Equipment control method applied to charging of power adapter

Abstract

The invention relates to the field of power adapters, and discloses an equipment control method applied to charging of a power adapter, which comprises the following steps of: 1, acquiring charging equipment of the power adapter, and identifying the type of the equipment and battery parameters; 2, analyzing an equipment application scene and collecting real-time electric quantity information; 3, setting an initial charging sequence based on the scene priority and the electric quantity gap; step 4, calculating transmission loss according to the resistance model of the cable temperature dynamic change acquired in the step 2; 5, constructing a multi-objective optimization function in combination with the battery health degree and the transmission loss; 6, recording historical charging and discharging data of the equipment; and 7, calculating a battery aging coefficient and a capacity reduction rate according to historical data. By identifying the type of the charging equipment and the battery parameters, an exclusive charging strategy can be formulated according to different equipment and battery characteristics, the problems of over-charging, under-charging and the like can be effectively avoided, and the service life of the battery is prolonged.

Description

Technical Field

[0001] The present invention relates to the technical field of power adapters, and in particular to a device control method applied to charging of a power adapter. Background Art

[0002] In today's digital age, various electronic devices are widely used. As a key component for device charging, the effectiveness and intelligence of the power adapter's charging control have a profound impact on device performance and user experience.

[0003] Traditional power adapter charging control methods are relatively simple and crude. They often only identify device type and battery parameters through simple, basic parameter recognition, failing to accurately distinguish between battery characteristics of different models and ages within the same device. This makes it difficult to provide personalized and accurate charging solutions based on the actual battery condition during subsequent charging.

[0004] The analysis of device application scenarios lacks depth and real-time performance. Most traditional methods fail to fully consider the dynamic changes in the device's power demand in different usage scenarios (such as office scenarios, outdoor mobile scenarios, gaming scenarios, etc.). In office scenarios, the device may be in light use for a long time, and the power consumption is relatively stable; in outdoor mobile scenarios, the device may need to frequently use high-energy consumption functions (such as GPS positioning, mobile data transmission, etc.), and the power consumption is rapid. However, traditional charging control methods are "equal" and cannot flexibly adjust the charging strategy according to the differences in these scenarios, resulting in low charging efficiency and an inability to effectively guarantee the battery life of the device.

[0005] Traditional methods for calculating transmission loss often use a fixed resistance model, ignoring the impact of dynamic cable temperature changes on resistance. During actual charging, as the charging current continues to flow, the cable temperature gradually rises. This temperature increase causes the cable resistance to increase, significantly increasing transmission loss. This fixed resistance model-based transmission loss calculation method significantly deviates from actual energy loss estimates during charging control, resulting in energy waste and potentially impacting device charging speed and battery life. Summary of the Invention

[0006] In response to the shortcomings of the existing technology, the present invention provides a device control method for charging with a power adapter, which solves the problem that the power adapter charging control method is relatively simple, easily causes overcharging, and reduces battery life and performance.

[0007] To achieve the above objectives, the present invention is implemented through the following technical solutions: A device control method for charging with a power adapter, comprising the following steps: Step 1: Get the charging device of the power adapter and identify the device type and battery parameters; Step 2: Analyze device application scenarios and collect real-time power information; Step 3: Set the initial charging sequence based on scene priority and power gap; Step 4: Calculate the transmission loss based on the resistance model of the cable temperature dynamic change collected in step 2; Step 5: Construct a multi-objective optimization function based on battery health and transmission loss; Step 6: Record the device charging and discharging history data; Step 7: Calculate the battery aging coefficient and capacity reduction rate based on historical data; Step 8: Evaluate battery health and update charging sequence priority; Step 9: Solve the optimal charge distribution through variational optimization algorithm; Step 10: Dynamically adjust the charging strategy and generate control instructions; Step 11: Execute charging control management and monitor safety status in real time.

[0008] Preferably, in the step 4: In the step 4: The calculation of the dynamic transmission loss is based on the temperature dependence model of the cable resistance, and the temperature R(T) dependence model satisfies: R(T)=R0[1+α(TT env )]; Among them, α is the temperature coefficient of the cable material, R0 is the resistance value, T is the real-time temperature, T env is the ambient temperature.

[0009] Preferably, in step seven: The calculation of the battery aging coefficient is based on the Arrhenius equation, and the activation energy parameter in the Arrhenius equation is calibrated through a battery cycle aging experiment.

[0010] Preferably, in step nine: The variational optimization algorithm involves constructing a functional that incorporates transmission loss, battery aging, and temperature constraints, and applying the Pontryagin minimum principle to solve the analytical solution for current distribution.

[0011] Preferably, the weight factor of the functional is dynamically adjusted according to the device application scenario, and the transmission loss weight ratio in the emergency charging scenario is increased to more than 70%.

