An anti-interference compensation method for an intelligent driving circuit with nonlinear load characteristics

Abstract

The present disclosure relates to the technical field of intelligent drive control, and particularly relates to an anti-interference compensation method for an intelligent drive circuit with nonlinear load characteristics, which comprises the following steps: collecting drive circuit operation data, constructing load characteristic description information representing the electrical characteristics of the nonlinear load, determining the current load working section based on the load characteristic description information, constructing comprehensive disturbance information by combining the predicted output and the actual output, performing component analysis on the comprehensive disturbance, generating a drive adjustment parameter set, modifying the drive control signal and outputting it to the power drive module; at the same time, forming a drive evaluation parameter according to the drive response, performing closed-loop adjustment on the drive adjustment parameter to generate a drive adjustment parameter set corresponding to the next drive cycle. The present disclosure combines nonlinear load characteristic identification, disturbance analysis and drive adjustment process, and realizes dynamic anti-interference compensation for different nonlinear load operating states.

Description

Technical Field

[0001] This disclosure relates to the field of intelligent drive control technology, and more specifically to an anti-interference compensation method for intelligent drive circuits with nonlinear load characteristics. Background Technology

[0002] Intelligent drive circuits are widely used in motor drives, electromagnetic actuators, piezoelectric devices, LED drivers, and industrial power supplies. However, during actual debugging, several problems often arise. For example, during the cold start phase of an LED driver, the output current may fluctuate briefly within hundreds of milliseconds, causing the LED string brightness to flicker within a visible range. At the end of the electromagnetic actuator's engagement, although the coil current has reached the set value, the push rod may still exhibit a slight rebound. When a piezoelectric device approaches its resonance point, the drive voltage waveform may appear normal, but the actual displacement may suddenly deviate from the original calibration value. In some industrial power supplies, when the load switches from a light load to a pulse load, the output terminal may first experience a very narrow overshoot, followed by several damped oscillations. These phenomena are often not caused by a single parameter drift, but rather by the combined effects of the load's impedance, energy storage state, switching transients, and external coupling interference within a short period.

[0003] Currently, techniques such as disturbance observers, active disturbance rejection control, load observers, or model predictive control are commonly used for nonlinear load drive systems. These techniques involve establishing a model of the controlled object, estimating the disturbance between the system output and the model prediction, and combining feedforward compensation and feedback compensation to suppress the impact of power supply fluctuations, load changes, and external disturbances on drive performance. While these methods can indeed reduce output fluctuations to a small range on some experimental platforms, in field equipment, it is often found that even with the same output deviation, it can be caused by various factors. Sometimes this is due to changes in the internal state of the load, such as increased resistance after coil temperature rise or local saturation of the magnetic core; other times it is caused by external factors, such as a momentary voltage dip in the upstream bus. Existing methods typically treat these deviations uniformly as a total disturbance. Although the compensation amount can change with the output error, it cannot distinguish whether the change is due to a change in the load's own state or an external disturbance. This leads to lag, overcompensation, or deviation in the compensation direction. Especially when the load is in a stage of magnetic saturation, increased back EMF, capacitive-inductive conversion, or significant changes in nonlinear current-voltage characteristics, the manifestation of disturbances is not fixed: sometimes it is mainly manifested as a steeper current rise edge, sometimes as a sudden increase in output ripple, and sometimes as abnormal temperature rise of the driver switching transistor and forced frequent correction of the PWM duty cycle. For these phenomena, existing control methods often can only rely on empirical parameter tuning, adding filtering links, or repeated experiments to mitigate the phenomenon, but it is difficult to separate and compensate for disturbances from different sources without changing the hardware structure. Summary of the Invention

[0004] This disclosure provides an anti-interference compensation method for intelligent drive circuits with nonlinear load characteristics. The aim is to construct a disturbance characterization and compensation control mechanism for nonlinear load driving processes. By describing the dynamic electrical behavior of the load under driving conditions, analyzing the sources of disturbance, and establishing corresponding compensation control relationships, the method realizes adaptive adjustment of the drive control signal and dynamic optimization of compensation parameters, thereby improving the anti-interference capability, drive stability, and control reliability of intelligent drive circuits under complex nonlinear load conditions.

[0005] This disclosure provides an anti-interference compensation method for intelligent drive circuits with nonlinear load characteristics, including: The system acquires operational data during the process of an intelligent drive circuit driving a nonlinear load. The operational data includes at least the drive control signal, input voltage, output voltage, output current, power device operating status, and load status information. Based on the operating data, load characteristic description information describing the dynamic electrical behavior of nonlinear loads under the current driving condition is constructed, and disturbance characterization information corresponding to the current driving condition is established based on the load characteristic description information. The disturbance characterization information includes the driving output prediction result, the actual driving output result, and the comprehensive disturbance information formed by the driving output prediction result and the actual driving output result. Based on the disturbance characterization information, the disturbance source is analyzed for the comprehensive disturbance information to generate disturbance composition information including load-intrinsic disturbances, power supply disturbances, switching disturbances, and externally coupled disturbances; Based on the disturbance component information, a compensation control relationship is established between each disturbance component and the drive adjustment object, and compensation control information corresponding to the current drive condition is generated based on the compensation control relationship. Based on the compensation control information, the drive control signal is corrected, a compensated drive control signal is generated, and the nonlinear load is driven. The disturbance characterization information is updated based on the compensated drive output, and the compensation control relationship is adjusted based on the updated disturbance characterization information.

[0006] Compared with existing technologies, the intelligent drive circuit anti-interference compensation method with nonlinear load characteristics provided by this invention constructs nonlinear load characteristic parameters from the operating data during the drive process, and participates the nonlinear load characteristic parameters in the load state identification, disturbance analysis, and drive adjustment process. This allows the drive adjustment process to be controlled in conjunction with changes in the electrical characteristics of the load, no longer limited to fixed parameters or uniform compensation methods, which is beneficial to improving the consistency of drive control under different operating conditions. At the same time, this invention distinguishes the comprehensive disturbances generated during the drive process into load disturbances, power supply disturbances, switching disturbances, and external coupling disturbances, and participates in the generation of drive adjustment parameters respectively. This allows disturbances from different sources to correspond to different adjustment objects, reducing the impact of the superposition of different disturbances on the drive control process, and facilitating the maintenance of drive output continuity during nonlinear load operation. In addition, this invention continues to feed back the drive response results to the subsequent drive adjustment parameter formation process, so that the drive adjustment parameters can be adjusted in conjunction with the response of the current drive cycle, forming a continuous closed-loop adjustment process. This is beneficial to adapting to changes in load electrical characteristics and operating conditions, and is applicable to nonlinear load drive scenarios such as motors, electromagnetic actuators, piezoelectric actuators, and LED arrays. Attached Figure Description

[0007] To more clearly illustrate the technical solutions in the embodiments of this disclosure or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this disclosure. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort, wherein: Figure 1 The flowchart of the present disclosure is illustrated schematically; Figure 2 This schematically illustrates a timing diagram of data processing within the drive cycle provided in this disclosure; Figure 3 A schematic diagram illustrating the nonlinear load characteristic construction provided in this disclosure is shown. Figure 4 The schematic diagram illustrates the integrated disturbance composition provided in this disclosure; Figure 5 A schematic diagram of the drive control signal correction provided in this disclosure is shown. Detailed Implementation

[0008] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this disclosure, and not all embodiments. Based on the embodiments of this disclosure, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this disclosure.

[0009] Before providing a detailed description of the technical solution of this disclosure, it should be noted that the anti-interference compensation method for intelligent drive circuits with nonlinear load characteristics described in this disclosure is designed for load objects with nonlinear electrical characteristics. It establishes a disturbance sensing, disturbance analysis, and compensation control mechanism on the basis of existing intelligent drive control architecture, capable of adapting to the dynamic changes in the load. Existing engineering implementation methods such as drive control, current closed-loop control, power drive, voltage sampling, analog-to-digital conversion, and PWM drive generation all have relatively mature technical solutions. Those skilled in the art can select and configure these methods according to specific application scenarios. The relevant content has been described in detail in the existing technology and will not be repeated here.

[0010] The specific application environment architecture or specific hardware architecture on which the implementation of the intelligent drive circuit anti-interference compensation method with nonlinear load characteristics depends is first described in the specific application environment architecture of this disclosure. This disclosure can be deployed in industrial servo drives, permanent magnet synchronous motor controllers, stepper motor drivers, solenoid valve controllers, relay drive modules, piezoelectric actuator drive units, LED constant current drive modules, industrial power supply modules, and intelligent power drive systems built with MOSFET, IGBT, GaN, or SiC power devices. In the above application scenarios, nonlinear loads typically exhibit continuous changes in load impedance with varying drive voltage, current, operating frequency, magnetic flux state, mechanical motion state, device temperature rise, or operating stage. For example, during the solenoid valve drive process, the equivalent inductance and excitation current change significantly when the coil transitions from the non-engaged stage to the engaged stage. During the start-up, acceleration, and steady-state operation of a permanent magnet synchronous motor, the continuous change in back electromotive force leads to continuous changes in the output current response characteristics. During the piezoelectric actuator drive process, different drive voltage ranges correspond to different capacitive reactance characteristics. During the LED array constant current drive process, the rise in device junction temperature will cause the volt-ampere characteristics to drift. All of these changes will cause the traditional drive control model based on fixed parameters to gradually deviate from the actual load state, making problems such as output ripple, current distortion, drive overshoot, and transient oscillations more pronounced.

