Charger charging current control system and method based on feedback analysis

Abstract

The present invention discloses a charger charging current control system and method based on feedback analysis, which relates to the technical field of chargers. The method relies on multiple built-in modules such as USB / PD handshake monitor, ADC sampling module, current sampler, etc. to collect key charging behavior parameters including connection time, voltage waveform, steady-state establishment time, etc. in real time, and performs data cleaning and standardization in the central processing unit to construct a behavior stability scoring function Wdev for characterizing the stability of user plug-in and unplug behavior. By comparing this function with the behavior stability threshold Wth, it is determined whether the current state is unstable. Compared with traditional technologies based on current limiting based on a single current and voltage signal, this solution has higher behavior recognition capabilities and pre-decision-making capabilities, and can perform stability modeling and risk prediction on the user's charging behavior without changing the hardware structure.

Description

Technical Field

[0001] The present invention relates to the technical field of chargers, and in particular to a charger charging current control system and method based on feedback analysis. Background Art

[0002] With the widespread adoption of new interface standards such as the USB-PD protocol and Type-C, user demand for high-power, fast charging of electronic devices continues to rise in a variety of scenarios. This is especially true in complex environments like offices, public spaces, and shared charging stations, where users frequently connect and disconnect devices in complex situations, often experiencing discontinuous behaviors such as multiple plugging and unplugging and temporary access. These behaviors directly impact the stability of current output and the physical durability of the interface. Therefore, intelligently adjusting charging current based on real-time user behavior without increasing hardware complexity has become a key research direction in intelligent power control.

[0003] In the invention application with application number CN201610803734.2, a charger output current control method and device are disclosed. This disclosure discloses a charger output current control method and device, which belongs to the field of charger design. The charger output current control method includes: when the first charger and the second charger are charging the battery, obtaining the temperature of the first charger and the temperature of the second charger; calculating the difference between the temperature of the first charger and the temperature of the second charger; when the absolute value of the difference is greater than a predetermined difference threshold, lowering the output current of the charger with the higher temperature among the first charger and the second charger, and increasing the output current of the charger with the lower temperature among the first charger and the second charger, wherein the sum of the output current of the first charger and the output current of the second charger is equal to a constant total output current. This solves the technical problem of local overheating of the terminal due to the large temperature difference between the two chargers, and achieves the technical effect of keeping the two chargers at similar temperatures.

[0004] In combination with the existing technology, the above application still has the following deficiencies:

[0005] At present, most charging control systems implement current limiting, power-off and other control measures based on static current limiting rules or protocol layer voltage and current feedback mechanisms, but ignore the actual user behavior patterns, such as frequent plugging and unplugging, short-time connection, and accidental access. In the absence of behavioral perception capabilities, the control system cannot determine whether the user behavior is an unstable access behavior, which may cause a higher current output to be released immediately when the device is just plugged in, causing a surge shock to the interface or power chip. In addition, existing systems usually do not adjust the current strategy in accordance with the time rhythm law, and lack an adaptive rhythm limitation mechanism for low power consumption periods such as "nighttime and lunch break". This neglect of "operating frequency" and "usage time" has led to significant limitations and potential risks in the actual operation of the current charging strategy;

[0006] The root cause of the above defects is that the existing technology has not established a closed-loop mechanism for behavioral feedback parameters, strategy adjustment, and control execution. The current shock caused by frequent plugging and unplugging may cause phenomena such as excessive temperature rise of the interface, contact point burning, and cracking of the connector solder joints at the physical level; at the protocol level, it may also trigger communication anomalies such as misidentification, power interruption, and device rejection. Long-term accumulation will directly affect the service life of the interface and the stability of the device, especially in high-power fast charging scenarios, where the transient current shock is more significant. In addition, in public application scenarios such as shared chargers, bus stations, and power desktops, the superposition of non-standard behaviors of multiple users will further aggravate power fluctuations, resulting in frequent restarts of the power module and abnormal output, seriously affecting the user experience and the stability of the power supply system. Therefore, it is urgent to introduce a behavior perception mechanism to model and analyze user plugging and unplugging behaviors, and drive dynamic adjustment of the current strategy based on the behavior results to achieve a safer, rhythmic, and adaptive smart charging process. Summary of the Invention

[0007] In view of the deficiencies in the prior art, the present invention provides a charger charging current control system and method based on feedback analysis, which solves the problems mentioned in the background art.

[0008] To achieve the above objectives, the present invention is implemented through the following technical solutions: comprising the following steps:

[0009] S1. Use the charger's built-in module and set the collection frequency to collect charging data in real time and transmit the charging data to the charger's storage center for data storage;

[0010] S2. Extract charging data from the storage center and send it to the central processor, where it is aggregated and preprocessed to obtain a standardized charging data set.

[0011] S3. Based on the standardized charging data set, preliminarily calculate and output the behavior stability scoring function Wdev, and set the behavior stability threshold Wth to perform preliminary comparative evaluation with the behavior stability scoring function Wdev;

[0012] S4. After the preliminary comparison and evaluation determines that the plugging and unplugging is unstable, the beat current regulation mechanism is triggered. The beat current regulation mechanism sets the rhythm segment and calculates and outputs the beat current duty cycle function Aduty based on the current behavior stability score function Wdev;

[0013] S5. Perform comprehensive calculation based on the behavior stability scoring function Wdev and the beat current duty cycle function Aduty, output the comprehensive control scoring function Zctrl, and perform secondary comparative evaluation based on the output result of the comprehensive control scoring function Zctrl, and execute the corresponding control strategy based on the secondary comparative evaluation result.

[0014] Preferably, said S1 includes S11 and S12;

[0015] S11, collect charging data in real time through the charger's built-in module and set the collection frequency;

[0016] The charger's built-in module includes a USB / PD handshake monitor, an ADC sampling module, a built-in current sampler, an active chip RTC module, and a work log;

[0017] The charging data includes connection start timestamp, disconnection timestamp, voltage waveform, current delivery time, charging current stabilization time, system time and connection duration;

[0018] The connection start timestamp, disconnection timestamp and connection duration are triggered by plug-in and unplug events, and the collection frequency is set to 1ms each time;

[0019] The voltage waveform is sampled instantaneously by access, and the sampling frequency is set to 1kHz;

[0020] The current delivery time and charging current stabilization time are periodically sampled after connection, and the sampling frequency is set to 500Hz;

[0021] The collection frequency of the system time is updated once every minute;

[0022] The connection start timestamp and disconnection timestamp are monitored in real time by the USB / PD handshake monitor, and each time the charger is connected and disconnected, the timestamp of the USB / PD handshake completion / disconnection event is recorded;

[0023] The voltage waveform is obtained by using the built-in ADC sampling module to collect the voltage time series and capture the voltage waveform 200ms before and after the access moment;

[0024] The current delivery time and charging current stabilization time are detected by a built-in current sampler to detect the time it takes for the current to enter a stable state from fluctuation;

[0025] The system time is directly read by the active chip RTC module;

[0026] The connection duration is extracted in real time by accessing the work log;

[0027] S12. Setting the internal Bluetooth channel of the charger, wirelessly connecting the storage center with each built-in module of the charger, and transmitting the charging data collected in real time to the storage center for data storage.