[0012] Preferably, in the step 10: The dynamic adjustment of the charging strategy includes real-time estimation of battery health through an extended Kalman filter algorithm, wherein the state variables in the Kalman filter algorithm include battery temperature and health; The ratio of power reduction is linearly positively correlated with the extent to which the temperature exceeds the threshold.

[0013] Preferably, in the step 11: The execution of the charge control management includes activating the vortex fan and reducing the power to 80% of the current value when the predicted temperature exceeds 90% of the safety threshold.

[0014] Preferably, the step eleven further comprises: If a metal foreign object is detected in the charging device, the output is cut off and an audible and visual alarm signal is activated, which lasts until manually reset.

[0015] The present invention provides a device control method for charging a power adapter. It has the following beneficial effects: 1. The present invention can formulate exclusive charging strategies based on different devices and battery characteristics by identifying the type of charging device and battery parameters. Compared with traditional charging methods, it can effectively avoid problems such as overcharging and undercharging, extend battery life, improve device performance stability, and meet the charging needs of diverse devices. By analyzing device application scenarios and collecting real-time power information, the initial charging sequence is set according to the scenario priority and power gap. A stable and energy-saving charging mode can be used in office scenarios; when electricity is urgently needed outdoors, priority is given to quickly replenishing the power. Compared with charging methods that do not consider scenario differences, the device's endurance and user experience in different scenarios are greatly improved; 2. This invention calculates transmission loss based on a resistance model that dynamically changes with cable temperature, accurately understanding energy loss during charging. Compared to using a fixed resistance model, this allows for more reasonable adjustment of charging parameters, reduces additional energy consumption caused by cable resistance changes, improves energy efficiency, and reduces electricity costs. 3. This invention combines battery health and transmission loss to construct a multi-objective optimization function. It calculates the battery aging coefficient and capacity reduction rate based on historical charging and discharging data, accurately assesses battery health, and updates the charging sequence priority. This effectively slows down battery aging and significantly extends battery life compared to charging methods that ignore battery health management, reducing the frequency and cost of battery replacement. 4. This invention uses an extended Kalman filter algorithm to estimate battery health in real time, dynamically adjust charging strategies, and generate control instructions. This allows for timely optimization of the charging process based on the battery's real-time status. This is more flexible and intelligent than fixed charging strategies, effectively responding to changes in battery status and ensuring charging quality and device safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a flow chart of the device control method of the present invention. DETAILED DESCRIPTION

[0017] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the drawings in the present specification. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0018] Please see the attached Figure 1 An embodiment of the present invention provides a device control method for charging with a power adapter, comprising the following steps: Step 1: Get the charging device of the power adapter and identify the device type and battery parameters; Step 1, the first step in this method, aims to identify the type of charging device currently connected and its key battery parameters through the power adapter's communication interface and physical connection status, providing basic data input for subsequent dynamic optimization control. This step must be compatible with multiple fast-charging protocols and handle abnormal connection scenarios to ensure reliable acquisition of device information under complex operating conditions. In its implementation, hardware protocol parsing and software state machine logic must be combined to achieve the coordinated operation of device enumeration, parameter reading, and exception handling modules.

[0019] Specifically, the implementation of step one includes the following technical contents: Device enumeration and protocol parsing: In some embodiments, a multi-protocol compatible power management chip is used as the communication interface core, and an enumeration request signal is sent through the CC pin of the USB Type-C interface: Send the SourceCapabilities message of the PD protocol (including 5V / 9V / 15V / 20V voltage levels); Detect the device's Request message response and analyze its maximum requested power P req :P req =V req ×I req ; Among them, V req is the maximum requested voltage, I req is the maximum requested current.

[0020] If the device does not respond to the PD protocol, it switches to the QC3.0 protocol and negotiates the voltage and current by tuning the D+ / D- pin voltage (0.325V to 3.3V).

[0021] Specifically, the protocol parsing module has a built-in state machine, and its state transition logic is as follows: Among them, V CC is the CC pin voltage of the USB Type-C interface (unit: V), VD+ USBD+ pin voltage (unit: V).

[0022] Device type identification and parameter acquisition: In one possible implementation, the device type is determined by matching the manufacturer information (such as VID / PID) returned by the device with a pre-set database: Mobile devices: Read the battery rated capacity C nominal (Unit: mAh), charging cut-off voltage V end ; Laptop devices: Get the maximum input power P max,in =min(P req ,P adapter )(unit: W); Headphones / wearable devices: Marked as low-priority devices, with a default current limit of I limit =0.5A.

[0023] Specifically, the parameters are obtained through I 2 C bus to access the device's battery management chip (such as BQ25895) and read the following registers: Reg0x0B: Battery voltage V bat (12-bitADC, accuracy ±10mV); Reg0x0D: Battery capacity C remaining (Unit: mAh); Reg0x12: Battery health status SOH (8-bit, 0x00 to 0xFF corresponds to 0% to 100%).

