risk of thermal runaway, and ensures safety throughout the charging process.

[0134] 3. Dynamically optimize overall charging efficiency: Adapt differentiated charging strategies for batteries with different aging levels. New batteries (early aging) adopt a stepped current ramp mode (0.5C→2C) to minimize fast charging time. Reinforcement learning optimization dynamically balances the three goals of charging speed, battery life and temperature rise control to avoid ineffective energy consumption and improve overall charging efficiency.

[0135] 4. Effectively reduce operation and maintenance costs: By using real-time EIS online diagnostic technology during charging, the traditional method of manually disassembling and inspecting battery aging status is replaced, significantly improving the efficiency of operation and maintenance inspection; the extension of battery life and the reduction of failure risk not only reduce the frequency of electric vehicle battery replacement, but also reduce the equipment maintenance costs in scenarios such as energy storage power stations, optimizing the economics of the entire life cycle.

[0136] 5. Possesses broad hardware compatibility: It can directly reuse mature power monitoring modules such as TI BQ27411-G1 without additional hardware architecture reconstruction, reducing adaptation costs; at the same time, it supports impedance model migration for different types of batteries such as lithium-ion batteries and solid-state batteries, adapting to various high-power battery application scenarios, and has strong technical scalability and feasibility for implementation.

[0137] In summary, this invention, through its core logic of real-time aging perception, dynamic strategy matching, multi-objective optimization, and end-to-end safety protection, not only solves the pain points of traditional charging solutions that ignore battery aging, such as shortened lifespan and safety hazards, but also achieves comprehensive optimization of charging efficiency, maintenance costs, and hardware compatibility, providing a more efficient, safe, and economical charging solution for high-power battery application scenarios.

[0138] A third embodiment of the present invention also provides an electronic device, the electronic device comprising:

[0139] At least one processor; and,

[0140] The memory is communicatively connected to the at least one processor; wherein,

[0141] The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the battery aging-based dynamic impedance matching charging method of any of the foregoing embodiments.

[0142] The fourth embodiment of the present invention also provides a non-transitory computer-readable storage medium storing computer instructions for causing the computer to execute the dynamic impedance matching charging method based on battery aging as described in any of the foregoing embodiments.

[0143] The fifth embodiment of the present invention also provides a computer program product, which includes a computing program stored on a non-transitory computer-readable storage medium. The computer program includes program instructions that, when executed by a computer, cause the computer to perform the dynamic impedance matching charging method based on battery aging of any of the foregoing embodiments.

[0144] The sixth embodiment of the present invention also provides a computer program, which includes program instructions that, when executed by a computer, cause the computer to perform the dynamic impedance matching charging method based on battery aging of any of the foregoing embodiments.

[0145] Figure 3 The diagram illustrates a method or device 1000 that can implement embodiments of the present invention. In some embodiments, it may include more or fewer devices than illustrated. In some embodiments, it may be implemented using a single or multiple devices. In some embodiments, it may be implemented using cloud-based or distributed devices.

[0146] like Figure 3As shown, device 1000 includes a processor 1001, which can perform various appropriate operations and processes based on programs and / or data stored in read-only memory (ROM) 1002 or programs and / or data loaded from storage portion 1008 into random access memory (RAM) 1003. Processor 1001 may be a multi-core processor or may contain multiple processors. In some embodiments, processor 1001 may include a general-purpose main processor and one or more special coprocessors, such as a central processing unit (CPU), graphics processing unit (GPU), neural network processor (NPU), digital signal processor (DSP), etc. Random access memory 1003 also stores various programs and data required for the operation of device 1000. Processor 1001, read-only memory 1002, and random access memory 1003 are interconnected via bus 1004. Input / output (I / O) interface 1005 is also connected to bus 1004.

[0147] The processor and memory described above are used together to execute programs stored in the memory. When the program is executed by a computer, it can implement the methods, steps, or functions described in the above embodiments.