[0028] Preferably, said S2 includes S21;

[0029] S21. Wirelessly connect the storage center to the central processing unit via a set internal Bluetooth channel, extract charging data from the storage center in real time, and perform preprocessing in the central processing unit. The preprocessing includes data cleaning, data calculation, and standardization to obtain a standardized charging data set.

[0030] The S21 includes S211, S212 and S213;

[0031] S211, the data cleaning is performed by performing abnormal elimination, smoothing time window processing and smoothing filtering on all parameters in the charging data;

[0032] The abnormal elimination is to mark the flash connection behavior within less than 0.5 seconds as invalid plug-in and removal, eliminate the data of flash connection behavior within less than 0.5 seconds, and calculate the average value of valid connections within the sliding window of 30 seconds through the smoothing time window;

[0033] The smoothing filter eliminates the peak interference of current and voltage by using a smoothing filter;

[0034] S212: Perform data calculation based on the cleaned charging data to obtain a charging data set, wherein the charging data set includes the duration Td of the i-th connection. i , the i-th connection voltage disturbance ratio Rd i , the steady-state delay of the i-th connection Yc i , the current system time Ttime and the duration of the last successful connection w;

[0035] The charging data set includes the i-th connection duration Td i Obtained by calculating the difference between the connection start timestamp and the disconnection timestamp;

[0036] The i-th connection voltage disturbance ratio Rd i The maximum voltage drop amplitude at the moment of connection is extracted from the voltage waveform, and the time required for the voltage to recover to the steady state is analyzed. The ratio of the maximum voltage drop amplitude at the moment of connection to the time required for the voltage to recover to the steady state is calculated to obtain the value.

[0037] The i-th connection steady-state delay Yc i The value is obtained by calculating the difference between the current delivery time and the charging current stabilization time;

[0038] The current system time Ttime and the last successful connection duration w are both obtained by direct extraction;

[0039] S213 , using a Z-Score normalization method based on the acquired charging data set, standardize all parameters in the charging data set, eliminate the dimensional units between all parameters, and acquire a standardized charging data set.

[0040] Preferably, said S3 includes S31 and S32;

[0041] S31. Based on the obtained standardized charging data set, calculate and output a charger behavior stability scoring function Wdev during the charging process, and analyze the overall behavior trend of the charger during the charging process;

[0042] The behavior stability scoring function Wdev is calculated and outputted by the following algorithm formula:

[0043] ;

[0044] Where N represents the total number of plugging and unplugging times, Represents a very small positive number, with a value of 10 -3 , to avoid division by zero, and denote the voltage disturbance weight value and the steady-state delay weight value respectively, and + =1, the specific value is set by the user.

[0045] Preferably, S32, based on the user setting a behavior stability threshold Wth according to the charger engineering tolerance boundary design principle, a preliminary comparison and evaluation analysis of the charger behavior stability is performed by comparing the behavior stability scoring function Wdev obtained in real time with the behavior stability threshold Wth. The specific evaluation content is as follows;

[0046] When the behavior stability scoring function Wdev is less than the behavior stability threshold Wth, it means that the charger's plug-in charging behavior is stable, and fast charging is allowed at this time;

[0047] When the behavior stability scoring function Wdev ≥ the behavior stability threshold Wth, it means that the charger's plug-in charging behavior is unstable, and the beat current regulation mechanism is triggered.

[0048] Preferably, said S4 includes S41 and S42;

[0049] S41. After the beat current regulation mechanism is triggered based on the preliminary evaluation, a 15-minute strategy update cycle is configured in the charger to obtain the rhythm control cycle length Tcycle;

[0050] At the same time, the user configures fixed rhythm segments in the charger based on charging needs. The fixed rhythm segments include active segments and inactive segments. When the current system time Ttime falls within the fixed rhythm segment, the user analyzes how many minutes of the current rhythm control cycle length Tcycle falls within the active segment, obtains the active period time percentage Tactive, and uses the Z-Score normalization method to eliminate the dimensionality effect of the rhythm control cycle length Tcycle and the active period time percentage Tactive.

[0051] S42. Extract the current charger's behavior stability score function Wdev, combine it with the dimensionless rhythm control cycle length Tcycle and the active period time proportion Tactive, and perform a comprehensive calculation to output the beat current duty cycle function Aduty. This function determines the proportion of power-on time in the beat charging mechanism when the behavior is unstable.

[0052] The beat current duty cycle function Aduty is calculated and output by the following algorithm formula:

[0053] ;

[0054] Where Wdev max Represents the maximum tolerance threshold of the behavior stability scoring function, which is used to ensure that the duty cycle is not negative.

[0055] Preferably, said S5 includes S51, S52 and S53;

[0056] S51, performing comprehensive calculation based on the obtained behavior stability score function Wdev and the beat current duty cycle function Aduty, outputting a comprehensive control score function Zctrl, and coupling the user's charging behavior and time rhythm;

[0057] The comprehensive control scoring function Zctrl is calculated and output by the following algorithm formula:

[0058] ;

[0059] Wherein, the comprehensive control scoring function Zctrl∈(0,1), e represents the exponential function, and k represents the inhibition factor.

[0060] Preferably, S52, performing a secondary comparative evaluation based on the output result of the comprehensive control scoring function Zctrl, and dividing the current charger charging into four levels of control status based on the secondary comparative evaluation result, with specific evaluation contents as follows;

[0061] When the comprehensive control score function Zctrl>0.75, it is divided into the first-level control state;

[0062] When 0.55<comprehensive control score function Zctrl≤0.75, it is divided into the secondary control state;

[0063] When 0.35<comprehensive control score function Zctrl≤0.55, it is divided into the third level control state;

[0064] When the comprehensive control score function Zctrl≤0.35, it is divided into four levels of control state.

[0065] Preferably, S53, based on the four levels of control states divided by the secondary evaluation results, executing corresponding control strategies, controlling the power module, PWM controller, MOS controller, S-curve array and DAC controller through the MCU, and adaptively controlling the power supply, voltage and current respectively;

[0066] When it is divided into the first-level control state, the MCU controls the power module to limit the contact current output, using the conventional constant current and constant voltage mode without the need for intermittent beat processing;

[0067] When the secondary control state is reached, the MCU controls the power module to enable timed interrupts and limit the current limit to 1.2A. The MOS controller then initiates light-beat control, which uses a 5-second power-on and 1-second power-off control, gradually increasing the current to the target within 5 seconds using a slow-rise current curve.