[0024] Abnormal connection handling mechanism: In some embodiments, protection logic is triggered if the following abnormal conditions are detected: Foreign object detection (FOD): by resonant frequency offset Δf=|f meas -f ref |>5%f ref If a foreign object is detected, the output will be turned off immediately.

[0025] Among them, f ref is the inherent resonant frequency of the wireless charging coil, f meas The resonant frequency is measured in real time.

[0026] Communication timeout: If the device is in communication time t timeout = If no response is received within 500ms, the system switches to the default 5V / 1A mode and records the error code (ErrorCode 0x01).

[0027] Data storage and state synchronization: As an option, the identified device information is stored in non-volatile memory.

[0028] In one possible implementation, the device connection status is updated in real time through an interrupt service routine (ISR) to ensure that subsequent steps (such as the scenario analysis in step 2) are executed based on the latest data.

[0029] Step 2: Analyze device application scenarios and collect real-time power information; Step 2 is based on the device type and basic parameters obtained in step 1, combined with real-time power data and historical behavior analysis, to dynamically divide the charging scenarios and provide a decision-making basis for the charging sequence setting in the subsequent step 3 and the multi-objective optimization in steps 4 to 9. This step builds a scenario classification model by integrating the physical state of the device with the user's usage habits, solving the problem of single scenario determination in traditional methods and ensuring that the charging strategy is accurately matched with user needs. In specific implementation, it is necessary to realize the coordinated operation of power collection, scenario weight calculation and abnormal data processing modules to ensure that the data input to the downstream steps is timely and reliable.

[0030] In this embodiment, the implementation of step 2 includes the following technical contents: Real-time power collection and verification: In some embodiments, the power adapter I 2 C bus to access the device's battery management chip and periodically read the following parameters: State of charge: Get the remaining capacity C from Reg0x0D remaining , and based on the rated capacity C nominal calculate: Among them, SOC i is the real-time power percentage of the i-th device, C remaining,i is the remaining capacity value reported by the device (accumulated by coulomb counter), C nominal,i is the nominal capacity of the battery (obtained from step 1).

[0031] Battery temperature T bat,i : Read the thermistor ADC value from Reg0x0F and convert it to a temperature value (unit: °C) after linearization. The formula is: Among them, R NTC is the current resistance of the thermistor (unit: Ω), R 25 is the nominal resistance at 25°C, and β is the thermal sensitivity index.

[0032] In one possible implementation, if an SOC data jump is detected (such as the difference between two adjacent samples ΔSOC>5%), the data verification mechanism is triggered: the register is re-read and three consecutive sampling values are compared; if there is still no convergence, it is marked as "power data abnormality" and historical average replacement is enabled.

[0033] Scenario classification model construction: In general, scenario classification is based on the following multi-dimensional features: device type (obtained in step 1): mobile phone, laptop, wearable device, etc.; Real-time power SOC i : Divide the threshold interval (such as <20%, 20% to 80%, >80%); Historical charging behavior: Read the charging period distribution (e.g., nighttime charging percentage) of the last 30 days from non-volatile memory; User preset modes: Priority tags (e.g., "Fast Charge," "Maintenance Mode") synced via the mobile app.

[0034] Specifically, a weighted decision tree model is used for scene judgment, and the weight coefficient is dynamically adjusted: Among them, w1, w2, w3 are weight coefficients, θ 紧急 ,θ 常规1 ,θ 常规2 The time period coefficient is 1 when the current time is in the user's usual charging time period, otherwise it is 0.5.

[0035] Dynamic priority weight adjustment: As an option, introduce a battery aging compensation factor in the maintenance scenario and modify the scenario judgment weight: w ′ 1=w1·(1+α); Among them, α is the aging sensitivity coefficient (default is 0.2), which is calibrated through experiments.

[0036] In one possible implementation, if the device temperature T bat,i >40℃, forced to switch to the "Temperature Control Priority" sub-scene, limit the charging current and increase the temperature weight: w ′ 1 ′ =w1·0.5, w4=0.5; Abnormal data processing and feedback: In some embodiments, processing logic is designed for the following abnormal situations: Power data failure: If three consecutive samples are invalid, the re-enumeration process of step one is triggered, and the fault log is recorded in the historical database of step six; Scenario conflict: If the same device meets multiple scenario conditions at the same time (such as SOC = 15% and is in the maintenance period), arbitration is carried out according to preset rules (such as "emergency first"), and a policy update instruction is sent to step ten.

[0037] Step 3: Set the initial charging sequence based on scene priority and power gap; Step three dynamically generates an initial charging sequence based on the scenario classification results and real-time power data output from step two, providing a priority benchmark for subsequent power allocation optimization (steps four to nine). This step solves the inefficiency problem caused by resource competition in multi-device charging scenarios by integrating device status, user needs, and system constraints, ensuring that high-priority devices receive timely power supply. In the specific implementation, it is necessary to build a dynamic scoring model, handle conflict logic between devices, and establish a data interface with downstream modules to ensure that sequence updates are synchronized with real-time working conditions.

