A battery state-based wireless charging control method and system

Abstract

The application discloses a wireless charging control method and system based on battery state, relates to the technical field of battery state monitoring and wireless charging control, and comprises the following steps: acquiring a capacity increment curve peak position sequence, an internal resistance change rate sequence and a terminal voltage drop amplitude value sequence of previous charging of a battery as an aging multi-dimensional feature trajectory; predicting a capacity attenuation trajectory after a current aging stage; calculating an optimal charging acceptance capacity according to the predicted capacity attenuation trajectory, converting the optimal charging acceptance capacity into a current-time control curve and a cutoff voltage value; issuing the control curve and the cutoff voltage value to a transmitting end controller through a bidirectional communication link in a wireless charging process, dynamically adjusting output power by the transmitting end controller, synchronously adjusting rectification parameters and voltage stabilization parameters by a receiving end controller; and appending actual data to a historical database and updating prediction model parameters after the charging is completed. The application can realize forward-looking dynamic compensation of a charging strategy and delay battery capacity attenuation.

Description

Technical Field

[0001] This invention relates to the field of battery state monitoring and wireless charging control technology, specifically to a wireless charging control method and system based on battery state. Background Technology

[0002] Wireless charging technology has been widely used in smartphones, wearable devices, and electric vehicles due to its advantages such as eliminating the need for physical connections and ease of use. Lithium-ion batteries inevitably age during repeated charge-discharge cycles, resulting in capacity decay and increased internal resistance, directly impacting charging efficiency and safety. Optimizing wireless charging strategies based on battery condition throughout the battery's entire lifespan remains a key technical challenge for the industry.

[0003] In the prior art, invention patent publication number CN112018906B proposes an adaptive wireless charging system for the entire life cycle of a lithium battery. This system plans the wireless charging method throughout its entire life cycle based on the battery's health status, employing three different charging modes at different life cycle stages. This solution reduces the risk of late-life battery failure to some extent by dividing the battery life into several stages and pre-setting corresponding charging modes for each stage. However, this solution only performs discrete mode switching when the battery health status falls within the preset range, which is a "reactive response after the fact" and cannot achieve continuous and smooth adjustment of the charging strategy during the battery aging process. Another invention patent publication number CN121965924A discloses a battery control method, device, electronic device, and storage medium. This method acquires features affecting the battery aging status corresponding to multiple cycles during terminal use, determines a third feature characterizing the battery aging trend based on the aging status corresponding to each cycle, and controls the battery based on target parameters determined by the third feature. This solution involves determining the aging trend, but its target parameters are battery parameters corresponding to future cycles, and adjustments are made on a cycle-by-cycle basis, i.e., evaluation after one charge and adjustment at the next charge. The solution lacks the ability to perform real-time dynamic compensation based on aging trend predictions during the same wireless charging process.

[0004] In summary, existing wireless charging control methods have the following shortcomings: First, switching charging modes based on discrete thresholds of battery health status results in strategy adjustments lagging behind the aging process, failing to achieve continuous dynamic compensation. Second, parameter adjustments are made on a per-charging-cycle basis, lacking the ability to optimize the charging strategy in real-time based on aging trajectory prediction during the same charging cycle. These shortcomings cause the charging strategy during the battery's aging transition phase to consistently lag behind the aging process, accelerating battery capacity degradation. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of the prior art and provide a wireless charging control method and system based on battery state. By acquiring battery charging data from previous charging cycles to construct an aging trajectory prediction model, the aging trend is predicted in real time during a single wireless charging process, and charging parameters are dynamically adjusted to achieve forward-looking continuous compensation control of the charging strategy, thereby delaying battery capacity decay.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In one aspect, a wireless charging control method based on battery state, the method comprising the following steps:

[0007] S1. Obtain the peak position sequence of the capacity increment curve, the internal resistance change rate sequence, and the voltage drop value sequence at the end of charging during each complete charging process of the battery, as a multi-dimensional characteristic trajectory of battery aging.

[0008] S2, Based on the aging multidimensional feature trajectory, construct an aging trajectory prediction model that combines a double exponential model with Kalman filtering to predict the capacity decay trajectory after a preset number of charging cycles following the current aging stage.

[0009] S3, calculate the optimal charging acceptance capability of the current aging stage based on the predicted capacity decay trajectory, and convert the optimal charging acceptance capability into the current-time control curve and cutoff voltage value of the wireless charging system.

[0010] S4. During this wireless charging process, the current-time control curve and cutoff voltage value are sent to the transmitter controller through the bidirectional communication link of the wireless charging system. The transmitter controller dynamically adjusts the output power according to the current-time control curve, and the receiver controller synchronously adjusts the rectification parameters and voltage regulation parameters.

[0011] S5, after the charging is completed, the actual data of the charging is added to the historical database and the parameters of the aging trajectory prediction model are updated;

[0012] S2-S5 are executed continuously within the current charging cycle, enabling the charging strategy to respond in real time to changes in the battery aging trajectory during the current charging process.

[0013] Furthermore, in S1, the sequence of peak positions of the capacity increment curve is obtained in the following way:

[0014] During the constant current charging phase, the battery terminal voltage and cumulative charge were recorded at fixed time intervals, and plotted. Curve extraction The voltage and peak amplitude corresponding to the peak positions of the curve are arranged according to the charging cycle number to obtain the peak position sequence of the capacity increment curve.

[0015] The drift direction and drift rate of the voltage value corresponding to the peak position are used to characterize the evolution trend of battery polarization.

[0016] Furthermore, in S1, the sequence of internal resistance change rates is obtained in the following manner:

[0017] The battery terminal voltage is measured and the first DC internal resistance is calculated during the first 30 seconds of the constant current charging start phase. The battery terminal voltage is measured and the second DC internal resistance is calculated during the last 30 seconds of the constant voltage charging end phase. The rate of change of the first DC internal resistance and the second DC internal resistance with the number of charging cycles is recorded to obtain the internal resistance change rate sequence.

[0018] The internal resistance change rate sequence is used to characterize the rate of degradation of battery conductivity as it ages.