[0068] When the three-level control state is reached, the S-curve array and DAC controller are used to refine the S-shaped current output point by point, while forcibly limiting the upper current limit to 0.8A and starting a strong beat. The strong beat is controlled by powering on for 3 seconds and powering off for 2 seconds.

[0069] When the control state is divided into the fourth level, the maximum current limit is reduced to 0.5A, and the protection beat is started. The protection beat is controlled by powering on for 2 seconds and powering off for 3 seconds.

[0070] A charger charging current control system based on feedback analysis, including a charging behavior perception module, a charging data processing module, a charging behavior analysis module, a rhythm analysis module and a charging control module;

[0071] The charging behavior sensing module collects charging data in real time through the charger's built-in module and sets the collection frequency, and transmits the charging data to the charger's storage center for data storage;

[0072] The charging data processing module extracts the charging data from the storage center and sends it to the central processing unit, where it is aggregated and pre-processed to obtain a standardized charging data set.

[0073] The charging behavior analysis module preliminarily calculates and outputs a behavior stability scoring function Wdev based on a standardized charging data set, and sets a behavior stability threshold Wth to perform a preliminary comparative evaluation with the behavior stability scoring function Wdev;

[0074] The rhythm analysis module triggers the beat current regulation mechanism after determining that the plugging and unplugging is unstable through preliminary comparison and evaluation. The beat current regulation mechanism sets the rhythm segment and calculates and outputs the beat current duty cycle function Aduty based on the current behavior stability score function Wdev;

[0075] The charging control module performs a comprehensive calculation based on the behavior stability scoring function Wdev and the beat current duty cycle function Aduty, outputs a comprehensive control scoring function Zctrl, performs a secondary comparative evaluation based on the output result of the comprehensive control scoring function Zctrl, and executes a corresponding control strategy based on the secondary comparative evaluation result.

[0076] The present invention provides a charger charging current control system and method based on feedback analysis. It has the following beneficial effects:

[0077] (1) This method relies on multiple built-in modules such as the USB / PD handshake monitor, ADC sampling module, and current sampler to collect key charging behavior parameters including connection time, voltage waveform, and steady-state establishment time in real time, and performs data cleaning and standardization in the central processing unit to construct a behavior stability scoring function Wdev to characterize the stability of the user's plug-in and unplug behavior. By comparing this function with the behavior stability threshold Wth, it can accurately identify whether the user has a behavior pattern of frequent plug-in and unplugging, excessive disturbance, or poor stability, and thus determine whether the current charging state is unstable. Compared with traditional current limiting technologies based only on a single current or voltage signal, this solution has higher behavior recognition capabilities and pre-decision-making capabilities. It can perform stability modeling and risk prediction on the user's charging behavior without changing the hardware structure, thereby improving the overall safety and response accuracy of the system.

[0078] (2) This method further introduces a beat current regulation mechanism based on the judgment of unstable behavior. By configuring the rhythm control cycle and the fixed active period segment, the rhythm control cycle length Tcycle and the active period time proportion Tactive are obtained, and the beat current duty cycle function Aduty is output in combination with the behavior stability scoring function Wdev. This function is used to dynamically adjust the proportion of intermittent power-on time, so that when the user behavior is unstable or in an inactive period, the power-on cycle is automatically shortened, the power-off beat is extended, and a rhythmic current limiting strategy is implemented. This design not only solves the problems of lack of rhythm recognition ability and rough beat regulation in the existing technology, but also significantly improves the charging power supply rhythm control accuracy of the charger in cross-time and multi-user scenarios, and has good adaptability and energy saving.

[0079] (3) This method constructs a comprehensive control scoring function Zctrl by comprehensively calculating the behavior stability scoring function Wdev and the beat current duty cycle function Aduty, and divides the current state of the charger into four control levels based on the output of the scoring function. Each control level is associated with different current limiting values, current rising trajectories and beat cycle settings, and is accurately executed by MCU in conjunction with PWM, MOS, DAC and other control devices. This multi-dimensional scoring model combines the dynamic factors of user behavior risk and time period rhythm, avoiding the problem that the fixed current limiting strategy in the existing solution cannot distinguish between usage scenarios, and realizes a flexible transition from fast charging to micro charging and hierarchical power supply control, greatly improving the stability and interface protection capabilities of the system in complex charging environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0080] Figure 1 This is a schematic diagram of the steps of the charger charging current control method based on feedback analysis of the present invention;

[0081] Figure 2 This is a flow chart of a charger charging current control system based on feedback analysis according to the present invention;

[0082] Figure 3 Score trend chart for user behavior stability. DETAILED DESCRIPTION

[0083] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. 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.

[0084] Example 1

[0085] See also Figure 1 The present invention provides a charger charging current control method based on feedback analysis. To achieve the above object, the present invention is implemented through the following technical solutions: comprising the following steps:

[0086] S1. Use the charger's built-in module and set the collection frequency to collect charging data in real time and transmit the charging data to the charger's storage center for data storage;

[0087] S2. Extract charging data from the storage center and send it to the central processor, where it is aggregated and preprocessed to obtain a standardized charging data set.

[0088] S3. Based on the standardized charging data set, preliminarily calculate and output the behavior stability scoring function Wdev, and set the behavior stability threshold Wth to perform preliminary comparative evaluation with the behavior stability scoring function Wdev;

[0089] S4. After the preliminary comparison and evaluation determines that the plugging and unplugging is unstable, the beat current regulation mechanism is triggered. The beat current regulation mechanism sets the rhythm segment and calculates the output beat current duty cycle function Aduty based on the current behavior stability score function Wdev;

[0090] S5. Perform comprehensive calculation based on the behavior stability scoring function Wdev and the beat current duty cycle function Aduty, output the comprehensive control scoring function Zctrl, and perform secondary comparative evaluation based on the output result of the comprehensive control scoring function Zctrl, and execute the corresponding control strategy based on the secondary comparative evaluation result.