[0038] In this embodiment, the implementation of step three includes the following technical contents: Dynamic priority scoring model: In some embodiments, a priority score Score is calculated for each access device. i , its function form is: Score i =w soc ·(1-SOC i )+w soh ·(1-SOH i )+w soene SceneWeight i ; Among them, SOC i is the real-time power of the device (obtained from step 2, unit: %), SOH i is the battery health (obtained from step 1 or step 8, unit: %), SceneWeight i is the scenario weight factor (emergency fast charging = 1.0, regular charging = 0.5, trickle maintenance = 0.2), w soc 、w soh 、w soene is the weight coefficient (default values 0.6, 0.3, 0.1), satisfying w soc +w soh +w soene =1.

[0039] Specifically, the weight coefficient is dynamically adjusted according to the device type: Laptop devices: Improve w soc to 0.7, reducing w soh to 0.2; Wearable device: Set w soene =0.4, strengthen scene dependence.

[0040] Multi-device sorting and conflict resolution: In one possible implementation, a Max-Heap data structure is used to maintain the device priority queue in real time: Initialize the heap: Score i Insert into the heap as a key value; Heap update triggering condition: new device access (step 1 triggers interruption); Device power change ΔSOC i ≥2% (Step 2, periodic testing).

[0041] Conflict handling rules: If the scores are the same: sort in ascending order by device access timestamp; Power exceeds the limit: If the total required power ∑I i V i >P max , remove the top devices until the constraints are met.

[0042] Charging mode selection and parameter binding: Typically, each device is assigned an initial charging mode: Scenario Charging mode Current Strategy Emergency fast charging Constant Current-Constant Voltage (CC-CV) <![CDATA[Maximum allowable current I max,i > Regular charging Pulse charging <![CDATA[0.7I max,i , duty cycle 50%]]> Trickle maintenance Small current constant voltage <![CDATA[0.2I max,i ]]> As an option, for health SOH i <70% of devices will be forced to override the scene settings, switch to maintenance mode and limit current: I limit,i =min(I max,i ,0.5·I nom,i ); Among them, I nom,i is the nominal charging current of the battery (obtained from step 1, unit: A), I max,i The maximum current supported by the protocol.

[0043] Abnormal state injection and sequence correction: In some embodiments, if the following abnormalities are detected, the charging sequence is dynamically corrected: Temperature limit: When T bat,i >45℃, (obtained from step 2), reduce the priority score i ′ : Score i ′ =Score i ·max(0,1-0.05(T bat,i -40); Cable overload: If a port triggers overcurrent protection three times in a row (feedback in step 11), its associated device is removed from the queue and marked as a disabled port.

[0044] Step 4: Calculate the transmission loss based on the resistance model of the cable temperature dynamics; In step 4: The calculation of dynamic transmission loss is based on the temperature dependence model of cable resistance. The temperature dependence model R(T) satisfies: R(T)=R0[1+α(TT env )]; Among them, α is the temperature coefficient of the cable material, R0 is the resistance value, T is the real-time temperature, T env is the ambient temperature.

[0045] Step 4 dynamically quantifies the energy loss during the charging process based on the charging sequence and real-time device status parameters generated in step 3, providing the core input for subsequent comprehensive loss optimization (steps 5 to 9). This step solves the energy inefficiency problem caused by static loss estimation in traditional solutions by establishing a charging loss model that couples a time-varying transmission loss model with battery aging. In the specific implementation, it is necessary to integrate hardware sensor data and dynamic calculation algorithms to achieve high-precision loss prediction and ensure coordinated operation with upstream priority data and downstream optimization modules.

[0046] In some embodiments, a four-wire Kelvin measurement method is used to obtain the cable resistance R of each charging port in real time. i (T), and combined with the infrared temperature sensor data to build a resistance-temperature model: R i (T)=R 0,i ·[1+α i ·(T line,i (t)-T env )]; Among them, R 0,i The cable at port i is at the reference temperature T env Initial resistance (unit: Ω) at (usually 25°C), α i is the temperature coefficient of the cable material, T line,i (t) is the real-time cable temperature (unit: °C), T env is the ambient temperature (unit: °C).

[0047] In one possible implementation, the transmission loss power Calculated as: Among them, I i (t) is the real-time charging current of port i (unit: A, obtained from step 3), V drop,i is the on-state voltage drop of a switching device (such as a MOSFET) (unit: V, obtained by looking up the table in the device manual).

[0048] Modeling battery internal losses: Specifically, battery charging losses Including internal resistance loss and polarization loss: Among them, R bat,i is the battery internal resistance (unit: Ω), and the health SOH i The relationship is: R bat,i =R bat0,i ·(1.8-0.8·SOH i ), R bat0,i is the internal resistance of the new battery, obtained from step 1, kp is the polarization loss coefficient (unit: W·Ah), calibrated by HPPC test.