[0019] Furthermore, in S2, the double-exponential model is:

[0020]

[0021] in, Let this be the remaining battery capacity after the nth charging cycle. This is the initial capacity of the battery. , , , Here are the model parameters, where n is the charging cycle number and exp is the natural exponential function.

[0022] The first exponential term in the double exponential model The second exponential term is used to fit the rapid capacity decay stage in the early stages of battery aging. Used to fit the slow capacity decay stage in the later stages of battery aging.

[0023] Furthermore, in S2, the aging trajectory prediction model, which integrates the dual exponential model and the Kalman filter, is constructed, specifically including:

[0024] Using the peak amplitude of the capacity increment curves from the first 20 charging cycles, the initial parameters of the double exponential model are determined by least squares fitting. , , , ;

[0025] Using a double exponential model as the state transition equation, and taking the peak amplitude of the capacity increment and the internal resistance value from each newly acquired charging data as observations, the model parameters are adjusted using a Kalman filter. , , , Perform real-time updates;

[0026] At the start of each charge, utilize the currently updated model parameters. , , , Predict the capacity decay trajectory over the next N charging cycles. , Where N is a preset positive integer;

[0027] The Kalman filter recursively corrects the model parameters in real time based on the residual between the actual observation and the model prediction after each charge, so that the predicted trajectory gradually converges to the real degradation path as the battery ages.

[0028] Furthermore, in S3, the optimal charge acceptance capability for the current aging stage is calculated based on the predicted capacity decay trajectory, specifically including:

[0029] A mapping table between battery remaining capacity and maximum acceptable charging rate was established through pre-experimental calibration.

[0030] The predicted capacity decay trajectory Substitute each future capacity value into the mapping table for interpolation to obtain the corresponding optimal charging rate. ;

[0031] Optimal charging rate The sequence is converted into the current-time control curve of this charging process. ;

[0032] The cutoff voltage compensation value is calculated based on the predicted capacity decay trajectory. When the battery is predicted to enter the accelerated aging stage, the cutoff voltage value for that charge is reduced.

[0033] The current-time control curve The charging process continuously changes over time, ensuring that the actual charging rate received by the battery always follows the optimal charging rate obtained by mapping the predicted aging trajectory.

[0034] Furthermore, in S3, the optimal charging rate is... The sequence is converted into the current-time control curve of this charging process. Specifically, it includes:

[0035] Based on model parameters and Estimate the effective duration of the current aging phase;

[0036] The aging process of the next N charging cycles is compressed and mapped onto the timeline of this charging cycle;

[0037] Generate a current curve that varies continuously with charging time. The current curve The first current value is used in the initial stage of charging, and it continuously decreases to the second current value in the later stage of charging as the predicted aging degree increases. The first current value is greater than the second current value.

[0038] Current curve Deceleration rate and model parameters and The aging acceleration is positively correlated; the greater the aging acceleration, the faster the current decrease rate during the transition from the initial stage of charging to the later stage of charging.

[0039] Furthermore, in step S4, the current-time control curve and cutoff voltage value are transmitted to the transmitter controller via the bidirectional communication link of the wireless charging system, specifically including:

[0040] The receiver controller will use the current-time control curve. The cutoff voltage value is sent to the transmitter controller via the wireless charging communication protocol;

[0041] The transmitter controller adjusts its parameters in 100ms cycles based on the received current-time control curve. The inverter's PWM duty cycle or operating frequency is adjusted in real time so that the output power changes with time according to the current-time control curve.

[0042] The receiver controller synchronously adjusts the output voltage setpoint of the rectifier circuit to ensure that the actual charging current received by the battery follows the current-time control curve. ;

[0043] The transmitter controller and receiver controller form a closed-loop control structure through a bidirectional communication link. While the transmitter controller adjusts the output power, the receiver controller feeds back the measured charging current to the transmitter controller to correct the power output value in the next adjustment cycle.

[0044] Furthermore, in step S5, updating the parameters of the aging trajectory prediction model specifically includes:

[0045] The peak position of the actual capacity increment curve, the internal resistance value, and the final voltage drop value of this charge are added to the historical database.

[0046] The actual capacity decay during this charge is compared with the predicted value of the aging trajectory prediction model to calculate the prediction error.

[0047] Update the model parameters of the double exponential model using Kalman filtering based on the prediction error. , , , ;

[0048] Save the updated model parameters to a non-volatile memory chip;

[0049] The updated model parameters are used as the initial model parameters for S2 at the start of the next charging cycle, forming an iterative learning closed loop across charging cycles.

[0050] On the other hand, a wireless charging control system based on battery state is applicable to a wireless charging control method based on battery state, the system comprising:

[0051] The wireless charging receiver includes a receiving coil, a rectifier circuit, a voltage regulator circuit, a receiver controller, a non-volatile memory chip, a voltage sampling circuit, a current sampling circuit, and a temperature sensor.

[0052] The wireless charging transmitter includes a transmitting coil, an inverter circuit, a transmitter controller, and a communication module. The transmitter controller and the receiver controller are connected via a bidirectional communication link.

[0053] The non-volatile memory chip is used to store the peak position sequence of the capacity increment curve, the internal resistance change rate sequence, and the voltage drop value sequence at the end of charging during each complete charge of the battery.

[0054] The receiver controller is used to construct an aging trajectory prediction model that combines a dual exponential model with a Kalman filter based on the peak position sequence of the capacity increment curve, the internal resistance change rate sequence, and the voltage drop value sequence at the end of charging. It predicts the capacity decay trajectory after the current aging stage for a preset number of charging cycles, calculates the current-time control curve and the cutoff voltage value based on the predicted capacity decay trajectory, and sends the current-time control curve and the cutoff voltage value to the transmitter controller through a bidirectional communication link.

[0055] The transmitter controller is used to dynamically adjust the output power of the inverter circuit according to the current-time control curve.

[0056] The receiver controller is also used to synchronously adjust the rectification parameters of the rectifier circuit and the voltage regulation parameters of the voltage regulator circuit, and after the charging is completed, to append the actual data of the charging to the non-volatile memory chip and update the parameters of the aging trajectory prediction model.