[0011] Based on the aforementioned engineering background, the intelligent drive control system upon which this disclosure relies mainly includes a power input unit, a power drive unit, a nonlinear load execution unit, a status acquisition unit, and a control processing unit. The power input unit provides stable power to the entire drive system and can be configured with surge protection, EMI filtering, power isolation, and overvoltage protection circuits according to actual engineering needs. The power drive unit controls the on / off state of the bridge arm power devices based on drive control signals to output corresponding drive energy to the nonlinear load. The status acquisition unit includes a voltage sampling module, a current sampling module, a temperature sampling module, and a power device operating status acquisition module. Voltage sampling can be achieved using isolated sampling, resistor voltage divider sampling, or differential sampling. Current sampling can be achieved using sampling resistors, Hall current sensors, or magnetoresistive sensors. Temperature sampling can be achieved using NTC thermistors, digital temperature sensors, or on-chip temperature feedback from power devices. Each sampled signal is input to the control processing unit after analog conditioning, filtering, and analog-to-digital conversion. The control processing unit can be implemented using an MCU, DSP, FPGA, SoC, or a combination thereof, and executes the control logic described in this disclosure, including load characteristic construction, disturbance characterization establishment, disturbance source analysis, compensation control generation, and parameter dynamic updating.

[0012] It should be noted that this disclosure does not add a new power drive structure, but rather builds an information processing link for the dynamic drive process of nonlinear loads based on the aforementioned hardware. During operation, the control processing unit first receives the drive operation data acquired by the status acquisition unit and, in conjunction with the drive control signal, establishes load characteristic description information that can describe the dynamic electrical behavior of the load under the current drive condition. Subsequently, based on the load characteristic description information and the drive control process, it establishes disturbance characterization information corresponding to the current drive condition. By predicting the comprehensive disturbance information formed between the drive output and the actual drive output, it further analyzes the cause of the disturbance. Then, it analyzes the comprehensive disturbance according to the disturbance formation mechanism to obtain different disturbance components such as load-intrinsic disturbance, power supply disturbance, power device switching disturbance, and external coupling disturbance. Finally, it establishes corresponding compensation control relationships according to different disturbance components, performs targeted corrections to the drive control signal, and continuously adjusts the compensation control relationship based on the compensated drive results, so that the entire drive system can maintain high drive stability even when the load characteristics continue to change.

[0013] Compared with existing intelligent drive control schemes, it is necessary to further explain that the closed-loop control, disturbance observation and compensation control schemes widely used in engineering can usually generate compensation control quantities based on output deviations and achieve drive stability through feedback control. The main difference lies in the type of controller or the selection method of control parameters. However, such schemes usually reflect load parameter changes, external electromagnetic interference, power fluctuations and switching transients as a unified output error, and then generate compensation control quantities based on the unified error. When the equivalent impedance, energy storage characteristics and dynamic response law of the nonlinear load continue to change, due to the superposition of disturbances from different sources, the unified error is difficult to accurately reflect the formation mechanism of various disturbances. This makes the compensation control quantity simultaneously undertake the functions of load change compensation and external disturbance suppression, which can easily lead to insufficient compensation, overcompensation or control lag.

[0014] This disclosure adopts a different implementation method from the above-mentioned technical routes. Instead of directly using output error as the sole basis for compensation, it first establishes load characteristic description information that can reflect the dynamic electrical behavior of the load, then establishes disturbance characterization information based on the load characteristic description information, and further analyzes the comprehensive disturbance into disturbance components from different sources. Finally, it establishes corresponding compensation control relationships for different disturbance components, realizing the transformation of the compensation control basis from a single output deviation to disturbance component information. Since the compensation control relationship directly corresponds to disturbances from different sources, the drive control process can dynamically adjust different control objects according to the current drive conditions, so that load state changes, power fluctuations, power device switching processes, and external coupling interference correspond to different control and adjustment strategies, rather than using a uniform compensation method for processing.

[0015] Furthermore, the technical features disclosed herein are not simply superimposed, but rather form a hierarchically related information processing process. Load characteristic description information is used to characterize the dynamic electrical behavior of the nonlinear load under the current driving condition, providing a foundation for establishing disturbance characterization information. Disturbance characterization information is used to establish the correlation between predicted drive output and actual drive output, providing a unified analysis object for disturbance source analysis. The disturbance source analysis results further determine the control object and control method corresponding to the compensation control relationship, and the compensated drive result serves as the basis for updating the disturbance characterization information and compensation control relationship in the next drive cycle, thus forming a closed-loop dynamic optimization mechanism for the entire drive process. The above stages have clear data dependencies and logical progression relationships; changes in information at any stage can affect subsequent compensation control processes. Therefore, the entire control flow can continuously adapt to changes in the nonlinear load's operating state, rather than relying on fixed parameters or preset compensation tables to complete drive control.

[0016] Based on the above technical approach, this disclosure can not only reduce the compensation deviation caused by load impedance changes during nonlinear load operation, but also distinguish between dynamic load changes and external disturbances, so that disturbances from different sources correspond to different compensation objects, thereby improving the pertinence of compensation control. At the same time, since the compensation control relationship can be continuously updated with the driving conditions, it can adapt to complex engineering scenarios such as load aging, device parameter drift, ambient temperature changes, and power quality fluctuations during long-term operation, reducing the probability of output ripple, current distortion, drive overshoot, and control oscillation, and improving the drive stability, control accuracy, and operational reliability of the intelligent drive system under complex nonlinear load conditions.

[0017] The following detailed explanation, in conjunction with the accompanying drawings and specific embodiments, describes the detailed implementation process of the anti-interference compensation method for intelligent drive circuits with nonlinear load characteristics as disclosed in this disclosure. For details, please refer to... Figure 1 As shown: A method for anti-interference compensation of intelligent drive circuits with nonlinear load characteristics, comprising: S1. Obtain operational data during the process of the intelligent drive circuit driving a nonlinear load. The operational data includes at least the drive control signal, input voltage, output voltage, output current, power device operating status, and load status information. For example, see [link to example]. Figure 2 The diagram illustrates the data processing timing within a drive cycle. A single drive cycle includes sequential processing stages such as signal sampling, feature extraction, disturbance analysis, drive adjustment, and PWM output. At the start of the drive cycle, the controller synchronously samples the input voltage, output voltage, output current, load temperature, and the switching state of the power devices, generating corresponding operating data for the drive cycle based on the sampling results. Subsequently, load features are extracted from the operating data, and disturbance analysis is performed based on the extracted load feature parameters to generate drive adjustment parameters corresponding to the current drive cycle. These drive adjustment parameters are converted into drive control signals, which are output to the power drive module within the current drive cycle to complete load drive. Sampling processing is repeated in the next drive cycle to form a continuous closed-loop control process. Note... Figure 2 The data processing relationships within the same driving cycle shown for each processing stage are only used to illustrate the temporal correlation between the processing objects of this invention, and do not limit the specific execution time of each processing stage.

[0018] S2. Based on the operating data, construct load characteristic description information describing the dynamic electrical behavior of the nonlinear load under the current driving condition, and establish disturbance characterization information corresponding to the current driving condition based on the load characteristic description information. The disturbance characterization information includes the driving output prediction result, the actual driving output result, and the comprehensive disturbance information formed based on the driving output prediction result and the actual driving output result. It should be noted that this disclosure establishes a unified data acquisition foundation for information from three different levels that exist simultaneously during the drive control process: power supply changes, power device switching processes, and load dynamic responses. Therefore, the control processing unit synchronously acquires drive control signals, input voltage, output voltage, output current, power device operating status, and load status information in each drive control cycle, and uses the above data as the raw operating data for subsequently constructing load characteristic description information. Those skilled in the art will understand that although different types of intelligent drive systems have different load objects, the drive control link all have a basic closed-loop structure of "controller outputs drive command - power device executes conduction control - nonlinear load response - sampling feedback controller". Therefore, this disclosure does not limit the specific controller model or the specific power drive form. For example, in applications using MOSFET full-bridge driven solenoid valves, the control processing unit can output drive control signals at a PWM control frequency of 20kHz, with each PWM cycle corresponding to 50μs. The current sampling module can use a Hall current sensor in conjunction with a 12-bit analog-to-digital converter for synchronous sampling, with the sampling frequency set to 100kHz, meaning a current sample is completed every 10μs. The output voltage can be obtained using isolated differential sampling to acquire the voltage at the bridge arm output. For the operating status of the power devices, the bridge arm on / off status, fault status, and other information can be directly read from the driver chip feedback, or the current bridge arm conduction interval and dead time information can be obtained through the PWM counter status register. For temperature information, it can be obtained using an NTC thermistor, a digital temperature sensor, or the temperature feedback signal integrated within the power device, depending on the actual engineering configuration. All of the above acquisition methods are existing mature technologies and will not be elaborated further here. It should be noted that, due to the significant differences in the physical processes corresponding to various sampling signals, their change rates and update cycles are not consistent. For example, PWM drive signals and bridge arm switching states are typically updated using one PWM cycle as the update unit, while the output current may rapidly establish itself within microseconds in the initial stage of bridge arm conduction, and the load temperature usually takes tens of milliseconds or even several seconds to change significantly. If data from different time scales are directly used together in subsequent analysis, it is easy to mix data generated under different drive conditions, resulting in the load state constructed at the same moment not accurately reflecting the current drive process. Therefore, after completing the acquisition of various operating data, this disclosure first establishes a data organization mechanism under a unified sampling period. In one optional embodiment, the control processing unit uses the drive control cycle as a unified time reference, assigns corresponding timestamps to various operating data acquired within the current control cycle, and establishes corresponding data caches. In the driving control cycle, when the update frequency of a certain type of data is higher than that of the current control cycle (e.g., high-speed sampling data such as output current and output voltage), all sampling results within the current control cycle are retained, and the data closest to the end of the control cycle is extracted according to a preset time window as the valid sampling value corresponding to the current driving condition. When the update frequency of a certain type of data is lower than that of the current control cycle (e.g., load temperature information), the most recent valid update value is retained until the next sampling is completed, ensuring that all operating data entering the subsequent calculation process corresponds to the same driving condition. For control systems using DMA continuous sampling in the field, a double-buffered data caching mechanism can also be used to complete data acquisition for the next control cycle while completing data processing in the current control cycle, avoiding data overwriting that could lead to inconsistent states.