[0091] In this embodiment, the method uses multiple built-in acquisition units such as the USB / PD handshake monitor, ADC voltage sampling module, current detection module, RTC system clock module, etc. to collect various charging behavior parameters including connection duration, voltage disturbance, current steady-state establishment time, system time, etc. at different sampling frequencies, and transmits the original data to the local storage center of the charger through the Bluetooth channel. The data extracted from the storage center is cleaned, anomalies are eliminated and standardized by the central processing unit to construct a standardized charging data set to ensure that subsequent calculations are stable and reliable. The behavior stability scoring function Wdev is calculated based on the standardized data set, and a preliminary comparison and judgment is made in combination with the preset behavior stability threshold Wth to realize the identification of whether the user's plugging and unplugging is stable. When it is determined to be an unstable behavior, the beat current regulation mechanism is started, and the beat current duty cycle function Aduty is calculated based on the set rhythm cycle and active period division to determine the acceptable power-on time ratio. The comprehensive control scoring function Zctrl is output from the behavior stability scoring function Wdev and the beat current duty cycle function Aduty, and the current charging behavior is divided into four levels of control states based on this score, corresponding to the execution of different current limiting amplitudes, beat rhythms and current rising curve control strategies. Through the above implementation, the present invention realizes the fusion perception, real-time analysis and strategy linkage control of the user's plugging and unplugging behavior and rhythmic state without changing the hardware structure, and constructs a charging current management method with behavior modeling capabilities and hierarchical response capabilities. This method not only effectively suppresses the current shock risk caused by frequent plugging and unplugging, but also dynamically adapts the power supply strategy in different time periods and different behavior states through the coordinated control of the beat mechanism and the current limiting strategy.

[0092] Example 2

[0093] See also Figure 1,Specifically: S1 includes S11 and S12;

[0094] S11, collect charging data in real time through the charger's built-in module and set the collection frequency;

[0095] The charger's built-in modules include a USB / PD handshake monitor, an ADC sampling module, a built-in current sampler, an active chip RTC module, and a work log.

[0096] Charging data includes connection start timestamp, disconnection timestamp, voltage waveform, current delivery time, charging current stabilization time, system time, and connection duration;

[0097] The connection start timestamp, disconnection timestamp and connection duration are triggered by plug-in and unplug events, and the collection frequency is set to 1ms each time;

[0098] The voltage waveform is sampled instantaneously by accessing it, and the sampling frequency is set to 1kHz;

[0099] The current delivery time and charging current stabilization time are periodically sampled after connection, and the sampling frequency is set to 500Hz;

[0100] The collection frequency of system time is updated once every minute;

[0101] The connection start timestamp and disconnection timestamp are monitored in real time by the USB / PD handshake monitor. Each charger connection start and disconnection is recorded with the timestamp of the USB / PD handshake completion / disconnection event.

[0102] The voltage waveform is obtained by using the built-in ADC sampling module to collect the voltage time series and capture the voltage waveform 200ms before and after the access moment;

[0103] The current delivery time and charging current stabilization time are detected by the built-in current sampler to detect the time it takes for the current to enter a stable state from fluctuation;

[0104] The system time is directly read through the active chip RTC module;

[0105] The connection duration is extracted in real time through access work logs;

[0106] S12. Setting the internal Bluetooth channel of the charger, wirelessly connecting the storage center with each built-in module of the charger, and transmitting the charging data collected in real time to the storage center for data storage.

[0107] In this embodiment, the method configures a variety of acquisition modules inside the charger, and each module corresponds to the behavioral parameter acquisition frequency of different dimensions. Charging data. By configuring a Bluetooth wireless communication channel inside the charger, high-speed wireless synchronization of data between each built-in module and the local storage center is achieved. Through the Bluetooth channel, each collected plug-in and unplug event, voltage and current fluctuations, stability parameters and timing identifiers can be uploaded to the storage center in real time to achieve localized distributed management of data and subsequent unified calls. Through the above implementation, the present invention establishes a high-frequency multi-source, multi-module collaborative charging behavior acquisition method with wireless integrated communication capabilities, which realizes comprehensive sampling and efficient convergence of user plug-in and unplug behavior, electrical response status and rhythmic time characteristics without increasing additional hardware costs. Compared with the traditional method of triggering current limiting protection only by instantaneous voltage or current fluctuations, this structural design has higher behavior recognition granularity and data completeness.

[0108] Example 3

[0109] See also Figure 1 and Figure 3 , specifically: S2 includes S21;

[0110] S21. Wirelessly connect the storage center to the central processing unit via a set internal Bluetooth channel, extract charging data from the storage center in real time, and perform preprocessing in the central processing unit. The preprocessing includes data cleaning, data calculation, and standardization to obtain a standardized charging data set.

[0111] S21 includes S211, S212 and S213;

[0112] S211, data cleaning by performing abnormal elimination, smoothing time window processing and smoothing filtering on all parameters in the charging data;

[0113] Abnormal elimination is achieved by marking flash connections within 0.5 seconds as invalid plug-ins and removals. Data with flash connections within 0.5 seconds are then eliminated. The average value of valid connections within a 30-second sliding window is calculated using a smoothing time window.

[0114] Smoothing filter eliminates the peak interference of current and voltage by using smoothing filter;

[0115] S212: Perform data calculation based on the cleaned charging data to obtain a charging data set, where the charging data set includes the duration Td of the i-th connection. i , the i-th connection voltage disturbance ratio Rd i , the steady-state delay of the i-th connection Yc i , the current system time Ttime and the duration of the last successful connection w;

[0116] The charging data set includes the duration of the i-th connection Td i Obtained by calculating the difference between the connection start timestamp and the disconnection timestamp;

[0117] The i-th connection voltage disturbance ratio Rd i The maximum voltage drop amplitude at the moment of connection is extracted from the voltage waveform, and the time required for the voltage to recover to the steady state is analyzed. The ratio of the maximum voltage drop amplitude at the moment of connection to the time required for the voltage to recover to the steady state is calculated to obtain the value.

[0118] The steady-state delay of the i-th connection is Yc i The value is obtained by calculating the difference between the current delivery time and the charging current stabilization time;

[0119] The current system time Ttime and the duration of the last successful connection w are both obtained by direct extraction;

[0120] S213 , using a Z-Score normalization method based on the acquired charging data set, standardize all parameters in the charging data set, eliminate the dimensional units between all parameters, and acquire a standardized charging data set.

[0121] In this embodiment, the method wirelessly connects the storage center to the central processing unit (CPU) via an internal Bluetooth channel, extracts charging data in real time, and performs phased data preprocessing within the CPU. An anomaly rejection strategy identifies and removes flashing connection behavior data of less than 0.5 seconds, marking them as invalid plug-ins. The fluctuations are then smoothed by combining the average value of the connection within a 30-second sliding window. A filtering algorithm is then used to smooth short-period noise, such as current and voltage spikes, thereby eliminating spurious fluctuations caused by hardware jitter and transient interference. Structured parameter extraction and logical calculations are performed based on the cleaned data to form a core charging dataset for scoring analysis. Finally, the Z-Score normalization method is used to normalize the parameters to a unified dimension, ensuring the comparability and numerical stability of the impact of each parameter on the scoring function across different dimensions. This results in a standardized charging dataset, which serves as the input for subsequent behavior scoring and rhythm control functions. Through this implementation, the present invention establishes a multi-level data cleaning and feature extraction mechanism for charging behavior, which continuously outputs a standardized charging dataset with a standardized structure, clear feature expression, no noise interference, and unified dimensions during charger operation.