[0057] The receiver controller and transmitter controller form a real-time closed-loop control structure during the charging process and an iterative learning structure across charging cycles through a bidirectional communication link.

[0058] Compared with existing technologies, this wireless charging control method and system based on battery state has the following advantages:

[0059] I. This invention constructs a multi-dimensional aging characteristic trajectory by acquiring the peak position of the capacity increment curve, the rate of change of internal resistance, and the magnitude of the voltage drop at the end of each battery charge. It establishes a prediction model that integrates a dual exponential model and a Kalman filter. During a single charge, the current-time control curve and the cutoff voltage value are dynamically adjusted according to the predicted aging trend. This allows the charging strategy to proactively follow the battery aging process and continuously compensate, thereby solving the problem of the charging strategy lagging behind the battery aging process in the prior art. It can effectively delay battery capacity decay and improve the charging safety of the battery throughout its entire life cycle.

[0060] Second, this invention achieves collaborative closed-loop control between the transmitter and receiver through a two-way communication link in the wireless charging system. After each charging cycle, the actual data is fed back to the prediction model for parameter updates, so that the control accuracy gradually improves with the number of uses. At the same time, the sensing, calculation and control links in the charging process are integrated into an integrated architecture, eliminating the need for additional complex battery status monitoring equipment, thereby reducing the system hardware cost and improving the adaptability and reliability of the wireless charging system throughout the battery's entire life cycle.

[0061] Other advantages, objectives and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination or study, or may be learned from the practice of the invention. Attached Figure Description

[0062] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0063] Figure 1 This is a schematic diagram of the overall architecture of the system of the present invention;

[0064] Figure 2 This is an overall flowchart of the method of the present invention;

[0065] Figure 3 This is a schematic diagram of the aging trajectory prediction model update and closed-loop control process in this invention. Detailed Implementation

[0066] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0067] like Figure 1As shown, this invention provides a wireless charging control system based on battery status. The system mainly comprises two parts: a wireless charging receiver and a wireless charging transmitter. The wireless charging receiver is located inside the device to be charged or serves as an external accessory to the device, and is electrically connected to the battery to be charged. Specifically, the receiver includes a receiving coil, a rectifier circuit, a voltage regulator circuit, a receiver controller, a non-volatile memory chip, a voltage sampling circuit, a current sampling circuit, and a temperature sensor.

[0068] The receiving coil is used to couple and receive alternating electromagnetic energy from the transmitting end and output high-frequency alternating current.

[0069] The rectifier circuit is connected to the output of the receiving coil to convert high-frequency alternating current into pulsating direct current.

[0070] The voltage regulator circuit is connected to the output of the rectifier circuit to regulate the pulsating DC current to a DC voltage suitable for battery charging.

[0071] The voltage sampling circuit is connected in parallel across the positive and negative terminals of the battery to collect the battery terminal voltage in real time.

[0072] The current sampling circuit is connected in series in the charging circuit to collect the charging current in real time.

[0073] Temperature sensors are attached to the surface of the battery or placed inside the battery pack to collect battery temperature data.

[0074] The receiver controller is connected to the control terminals of the voltage sampling circuit, current sampling circuit, temperature sensor, non-volatile memory chip, and voltage regulator circuit, respectively, and is used to receive sampled data, execute control algorithms, and output control signals.

[0075] The non-volatile memory chip uses a serial Flash memory, model W25Q or similar, with a storage capacity of at least 2MB. This capacity is sufficient to store complete historical data for at least 500 charging cycles. The data recorded for each charging cycle includes the charging cycle number, peak voltage value of the capacity increment curve, peak amplitude, first DC internal resistance, second DC internal resistance, rate of change of internal resistance, final voltage drop, and double exponential model parameters. , , , Each record contains approximately 128 bytes of data, and the total data size for 500 records is approximately 64KB. The remaining storage space is used to store the mapping table and program code. When the number of stored historical data records exceeds 500, the oldest record is overwritten according to the first-in, first-out (FIFO) principle.

[0076] The wireless charging transmitter is located in the charging base or charging station and specifically includes a transmitting coil, an inverter circuit, a transmitter controller, and a communication module. The transmitting coil converts alternating electrical energy into an alternating electromagnetic field for outward radiation. The output of the inverter circuit is connected to the transmitting coil, converting the DC input power into high-frequency AC power to supply the transmitting coil. The transmitter controller is connected to the control terminal of the inverter circuit, used to adjust the operating parameters of the inverter circuit. The communication module is connected to the transmitter controller, enabling bidirectional data interaction with the receiver. The transmitter controller and receiver controller are connected via a bidirectional communication link. This bidirectional communication link can be implemented based on existing communication channels in wireless charging protocols, such as the communication link established through load modulation in the Qi standard, or based on an additional Bluetooth Low Energy communication module or near-field communication module. Through this bidirectional communication link, the receiver controller can send control parameters to the transmitter controller, and the transmitter controller can also feed back operating status information to the receiver controller. The receiver controller uses a Cortex-M4 core microcontroller, specifically the STM32F407, with a clock frequency of 100MHz and a built-in floating-point unit, capable of meeting the computational requirements for double exponential model parameter fitting and Kalman filter matrix operations. The transmitter controller can use a microcontroller of the same model or with equivalent performance.

[0077] like Figure 2 As shown, the method includes the following steps.

[0078] Step S1: Obtain the multi-dimensional characteristic trajectory of battery aging. After each complete charging process, the receiver controller collects data from the entire charging process through the voltage sampling circuit, current sampling circuit, and temperature sensor, and extracts three key aging characteristic parameters: the peak position sequence of the capacity increment curve, the internal resistance change rate sequence, and the voltage drop value sequence at the end of charging. These three sequences are used together as the multi-dimensional characteristic trajectory of battery aging.

[0079] The sequence of peak positions of the capacity increment curve is obtained in the following way.