[0019] After completing the unified organization of the operating data, the control processing unit constructs load characteristic description information. The load characteristic description information described in this disclosure is a set of characteristic parameters with unified time attributes and physical correlations used to describe the dynamic electrical behavior of the nonlinear load under the current driving condition. To facilitate subsequent disturbance prediction, this disclosure divides it into a first characteristic parameter group, a second characteristic parameter group, and a third characteristic parameter group according to different information sources. The first characteristic parameter group mainly reflects the electrical response relationship of the load itself, and its calculation objects are derived from input voltage, output voltage, and output current. In some more specific application scenarios, such as driving a solenoid valve or relay coil, because the current rises rapidly in the initial stage of coil excitation, and the rate of current change gradually decreases after the magnetic circuit approaches saturation, the control processing unit can... The equivalent impedance of the load is calculated based on the currently sampled output voltage and output current, and the dynamic incremental impedance is calculated based on the changes in voltage and current within two adjacent sampling periods. For inductive loads, the current equivalent inductance can also be estimated based on the rate of change of output voltage and current. For capacitive loads such as piezoelectric actuators and capacitive actuators, the equivalent capacitance can be estimated based on the rate of change of output current and voltage. This disclosure does not limit all of the above parameters to be included in the calculation, but rather selects one or more parameters to form the first characteristic parameter group based on the actual electrical characteristics of the load. For example, for LED constant current drives, the equivalent impedance and dynamic incremental impedance can be used to describe the changes in its volt-ampere characteristics. For permanent magnet synchronous motor drives, the equivalent impedance change information corresponding to the back electromotive force can be added to more accurately describe the current load operating state.

[0020] The second set of characteristic parameters mainly reflects the drive switching characteristics generated by the power drive process itself. Its data comes from the drive control signal and the operating status of the power device. Unlike the first set of characteristic parameters, this part of the parameters does not directly describe the load, but rather describes the way drive energy is injected into the load. For example, the control processing unit can obtain the on and off states of the bridge arm according to the PWM drive signal, obtain the switching transient characteristics according to the voltage change of the bridge arm drive node, calculate the output ripple amplitude according to the output waveform of multiple consecutive PWM cycles, and calculate the harmonic characteristics generated in the current drive process according to the output current spectrum. For high-speed drive systems using GaN or SiC power devices, due to the fast switching speed of the devices, the parasitic parameters of the bridge arm are more likely to cause high-frequency spikes. In this case, the overshoot amplitude of the switching node or the ringing duration can be added as switching transient characteristics to participate in the construction of the second set of characteristic parameters, so as to improve the ability of subsequent disturbance analysis to characterize the switching process.

[0021] The third set of characteristic parameters is mainly used to reflect the changing trend of the nonlinear load's operating state, and its data comes from load state information. Compared with the first two types of parameters, this part of the parameters focuses more on the load's operating stage rather than instantaneous electrical quantities. For example, during the solenoid valve's engagement, as the coil is continuously energized, the iron core position gradually changes, and the excitation current build-up rate changes. During motor driving, as the speed increases, the back electromotive force gradually increases, and the current build-up time prolongs. During LED driving, as the device junction temperature continues to rise, its on-state voltage drop gradually decreases. Therefore, the control processing unit can extract parameters such as load temperature, current change rate, voltage change rate, and output response delay as the third set of characteristic parameters to describe the current operating stage of the load and its dynamic changing trend.

[0022] Furthermore, after the first, second, and third characteristic parameter groups are constructed, the control processing unit performs time alignment on each parameter group according to a unified sampling period, and organizes them into a set of characteristic parameters corresponding to the same driving condition according to a pre-set parameter arrangement order, forming the load characteristic description information corresponding to the current driving cycle. It should be noted that this load characteristic description information not only retains the numerical information of each characteristic parameter, but also maintains the unified time attribute and driving condition attribute corresponding to each parameter, so that data from different physical processes can jointly describe the dynamic electrical behavior of the nonlinear load within the same driving cycle, and there is no need to perform time matching or state correction on data from different sources again.

[0023] As an example, see Figure 3 The diagram illustrates the construction of nonlinear load characteristics, where the horizontal axis represents the load terminal voltage and the vertical axis represents the load terminal current. Different operating points correspond to the load's volt-ampere characteristics under different driving states. The controller determines the current operating point based on the voltage and current sampling data collected within the same driving cycle, and extracts equivalent impedance, dynamic incremental impedance, and inductance or capacitance-related characteristics reflecting the load's energy storage characteristics by combining the changing relationships between adjacent operating points, thus forming a set of load characteristic parameters corresponding to the current driving cycle. After completing the construction of the load characteristic description information, the control processing unit establishes the predicted drive output result under the current driving condition, reflecting the theoretically achievable output result based on the current drive control signal and the current load state when there are no additional disturbances. The deviation between the subsequent actual sampled output and this reference output no longer only represents control error, but can be further used as basic data for disturbance source analysis.

[0024] During implementation, the control processing unit first performs parameter normalization processing on the load characteristic description information obtained in the previous stage. It should be noted that since equivalent impedance, dynamic incremental impedance, equivalent inductance, and equivalent capacitance reflect different physical quantities, their dimensions and numerical ranges are significantly different. For example, in the case of solenoid valve drive, the equivalent impedance is usually several ohms to tens of ohms, while the equivalent inductance is usually in the millihenry range. For piezoelectric actuators, the equivalent capacitance may reach several microfarads or even higher. If all parameters are directly used in subsequent calculations, the parameters with larger values ​​may easily mask the load state reflected by the changes in other parameters. Therefore, the control processing unit first performs normalization processing based on the calibration range corresponding to each parameter, so that each characteristic parameter is converted into a unified numerical range, such as being mapped to the range of 0 to 1 or -1 to 1. The calibration range can be determined based on typical operating data collected during the product debugging stage, or a corresponding parameter range table can be established according to the actual load type. This disclosure does not limit this.

[0025] After normalization, the control processing unit organizes the equivalent impedance characteristic parameters, dynamic incremental impedance characteristic parameters, equivalent inductance characteristic parameters, and equivalent capacitance characteristic parameters in a pre-set order to form the load state vector corresponding to the current driving cycle. It should be noted that the fixed order in which the load state vector is constructed is not only for data organization, but also to ensure that the state expressions formed in different driving cycles have a consistent data structure, so that each subsequent working segment can use a unified data entry point for processing, avoiding the problem of different state expressions corresponding to the same operating state due to changes in parameter arrangement.

[0026] Meanwhile, the control processing unit calculates the actual driving capability of the power drive module within the current drive cycle based on the current drive control signal. Specifically, for intelligent drive circuits using PWM drive, the control processing unit reads the current PWM duty cycle, drive switching frequency, and bridge arm dead time. Based on the bridge arm conduction sequence, it calculates the effective on-time and effective off-time within the current switching cycle and further combines this with the bus input voltage to calculate the average output voltage of the bridge arm. As those skilled in the art will understand, the average output voltage of the bridge arm can more accurately reflect the average driving energy actually applied to both ends of the load within the current drive cycle. For example, when the drive frequency is maintained at 20kHz, the PWM duty cycle is 60%, and the dead time is 500ns, the actual average drive voltage obtained by the load is usually slightly lower than the average output voltage corresponding to the theoretical PWM duty cycle because the dead time occupies part of the on-time. By incorporating the dead time effect into the calculation process, the subsequent predicted output can be prevented from deviating from the actual drive state for a long time.