[0049] Loss data fusion and verification: As an option, a sliding window filtering algorithm is used to fuzzy the loss data. Perform smoothing with a window size of N = 5: Outlier elimination rule: If a sampling value deviates from the window mean by more than 20%, it will be replaced by the previous valid value.

[0050] In one possible implementation, when a sudden temperature change (ΔT line,i >5℃ / s), trigger the emergency retest mechanism: suspend charging for 1ms and re-measure R i (T); If the measurement is abnormal for three consecutive times, mark the port as faulty and notify step 11 to perform isolation.

[0051] Multi-port loss aggregation and constraint checking: In general, the total loss power Calculated as: Constraints: Among them, η is the maximum loss ratio allowed by the system, P input (t) is the adapter input power (unit: W, obtained through the AC / DC front-end detection circuit).

[0052] In some embodiments, if the total loss exceeds the limit, a load reduction request is sent to step nine, and the current is gradually reduced according to the priority queue (generated in step three).

[0053] In this embodiment, step 4 provides the following key outputs for subsequent steps through dynamic modeling and real-time verification: Itemized loss data: By separating transmission loss and battery loss in step 4, this supports the weight allocation of the comprehensive loss function in step 5; Aggregation loss constraint: serves as the boundary condition for the optimization algorithm in step nine to prevent system overload; Abnormal event flag: The cable fault signal is transmitted to step 11 to trigger the hardware protection action.

[0054] Data flow connectivity: Input dependency: I i (t) Charging mode setting from step 3, SOH i Inherited from step 1 or step 8; Output transfer: Transmission loss power Battery charging loss Input to step 5 to construct the comprehensive loss function; Feedback mechanism: The loss exceeding limit signal triggers the load reduction strategy iteration in step nine.

[0055] Step 5: Construct a multi-objective optimization function based on battery health and transmission loss; Step 5, based on the transmission loss and battery loss data calculated in step 4, integrates the device health status and temperature constraints to construct a dynamic comprehensive loss function, providing a global optimization target for the charge distribution optimization in the subsequent step 9. This step introduces a multi-factor weight adjustment mechanism to address the problem that a single loss indicator in traditional solutions cannot balance efficiency and device life, ensuring that the charging strategy achieves the optimal trade-off between energy loss, battery aging, and thermal safety. In specific implementation, it is necessary to dynamically correct the model parameters in combination with real-time operating conditions, and establish a data interface with upstream loss data and downstream optimization algorithms.

[0056] In one possible implementation, the multi-port comprehensive loss is aggregated: In general, the global comprehensive loss J total (t) is calculated as the loss J at each port i (t) Sum: Constraints: J total (t)≤J max ; Among them, J max The maximum comprehensive loss allowed by the system (unit: W), dynamically set according to the heat dissipation capacity of the adapter: J max =k cool ·P rated ; where k cool =0.2 is the heat dissipation coefficient, P rated is the rated power of the adapter.

[0057] In some embodiments, if J total (t)>J max , send optimization instructions to step nine, requesting the redistribution of charge.

[0058] Step 5 provides the following core outputs for subsequent steps through dynamic weight adjustment and multi-factor integration: Comprehensive loss index: used as the objective function input of the multi-objective optimization algorithm in step nine; Dynamic weight parameters: passed to the strategy adjustment module in step 10 to achieve scenario-adaptive optimization; Global constraint signal: triggers the iterative calculation of step nine or the protection action of step eleven.

[0059] Data connection: Input dependencies: From step 4, Dependency step seven; Output transfer: global comprehensive loss J total (t) Input to the optimization engine in step nine.

[0060] Step 6: Record the device charging and discharging history data; Step 6 continuously collects and stores key parameters during the device's charge and discharge processes to build a historical database, providing a data foundation for subsequent battery aging analysis (steps 7 and 8) and optimization strategy iteration. This step addresses the data redundancy and integrity issues inherent in traditional solutions by designing efficient data compression algorithms and exception verification mechanisms, ensuring efficient utilization of storage resources in long-term operation. Specific implementation requires multi-source data fusion, non-volatile storage management, and data interface synchronization with upstream and downstream modules.

[0061] In this embodiment, the implementation of step six includes the following technical contents: Data collection and timestamp synchronization: In some embodiments, the following parameters are recorded periodically: sampling period T s =1s: Charging stage: real-time current I i (t), voltage V bat,i (t) (obtained from step 4); Cable temperature T line,i (t), battery temperature T bat,i (t) (obtained from step 2 or step 4); Charge mode flag (e.g. CC / CV, obtained from step 3).

[0062] Discharge cycle: Depth of discharge DoD i : Average discharge rate

[0063] Among them, t end is the discharge end point, t start The discharge start point.

[0064] In one possible implementation, a GPS module or an RTC chip is used to add precise timestamps to the data, with an error of less than ±1ms.