[0080] During the constant current charging phase, the receiver controller records the battery terminal voltage and cumulative charge at fixed time intervals. This fixed time interval ranges from 1 second to 30 seconds; a shorter interval allows for more data points to be collected, resulting in a higher resolution capacity increment curve, but also increases storage capacity requirements. As a specific example in this embodiment, the time interval is set to 10 seconds. For each data point recorded, the controller calculates the capacity increment at that moment. With voltage increment ratio Taking constant current charging as an example, the charging current... For a constant value, the change in capacity ,but ,in The sampling time interval, This represents the change in battery terminal voltage during this time interval. The calculated value... By plotting the value against the corresponding voltage value V, the capacity increment curve can be obtained. The curve represents the battery terminal voltage on the horizontal axis and the ratio of capacity increment to voltage increment on the vertical axis. This curve typically exhibits multiple peaks, each corresponding to a phase transition plateau in an electrochemical reaction within the battery. The main peak corresponds to the phase transition reaction of the battery's positive electrode material. The position, voltage, and amplitude of the main peak exhibit a regular change with battery aging. The controller extracts the voltage and amplitude corresponding to the main peak using the following method: Find the point with the maximum vertical coordinate on the curve, record the horizontal coordinate (voltage value) of that point as the peak position, and record the vertical coordinate value as the peak amplitude. The voltage value corresponding to the peak position of the capacity increment curve will shift as the battery ages, and the peak amplitude will gradually decrease as the battery capacity decays. Arrange the voltage values ​​and peak amplitudes corresponding to the peak positions of each charge according to the charging cycle number to obtain the peak position sequence of the capacity increment curve. The drift direction and drift rate of the peak voltage in this sequence can be used to characterize the evolution trend of the battery polarization, and the decay rate of the peak amplitude can be used to characterize the process of loss of battery active material or active lithium.

[0081] The internal resistance change rate sequence was obtained in the following way.

[0082] During each charging process, the receiver controller measures the battery's DC internal resistance in two specific time periods. The first time period is the first 30 seconds of the constant current charging initial stage. At this time, battery polarization has not yet been fully established, and the measured terminal voltage mainly reflects the influence of ohmic internal resistance. The controller records the battery terminal voltage and charging current at the beginning and end of this period and calculates the first DC internal resistance using Ohm's law. The second time period is the last 30 seconds of the constant voltage charging termination stage. At this time, the charging current has dropped to a lower level, and the battery is close to full charge. The measured terminal voltage better reflects the battery's internal resistance characteristics at the end of the charging process. The controller records the battery terminal voltage and charging current at the beginning and end of this period and calculates the second DC internal resistance. The first and second DC internal resistances of each charge are arranged according to the charging cycle number, and the rate of change of internal resistance between adjacent cycles is calculated to obtain an internal resistance change rate sequence. This sequence can be used to characterize the rate of degradation of battery conductivity with aging. An increase in the rate of change of internal resistance indicates that more irreversible structural changes have occurred inside the battery.

[0083] The sequence of voltage drop values ​​at the end of charging is obtained in the following way.

[0084] During the constant-voltage charging phase, the controller records the battery terminal voltage at fixed time intervals from the start of constant-voltage charging until the charging current drops to the cutoff current. The controller records the battery terminal voltage change trajectory throughout the entire process, from the start of constant-voltage charging to the drop in charging current to the cutoff current, and extracts the voltage drop amplitude during this process—the difference between the voltage value at the start of constant-voltage charging and the voltage value at the end of charging—as the final voltage drop value. The final voltage drop values ​​for each charging cycle are arranged according to the charging cycle number to obtain a final voltage drop value sequence. This sequence reflects the accumulation and release characteristics of the battery polarization voltage during the constant-voltage charging phase, and its trend can be used to help determine the battery's aging state.

[0085] Step S2: Construct an aging trajectory prediction model and predict the capacity decay trajectory. Based on the aging multidimensional feature trajectory obtained in step S1, the receiver controller constructs an aging trajectory prediction model that fuses a dual exponential model with a Kalman filter. This model is then used to predict the capacity decay trajectory after a preset number of charging cycles following the current aging stage.

[0086] The double exponential model takes the form of:

[0087]

[0088] in, Let this be the remaining battery capacity after the nth charging cycle. This is the initial capacity of the battery. and All are expressed as a percentage of the battery's rated capacity. For example... Corresponding to the rated capacity value of the battery at the time of manufacture, This indicates that after n charging cycles, the battery's remaining capacity is 88% of its rated capacity. (Parameter) and This is the capacity attenuation amplitude parameter, in units of... Same, both are percentages. Parameters and Here, is the decay rate parameter, measured in charge cycles. n is the charge cycle number, and exp is the natural exponential function. This model can fit well the typical trajectory of lithium-ion battery capacity decay, which is initially rapid and then slows down. The first exponential term in the model... The second exponential term is used to fit the rapid capacity decay stage in the early stages of battery aging. Used to fit the slow capacity decay stage in the later stages of battery aging. Parameters and The magnitudes of the two exponential terms are controlled separately, and the parameters are... and The decay rates of the two exponential terms are controlled separately.

[0089] The specific process of constructing an aging trajectory prediction model that integrates a dual exponential model and a Kalman filter is as follows.

[0090] The model parameters are initialized using the peak amplitudes of the capacity increment curves from the first 20 charging cycles. A strong positive correlation exists between the peak amplitudes of the capacity increment curves and the remaining battery capacity; therefore, the peak amplitude sequence can be used as a proxy for the remaining capacity. The receiver controller reads the peak amplitudes of the capacity increment curves from the non-volatile memory chip and determines the initial parameters of the bi-exponential model using least-squares fitting. , , , The goal of least squares fitting is to minimize the sum of squared errors between the model-predicted capacity decay trajectory and the actual observed peak magnitude decay trajectory.

[0091] The double exponential model is used as the state transition equation of the Kalman filter. The peak amplitude of the capacity increment and the internal resistance value in each newly acquired charging data are used as observation values. The model parameters are then adjusted by the Kalman filter. , , , Real-time updates are performed. The Kalman filter is a recursive state estimation algorithm that recursively calculates the current state estimate based on the state estimate from the previous time step and the current observation. In this invention, the state vector is... , where the subscript k represents the state after the k-th charging cycle. The state transition equation is described by a double exponential model, which describes the model parameters from the k-th cycle to the k-th cycle. The variation pattern of each cycle is described by the observation equation, which illustrates the relationship between the capacity decay value predicted by the current model parameters and the actual observed value. After each charge, the controller compares the peak amplitude of the actual capacity increment and the internal resistance value with the model prediction, calculates the residual, and uses this residual to correct the model parameters through the Kalman filter update equation.