[0027] Furthermore, this disclosure determines the operating segment based on the current load characteristics. It is understood that although different types of nonlinear loads all exhibit voltage and current nonlinear relationships, the dominant factors determining the drive output differ across different operating stages. For example, in the initial excitation stage of a relay coil, current build-up is primarily influenced by inductance; after the coil magnetic circuit approaches stability, impedance changes gradually become the main influencing factor; for piezoelectric actuators, capacitive reactance changes are more pronounced in the high-voltage region. Therefore, this disclosure uses the ratio between the equivalent impedance characteristic parameter and the dynamic incremental impedance characteristic parameter to reflect the current resistive change, and simultaneously uses the ratio between the equivalent inductance characteristic parameter and the equivalent capacitance characteristic parameter to reflect... The current energy storage characteristic ratio is compared with preset segment thresholds to divide the current driving condition into resistive-dominated, inductive-dominated, or capacitive-dominated operating segments. Note that the above segment thresholds can be obtained through offline calibration during the commissioning phase of a specific product. For example, for the same model of industrial electromagnetic actuator, a large amount of operating data can be collected during the no-load, engagement, holding, and release phases, and the distribution range of characteristic parameters corresponding to different phases can be statistically analyzed to determine the judgment thresholds corresponding to each operating segment. When the corresponding ratio crosses the preset threshold during actual operation, the control processing unit completes the operating segment switching without having to rebuild the entire prediction model, allowing the entire prediction process to run continuously. During implementation, to avoid state jitter when characteristic parameters approach the segmentation threshold, a dual-threshold hysteresis judgment method can be adopted. For example, when the resistive judgment ratio is greater than 0.8 for three consecutive sampling periods, the system enters the resistive-dominated working segment; when the ratio is less than 0.7 for three consecutive sampling periods, the system exits the working segment. When multiple criteria are met simultaneously, the working segment with stronger equivalent energy storage characteristics is selected first. When the absolute value of the denominator of the calculated ratio is less than the preset lower limit (e.g., 0.01Ω or 0.1μF), the valid parameters of the previous sampling period are used for judgment. If a valid parameter cannot be obtained for 10 consecutive sampling periods, the system switches to the preset default working segment.

[0028] Once determined, the state update coefficient matrix corresponding to the working segment is used to establish the drive output prediction result. It should be noted that the state update coefficient matrix is ​​stored in the controller's storage space after offline identification based on historical sampling data during product development or system debugging. Historical data should cover each working segment and typical power fluctuations and load changes, with the minimum prediction voltage and current error as the identification target. This data is stored in the controller after confirmation by an independent verification dataset. For different working segments, corresponding state update coefficient matrices can be established separately. Each matrix includes state transition coefficients, conduction correction coefficients, turn-off correction coefficients, and voltage mapping coefficients. The state transition coefficients describe the current load state vector in adjacent drive cycles. The relationship between the changes in the two states is as follows: the conduction correction coefficient and the turn-off correction coefficient are used to compensate for the dynamic effects brought about by the conduction and turn-off processes of the bridge arm, respectively; the voltage mapping coefficient is used to establish the correspondence between the average output voltage of the bridge arm and the load output response; the control processing unit performs unified calculations on the current load state vector, effective conduction time, effective turn-off time and average output voltage of the bridge arm according to the selected state update coefficient matrix, and obtains the predicted output voltage value and predicted output current value corresponding to the current drive cycle, which are used as the drive output prediction results; it should be noted that the prediction results always correspond to the current drive cycle, and are not the prediction values ​​after multiple control cycles, so they can be used as the direct comparison objects of the current actual sampling results.

[0029] After obtaining the predicted drive output, the control processing unit continues to establish the actual drive output and comprehensive disturbance information corresponding to the current drive condition. In specific implementation, the control processing unit reads the actual output voltage value and actual output current value synchronously acquired in the current drive cycle, and compares them with the predicted output voltage value and predicted output current value respectively to calculate the voltage deviation and current deviation. Those skilled in the art will understand that the above deviations not only include the output differences caused by changes in load parameters, but may also include the combined effects of multiple factors such as power supply ripple, bridge arm switching transients, electromagnetic coupling, and external random interference. Therefore, this disclosure does not directly compensate based on a single deviation, but further establishes a disturbance sampling sequence.

[0030] Specifically, the control processing unit organizes the voltage deviation, current deviation, and sampling time corresponding to each sampling moment into a disturbance sampling unit, and writes them into the disturbance sampling sequence in the order of sampling time. In some more specific application scenarios, such as when the control period is 50μs and the sampling period is 10μs, multiple disturbance sampling units can be formed in each control period. The control processing unit continues to perform time-series combination of multiple consecutive disturbance sampling units according to the preset sampling window. For example, a sliding sampling window covering four consecutive control periods can be used, or the sampling window length can be dynamically adjusted according to the load switching stage to ensure that the disturbance information can reflect both short-term switching disturbances and retain the continuous change characteristics during the load state change process.

[0031] Furthermore, the control processing unit forms comprehensive disturbance information based on each disturbance sampling unit within the disturbance sampling window. The disturbance amplitude information describes the degree of disturbance change, the disturbance change direction information reflects the trend of disturbance increasing or decreasing over time, and the disturbance duration information describes the persistence characteristics of the current disturbance in the continuous driving cycle. It should be noted that this disclosure does not determine the source of the disturbance at this stage. Instead, it first establishes comprehensive disturbance information with unified time attributes and continuous evolution relationships, providing a complete data foundation for subsequent source analysis based on switching disturbances, load-inherent disturbances, power supply disturbances, and externally coupled disturbances. Thus, the subsequent disturbance analysis process no longer deals with discrete instantaneous errors, but with disturbance characterization objects possessing time continuity, amplitude variation characteristics, and persistence characteristics, effectively improving the distinguishability between different types of disturbances.

[0032] S3. Based on the disturbance characterization information, the disturbance source is analyzed for the comprehensive disturbance information to generate disturbance composition information including load-intrinsic disturbance, power supply disturbance, switching disturbance, and external coupling disturbance. It should be noted that during the operation of the intelligent drive circuit, the comprehensive disturbance formed between the predicted output and the actual output usually includes multiple influencing factors such as load nonlinear changes, power supply ripple, power device switching transients, and external electromagnetic coupling. If unified compensation is directly implemented based on the comprehensive disturbance information, the dynamic response generated during the normal operation of the load itself may be mistaken for external interference, resulting in a deviation in the compensation direction. Therefore, this disclosure establishes a multi-level disturbance analysis mechanism with physical reference significance to gradually complete the analysis of various disturbance sources under a unified time reference. In practice, the control processing unit first establishes a continuous disturbance analysis window based on the sampling time corresponding to the comprehensive disturbance information. Those skilled in the art will understand that the above-mentioned comprehensive disturbance information already has a unified sampling time and continuous time attribute, so there is no need to reorganize the original sampling data, but to directly divide the analysis window based on the comprehensive disturbance information. Note that the purpose of establishing a disturbance analysis window is not merely for frequency domain calculations, but to establish a unified time reference coordinate for disturbances with different duration characteristics. For example, transient disturbances in MOSFET or IGBT switching typically last only a few hundred nanoseconds to a few microseconds, while disturbances corresponding to load magnetic saturation processes may last for dozens of PWM drive cycles, and power frequency ripple exhibits a fixed periodic variation. If various disturbances are directly mixed for analysis, it is difficult to reflect the differences in time scale between different disturbances. Therefore, this disclosure divides the comprehensive disturbance information into multiple continuous disturbance analysis windows according to the sampling time sequence. Each disturbance analysis window covers several consecutive sampling points, and its window length can be set according to the PWM drive frequency, control cycle, or load dynamic response time. For example, in a drive system with a PWM drive frequency of 20kHz and a control cycle of 50μs, each disturbance analysis window can cover 4 to 8 consecutive control cycles to take into account the common analysis needs of transient and continuous disturbances. The window can be shortened during the load change phase and extended during the steady-state phase, depending on the situation. Non-essential features are not limited.

[0033] As an optional implementation, within each disturbance analysis window, the control processing unit extracts the disturbance amplitude sequence, disturbance frequency sequence, and disturbance phase sequence, respectively. The disturbance amplitude sequence reflects the trend of disturbance energy change over time; the disturbance frequency sequence describes the main frequency distribution of the disturbance; and the disturbance phase sequence describes the time correspondence between the disturbance and the drive control process. Existing technologies have already disclosed frequency domain analysis methods, such as Fast Fourier Transform, Discrete Fourier Transform, or Sliding Spectrum Analysis, all of which can complete the above frequency extraction process, and will not be elaborated upon here. This disclosure focuses more on the correspondence between different disturbance sequences and different physical reference sequences, rather than specific spectrum calculation methods. In practical implementation, the control processing unit prioritizes identifying switching disturbance components. Disturbances generated during the turn-on and turn-off processes of power devices have clear driving rhythms, and their main frequencies are usually distributed around the PWM switching frequency and its corresponding harmonics. Therefore, the control processing unit first reads the switching frequency of the power device corresponding to the current driving cycle and establishes a reference frequency sequence containing the fundamental frequency and multiple harmonic components. For example, when the driving frequency is set to 20kHz, reference frequency sequences of 20kHz, 40kHz, 60kHz, and 80kHz can be established. The control processing unit performs frequency matching between the disturbance frequency sequence corresponding to the disturbance analysis window and the above reference frequency sequence. When the corresponding frequency deviation is within a preset allowable range, such as when the frequency deviation is less than 3% to 5% of the corresponding reference frequency, it is determined that the disturbance data has the same frequency characteristics as the switching process of the power device, and the corresponding disturbance data is marked as a switching disturbance component. Since this type of disturbance already has a clear physical source, it will not participate in the determination of other types of disturbances in the subsequent disturbance analysis process.