[0065] Data compression and storage optimization: Specifically, a hybrid algorithm combining differential coding and Huffman compression is used: Differential encoding: Calculate the difference ΔX = X(t)-X(tT) for continuous sampling data s ), only the change is stored, where X(t) is the sampled data at time t; Huffman coding: Generates the optimal prefix code table based on parameter statistical distribution, with a compression rate of 40% to 60%.

[0066] Abnormal data detection and repair: As an option, define data validation rules: Current rationality: |I i (t)|≤1.2×I max,i (I max,i Obtained from step 1); Voltage monotonicity: When charging, V bat,i (t)≥V bat,i (tT s ); Temperature continuity: |T line.i (t)-T line.i (tT s )|≤5℃.

[0067] In one possible implementation, if abnormal data is detected, it is marked as invalid and triggers the re-enumeration process of S1; the missing value X(t) is filled using linear interpolation: Storage media management and cyclic overwriting: Generally, non-volatile memory is used and divided into multiple logical sectors: Metadata area: storage device ID mapping table, compression code table; Data area: written cyclically in chronological order, a single log entry occupies 12 bytes; Index area: records the latest data pointer and checksum.

[0068] In some embodiments, when the remaining storage space is less than 10%, the aging data elimination strategy is activated: Delete the earliest 30% of data blocks; Updates the index table and performs defragmentation.

[0069] Step 6 provides the following core support for subsequent steps through efficient data management: Historical data pool: charge and discharge records are used as input for aging coefficient calculation in step seven; Abnormal event tracing: invalid data mark is passed to step 11 to trigger fault diagnosis; Storage resource optimization: Compression and elimination mechanisms ensure long-term operational feasibility.

[0070] Data flow connectivity: Input sources: Current and voltage come from step 4, temperature comes from step 2 or step 4, and device ID comes from step 1; Output flow: Raw data is used for offline analysis in step 7, and statistical features (such as average discharge rate) are input into step 8; Abnormal feedback: The data verification failure signal triggers the collaborative recovery process of step 1 or step 11.

[0071] Step 7: Calculate the battery aging coefficient and capacity reduction rate based on historical data; In step seven: The calculation of the battery aging coefficient is based on the Arrhenius equation, in which the activation energy parameter is calibrated through battery cycle aging experiments.

[0072] Step 7, based on the charge and discharge history data recorded in step 6 and combined with real-time operating parameters, quantifies the battery aging process and calculates the capacity reduction rate, providing a basis for aging dynamics for the subsequent health assessment and charging strategy optimization in step 8 (steps 9 to 11). This step solves the problem of static aging prediction in traditional solutions by integrating electrochemical models with statistical learning methods, achieving accurate modeling of battery life degradation. Specific implementation requires the fusion of multi-source data, online calibration of model parameters, and abnormal data processing mechanisms.

[0073] In this embodiment, the implementation of step seven includes the following technical contents: Construction of aging kinetic model: In some embodiments, the aging coefficient η is constructed based on the Arrhenius equation and the Peukert effect. aging,i : η aging,i =A i exp; Among them, A i is the battery chemical attenuation factor (unit: s -1 ·A -2 ), calibrated by cyclic aging experiment, is.

[0074] Capacity reduction rate calculation: Specifically, the capacity reduction rate ΔC i Correlation with Coulombic efficiency through capacity fading model: ΔC i =C; Where C is the nominal capacity of the battery.

[0075] Online parameter calibration and correction: As an option, the extended Kalman filter is used to dynamically update the battery chemical attenuation factor A i and activation energy E a,i : Equation of state: Among them, w A , w E is process noise, which obeys Gaussian distribution

[0076] Observation equation: ΔC meas,i =C nominal,i ·(1-exp(-η aging,i ·t cycle ))+v Where v is the observation noise, obeying t cycle is the loop duration.

[0077] In one possible implementation, if the following anomalies are detected, a model reset is triggered: Capacity jump: adjacent cycle: Temperature exceeds the limit: T in historical data bat,i (t)>T critical , where T critical is the critical temperature; Reset operation: Clear the current A i ,E a,i Estimated values; reload initial parameters (factory calibration values obtained from step 1).

[0078] Step 7 provides the following core outputs for subsequent steps through dynamic modeling and online calibration: Aging coefficient: input to the battery health assessment model in step 8; Capacity reduction rate: capacity constraint used in the optimization algorithm in step nine; Parameter update signal: triggers the dynamic adjustment of the charging strategy in step 10.

[0079] Data flow connectivity: Input dependency: Charge and discharge data comes from step 6, and temperature data comes from step 2 or step 4; Output transfer: The aging coefficient and capacity reduction rate are input to step 8; Abnormal feedback: The model reset instruction is passed to step 11 to perform the protection action.