[0122] Example 4

[0123] See also Figure 1 and Figure 3 ,Specifically: S3 includes S31 and S32;

[0124] S31. Based on the obtained standardized charging data set, calculate and output a charger behavior stability scoring function Wdev during the charging process, and analyze the overall behavior trend of the charger during the charging process;

[0125] The behavior stability scoring function Wdev is calculated and output by the following algorithm formula;

[0126] ;

[0127] Where N represents the total number of plugging and unplugging times, Represents a very small positive number, with a value of 10 -3 , to avoid division by zero, and denote the voltage disturbance weight value and the steady-state delay weight value respectively, and + =1, the specific value is set by the user;

[0128] It represents the inverse of the connection duration, indicating that the user frequently plugs and unplugs or charges discontinuously. This is a risk signal and must be penalized with a high score.

[0129] Indicates the voltage disturbance amplification item. If the user causes a large voltage fluctuation when inserting, it indicates poor physical contact or capacitance / protocol mutation; with the help of the voltage disturbance weight value , increase the weight of the disturbance;

[0130] Represents the steady-state penalty term. The longer the steady-state establishment delay, the longer it takes for the current to stabilize. This may be due to slow protocol handshake, contact problems, or matching failure. It increases the behavior score and is judged as unstable.

[0131] Finally, the entire cycle is averaged to conduct a comprehensive analysis and evaluation of all the user's charging connection behaviors.

[0132] S32. Based on the user setting a behavior stability threshold Wth according to the charger engineering tolerance boundary design principle, a preliminary comparison is performed between the behavior stability scoring function Wdev obtained in real time and the behavior stability threshold Wth to evaluate and analyze the charger behavior stability. The specific evaluation content is as follows;

[0133] When the behavior stability scoring function Wdev is less than the behavior stability threshold Wth, it means that the charger's plug-in charging behavior is stable, and fast charging is allowed at this time;

[0134] When the behavior stability scoring function Wdev ≥ the behavior stability threshold Wth, it means that the charger's plug-in charging behavior is unstable, and the beat current regulation mechanism is triggered;

[0135] The design principle of the charger engineering tolerance boundary is: the behavior stability threshold Wth is lower than 0.6, indicating that the behavior is generally reliable, and the interface impact and electrical fluctuations are within the "controllable tolerable range";

[0136] Setting it to 0.6 can effectively intercept the following typical scenarios:

[0137] The user plugs and unplugs multiple times within 1 minute;

[0138] Voltage disturbance amplitude>0.3V, recovery time>100ms;

[0139] The current stabilization delay exceeds 1.5 seconds;

[0140] If the setting is too low, such as 0.4, the protection mechanism may be mistakenly triggered, resulting in a poor user experience;

[0141] If the setting is too high, such as 0.8, some dangerous operations will not be intercepted.

[0142] In this embodiment, the method calculates and outputs a behavioral stability scoring function Wdev, which is used to describe the overall behavioral characteristics of the charger, based on the standardized charging data set obtained in the previous step. This function forms a unified scoring index by weighted calculation of multiple parameters such as the connection duration, voltage disturbance intensity, and current steady-state establishment delay during multiple plug-in and unplug processes. The formula structure adopts a three-term weighted average model, in which: the inverse term of the connection duration is used to penalize frequent plug-in and unplug behavior; the voltage disturbance term reflects the degree of physical contact fluctuation through the ratio of the disturbance amplitude to the recovery time; and the steady-state delay term quantifies the time required for current stabilization to determine whether the device has protocol delays or electrical compatibility issues. The average calculation is performed through N plug-in and unplug behavior cycles to fully reflect the overall usage stability trend of the user in the current cycle. Based on the engineering tolerance boundary design principle, a behavioral stability threshold Wth is set and compared with the behavioral stability scoring function Wdev calculated in real time. The setting of this threshold refers to the actual engineering tolerance, and has a good recognition rate near the value of 0.6. It can effectively intercept typical high-risk behaviors such as multiple plugging and unplugging within one minute, voltage disturbance greater than 0.3V, and current stabilization delay exceeding 1.5 seconds, and realize adaptive policy control in different environments. Through this implementation, the present invention constructs a quantitative risk identification mechanism based on plug-in behavior data. Through the combination of the behavior stability scoring function Wdev and the behavior stability threshold Wth, it realizes quantitative analysis and trigger judgment of user behavior stability without relying on complex hardware structure. Compared with the traditional method of directly triggering current limiting based on current threshold, this solution has higher behavior recognition granularity and more flexible control response entrance, which effectively improves the behavior perception ability and strategy accuracy of charging in multiple users and multiple scenarios, and significantly enhances the risk prevention ability and interface security.

[0143] Example 5

[0144] See also Figure 1 , specifically: S4 includes S41 and S42;

[0145] S41. After the beat current regulation mechanism is triggered based on the preliminary evaluation, a 15-minute strategy update cycle is configured in the charger to obtain the rhythm control cycle length Tcycle;

[0146] At the same time, users configure fixed rhythm segments in the charger based on charging needs. Fixed rhythm segments include active segments and inactive segments. When the current system time Ttime falls within a fixed rhythm segment, the system analyzes how many minutes of the current rhythm control cycle length Tcycle falls within the active segment, obtains the active period time percentage Tactive, and uses the Z-Score normalization method to eliminate the dimensionality effect of the rhythm control cycle length Tcycle and the active period time percentage Tactive.

[0147] Example table of fixed rhythm segments:

[0148]

[0149] Example calculation:

[0150] The current time is 11:53, and the rhythm control cycle length Tcycle = 15 minutes;

[0151] The cycle starts at 11:45 and ends at 12:00;

[0152] The effective active time is from 11:45 to 12:00, a total of 15 minutes;

[0153] Then: the rhythm control cycle length Tcycle=15, the active period time proportion Tactive=15;

[0154] S42. Extract the current charger's behavior stability score function Wdev. Combined with the dimensionless rhythm control cycle length Tcycle and the active period time ratio Tactive, a comprehensive calculation is performed to output the beat current duty cycle function Aduty. This determines the proportion of power-on time in the beat charging mechanism when the behavior is unstable, thereby achieving intermittent current output control for unstable behavior.