[0034] As an example, for data not marked as switching disturbances, the control processing unit continues to identify load disturbances. The load disturbances described in this disclosure are determined based on whether the time of disturbance occurrence corresponds to changes in load state. This is because, during magnetic saturation, back EMF establishment, capacitive charging and discharging, and sudden load changes, although the disturbance amplitude of a nonlinear load may change significantly, its frequency is not necessarily fixed; its time correlation characteristics are more pronounced than frequency characteristics. Therefore, the control processing unit calls upon the load state change information established in the aforementioned steps to generate a state switching sequence corresponding to the driving process. The switching sequence includes the transition times of various states, such as the transition from the light load linear region to the nonlinear region, the establishment of magnetic saturation, the formation of motor back EMF, and the conversion of capacitive charging and discharging. The control processing unit aligns the disturbance phase sequence with the state switching sequence in time. When the disturbance occurs within the preset time range of the corresponding state switching time, such as within one control cycle before or after, the disturbance is considered to originate from the dynamic change process of the load itself, and the corresponding disturbance data is marked as an endogenous disturbance component of the load. Only when the voltage deviation, current deviation, or response delay during the state switching period exceeds the allowable envelope of the corresponding working segment is it marked as an endogenous disturbance of the load.

[0035] In particular, for disturbance data that is still not marked after completing the above two-level analysis, the power supply disturbance component is further identified. In practice, the input power supply usually contains periodic variations such as power frequency ripple, rectification ripple, or switching power supply output ripple. Therefore, the control processing unit establishes an input power supply ripple reference frequency sequence based on the input voltage sampling results and compares the disturbance frequency sequence corresponding to the current disturbance analysis window with the input power supply ripple reference frequency sequence. For example, in the AC rectification power supply scenario, 100Hz, 120Hz, or the corresponding rectification multiplier can be established as reference frequencies. When the DC bus is powered by a DC / DC converter, a corresponding reference frequency sequence can be established based on the DC / DC operating frequency. When the disturbance frequency meets the preset frequency deviation range, it is considered that the disturbance mainly originates from the input power supply fluctuation, and the corresponding disturbance data is marked as a power supply disturbance component.

[0036] It should be noted that after the above three-level analysis, there may still be some data in the comprehensive disturbance information that has not been fully source-marked. Those skilled in the art will understand that this part of the disturbance usually comes from changes in the on-site electromagnetic environment, such as the start-up and shutdown of adjacent high-power equipment, cable coupling, electromagnetic radiation, or grounding loop interference, etc. These random factors do not have a fixed switching cycle, nor are they synchronized with changes in load status, and they do not conform to the input power supply ripple frequency characteristics. Therefore, this disclosure does not need to continue to establish a new identification model, but instead uniformly collects all disturbance data with incomplete source marking into external coupling disturbance components. Since this type of disturbance has excluded the physical sources corresponding to switching disturbances, load disturbances, and power supply disturbances, it can more accurately reflect the random coupling interference in the actual operating environment. It should be noted that the various disturbance components are not mutually exclusive. Multiple disturbance components can be identified simultaneously within the same sampling window. Each disturbance component establishes a corresponding compensation control relationship and performs superimposed compensation based on disturbance priority or control timing. When multiple compensations are applied to the same drive cycle, they can be configured in the following order: bridge arm timing compensation first, drive output compensation second, drive cycle time compensation third, and drive constraint compensation last, in order to avoid mutual cancellation or control conflicts between different compensations. S4. Establish a compensation control relationship between each disturbance component and the drive adjustment object based on the disturbance component information, and generate compensation control information corresponding to the current drive condition based on the compensation control relationship. In practice, the compensation control relationship is not a fixed mapping relationship, but is dynamically generated based on the current load characteristic description information, disturbance composition information and drive conditions. When the drive conditions change, the same disturbance composition can correspond to different drive adjustment objects, or to compensation control quantities with different amplitudes or durations. Therefore, the compensation control relationship has a dynamic correlation characteristic that changes in real time with the drive state.

[0037] As an example, see Figure 4 The diagram illustrates the composition of the overall disturbance. The overall disturbance is formed by the superposition of multiple disturbance components in the drive output, including low-frequency disturbances corresponding to power supply ripple, dynamic disturbances caused by load state changes, transient spike disturbances generated during power device switching, and high-frequency disturbances caused by external electromagnetic coupling. The controller analyzes the different components of the overall disturbance based on its time-domain variation characteristics, frequency distribution characteristics, and correlation with the drive timing, and generates corresponding disturbance information to serve as the basis for subsequently generating drive adjustment parameters. Note that... Figure 4 The different disturbance components are only used to illustrate the source composition of the overall disturbance and do not indicate that the various disturbances exist independently of each other.

[0038] S5. Based on the compensation control information, the drive control signal is corrected to generate a compensated drive control signal and drive the nonlinear load. It should be noted that this disclosure believes that the physical location and action stage of disturbances from different sources on the intelligent drive circuit are not the same. If a unified control variable is used for compensation, it is easy for one type of disturbance to be effectively suppressed, while another type of disturbance is amplified. Therefore, this disclosure uses the disturbance source as the control entry point and the drive adjustment object as the compensation exit point, mapping different disturbances to different drive control links to achieve a one-to-one correspondence between the compensation object and the disturbance source.

[0039] In practice, the information about the components of the disturbance can be represented as ,in, This represents the switching disturbance component. This represents the intrinsic disturbance component of the load. This represents the power supply disturbance component. This represents the externally coupled disturbance component; correspondingly, this disclosure establishes a set of driving regulation objects. ,in, Indicates the object to be adjusted in the timing of the bridge arm. Indicates the object to be controlled by the output drive. Indicates the object to be adjusted in the driving cycle. This represents the object being driven and constrained; therefore, a mapping relationship can be established between disturbance composition information and the object being driven and constrained. , where the mapping function This is used to establish the correspondence between disturbance sources and the controlled objects, rather than to calculate specific control quantities. Those skilled in the art will understand that once the above mapping relationship is established, each subsequent disturbance can directly act on the corresponding controlled object without the need for unified compensation variable conversion, ensuring that disturbances from different sources can be regulated using the most suitable control method.

[0040] As an optional implementation, the control processing unit first establishes bridge arm timing adjustment parameters based on the switching disturbance components. The switching disturbances mainly originate from transient processes such as parasitic oscillations, reverse recovery, current spikes, and bridge arm cross-conduction generated during the turn-on and turn-off processes of MOSFET, IGBT, or GaN power devices. The duration of these disturbances is usually much shorter than the entire PWM control cycle. Therefore, this disclosure does not adjust the entire drive output reference value, but only adjusts the local control timing corresponding to the turn-on and turn-off of the bridge arms.

[0041] In practice, the control processing unit first reads the time interval corresponding to the switching disturbance component and establishes a corresponding time mapping relationship based on the current PWM drive cycle, locating the disturbance occurrence interval to the specific time position in the corresponding bridge arm drive cycle. In particular, when the switching disturbance is concentrated in the initial stage of bridge arm conduction, the bridge arm conduction control node can be appropriately delayed; when the disturbance is concentrated in the turn-off stage, the bridge arm turn-off control node can be appropriately advanced; for cases where there are significant disturbances in both the conduction and turn-off processes, the conduction and turn-off nodes are adjusted independently without simultaneously changing the entire PWM duty cycle. This yields the bridge arm timing adjustment parameters, which characterize the correction relationship between the bridge arm conduction timing and turn-off timing.

[0042] In practical implementation, this disclosure adopts a different processing method for load-inherent disturbances than the bridge arm timing adjustment. It should be noted that load-inherent disturbances usually originate from the dynamic characteristics of the load itself, such as the establishment of magnetic saturation, changes in motor back electromotive force, changes in piezoelectric actuator capacitive reactance, or changes in LED volt-ampere characteristics. Their effect is mainly manifested in the average drive output, rather than the local switching process. Therefore, this disclosure maps the time interval corresponding to the load-inherent disturbance to the drive output control cycle, rather than to the instant of bridge arm switching.

[0043] In practice, the control processing unit covers the corresponding disturbance range to one or more consecutive drive control cycles based on the duration of the load's intrinsic disturbance, and configures the drive output control quantity according to the range of action. Those skilled in the art will understand that the drive output control quantity can be the PWM duty cycle, or a current closed-loop reference value, voltage reference value, or other drive output reference quantity. When the driven object uses PWM control, the PWM duty cycle within the corresponding control cycle can be appropriately increased or decreased according to the direction of the load's intrinsic disturbance. For drive systems using analog output control, the drive reference voltage can be adjusted synchronously. This forms the drive output adjustment parameters, enabling the drive output to adjust synchronously with dynamic load changes without altering the bridge arm drive timing.