[0092] The Kalman filter needs to set the initial covariance matrix of the state vector during its first run. In this embodiment, Set as one The diagonal matrix is ​​set to 100 for all diagonal elements, indicating significant uncertainty in the initial model parameter estimation. As the number of charging cycles increases, the Kalman filter continuously updates and corrects the covariance matrix, gradually reducing the uncertainty in state estimation. The process noise covariance matrix Q and the measurement noise covariance matrix R are calculated by performing at least 50 consecutive charge-discharge cycles on the same type of battery under laboratory conditions. The residual sequence between the model prediction and the actual observation is statistically analyzed, and the variance of the residuals is used as the diagonal element value of Q and R, while the off-diagonal elements are all set to 0. The Kalman filter updates the model parameters in real time based on the residual between the actual observation and the model prediction after each charge, recursively correcting the predicted trajectory to gradually converge to the actual degradation path as the battery ages.

[0093] At the start of each charge, the receiver controller uses the currently updated model parameters. , , , Predict the capacity decay trajectory over the next N charging cycles. , The value of N is set based on the expected frequency of battery use and the rate of aging. In electric vehicle applications, the battery is charged an average of approximately 8 to 10 times per month, and the setting is... This corresponds to an aging prediction window of approximately 5 to 6 months, which is sufficient to cover short-term cycles of battery aging trends. In consumer electronics applications, batteries are charged 1 to 2 times daily, and the settings are... This corresponds to a prediction window of approximately 15 to 30 days, matching the typical cycle for battery aging assessment of this type of device. The value of N is also limited by the storage capacity of the non-volatile memory chip; a larger storage capacity allows for the storage of more historical data records, and the upper limit of N is determined by the storage chip capacity.

[0094] Step S3: Calculate the optimal charging acceptance capability and convert it into a control curve. Based on the capacity decay trajectory predicted in step S2, the receiver controller calculates the optimal charging acceptance capability for the current aging stage and converts the optimal charging acceptance capability into the current-time control curve and cutoff voltage value of the wireless charging system.

[0095] A mapping table between battery remaining capacity and maximum acceptable charging rate is established through pre-experimental calibration. This mapping table is generated using laboratory aging test data of the same type of battery and pre-stored in a designated storage area of ​​a non-volatile memory chip before the wireless charging receiver device containing the control system of this invention leaves the factory. The method for establishing the mapping table is as follows:

[0096] Several fresh batteries of the same model as the batteries to be recharged were selected and subjected to accelerated aging tests at standard ambient temperature. The batteries were then subjected to incremental charging rate tests at different aging stages. At each aging stage, the charging rate was gradually increased from a lower rate, while simultaneously monitoring changes in battery terminal voltage, temperature, and capacity increment curves. When the critical characteristic of lithium plating initiation was detected, the critical charging rate at that aging stage was recorded as the maximum acceptable charging rate. This critical characteristic manifests as: the appearance of a new shoulder peak or small peak to the left of the main peak on the capacity increment curve, or in the charging voltage differential curve... A negative inflection point appears. The appearance of the above characteristic indicates that lithium metal deposition has begun on the surface of the battery's negative electrode, i.e., the lithium plating reaction has started. Lithium plating detection is achieved by performing second derivative analysis on the real-time acquired capacity increment curve. When the second derivative shows a sign reversal within a specific voltage range, it is determined to be the starting point of lithium plating. A mapping table is obtained by creating a table of the remaining capacity values ​​at each aging level and their corresponding maximum acceptable charging rates. The preset mapping table contains multiple calibration points in 1% increments within the range of 60% to 100% of the battery's remaining capacity, with each calibration point corresponding to a maximum acceptable charging rate value. The mapping table does not change with the number of charging cycles during equipment use and serves as a fixed reference table for interpolation calculations in step S3.

[0097] The predicted capacity decay trajectory Substitute each future capacity value into the mapping table for interpolation to obtain the corresponding optimal charging rate. Interpolation methods can employ linear interpolation or spline interpolation to ensure the accuracy of the results. The sequence is smooth and continuous.

[0098] Optimal charging rate The sequence is converted into the current-time control curve of this charging process. The specific conversion method is as follows. Based on the parameters of the double exponential model... and Estimate the effective duration of the current aging phase. Parameters and It characterizes the rate of battery aging. The larger the value, the faster the initial aging process. A larger value indicates faster aging in the later stages. The effective duration refers to the time interval from the current moment until a significant change in the battery's aging characteristics occurs. The aging process of the next N charging cycles is compressed and mapped onto the timeline of the current charge. The specific rule for this compression mapping is: let the estimated total duration of this charge be... , timeline The time is divided into N equal intervals, each interval corresponding to a future charging cycle number n. Within the nth time interval, the current-time control curve... The value of is equal to the optimal charging rate of the future cycle number n corresponding to the center time of the interval. In this way, the aging process over the next N cycles is linearly compressed and mapped onto the N time intervals of this charge. The estimated total duration is then calculated. The charging time limit is determined by both the battery's current remaining capacity and the preset charging cutoff conditions. ,in This represents the current remaining capacity of the battery. This is the estimated average current for this charge. Finally, a current curve that continuously varies with the charging time is generated. The curve shows the first current value at the beginning of charging, and then continuously decreases to a second current value as the predicted aging progresses in the later stages of charging. The first current value is greater than the second current value. (Current Curve) Deceleration rate and model parameters and The aging acceleration is positively correlated; the greater the aging acceleration, the faster the current decrease rate during the transition from the initial stage of charging to the later stage of charging.