[0155] The beat current duty cycle function Aduty is calculated and output by the following algorithm formula;

[0156] ;

[0157] Where Wth max It represents the maximum tolerance threshold of the behavior stability scoring function, which is used to ensure that the duty cycle is not negative;

[0158] Represents the behavior risk adjustment factor. If the behavior stability scoring function Wdev is close to 1, it means that the behavior is very unstable and the duty cycle is close to 0. If the behavior stability scoring function Wdev is close to 0, it means that the behavior is very stable and the duty cycle is close to 1. Function: Use user behavior stability to control the proportion of power-on time allowed;

[0159] Indicates the rhythm tolerance factor. If the current rhythm is active, such as office hours, this ratio is close to 1; if it is in an inactive period, such as lunch break or at night, the value is close to 0.2-0.4. Effect: Even if the user's behavior is good, the beat frequency should be reduced in the inactive period to improve protection.

[0160] In this embodiment, when the behavior stability scoring function Wdev determines unstable behavior, the method activates the rhythm adjustment mechanism. First, the policy update cycle is set to 15 minutes, denoted as the rhythm control cycle length Tcycle. Based on the charger's preconfigured fixed conclusion segments (e.g., the fixed rhythm segment example table), the current time Ttime is matched to the rhythm segments, identifying the number of minutes in the current cycle that fall within the active segment and calculating the active time percentage Tactive. These two parameters are then dimensionlessly normalized using the Z-Score method to eliminate interference from time segment length differences in the algorithm's scoring and ensure a consistent scoring benchmark across different cycle conditions. The current behavior score Wdev is extracted and input into the rhythm control cycle length Tcycle and the active time percentage Tactive into the rhythm current duty cycle function to calculate the output duty cycle value Aduty. This function comprehensively considers the behavior risk adjustment factor and the rhythm tolerance factor, and is composed of the active time percentage, implementing a rhythmic power supply control strategy based on a dual-factor feedback mechanism. Even if the user's behavior score is low, if they are in the inactive segment, the duty cycle will be actively compressed, implementing a more conservative power-on rhythm control. On the contrary, if the behavioral risk is low and it is during the active period during the day, the duty cycle is allowed to be higher to maximize the power supply efficiency. Through the above implementation, the present invention constructs a rhythm-responsive current regulation mechanism, which effectively integrates the user operation behavior and the usage rhythm for the first time, forming a power-on duty cycle control algorithm that takes into account both behavioral safety and time scenario adaptability. Compared with the traditional method of current limiting based only on static behavioral thresholds, this mechanism has the triple control capabilities of time sensitivity, risk perception, and adaptive allocation. It can not only effectively suppress the risk of continuous fast charging under unstable behavior, but also realize on-demand power supply strategy during peak load or weak load periods, thereby significantly improving the safety, stability and energy-saving efficiency of operation, and is especially suitable for high-frequency fast charging equipment environments in scenarios where multiple users share and use alternately day and night.

[0161] Example 6

[0162] See also Figure 1 , specifically: S5 includes S51, S52 and S53;

[0163] S51, performing comprehensive calculation based on the obtained behavior stability score function Wdev and the beat current duty cycle function Aduty, outputting a comprehensive control score function Zctrl, and coupling the user's charging behavior and time rhythm;

[0164] The comprehensive control scoring function Zctrl is calculated and output by the following algorithm formula;

[0165] ;

[0166] Where, the comprehensive control score function Zctrl∈(0,1), e represents the exponential function, and k represents the inhibition factor. The specific value is set by the user to control the nonlinear attenuation strength of the behavior score in the formula;

[0167] It represents the behavior score exponential decay function, which is used to map the behavior score exponential decay to a weakening factor. At the same time, multiplying it with the beat current duty cycle function Aduty is the current acceptable total output intensity score.

[0168] S52. Perform a secondary comparative evaluation based on the output result of the comprehensive control scoring function Zctrl, and classify the current charger charging into four levels of control status based on the secondary comparative evaluation result. The specific evaluation contents are as follows;

[0169] When the comprehensive control score function Zctrl>0.75, it is divided into the first-level control state;

[0170] When 0.55<comprehensive control score function Zctrl≤0.75, it is divided into the secondary control state;

[0171] When 0.35<comprehensive control score function Zctrl≤0.55, it is divided into the third level control state;

[0172] When the comprehensive control score function Zctrl≤0.35, it is divided into four levels of control state.

[0173] S53, based on the four levels of control state divided by the secondary evaluation result, executing the corresponding control strategy, controlling the power module, PWM controller, MOS controller, S-curve array and DAC controller through the MCU, and adaptively controlling the power supply, voltage and current respectively;

[0174] When it is divided into the first-level control state, the MCU controls the power module to limit the contact current output, using the conventional constant current and constant voltage mode without the need for intermittent beat processing;

[0175] Practical application: User behavior is very stable and is active during the day; for example, users use the device stably and for long periods of time in the office.

[0176] When the secondary control state is reached, the MCU controls the power module to enable timed interrupts and limit the current limit to 1.2A. The MOS controller then initiates light-beat control, which uses a 5-second power-on and 1-second power-off control scheme to gradually increase the current to the target within 5 seconds using a slow-rise current curve.

[0177] Practical application: The user frequently plugs and unplugs the device but the device successfully recognizes the user, or the user is taking a lunch break but their behavior is relatively stable.

[0178] When divided into three-level control states, the S-curve array and DAC controller refine the S-shaped current output point by point. The S-shaped curve avoids the inductive impact caused by the step-like current mutation. At the same time, it forcibly limits the current upper limit to 0.8A and starts a strong beat. The strong beat is controlled by powering on for 3 seconds and powering off for 2 seconds.

[0179] Practical application: Users have obvious plug-in and unplug disturbance behavior; it is during the morning or evening hours, not peak hours, and protection strategies are required.

[0180] When the control state is divided into the fourth level, the maximum current limit is reduced to 0.5A, and the protection beat is started. The protection beat is controlled by 2 seconds to power on and 3 seconds to power off.

[0181] Practical application: Users frequently plug and unplug, with voltage disturbances > 0.5V; during the low-peak period at night, the overall load is high and communication interference may be large; it is particularly suitable for public shared charging equipment and high-density office deployment.