[0044] As an example, for power supply disturbance components, this disclosure does not adjust the drive output control quantity, but instead establishes drive cycle adjustment parameters. It should be noted that power supply ripple typically has a fixed period, primarily affecting the energy build-up process within the drive control cycle; therefore, simply increasing the output reference value often cannot effectively suppress the ripple effect. This disclosure maps the power supply disturbance's duration to the drive modulation cycle and adjusts the drive modulation cycle according to the corresponding time interval. In practice, the drive modulation cycle may include one or more of the carrier cycle, PWM modulation cycle, and drive update cycle. For example, when periodic ripple exists on the input bus, the PWM carrier update position can be appropriately adjusted to ensure that the drive update time avoids the ripple peak region. For applications using digital control... The drive system that implements PWM output can also adjust the controller parameter refresh cycle to keep the drive parameter update time staggered with the input ripple period, reducing the probability of ripple directly affecting the drive output; thus forming drive cycle adjustment parameters; in particular, for external coupling disturbances, this disclosure adopts a drive constraint adjustment method for processing; it should be noted that external coupling disturbances usually originate from factors such as random electromagnetic radiation, start-up and shutdown of nearby equipment, cable crosstalk, or grounding loop current, and their occurrence time and duration are highly random, so they cannot be effectively compensated by modifying the drive target; based on this, this disclosure maps the action time interval corresponding to the external coupling disturbance to the drive protection cycle, and configures the drive constraint control quantity according to the disturbance duration characteristics.

[0045] In practice, the drive constraint control parameters may include at least one of the drive output slope, bridge arm dead time, and drive current settling time. For example, when a sustained increase in random coupling disturbance is detected, the drive output slope can be appropriately reduced to make the bridge arm current settling process smoother. For operating conditions where the risk of bridge arm cross-conduction increases, the bridge arm dead time can be appropriately extended. For situations with large current surges, the drive current settling time can be reconfigured to reduce the device surge risk caused by external disturbances, thereby forming drive constraint adjustment parameters. This disclosure organizes the bridge arm timing adjustment parameters, drive output adjustment parameters, drive cycle adjustment parameters, and drive constraint adjustment parameters in a unified manner according to the chronological order of the time intervals of each disturbance, forming a set of drive adjustment parameters, which can be expressed as: ,in, For bridge arm timing adjustment parameters, To drive the output adjustment parameters, To drive the beat adjustment parameters, To drive the constraint adjustment parameters, in practice, the above-mentioned set of drive adjustment parameters not only retains the values ​​of each adjustment parameter itself, but also records the corresponding time interval and execution order, so that the control processing unit can coordinate the execution process of each adjustment parameter according to a unified time axis. For example, when there are both switching disturbances and power supply disturbances in the same drive cycle, the control processing unit will first execute the bridge arm timing adjustment, and then execute the drive cycle adjustment, so as to avoid different adjustment parameters acting on the same controlled object and causing control conflicts. In this way, the compensation control information corresponding to the current drive condition is established.

[0046] It should be noted that the preset adjustment coefficient in this disclosure is used to characterize the mapping relationship between the disturbance constraint coefficient and each drive constraint control quantity. Its specific value can be set according to engineering conditions such as the device parameters of the power drive module, the type of nonlinear load, the drive frequency, and the allowable current variation range, and is not limited to a fixed value. In specific implementation, the preset adjustment coefficient can be obtained through prototype testing, historical operating data statistics, device specification parameter analysis, or experimental calibration to ensure that the drive constraint control quantity meets the stability requirements of the corresponding drive system.

[0047] After the drive adjustment parameter set is constructed, the control processing unit generates a compensated drive control signal based on various drive adjustment parameters and outputs the drive control signal to the power drive module to drive the corresponding nonlinear load to complete the energy output of the current control cycle. It should be noted that this disclosure does not directly superimpose the bridge arm timing adjustment parameters, drive output adjustment parameters, drive cycle adjustment parameters, and drive constraint adjustment parameters to form a unified compensation amount. Instead, according to the actual control link of the intelligent drive circuit, each adjustment parameter is applied to different control levels, so that each type of compensation corresponds to the actual hardware control object. Thus, the entire drive control process still maintains the original control structure, and only the corresponding parameters of each control level are dynamically corrected, which can avoid the problem of control object mixing caused by the direct superposition of multiple compensation amounts.

[0048] In practice, the control processing unit first reads the bridge arm timing adjustment parameters, drive output adjustment parameters, drive cycle adjustment parameters, and drive constraint adjustment parameters from the drive adjustment parameter set, and completes parameter scheduling according to the corresponding time interval of each parameter. Since the above adjustment parameters correspond to different control layers, the control processing unit completes the bridge arm timing construction, drive output configuration, drive cycle configuration, and drive constraint configuration sequentially according to the drive signal formation order, rather than executing them according to the parameter generation order. Those skilled in the art will understand that the above execution order corresponds to the actual signal formation process of the power drive module, thus ensuring that the time base between each control parameter remains consistent.

[0049] As an optional implementation, the control processing unit first establishes the turn-on and turn-off timing corresponding to each power bridge arm based on the bridge arm timing adjustment parameters, forming the bridge arm drive timing. It should be noted that the bridge arm drive timing belongs to the bottom-level timing of the entire drive control link, which directly determines when the power devices are turned on and when they are turned off. Therefore, other control parameters are all based on the bridge arm drive timing being determined.

[0050] In practice, for a half-bridge drive structure, the turn-on and turn-off nodes corresponding to the high-side and low-side bridge arms can be established separately; for a full-bridge drive structure, the drive nodes corresponding to the four power bridge arms are established separately; for a three-phase inverter drive system, the drive timing corresponding to the A-phase bridge arm, B-phase bridge arm, and C-phase bridge arm are established separately. Existing technologies have already disclosed bridge arm drive logic, such as complementary PWM drive, synchronous rectification drive, and space vector PWM drive, which will not be elaborated here. The focus of this disclosure is to dynamically correct the turn-on and turn-off nodes corresponding to each bridge arm based on the bridge arm timing adjustment parameters; for example, when the turn-on delay corresponding to the bridge arm timing adjustment parameters is... The turn-off lead time is At that time, the driving timing of each bridge arm can be expressed as: ,in, Indicates the initial conduction time. This indicates the original shutdown time, and the corrected bridge arm drive timing serves as the basis for subsequent drive parameter configuration.

[0051] In practical implementation, the control processing unit further configures the drive output adjustment parameters according to the bridge arm drive timing to form the bridge arm drive parameters. It should be noted that the bridge arm drive parameters correspond to the drive output capability, not the timing of the bridge arm movement. Therefore, their application is to the bridge arm drive timing that has already been established.

[0052] In practice, the drive output adjustment parameters may include at least one of the following: PWM duty cycle correction, output current reference value correction, output voltage reference value correction, or drive current peak limit. When PWM modulation is used as the drive method, the control processing unit corrects the corresponding conduction time while keeping the bridge arm conduction node unchanged. For example, suppose the original PWM duty cycle is... The duty cycle correction amount corresponding to the drive output adjustment parameter is: The compensated drive duty cycle then satisfies: It should be noted that the duty cycle correction only applies to the established bridge arm drive timing and does not change the bridge arm conduction sequence, thus ensuring that the PWM modulation logic remains continuous. When the drive system adopts analog control, the output reference voltage or output reference current can be reconfigured according to the drive output adjustment parameters to form the corresponding bridge arm drive parameters.

[0053] In particular, after the bridge arm drive parameters are established, the control processing unit continues to configure the corresponding drive modulation timing according to the drive timing adjustment parameters to form the bridge arm drive timing parameters. It should be noted that the drive timing mentioned in this disclosure is not the drive output size, but the time rhythm corresponding to the update of drive control parameters and PWM modulation operation, which is mainly used to coordinate the time relationship between drive output and power supply disturbance.

[0054] In practice, the drive cycle adjustment parameters can correspond to the PWM carrier period, modulation update period, and drive parameter refresh period, respectively. For example, for the PWM carrier period, the carrier start position can be adjusted appropriately according to the duration of the power supply disturbance. For the drive parameter refresh period, the drive parameter update time can be adjusted according to the ripple change law to make the drive output avoid the input ripple peak region. When the drive controller supports multi-level modulation, the parameter update time between different modulation layers can also be adjusted synchronously. This forms the bridge arm drive cycle parameters, so that the drive control process and the power supply change process are coordinated and consistent. The parameter update adopts a limited step size. If the evaluation index deteriorates continuously after the update or overcurrent, overvoltage, or overtemperature occurs, the update is canceled and the most recent stable parameters are restored.

[0055] As an example, after the boom drive cycle parameters are determined, the control processing unit further establishes boom drive constraint parameters based on the drive constraint adjustment parameters. It should be noted that the drive constraint parameters described in this disclosure do not participate in the drive target calculation, but rather apply safety boundary restrictions to the drive control process before the drive control signal is output, so as to avoid the drive device from entering an abnormal working state due to the compensation process.