[0148] The following components are connected to I / O interface 1005: an input section 1006 including a keyboard, mouse, touchscreen, etc.; an output section 1007 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and speakers, etc.; a storage section 1008 including a hard disk, etc.; and a communication section 1009 including a network interface card such as a LAN card, modem, etc. The communication section 1009 performs communication processing via a network such as the Internet. A drive 1010 is also connected to I / O interface 1005 as needed. A removable medium 1011, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on drive 1010 as needed so that computer programs read from it can be installed into storage section 1008 as needed. Figure 3 The diagram only shows a portion of the components and does not imply that the device 1000 only includes... Figure 3 The components shown.

[0149] The systems, devices, modules, or units described in the above embodiments can be implemented by a computer or its associated components. The computer may be, for example, a mobile terminal, smartphone, personal computer, laptop computer, in-vehicle human-machine interface device, personal digital assistant, media player, navigation device, game console, tablet computer, wearable device, smart TV, Internet of Things system, smart home, industrial computer, server, or a combination thereof.

[0150] Although not shown, in this embodiment of the invention, a computer-readable storage medium is provided having a computer program / instruction stored thereon, which, when executed by a processor, implements the dynamic impedance matching charging method based on battery aging as described in Embodiment 2.

[0151] Storage media in embodiments of the present invention include articles that are permanent and non-permanent, removable and non-removable, capable of storing information by any method or technology. Examples of storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transfer medium that can be used to store information accessible by a computing device.

[0152] Although not shown, embodiments of the present invention also provide a computer program product, including: a computer program / instructions that, when executed by a processor, implement the dynamic impedance matching charging method based on battery aging described in Embodiment 1.

[0153] The methods, programs, systems, apparatuses, etc., in embodiments of the present invention can be executed or implemented in one or more networked computers, or practiced in a distributed computing environment. In the embodiments of this specification, in these distributed computing environments, tasks can be performed by remote processing devices connected via a communication network.

[0154] Those skilled in the art will understand that the embodiments described in this specification can be provided as methods, systems, or computer program products. Therefore, those skilled in the art will realize that the functional modules / units or controllers and related method steps described in the above embodiments can be implemented in software, hardware, or a combination of both.

[0155] Unless explicitly stated otherwise, the actions or steps of the methods and procedures described in the embodiments of the present invention do not necessarily have to be performed in a specific order and can still achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0156] This document describes several embodiments of the present invention; however, for the sake of brevity, the descriptions of the embodiments are not exhaustive, and identical or similar features or parts between the embodiments may be omitted. In this document, "one embodiment," "some embodiments," "example," "specific example," or "some examples" refers to embodiments applicable to at least one, but not all, of the present invention. The above terms do not necessarily refer to the same embodiments or examples. Without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described herein, as well as the features of the different embodiments or examples.

[0157] The exemplary systems and methods of the present invention have been specifically shown and described with reference to the above embodiments, which are merely examples of the best mode for implementing the systems and methods. Those skilled in the art will understand that various changes can be made to the embodiments of the systems and methods described herein without departing from the spirit and scope of the invention as defined in the appended claims when implementing the systems and / or methods.

Claims

1. A dynamic impedance matching charger based on battery aging, characterized in that, The dynamic impedance matching charger includes: a battery status monitoring module, an aging level assessment module, a dynamic charging control module, a cell balancing module, an optimization control module, an anti-interference module, and a main control unit. The main control unit is electrically connected to the battery status monitoring module, the aging level assessment module, the dynamic charging control module, the cell balancing module, the optimization control module, and the anti-interference module, respectively, to collaboratively achieve closed-loop control of battery charging; The battery status monitoring module is used to inject an AC signal of a preset frequency into the battery pack and output the battery's impedance parameters, state of charge, and temperature parameters. The aging level assessment module is used to receive the impedance parameter, combine it with a preset aging assessment model, and output the battery aging level. The dynamic charging control module is used to match and output an appropriate charging mode command according to the battery aging level. The cell balancing module is used to receive the impedance parameters, calculate the impedance difference between cells, and generate and execute a balancing operation command when the impedance difference reaches a preset threshold. The optimization control module is used to construct a multi-dimensional state vector including battery impedance increment, temperature change rate, current margin and basic state parameters, and normalize the multi-dimensional state vector. It then uses a delayed update deep reinforcement learning algorithm that integrates historical monitoring data feature extraction mechanism to optimize the charging command in combination with a preset multi-dimensional reward function, and sends the optimized charging command to the dynamic charging control module. The anti-interference module is used to improve the signal-to-noise ratio of the signal collected by the battery status monitoring module to a preset standard value through a preset isolation and shielding design.