[0182] In this embodiment, the method calculates and outputs a comprehensive control scoring function Zctrl by combining the behavior stability scoring function WdevW obtained based on the previous calculation with the beat current duty cycle function Aduty. A secondary scoring evaluation is performed based on the output result of the comprehensive control scoring function Zctrl, and the current charging state is divided into four control levels. Each level corresponds to a different current limiting amplitude, current output curve and beat cycle control logic to ensure the implementation of accurate and differentiated power supply strategies under different risk levels. The division of the scoring boundaries is based on engineering tolerance and the statistical laws of behavior fluctuations, taking into account both safety and user experience. Based on the above-mentioned level results, the MCU is called to jointly execute the PWM controller, MOS controller, S-curve current rise array and DAC module to execute the corresponding voltage and current refined control strategies. For example, in the first-level state, the constant voltage / constant current fast charging mode is directly enabled; in the second-level state, a light beat of 5 seconds on and 1 second off is achieved with a slow-rising current curve; in the third-level state, an S-shaped current slow-rising curve is implemented with a medium beat strategy of 3 seconds on and 2 seconds off; and in the fourth-level protection state, a micro-charging rhythm of 2 seconds on and 3 seconds off is implemented, and the current is limited to below 0.5A to ensure interface safety and relieve bus load pressure. Through this implementation, the present invention establishes an intelligent current limiting system with multi-parameter fusion scoring drive, multi-level response strategy matching, and programmable current regulation execution. It can generate a comprehensive control score in real time according to user behavior stability and time rhythm factors, and intelligently decide the optimal charging strategy based on this. Without changing the existing charger hardware structure, this solution realizes continuous dynamic switching from fast charging to micro-charging, and has higher charging behavior adaptability, abnormal connection protection capability and interface protection accuracy. It is particularly suitable for complex application scenarios such as multi-user sharing, high plug-in and unplug frequency, and irregular rhythm.

[0183] Example 7

[0184] See also Figure 1 and Figure 2 , a charger charging current control system based on feedback analysis, including a charging behavior perception module, a charging data processing module, a charging behavior analysis module, a rhythm analysis module and a charging control module;

[0185] The charging behavior sensing module collects charging data in real time through the charger's built-in module and sets the collection frequency, and transmits the charging data to the charger's storage center for data storage;

[0186] The charging data processing module extracts the charging data from the storage center and sends it to the central processor, where it is aggregated and pre-processed to obtain a standardized charging data set.

[0187] The charging behavior analysis module preliminarily calculates and outputs the behavior stability scoring function Wdev based on the standardized charging data set, and sets the behavior stability threshold Wth for preliminary comparative evaluation with the behavior stability scoring function Wdev;

[0188] After the rhythm analysis module determines that the plugging and unplugging is unstable through preliminary comparison and evaluation, it triggers the beat current regulation mechanism. The beat current regulation mechanism sets the rhythm segment and calculates the output beat current duty cycle function Aduty based on the current behavior stability score function Wdev;

[0189] The charging control module performs comprehensive calculations based on the behavior stability scoring function Wdev and the beat current duty cycle function Aduty, outputs the comprehensive control scoring function Zctrl, and performs a secondary comparative evaluation based on the output result of the comprehensive control scoring function Zctrl, and executes the corresponding control strategy based on the secondary comparative evaluation result.

[0190] While the embodiments of the present invention have been shown and described, it will be apparent to those skilled in the art that various changes, modifications, substitutions, and alterations can be made to the embodiments without departing from the principles and spirit of the invention.

Claims

1. A charger charging current control method based on feedback analysis, characterized in that: The following steps are involved: S1. Use the charger's built-in module and set the collection frequency to collect charging data in real time and transmit the charging data to the charger's storage center for data storage; S2. Extract charging data from the storage center and send it to the central processor, where it is aggregated and preprocessed to obtain a standardized charging data set. S3. Based on the standardized charging data set, preliminarily calculate and output the behavior stability scoring function Wdev, and set the behavior stability threshold Wth to perform preliminary comparative evaluation with the behavior stability scoring function Wdev; Said S3 includes S31 and S32; S31. Based on the obtained standardized charging data set, calculate and output a charger behavior stability scoring function Wdev during the charging process, and analyze the overall behavior trend of the charger during the charging process; The behavior stability scoring function Wdev is calculated and outputted by the following algorithm formula: ； Where N represents the total number of plugging and unplugging times. Represents a very small positive number, with a value of 10 -3 , to avoid division by zero, and denote the voltage disturbance weight value and the steady-state delay weight value respectively, and + =1, the specific value is set by the user, Td i Indicates the duration of the i-th connection, Rd i Indicates the voltage disturbance ratio of the i-th connection, Yc i represents the steady-state delay of the i-th connection; S4. After the preliminary comparison and evaluation determines that the plugging and unplugging is unstable, the beat current regulation mechanism is triggered. The beat current regulation mechanism sets the rhythm segment and calculates and outputs the beat current duty cycle function Aduty based on the current behavior stability score function Wdev; Said S4 includes S41 and S42; S41. After the beat current regulation mechanism is triggered based on the preliminary evaluation, a 15-minute strategy update cycle is configured in the charger to obtain the rhythm control cycle length Tcycle; At the same time, the user configures fixed rhythm segments in the charger based on charging needs. The fixed rhythm segments include active segments and inactive segments. When the current system time Ttime falls within the fixed rhythm segment, the user analyzes how many minutes of the current rhythm control cycle length Tcycle falls within the active segment, obtains the active period time percentage Tactive, and uses the Z-Score normalization method to eliminate the dimensionality effect of the rhythm control cycle length Tcycle and the active period time percentage Tactive. S42. Extract the current charger's behavior stability score function Wdev, combine it with the dimensionless rhythm control cycle length Tcycle and the active period time proportion Tactive, and perform a comprehensive calculation to output the beat current duty cycle function Aduty. This function determines the proportion of power-on time in the beat charging mechanism when the behavior is unstable. The beat current duty cycle function Aduty is calculated and output by the following algorithm formula: ； Where Wdev max represents the maximum tolerance threshold of the behavior stability scoring function; S5. Perform comprehensive calculation based on the behavior stability scoring function Wdev and the beat current duty cycle function Aduty, output a comprehensive control scoring function Zctrl, perform a secondary comparative evaluation based on the output result of the comprehensive control scoring function Zctrl, and execute a corresponding control strategy based on the secondary comparative evaluation result; Said S5 includes S51; S51, performing comprehensive calculation based on the obtained behavior stability score function Wdev and the beat current duty cycle function Aduty, outputting a comprehensive control score function Zctrl, and coupling the user's charging behavior and time rhythm; The comprehensive control scoring function Zctrl is calculated and output by the following algorithm formula: ； Wherein, the comprehensive control scoring function Zctrl∈(0,1), e represents the exponential function, and k represents the inhibition factor.