[0099] Simultaneously, the receiver controller calculates the cutoff voltage compensation value based on the predicted capacity decay trajectory. When it is predicted that the battery is about to enter the accelerated aging phase, the cutoff voltage value for this charge is appropriately reduced. The method for determining whether the battery has entered the accelerated aging phase is: calculating the predicted capacity decay trajectory. exist to If, within the range of second-order differences, there exist several consecutive points where the second-order difference is positive and its absolute value exceeds a preset threshold, then the battery is determined to enter the accelerated aging phase within that time period. In this embodiment, the preset threshold is set to (0.02%). 2 The calculation method for cutoff voltage compensation is as follows: based on the capacity decay rate of the next few cycles in the predicted trajectory, determine whether the battery is about to enter the accelerated aging stage. If so, subtract a compensation amount from the standard cutoff voltage. The size of the compensation amount is positively correlated with the predicted degree of accelerated aging.

[0100] Step S4: Control parameters are sent and coordinated control is executed via a two-way communication link. During this wireless charging process, the receiver controller transmits the current-time control curve generated in step S3 via the two-way communication link of the wireless charging system. The cutoff voltage value is sent to the transmitter controller.

[0101] The receiver controller will use the current-time control curve. The cutoff voltage value is encapsulated into a data packet according to the frame format of the wireless charging communication protocol and sent to the transmitter controller via a bidirectional communication link. After receiving the data packet, the transmitter controller parses the current-time control curve. and cutoff voltage value.

[0102] The transmitter controller adjusts its parameters in 100ms cycles based on the received current-time control curve. The inverter's PWM duty cycle or operating frequency is adjusted in real time to make the output power change with time according to the current-time control curve. Within each 100ms adjustment cycle, the transmitter controller adjusts the current-time control curve according to the current charging time t. Find the corresponding target current value, then calculate the required output power based on the target current value, and then adjust the PWM duty cycle or operating frequency to make the actual output power approach the target value.

[0103] The receiver controller synchronously adjusts the output voltage setpoint of the rectifier circuit to ensure that the actual charging current received by the battery follows the current-time control curve. The receiver controller monitors the actual charging current value collected by the current sampling circuit in real time and compares it with the current-time control curve. The current is compared with the target current value at the current moment. When the actual current deviates from the target value, the charging current is corrected by adjusting the output voltage setting of the voltage regulator circuit.

[0104] The transmitter controller and receiver controller form a closed-loop control structure via a bidirectional communication link. While the transmitter controller adjusts its output power, the receiver controller feeds back the measured charging current to the transmitter controller via the bidirectional communication link to correct the power output value for the next adjustment cycle. The feedback correction uses a PID control algorithm. Within each 100ms adjustment cycle, the transmitter controller calculates the measured charging current and the current-time control curve. Deviation between target current values The PID controller calculates the correction amount for the output power based on this deviation. The calculation formula is:

[0105]

[0106] in , , These are the proportional coefficient, integral coefficient, and derivative coefficient, respectively. In this embodiment, , , The corrected output power value equals the reference power value for the current adjustment cycle plus the correction amount. Integral term The cumulative time window is the first 10 adjustment cycles, i.e., the first 1000ms, to prevent integral saturation.

[0107] The receiver controller also monitors the battery terminal voltage in real time, and terminates the charging process when the battery terminal voltage reaches the cutoff voltage value.

[0108] Step S5: Update the parameters of the aging trajectory prediction model. After this charging cycle is completed, the receiving controller appends the actual data from this charging cycle to the historical database and updates the parameters of the aging trajectory prediction model.

[0109] The receiver controller appends the peak position, internal resistance value, and final voltage drop value of the actual capacity increment curve for this charge to the historical database in the non-volatile memory chip. The actual capacity decay for this charge is compared with the predicted value at the start of this charge by the aging trajectory prediction model in step S2, and the prediction error is calculated. This prediction error is then used to update the model parameters of the double exponential model through Kalman filtering. , , , The Kalman filter update process is the same as described in step S2, that is, the state vector is corrected using the residual between the observed values ​​and the model predictions after this charge. The updated model parameters are saved to a non-volatile memory chip as the initial model parameters for step S2 at the start of the next charge, forming an iterative learning closed loop across charging cycles.

[0110] The Kalman filtering process incorporates a convergence monitoring mechanism. After each filter update, the trace of the covariance matrix of the residuals is calculated. If this trace value is greater than 10 times the initial trace value after three consecutive updates, it is considered filter divergence. In this case, the model parameters are reverted to their values ​​before the previous update, and the current charging data is marked as abnormal data and not included in the model parameter update. If filter divergence occurs for five consecutive charging cycles, a model reset is triggered, restoring the model parameters to the factory-preset initial parameters. Simultaneously, the data from the most recent 10 charging cycles in the historical database is cleared, and parameter identification restarts.

[0111] The bidirectional communication link is equipped with a link status monitoring and retransmission mechanism. The transmitting and receiving controllers confirm the communication status every 100ms adjustment period. After sending control parameters, the receiving controller waits for an acknowledgment frame from the transmitting controller. If no acknowledgment frame is received within 50ms, the communication link is considered interrupted. At this time, the receiving controller activates a local emergency control mode and continues charging according to a preset safe charging curve. The current value of this safe charging curve is a current-time control curve. The current value at the moment of interruption is multiplied by 0.7 to ensure that the charging process is not completely interrupted. After the communication link is restored, the receiving controller resends the complete control curve parameters and synchronizes the current charging progress, restoring normal control mode.

[0112] like Figure 3As shown, the entire control process consists of two levels: an inner control loop and an outer learning loop. The inner control loop occurs within a single charging cycle: at the start of charging, the receiving controller predicts the aging trajectory and calculates the current-time control curve based on the current model parameters. This curve is then transmitted to the transmitting controller via a bidirectional communication link. The transmitting and receiving controllers collaboratively execute this control curve. Simultaneously, the receiving controller feeds back the measured charging current to the transmitting controller to correct the power output, forming a real-time closed-loop control structure within the charging process. The outer learning loop spans multiple charging cycles: after each charging cycle, the receiving controller appends the actual data from that charge to the historical database, calculates the prediction error, and updates the model parameters using Kalman filtering. The updated model parameters are used for prediction at the start of the next charge, forming an iterative learning structure across charging cycles. The two loops work together: the inner loop ensures real-time response to aging trends during a single charging cycle, while the outer loop ensures the prediction model is continuously refined as the battery ages.

[0113] The implementation methods and effects of the present invention will be further explained below through three specific application scenarios.