[0056] In practice, drive constraint adjustment parameters may include one or more of the following: drive output slope limit, bridge arm dead time limit, and drive current settling time limit. For example, when external electromagnetic coupling disturbances continue to increase, the drive output change slope can be reduced so that the output current gradually builds up according to a preset slope. For bridge arm dead time, the dead time compensation value can be reconfigured according to the drive constraint adjustment parameters to reduce the risk of bridge arm cross-conduction. For drive current settling time, the current settling speed can be reset according to the duration of external coupling disturbances so that the drive device always operates within the allowable safe range. Those skilled in the art will understand that the above constraint parameters do not change the drive objective, but are only used to constrain the actual execution process of the drive control signal.

[0057] It should be noted that, to ensure that the above-mentioned adjustment parameters can act together on the drive control signal according to a unified control logic, this disclosure further establishes a drive control signal combination process; in practice, the control processing unit organizes the bridge arm drive timing, bridge arm drive parameters, bridge arm drive cycle parameters, and bridge arm drive constraint parameters in a unified manner according to the correspondence of each power bridge arm, which can be expressed as:

[0058] in, Indicates the bridge arm drive timing. Indicates the bridge arm drive parameters, This indicates the bridge arm drive cycle parameters. Indicates the bridge arm drive constraint parameters. The above functional relationship, which represents the process of generating drive control signals, does not limit the specific mathematical calculation form, but is used to illustrate that each drive parameter participates in the generation of drive control signals according to a fixed control link. When the drive system includes multiple power arms, the control processing unit establishes the corresponding arm drive timing, arm drive parameters, arm drive cycle parameters, and arm drive constraint parameters for each arm, and completes synchronous combination according to the arm correspondence, ultimately forming the drive control signal output to the power drive module.

[0059] As an example, see Figure 5 The diagram shown illustrates the correction of the drive control signal. The upper part represents the original drive signal, and the lower part represents the drive signal corrected according to the drive adjustment parameters. The controller corrects the target drive range in the drive control signal based on the drive adjustment parameters corresponding to the current drive cycle. When the drive control signal adopts PWM, the phase position, pulse width, or conduction time of the target pulse can be adjusted accordingly to form the corrected PWM drive signal. When the drive control signal adopts other drive forms, the corresponding drive parameters can be adjusted according to the same control relationship. The corrected drive control signal is output to the power drive module to drive the nonlinear load into the next drive cycle. In more specific application scenarios, such as a three-phase motor drive system, the control processing unit generates corresponding compensated drive control signals for phases A, B, and C, respectively, and ensures that the three-phase bridge arms maintain a preset phase relationship. For a full-bridge electromagnetic actuator drive system, it generates corresponding compensated drive control signals for the high-side bridge arm and the low-side bridge arm, respectively, to maintain the complementary conduction relationship of the bridge arms. Since the drive control signal already includes bridge arm timing correction, drive output correction, drive cycle correction, and drive constraint correction, after being output to the power drive module, it can act on the corresponding control level for disturbances from different sources, and realize the dynamic anti-interference compensation of the intelligent drive circuit for the operation of the nonlinear load without changing the original drive topology.

[0060] After completing the above drive control signal output, this implementation continues to use the operating data collected within the same drive cycle to perform online evaluation of the current compensation control effect, and dynamically corrects the compensation control relationship based on the evaluation results, so that the compensation control relationship can be continuously updated with changes in nonlinear load characteristics, that is: S6. Update the disturbance characterization information according to the compensated drive output, and adjust the compensation control relationship based on the updated disturbance characterization information; that is, the compensation control relationship established in this embodiment is not a fixed parameter, but forms a continuous iterative update process between each drive cycle, so that the controller can gradually adapt to long-term operating factors such as load aging, power quality changes, ambient temperature changes and power device parameter drift.

[0061] Those skilled in the art will understand that the output voltage, output current, and load temperature mentioned above have already been acquired during the aforementioned data acquisition process. Therefore, this step does not require the addition of new sampling hardware. It is sufficient to use the data already obtained at the end of the current drive cycle to evaluate the compensation effect. Existing technologies regarding sampling circuits and analog-to-digital conversion processes have been disclosed and will not be elaborated here.

[0062] As an optional implementation, after the current drive control signal is output, the controller first reads the output voltage, output current and load temperature data collected in the corresponding drive cycle, and extracts a set of drive evaluation parameters that can reflect the compensation control effect according to the preset evaluation rules.

[0063] In practice, different types of nonlinear loads require different evaluation metrics. For example, for inductive loads such as solenoid valves and relay coils, the focus is more on current settling time, current overshoot during pull-in, and ripple changes during freewheeling. For drive scenarios such as servo motors and brushless motors, the focus is more on steady-state error, current ripple corresponding to torque fluctuations, and response time. For LED driver circuits, the focus is more on output ripple, current stability, and harmonic distortion rate. For piezoelectric actuators, the focus is more on dynamic response speed and voltage settling accuracy. Therefore, this implementation allows the selection of at least two parameters from output ripple, harmonic distortion rate, steady-state error, dynamic response time, and current overshoot as the set of drive evaluation parameters based on the load type, to improve the relevance of the evaluation results to the current load condition. As an example, the controller can calculate the output ripple based on the continuously sampled output voltage sequence within the current drive cycle.

[0064] in, This indicates the maximum output voltage during the current drive cycle. This indicates the minimum output voltage. This represents the average output voltage; the smaller the ripple, the better the current compensation strategy suppresses periodic disturbances; furthermore, for the output current, the controller can calculate the current overshoot. ,in, This represents the peak current that occurs during the drive setup phase. This represents the target reference current; when this indicator continues to increase, it usually indicates that the drive output adjustment parameter corresponding to the current compensation control relationship has an overcompensation trend; for steady-state error, it can be represented by the average deviation between multiple consecutive sampling points and the target reference value during the stable phase of the drive cycle.

[0065] in, This represents the sampled output value during the stable phase. This represents the target reference output value. It should be noted that the aforementioned evaluation parameters do not participate in subsequent updates independently, but rather collectively form the drive evaluation parameter set corresponding to the current drive cycle, used to reflect the comprehensive control effect of the compensation control relationship under the current operating conditions. In specific implementation, the controller does not directly adjust the compensation control relationship based on the drive evaluation parameters, but first establishes a one-to-one correspondence between the drive evaluation parameters and the corresponding drive adjustment parameters when the current drive control signal is generated. In practice, the aforementioned S4 has already generated a set of drive adjustment parameters, including bridge arm timing adjustment parameters, drive output adjustment parameters, drive cycle adjustment parameters, and... Each of the drive constraint adjustment parameters has a unique corresponding drive cycle number. Therefore, in this implementation, according to the parameter type and drive cycle number, each evaluation index in the current drive evaluation parameter set is mapped to the drive adjustment parameter corresponding to the generation of the current drive control signal. For example, for the bridge arm timing adjustment parameter, the current overshoot and switching transient ripple can be closely correlated; for the drive output adjustment parameter, the steady-state error and dynamic response time can be closely correlated; for the drive cycle adjustment parameter, the output ripple and harmonic distortion rate can be closely correlated; and for the drive constraint adjustment parameter, the load temperature rise and peak current change can be closely correlated. In other words, each driving adjustment parameter has its own corresponding evaluation object, rather than using a unified evaluation index, thereby avoiding evaluation coupling between different adjustment parameters and improving the accuracy of subsequent parameter adjustment directions.

[0066] As a more specific implementation, for each set of drive adjustment parameters, the controller establishes a corresponding evaluation result: ,in, Indicates the first Class-driven parameter adjustment This represents the corresponding driving evaluation parameters. This indicates the evaluation result corresponding to the drive adjustment parameter.

[0067] It should be noted that the functional relationship in this embodiment is an evaluation relationship established based on preset engineering rules. For example, when the steady-state error continues to decrease and the dynamic response time shortens, it can be determined that the drive output adjustment parameters have a good adaptation effect. When the ripple increases and the harmonic distortion rate rises synchronously, it can be determined that the drive cycle adjustment parameters need to be readjusted. When a significant current spike appears during the bridge arm conduction phase, it can be determined that the bridge arm timing adjustment parameters exhibit premature conduction or delayed turn-off phenomena. After completing the above evaluation, the controller further determines whether the evaluation results meet the preset adjustment conditions; In practice, preset adjustment conditions can include at least one of the following situations: (1) The steady-state error continues to increase over multiple consecutive driving cycles; (2) The output ripple exceeds the preset ripple threshold; (3) The harmonic distortion rate exceeds the allowable range of the corresponding industry standard; (4) The current overshoot continuously exceeds the safety limit; (5) The dynamic response time exceeds the allowable setup time for the corresponding load.

[0068] If the evaluation result does not meet the adjustment conditions, the current set of drive adjustment parameters is retained and used in the next drive cycle to ensure the stability of the control parameters; if the evaluation result meets the adjustment conditions, the controller recalculates the corresponding drive adjustment parameters based on the corresponding drive evaluation parameters.