[0080] Step 8: Evaluate battery health and update charging sequence priority; In this embodiment, the implementation of step eight includes the following technical contents: Health comprehensive assessment model: In some embodiments, the battery health SOH is defined i SOH C and internal resistance health SOH R Weighted combination of: SOH i =w C ·SOH C,i +w R ·SOH R,i ; in: C current,i is the current measured capacity (obtained from step 7).

[0081] where R current,i is the current internal resistance (obtained from step 4); w C 、w R are weight coefficients, which are 0.7 and 0.3 by default, respectively, to meet wC +w R =1.

[0082] In one possible implementation, the weight coefficient is dynamically adjusted according to the battery type: Power battery (such as electric vehicles): w R Increase to 0.5 to strengthen the influence of internal resistance; Consumer electronics batteries (such as mobile phones): w C Fixed to 0.8.

[0083] Dynamic weight correction mechanism: Specifically, the temperature compensation factor and the cycle number compensation factor are introduced to correct the weight coefficient w C : Among them, T avg,i is the historical average temperature of the battery (unit: °C, obtained from step 6), N i is the number of charge and discharge cycles (statisticed from step 6), α T , α N is the compensation coefficient (default 0.2, 0.1), calibrated through aging experiments.

[0084] As an option, define the health data validity rules: Capacity-internal resistance consistency: If |SOH C,i -SOH R,i |>15%, marked as "data inaccuracy"; Temperature dependence: When T avg,i When >45℃, SOH is forced to decrease R Weight to 0.2.

[0085] In one possible implementation, the abnormal data is processed using a sliding window median filter: Window size: 3 consecutive sampling points, period T s =1h.

[0086] Health prediction and life span warning: Generally, the health decay in the next 30 days is predicted based on the ARIMA time series model: Among them, φ p is the autoregressive coefficient (trained by historical data), ∈ t is a white noise term, obeying t is the number of days.

[0087] In some embodiments, if the prediction Trigger the warning signal and push to step 11 to execute maintenance mode.

[0088] Step 8 provides the following core outputs for subsequent steps through multi-dimensional fusion and dynamic correction: Health quantification value: serves as the aging constraint for the optimization algorithm in step nine; Weight correction parameters: input to the dynamic strategy adjustment module in step 10; Life warning signal: triggers the maintenance agreement or user notification in step 11.

[0089] Data flow connectivity: Input sources: Capacity data comes from step 7, internal resistance data comes from step 4, and temperature data comes from step 6; Output transfer: The health value is input into the charging sequence update in step 3 and the loss function in step 9; Abnormal feedback: The data inaccuracy mark triggers the data re-collection process in step six.

[0090] Step 9: Solve the optimal charge distribution through variational optimization algorithm; In step nine: The variational optimization algorithm involves constructing a functional that incorporates transmission loss, battery aging, and temperature constraints, and applying the Pontryagin minimum principle to obtain an analytical solution for current distribution. The weight factor of the functional is dynamically adjusted according to the device application scenario, and the transmission loss weight in the emergency charging scenario is increased to more than 70%.

[0091] In this embodiment, the implementation of step nine includes the following technical content: Multi-objective optimization problem construction: In some embodiments, the optimization objectives are defined as minimizing the global comprehensive loss and maximizing the weighted health, subject to the following constraints: Power Constraints: (P max is the maximum output power of the adapter); Temperature constraint: T line,i (T line,i is the cable safety temperature, obtained from step 4); Current constraint: 0≤I i ≤I max,i (I max,i The maximum current supported by the device protocol).

[0092] Optimization algorithm selection and implementation: Specifically, a non-dominated sorting genetic algorithm with an elitist strategy is used to solve the Pareto optimal solution set.

[0093] Fitness function: f1 = J total , Among them, ω i is the weight coefficient, which is dynamically adjusted as the scene switches.

[0094] In one possible implementation, if the real-time requirement is high (such as frequent switching of charging modes), a sequential quadratic programming (SQP) algorithm is used to quickly solve the local optimal solution.

[0095] Constraint processing and feasible solution screening: As an option, a penalty function method is used to deal with solutions f that violate the constraints. penalty : Among them, g j (I) is the j-th constraint function, λ is the penalty factor (default is 1000), which increases linearly with the number of iterations.

[0096] Screening rules: prioritize solutions that satisfy all constraints; If there is no market, select the solution with the least constraint violations and trigger the load shedding strategy in step 10.

[0097] Dynamic weight adjustment and scene adaptation: In one possible implementation, the optimization target weight is dynamically adjusted based on the scene classification results in step 2: Emergency fast charging scenario: f1 weight is increased to 80%, and f2 weight is reduced to 20%; Maintenance scenario: f2 weight increased to 70%, increasing SOH i The lower limit of the constraint.

[0098] Optimization result output and verification: In general, the optimal current distribution value Write to shared memory for step 10 to call; Exception handling: If the calculation result oscillates (two consecutive ), enable historical mean smoothing; If convergence continues to fail, the safety protocol in step 11 is triggered and the fault code is recorded.