2. The charger charging current control method based on feedback analysis according to claim 1, characterized in that: Said S1 includes S11 and S12; S11, collect charging data in real time through the charger's built-in module and set the collection frequency; The charger's built-in module includes a USB / PD handshake monitor, an ADC sampling module, a built-in current sampler, an active chip RTC module, and a work log; The charging data includes connection start timestamp, disconnection timestamp, voltage waveform, current delivery time, charging current stabilization time, system time and connection duration; The connection start timestamp, disconnection timestamp and connection duration are triggered by plug-in and unplug events, and the collection frequency is set to 1ms each time; The voltage waveform is sampled instantaneously by access, and the sampling frequency is set to 1kHz; The current delivery time and charging current stabilization time are periodically sampled after connection, and the sampling frequency is set to 500Hz; The collection frequency of the system time is updated once every minute; The connection start timestamp and disconnection timestamp are monitored in real time by the USB / PD handshake monitor, and each time the charger is connected and disconnected, the timestamp of the USB / PD handshake completion / disconnection event is recorded; The voltage waveform is obtained by using the built-in ADC sampling module to collect the voltage time series and capture the voltage waveform 200ms before and after the access moment; The current delivery time and charging current stabilization time are detected by a built-in current sampler to detect the time it takes for the current to enter a stable state from fluctuation; The system time is directly read by the active chip RTC module; The connection duration is extracted in real time by accessing the work log; S12. Setting the internal Bluetooth channel of the charger, wirelessly connecting the storage center with each built-in module of the charger, and transmitting the charging data collected in real time to the storage center for data storage.

3. The charger charging current control method based on feedback analysis according to claim 2, characterized in that: Said S2 includes S21; S21. Wirelessly connect the storage center to the central processing unit via a set internal Bluetooth channel, extract charging data from the storage center in real time, and perform preprocessing in the central processing unit. The preprocessing includes data cleaning, data calculation, and standardization to obtain a standardized charging data set. The S21 includes S211, S212 and S213; S211, the data cleaning is performed by performing abnormal elimination, smoothing time window processing and smoothing filtering on all parameters in the charging data; The abnormal elimination is to mark the flash connection behavior within less than 0.5 seconds as invalid plug-in and removal, eliminate the data of flash connection behavior within less than 0.5 seconds, and calculate the average value of valid connections within the sliding window of 30 seconds through the smoothing time window; The smoothing filter eliminates the peak interference of current and voltage by using a smoothing filter; S212: Perform data calculation based on the cleaned charging data to obtain a charging data set, wherein the charging data set includes the duration Td of the i-th connection. i , the i-th connection voltage disturbance ratio Rd i , the steady-state delay of the i-th connection Yc i , the current system time Ttime and the duration of the last successful connection w; The charging data set includes the i-th connection duration Td i Obtained by calculating the difference between the connection start timestamp and the disconnection timestamp; The i-th connection voltage disturbance ratio Rd i The maximum voltage drop amplitude at the moment of connection is extracted from the voltage waveform, and the time required for the voltage to recover to the steady state is analyzed. The ratio of the maximum voltage drop amplitude at the moment of connection to the time required for the voltage to recover to the steady state is calculated to obtain the value. The i-th connection steady-state delay Yc i The value is obtained by calculating the difference between the current delivery time and the charging current stabilization time; The current system time Ttime and the last successful connection duration w are both obtained by direct extraction; S213 , using a Z-Score normalization method based on the acquired charging data set, standardize all parameters in the charging data set, eliminate the dimensional units between all parameters, and acquire a standardized charging data set.

4. The charger charging current control method based on feedback analysis according to claim 1, characterized in that: S32. Based on the user setting a behavior stability threshold Wth according to the charger engineering tolerance boundary design principle, a preliminary comparison is performed between the behavior stability scoring function Wdev obtained in real time and the behavior stability threshold Wth to evaluate and analyze the charger behavior stability. The specific evaluation content is as follows; When the behavior stability scoring function Wdev is less than the behavior stability threshold Wth, it means that the charger's plug-in charging behavior is stable, and fast charging is allowed at this time; When the behavior stability scoring function Wdev ≥ the behavior stability threshold Wth, it means that the charger's plug-in charging behavior is unstable, and the beat current regulation mechanism is triggered.

5. The charger charging current control method based on feedback analysis according to claim 1, characterized in that: Said S5 further includes S52; S52. Perform a secondary comparative evaluation based on the output result of the comprehensive control scoring function Zctrl, and classify the current charger charging into four levels of control status based on the secondary comparative evaluation result. The specific evaluation contents are as follows; When the comprehensive control score function Zctrl>0.75, it is divided into the first-level control state; When 0.55<comprehensive control score function Zctrl≤0.75, it is divided into the secondary control state; When 0.35<comprehensive control score function Zctrl≤0.55, it is divided into the third level control state; When the comprehensive control score function Zctrl≤0.35, it is divided into four levels of control state.

6. The charger charging current control method based on feedback analysis according to claim 5, characterized in that: Said S5 also includes S53; S53, based on the four levels of control state divided by the secondary evaluation result, executing the corresponding control strategy, controlling the power module, PWM controller, MOS controller, S-curve array and DAC controller through the MCU, and adaptively controlling the power supply, voltage and current respectively; When it is divided into the first-level control state, the MCU controls the power module to limit the contact current output, using the conventional constant current and constant voltage mode without the need for intermittent beat processing; When the secondary control state is reached, the MCU controls the power module to enable timed interrupts and limit the current limit to 1.2A. The MOS controller then initiates light-beat control, which uses a 5-second power-on and 1-second power-off control, gradually increasing the current to the target within 5 seconds using a slow-rise current curve. When the three-level control state is reached, the S-curve array and DAC controller are used to refine the S-shaped current output point by point, while forcibly limiting the upper current limit to 0.8A and starting a strong beat. The strong beat is controlled by powering on for 3 seconds and powering off for 2 seconds. When the control state is divided into the fourth level, the maximum current limit is reduced to 0.5A, and the protection beat is started. The protection beat is controlled by powering on for 2 seconds and powering off for 3 seconds.

7. A charger charging current control system based on feedback analysis, applied to the charger charging current control method based on feedback analysis according to any one of claims 1 to 6, characterized in that: It includes charging behavior perception module, charging data processing module, charging behavior analysis module, rhythm analysis module and charging control module; The charging behavior sensing module collects charging data in real time through the charger's built-in module and sets the collection frequency, and transmits the charging data to the charger's storage center for data storage; The charging data processing module extracts the charging data from the storage center and sends it to the central processor, where it is aggregated and pre-processed to obtain a standardized charging data set; The charging behavior analysis module preliminarily calculates and outputs a behavior stability scoring function Wdev based on a standardized charging data set, and sets a behavior stability threshold Wth to perform a preliminary comparative evaluation with the behavior stability scoring function Wdev; The rhythm analysis module triggers the beat current regulation mechanism after determining that the plugging and unplugging is unstable through preliminary comparison and evaluation. The beat current regulation mechanism sets the rhythm segment and calculates and outputs the beat current duty cycle function Aduty based on the current behavior stability score function Wdev; The charging control module performs a comprehensive calculation based on the behavior stability scoring function Wdev and the beat current duty cycle function Aduty, outputs a comprehensive control scoring function Zctrl, performs a secondary comparative evaluation based on the output result of the comprehensive control scoring function Zctrl, and executes a corresponding control strategy based on the secondary comparative evaluation result.

Images