2. The dynamic impedance matching charger based on battery aging according to claim 1, characterized in that, The frequency range of the AC signal is 0.1–1000Hz, and the impedance parameters include: solid electrolyte interface film resistance and charge transfer resistance.

3. The dynamic impedance matching charger based on battery aging according to claim 1, characterized in that, The aging assessment model is a multi-stage model, which classifies aging levels based on the increase in charge transfer resistance. The dynamic charging control module dynamically switches the charging mode according to the aging level, and matches and outputs the appropriate charging mode command.

4. The dynamic impedance matching charger based on battery aging according to claim 1, characterized in that, The impedance difference is the difference between the target cell aging impedance and the minimum aging cell impedance. The equalization operation command is calculated and generated based on the impedance difference and the current maximum single-cell state of charge of the battery.

5. The dynamic impedance matching charger based on battery aging according to claim 1, characterized in that, The delayed-update deep reinforcement learning algorithm, which utilizes a feature extraction mechanism based on fused historical monitoring data, optimizes charging instructions by combining a preset multi-dimensional reward function, including: A reinforcement learning model is constructed by using a dual-delay deep deterministic policy gradient algorithm to extract subnets based on temporal features. The reinforcement learning model is trained based on a multi-dimensional reward function that considers charging efficiency, battery aging suppression, temperature safety, parameter stability, and consistency of lifetime prediction. A quantization strategy is adopted for the trained reinforcement learning model, and online inference is achieved through a cache alternation mechanism; If the battery temperature exceeds the preset threshold or the single cell voltage exceeds the safety threshold, the hardware comparator will directly cut off the main relay within microseconds to achieve an emergency interruption of the charging main circuit.

6. The dynamic impedance matching charger based on battery aging according to claim 1, characterized in that, The anti-interference module includes: adding an isolation power supply at the power supply end of the current sampling circuit, and wrapping a μ-level metal shielding layer around the outer layer of the current sampling cable and the outer shell of the sampling module.

7. A dynamic impedance matching charging method based on battery aging, characterized in that, The method applicable to the dynamic impedance matching charger based on battery aging as described in any one of claims 1-6 includes: S1: Receives a charging request and initiates the charging process; S2: Inject an AC signal of a preset frequency into the battery pack and collect the battery's impedance parameters, state of charge, and temperature parameters. S3: Determine the battery aging level based on the impedance parameters; S4: Match the appropriate charging mode according to the battery aging level; S5: The charging command is optimized by using a delayed update deep reinforcement learning algorithm that integrates the feature extraction mechanism of historical monitoring data and a preset multi-dimensional reward function, and the optimized charging command is sent to the dynamic charging control module. S6: Calculate the impedance difference between cells in the battery pack. When the impedance difference reaches a preset threshold, start the impedance equalization operation. S7: Perform temperature control monitoring and execute corresponding charging adjustment or protection operations based on the temperature parameters; S8: Cycle through S2-S7 until charging is complete.

8. The dynamic impedance matching charging method based on battery aging according to claim 7, characterized in that, Step S7 also includes an anomaly diagnosis operation, which triggers a current reduction or charging stop operation when the rate of change of the impedance parameter per unit time exceeds a preset rate of change threshold, or when the voltage of a single cell exceeds a preset voltage range.

9. An electronic device, characterized in that, The electronic device includes: At least one processor; and a memory communicatively connected to said at least one processor; The memory stores instructions that can be executed by the at least one processor, which, when executed by the at least one processor, enables the at least one processor to perform the dynamic impedance matching charging method based on battery aging as described in any one of claims 7 to 8.

10. A non-transitory computer-readable storage medium, characterized in that, The non-transitory computer-readable storage medium stores computer instructions for causing the computer to execute the dynamic impedance matching charging method based on battery aging as described in any one of claims 7 to 8.

Images