[0114] In Scenario 1, a smartphone battery wireless charging application: A certain brand of smartphone uses a 4000mAh lithium-ion polymer battery and supports Qi standard wireless charging. When the phone leaves the factory, the non-volatile memory chip in the wireless charging receiver pre-stores the initial capacity of this battery model. The initial parameters of the double exponential model are also determined. Users perform wireless charging once a day during regular use. During the first 20 charging cycles, the receiver controller collects and stores 20 sets of aging characteristic data through step S1. At the start of the 21st charge, the controller uses the data from the previous 20 cycles to determine the initial parameters of the double exponential model through least squares fitting. , , , The controller uses these parameters to predict the capacity decay trajectory over the next 50 charging cycles. The prediction shows that the battery capacity will decrease from approximately 97% to approximately 94% in the next 50 cycles. Based on the predicted trajectory, the controller calculates a current-time control curve, which is approximately 0.7C (2800mA) in the initial charging phase and decreases to approximately 0.5C (2000mA) in the later stages. The controller sends the control curve and the cutoff voltage of 4.35V to the charging dock via the Qi communication link. The charging dock adjusts its output power at 100ms intervals, and the phone simultaneously adjusts its rectification parameters. Throughout the charging process, the deviation between the measured charging current and the target curve remains within ±50mA. After charging, the controller appends the actual charging data to the historical database and updates the model parameters. Measured data shows that using the control method of this invention, the battery capacity retention rate after 300 cycles is approximately 88% to 90%, while the capacity retention rate of the same model battery using conventional constant current / constant voltage charging is approximately 80% to 83% after the same number of cycles. The above capacity retention data comes from comparative tests of the same model of battery under laboratory conditions. The actual performance in the actual use environment may vary depending on the battery batch and usage conditions.

[0115] In Scenario 2, an electric vehicle (EV) battery wireless charging application is described: An EV uses a 75kWh lithium-ion battery pack, supporting 11kW wireless charging power. When the vehicle is parked in a parking space equipped with a wireless charging pad, the vehicle's wireless charging receiver establishes a communication connection with the ground-based wireless charging transmitter. The receiver controller reads all charging history data of the battery pack since the vehicle was new from a non-volatile memory chip, recording approximately 200 charging cycles. The controller uses this data to continuously update the parameters of the bi-exponential model through Kalman filtering; the current model parameters are... , , , The controller predicted the capacity degradation trajectory over the next 30 charging cycles and detected that the battery was about to enter an accelerated aging phase. Based on this, the controller set the current-time control curve for this charge to gradually decrease from approximately 165A at 0.5C to approximately 115A at 0.35C, and lowered the cutoff voltage from 4.20V to 4.17V. The control curve and cutoff voltage values ​​were transmitted to the ground-based transmitter controller via the wireless charging communication link. The transmitter controller adjusted the inverter's PWM duty cycle according to the control curve, gradually reducing the transmitted power from approximately 11kW to approximately 8kW. The receiver controller synchronously adjusted the output voltage of the on-board rectifier to ensure that the battery current followed the target curve. The entire charging process lasted approximately 6 hours, and charging was terminated when the battery voltage reached 4.17V. After charging, the controller added the charging data to the historical database and updated the model parameters. According to the BMS evaluation, the charging efficiency of this charge was approximately 91%, and the battery temperature at the end of charging was approximately 5°C to 7°C higher than the ambient temperature, which is within a safe range.

[0116] In Scenario 3, a power tool battery pack wireless charging application is described: A power tool brand uses an 18V lithium-ion battery pack with a capacity of 5000mAh, supporting contact-based wireless charging. The built-in wireless charging receiver communicates with a dedicated charging dock during each charging cycle. This battery pack has been used on-site for approximately 8 months, undergoing about 150 charge-discharge cycles, and currently has a remaining capacity of approximately 4200mAh, or 84% of its rated capacity. After reading historical data, the controller found that the battery's capacity decay rate accelerated in the last 30 cycles, with the internal resistance change rate increasing from approximately 0.5mΩ per 10 cycles to approximately 1.2mΩ per 10 cycles. The controller utilizes updated double-exponential model parameters... , , , The capacity degradation trajectory for the next 20 cycles was predicted. The prediction showed that without adjusting the charging strategy, the battery capacity would drop to approximately 3900mAh, or 78% of the rated capacity, after 20 cycles. Based on this, the controller generated a current-time control curve. The initial charging current was set at 0.8C (4000mA), decreasing to 0.55C (2750mA) in the later stages of charging, while the cutoff voltage decreased from 4.20V to 4.15V. After the control parameters were sent to the charging dock, the dock adjusted its output power according to the control curve. After charging, the battery terminal voltage stabilized at 4.15V without voltage rebound. The actual capacity increment curve for this charge showed that the voltage value at the peak position had a small offset from the previous charge, within the range of 5mV to 8mV, indicating that the rate of increase in battery polarization was suppressed to some extent. After adding this data to the historical database, the controller updated the model parameters. The updated parameters showed a slight decrease in the predicted aging rate, indicating that the forward-looking compensation control had a positive effect on delaying aging.

[0117] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A wireless charging control method based on battery state, characterized in that, The method includes the following steps: S1. Obtain the peak position sequence of the capacity increment curve, the internal resistance change rate sequence, and the voltage drop value sequence at the end of charging during each complete charging process of the battery, as a multi-dimensional characteristic trajectory of battery aging. S2, Based on the aging multidimensional feature trajectory, construct an aging trajectory prediction model that combines a double exponential model with Kalman filtering to predict the capacity decay trajectory after a preset number of charging cycles following the current aging stage. S3, calculate the optimal charging acceptance capability of the current aging stage based on the predicted capacity decay trajectory, and convert the optimal charging acceptance capability into the current-time control curve and cutoff voltage value of the wireless charging system. S4. During this wireless charging process, the current-time control curve and cutoff voltage value are sent to the transmitter controller through the bidirectional communication link of the wireless charging system. The transmitter controller dynamically adjusts the output power according to the current-time control curve, and the receiver controller synchronously adjusts the rectification parameters and voltage regulation parameters. S5, after the charging cycle is completed, the actual data of the charging cycle is added to the historical database to update the parameters of the aging trajectory prediction model.