[0099] Step 10: Dynamically adjust the charging strategy and generate control instructions; In step 10: Dynamically adjusting the charging strategy involves estimating the battery health in real time through an extended Kalman filter algorithm, where the state variables include battery temperature and health.

[0100] In this embodiment, the implementation of step 10 includes the following technical contents: Dynamic strategy adjustment: Specifically, the health prediction value is updated every 5 seconds through the extended Kalman filter (model of step 7 / step 8). If the SOH is detected i When the decline rate exceeds the limit (>0.5% / day), the following actions are triggered: Reduce the maximum allowable current; Send a charging sequence update request to step 3 to increase the priority of the device.

[0101] Command issuance and synchronization: As an option, the CAN bus or I2C bus is used to write the PWM duty cycle command to the multi-port controller, ensuring that the synchronization error of each port is less than ±1%.

[0102] Step 11: Execute charging control management and monitor safety status in real time.

[0103] In step 11: The execution of charge control management includes starting the vortex fans and reducing the power to 80% of the current value when the predicted temperature exceeds 90% of the safety threshold.

[0104] The power reduction ratio is linearly positively correlated with the extent to which the temperature exceeds the threshold.

[0105] Step 11 also includes: If a metal foreign object is detected in the charging device, the output will be cut off and the audible and visual alarm signals will be activated. The alarm signal will continue until it is manually reset.

[0106] In this embodiment, the implementation of step S11 includes the following technical contents: Power output execution: In some embodiments, a multi-phase parallel Buck-Boost topology is adopted, and each phase is composed of SiC MOSFET (model C3M0065090J) and magnetic components: Switching frequency: 200kHz (reduce ripple); Current sampling: Real-time feedback via ACS724 Hall sensor, accuracy ±1%.

[0107] Active thermal management: In one possible implementation, the cooling strategy is dynamically adjusted based on the S4 temperature rise prediction model: Level 1 temperature control: start the fan; Secondary temperature control: proportional load reduction.

[0108] Status feedback and logging: Generally, the execution results (actual current, temperature, fault code) are sent back to step 6 for storage, and a JSON format log is generated for the user app to read.

[0109] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A device control method for charging with a power adapter, characterized in that: The following steps are involved: Step 1: Get the charging device of the power adapter and identify the device type and battery parameters; Step 2: Analyze device application scenarios and collect real-time power information; Step 3: Set the initial charging sequence based on scene priority and power gap; Step 4: Calculate the transmission loss based on the resistance model of the cable temperature dynamic change collected in step 2; Step 5: Construct a multi-objective optimization function based on battery health and transmission loss; Step 6: Record the device charging and discharging history data; Step 7: Calculate the battery aging coefficient and capacity reduction rate based on historical data; Step 8: Evaluate battery health and update charging sequence priority; Step 9: Solve the optimal charge distribution through variational optimization algorithm; Step 10: Dynamically adjust the charging strategy and generate control instructions; Step 11: Execute charging control management and monitor safety status in real time.

2. The device control method for charging with a power adapter according to claim 1, characterized in that: In the step 4: The calculation of the dynamic transmission loss is based on the temperature dependence model of the cable resistance, and the temperature R(T) dependence model satisfies: R(T)=R0[1+α(TT env )]; Among them, α is the temperature coefficient of the cable material, R0 is the resistance value, T is the real-time temperature, T env is the ambient temperature.

3. The device control method for charging with a power adapter according to claim 1, characterized in that: In the step seven: The calculation of the battery aging coefficient is based on the Arrhenius equation, and the activation energy parameter in the Arrhenius equation is calibrated through a battery cycle aging experiment.

4. The device control method for charging with a power adapter according to claim 1, wherein: In the step nine: The variational optimization algorithm involves constructing a functional that incorporates transmission loss, battery aging, and temperature constraints, and applying the Pontryagin minimum principle to solve the analytical solution for current distribution.

5. The device control method for charging with a power adapter according to claim 4, characterized in that: The weight factor of the functional is dynamically adjusted according to the device application scenario, and the transmission loss weight ratio in the emergency charging scenario is increased to more than 70%.

6. The device control method for charging with a power adapter according to claim 1, characterized in that: In the step 10: The dynamic adjustment of the charging strategy includes real-time estimation of battery health through an extended Kalman filter algorithm, wherein the state variables in the Kalman filter algorithm include battery temperature and health.

7. The device control method for charging with a power adapter according to claim 1, characterized in that: In the step eleven: The execution of the charge control management includes starting the eddy current fan and reducing the power to 80% of the current value when the predicted temperature exceeds 90% of the safety threshold; The ratio of power reduction is linearly positively correlated with the extent to which the temperature exceeds the threshold.

8. The device control method for charging with a power adapter according to claim 1, characterized in that: The step eleven further comprises: If a metal foreign object is detected in the charging device, the output is cut off and an audible and visual alarm signal is activated, which lasts until manually reset.

Images