2. The wireless charging control method based on battery state according to claim 1, characterized in that, In S1, the sequence of peak positions of the capacity increment curve is obtained in the following way: During the constant current charging phase, the battery terminal voltage and cumulative charge were recorded at fixed time intervals, and plotted. Curve extraction The voltage and peak amplitude corresponding to the peak positions of the curve are obtained by arranging the voltage and peak amplitude corresponding to the peak positions of each charging cycle according to the charging cycle number, thus obtaining the peak position sequence of the capacity increment curve.

3. The wireless charging control method based on battery state according to claim 1, characterized in that, In S1, the sequence of internal resistance change rates is obtained in the following way: The battery terminal voltage is measured during the first 30 seconds of the constant current charging start phase to calculate the first DC internal resistance. The battery terminal voltage is measured during the last 30 seconds of the constant voltage charging end phase to calculate the second DC internal resistance. The rate of change of the first DC internal resistance and the second DC internal resistance with the number of charging cycles is recorded to obtain the internal resistance change rate sequence.

4. The wireless charging control method based on battery state according to claim 1, characterized in that, In S2, the double-exponential model is: in, Let this be the remaining battery capacity after the nth charging cycle. This is the initial capacity of the battery. , , , Here, n represents the charging cycle number, and exp represents the natural exponential function.

5. The wireless charging control method based on battery state according to claim 1, characterized in that, In S2, an aging trajectory prediction model is constructed by fusing a dual exponential model with a Kalman filter, specifically including: Using the peak amplitude of the capacity increment curves from the first 20 charging cycles, the initial parameters of the double exponential model are determined by least squares fitting. , , , ; Using a double exponential model as the state transition equation, and taking the peak amplitude of the capacity increment and the internal resistance value from each newly acquired charging data as observations, the model parameters are adjusted using a Kalman filter. , , , Perform real-time updates; At the start of each charge, utilize the currently updated model parameters. , , , Predict the capacity decay trajectory over the next N charging cycles. , , where N is a preset positive integer.

6. The wireless charging control method based on battery state according to claim 1, characterized in that, In step S3, the optimal charge acceptance capability for the current aging stage is calculated based on the predicted capacity decay trajectory, specifically including: A mapping table between battery remaining capacity and maximum acceptable charging rate was established through pre-experimental calibration. The predicted capacity decay trajectory Substitute each future capacity value into the mapping table for interpolation to obtain the corresponding optimal charging rate. ; Optimal charging rate The sequence is converted into the current-time control curve of this charging process. ; The cutoff voltage compensation value is calculated based on the predicted capacity decay trajectory. When the battery is predicted to enter the accelerated aging stage, the cutoff voltage value for that charge is reduced.

7. The wireless charging control method based on battery state according to claim 1, characterized in that, In step S3, the optimal charging rate will be... The sequence is converted into the current-time control curve of this charging process. Specifically, it includes: Based on model parameters and Estimate the effective duration of the current aging phase; The aging process of the next N charging cycles is compressed and mapped onto the timeline of this charging cycle; Generate a current curve that varies continuously with charging time. The current curve The first current value is used in the initial stage of charging, and it decreases continuously to the second current value in the later stage of charging as the predicted aging degree increases. The first current value is greater than the second current value.

8. The wireless charging control method based on battery state according to claim 1, characterized in that, In step S4, the current-time control curve and cutoff voltage value are transmitted to the transmitter controller via the bidirectional communication link of the wireless charging system, specifically including: The receiver controller will use the current-time control curve. The cutoff voltage value is sent to the transmitter controller via the wireless charging communication protocol; The transmitter controller adjusts its parameters in 100ms cycles based on the received current-time control curve. The inverter's PWM duty cycle or operating frequency is adjusted in real time so that the output power changes with time according to the current-time control curve. The receiver controller synchronously adjusts the output voltage setpoint of the rectifier circuit to ensure that the actual charging current received by the battery follows the current-time control curve. .

9. The wireless charging control method based on battery state according to claim 1, characterized in that, In step S5, updating the parameters of the aging trajectory prediction model specifically includes: The peak position of the actual capacity increment curve, the internal resistance value, and the final voltage drop value of this charge are added to the historical database. The actual capacity decay during this charge is compared with the predicted value of the aging trajectory prediction model to calculate the prediction error. Update the model parameters of the double exponential model using Kalman filtering based on the prediction error. , , , ; The updated model parameters are saved to a non-volatile memory chip.

10. A wireless charging control system based on battery state, applicable to the wireless charging control method based on battery state as described in any one of claims 1 to 9, characterized in that, The system comprises: The wireless charging receiver includes a receiving coil, a rectifier circuit, a voltage regulator circuit, a receiver controller, a non-volatile memory chip, a voltage sampling circuit, a current sampling circuit, and a temperature sensor. The wireless charging transmitter includes a transmitting coil, an inverter circuit, a transmitter controller, and a communication module. The transmitter controller and the receiver controller are connected via a bidirectional communication link. The non-volatile memory chip is used to store the peak position sequence of the capacity increment curve, the internal resistance change rate sequence, and the voltage drop value sequence at the end of charging during each complete charge of the battery. The receiver controller is used to construct an aging trajectory prediction model that combines a dual exponential model with a Kalman filter based on the peak position sequence of the capacity increment curve, the internal resistance change rate sequence, and the voltage drop value sequence at the end of charging. It predicts the capacity decay trajectory after the current aging stage for a preset number of charging cycles, calculates the current-time control curve and the cutoff voltage value based on the predicted capacity decay trajectory, and sends the current-time control curve and the cutoff voltage value to the transmitter controller through a bidirectional communication link. The transmitter controller is used to dynamically adjust the output power of the inverter circuit according to the current-time control curve. The receiver controller is also used to synchronously adjust the rectification parameters of the rectifier circuit and the voltage regulation parameters of the voltage regulator circuit, and after the charging is completed, to append the actual data of the charging to the non-volatile memory chip and update the parameters of the aging trajectory prediction model.

Images