[0069] In particular, this implementation adopts a constraint update method, that is, only the drive adjustment parameters corresponding to the evaluation results are partially updated, while the other drive adjustment parameters remain unchanged, thereby avoiding the decrease in drive system stability caused by the simultaneous changes of multiple compensation channels. As an example, when the output ripple continues to increase, only the carrier period or drive update period corresponding to the drive clock cycle adjustment parameter is adjusted; when the steady-state error increases, only the drive duty cycle correction amount corresponding to the drive output adjustment parameter is adjusted; when the bridge arm conduction spike increases, only the conduction node position and turn-off node position corresponding to the bridge arm timing adjustment parameter are adjusted; when the load temperature rise increases significantly, only the drive output slope or dead time corresponding to the drive constraint adjustment parameter is adjusted.

[0070] In some more specific application scenarios, in order to avoid the compensation control relationship from oscillating due to excessively rapid parameter updates, this implementation sets update constraints on the recalculated drive adjustment parameters. For example, it limits the parameter change ratio within a single drive cycle to no more than a preset change upper limit, or limits the cumulative adjustment range within multiple consecutive drive cycles to no more than a set range, so that the drive adjustment parameters are updated in a gradual manner.

[0071] After the above update is completed, the controller replaces the corresponding parameters in the drive adjustment parameter set with the recalculated drive adjustment parameters, generates the drive adjustment parameter set for the next drive cycle, and uses it as the input parameters for establishing the compensation control relationship and generating the drive control signal in the next round.

[0072] As an optional implementation, to further improve the applicability of this disclosure to different types of nonlinear loads, before constructing disturbance characterization information, the controller determines the load characteristic quantities participating in disturbance modeling based on the electrical characteristics corresponding to the current driven object, and adds specific electrical characteristics corresponding to the current nonlinear load on the basis of common load characteristic parameters, so that the disturbance characterization information subsequently established can more accurately reflect the dynamic electrical behavior of the current load during actual operation. It should be noted that this implementation does not change the aforementioned construction process of load characteristic description information, but rather, based on the common characteristic parameters such as equivalent impedance, dynamic incremental impedance, output current change rate, output voltage change rate, and load temperature already included in the load characteristic description information, it further introduces specific load characteristic quantities that can reflect the dynamic characteristics of the corresponding load according to the differences in electrical response mechanisms of different nonlinear loads, forming disturbance characterization information applicable to the current driven object. Therefore, the subsequent disturbance prediction, disturbance source analysis, and compensation control relationship establishment processes still adopt the control process disclosed in the aforementioned implementation, and only the core characteristic variables participating in modeling are changed, which will not be elaborated further.

[0073] Those skilled in the art will understand that drive controllers can typically obtain information about the current driven object during the system initialization phase. For example, the driven object may originate from device configuration parameters, driver initialization configuration, device type identifier issued by the fieldbus, device configuration file in non-volatile memory, or load parameters read when the controller starts. Therefore, the nonlinear load type in this embodiment can be identified without manual intervention and used as the basis for subsequent load characteristic quantity selection.

[0074] In practice, the controller first establishes a common load characteristic set. ,in, Indicates the characteristic parameters of the equivalent impedance; Indicates the characteristic parameters of dynamic incremental impedance; Indicates the rate of change of output current; represents the output voltage change rate; T represents the current load temperature; the above common feature set is applicable to various nonlinear loads and is also the basic feature that always participates in the calculation during the establishment of the aforementioned disturbance prediction model. Based on this, the controller further determines the corresponding dedicated load characteristic quantities according to the electrical response characteristics of the current load, and constitutes the final set of load characteristics participating in the construction of disturbance characterization information:

[0075] in, This represents the specific load characteristic quantity selected based on the current nonlinear load type; As a specific implementation, when the controller determines that the current driven object has inductive dynamic response characteristics, it calculates the equivalent inductance corresponding to the current drive cycle as a dedicated load characteristic quantity in response to the dynamic change process corresponding to the output current change rate. In practice, inductive dynamic response characteristics include, but are not limited to, relay coils, solenoid valves, electromagnets, and various inductive actuators. Since the energy storage of this type of load mainly comes from the magnetic field establishment process, the current establishment speed directly determines the drive response speed. Therefore, using the equivalent inductance can more accurately reflect the dynamic characteristics during the flux linkage establishment process. The equivalent inductance can be expressed as:

[0076] The controller adds the calculated equivalent inductance to the common load feature set and participates in the subsequent disturbance prediction model calculation. As another optional implementation, when the controller determines that the current driven object is a capacitive load, it calculates the equivalent capacitance corresponding to the current driving cycle as a dedicated load feature quantity in response to the charging and discharging process corresponding to the output voltage change rate.

[0077] In practice, when piezoelectric ceramic drives, capacitor energy storage modules, and some analog drive actuators all exhibit capacitive dynamic response characteristics, their dynamic process mainly manifests as voltage build-up speed and charging / discharging process. Therefore, using equivalent capacitance can more accurately describe the current load charging / discharging state; equivalent capacitance can be expressed as: Among them, the slower the output voltage build-up speed, the larger the equivalent capacitance is usually, and the corresponding disturbance prediction model can reflect the dynamic deviation caused by capacitive hysteresis in advance. In more specific application scenarios, when the driven object is a motor load and is affected by back electromotive force (EMF), the controller further extracts back EMF characteristics to participate in the construction of disturbance characterization information. It should be noted that the biggest difference between motor loads and ordinary inductive loads is that their output voltage is affected not only by winding impedance but also by the back EMF generated by rotor rotation. If this characteristic is ignored during disturbance prediction, the predicted output is likely to continuously deviate from the actual output. Therefore, this implementation calculates the estimated back EMF value based on the drive output voltage, winding current, and winding resistance.

[0078] Where R represents the equivalent resistance of the winding and L represents the equivalent inductance of the winding; the calculated back electromotive force characteristic is used as the dedicated load characteristic corresponding to the current motor load to participate in the disturbance prediction, so that the prediction output can synchronously reflect the changes in the motor operating state.

[0079] As another implementation, when the controller determines that the current driven object is a piezoelectric load such as a piezoelectric actuator and has voltage and current phase change characteristics, the phase difference between the output voltage and output current is converted into a change in capacitive reactance as a unique load characteristic. In practice, piezoelectric materials have obvious frequency-dependent characteristics at different driving frequencies, and their impedance changes are reflected not only in amplitude changes but also in phase changes between voltage and current. Therefore, the controller calculates the change in capacitive reactance based on the phase difference between the output voltage and output current in the current driving cycle, so that the disturbance prediction model can reflect the dynamic capacitive reactance change process of the piezoelectric material and improve the prediction accuracy under high-frequency driving conditions. Furthermore, when the driven object is a light-emitting diode load and has nonlinear current-voltage response characteristics, the controller extracts the change in current-voltage characteristics corresponding to the current driving cycle as a dedicated load characteristic.

[0080] In practice, due to the significant nonlinearity of the forward conduction characteristics of LEDs, changes in their junction temperature will cause the on-state voltage drop to drift continuously. It is difficult to accurately describe the actual conduction state based solely on the equivalent impedance. Therefore, the controller calculates the offset of the current operating point relative to the historical volt-ampere curve based on the volt-ampere curve formed by the output voltage and output current over multiple consecutive driving cycles, and adds this offset as a change in volt-ampere characteristics to the load characteristic set to reflect the changes in conduction characteristics corresponding to changes in LED junction temperature and aging process.

[0081] It should be noted that this implementation method only changes the specific load characteristics involved in disturbance modeling for different types of nonlinear loads, while the common load characteristic set remains consistent. Therefore, different load types can all adopt a unified disturbance prediction model, a unified disturbance source analysis process, and a unified compensation control relationship establishment mechanism. By simply adjusting the core electrical characteristics involved in modeling, the disturbance characterization information can be made more consistent with the actual operating state of the current load, thereby improving the accuracy of subsequent disturbance identification and compensation control, while ensuring that the entire compensation control framework has good engineering versatility and scalability.

[0082] In the above implementation, the drive output prediction result can be established based on a state-space model, a piecewise linear model, a circuit equivalent model, or a parameter identification model. Under different load types, different prediction models can be called to generate corresponding prediction output results, as long as they can reflect the dynamic response of the load under the current drive condition. This disclosure does not limit this. The disturbance analysis window can be a fixed-length window, or the window length can be dynamically adjusted according to the drive frequency, PWM period, load state change rate, or output ripple change to ensure that the disturbance analysis window covers the complete disturbance process.

[0083] In more specific application scenarios, compensation parameter updates can be performed after each drive cycle or updated uniformly after forming a statistical window of multiple drive cycles. For drive conditions with stable operation, the update frequency can be reduced to decrease the controller's computational resource consumption. For load abrupt changes or abnormal disturbances, the parameter update frequency can be increased. When output overcurrent, output overvoltage, device overtemperature, drive lockout, or compensation oscillation is detected and continues for more than a preset time, compensation parameter updates can be paused and the current stable drive parameters can be maintained, or the most recent stable parameters can be restored to avoid continuous deviation of compensation parameters due to abnormal conditions. When the driven object changes, such as switching from a motor to an electromagnetic actuator, or from an LED load to a piezoelectric actuator, or when a change in the current load type is detected, the corresponding load characteristic description information can be re-established, and the corresponding disturbance characterization information and compensation control relationship can be regenerated without relying on the original load historical data. In addition, since voltage sampling, current sampling, temperature sampling, and power device status sampling may come from different sampling channels, and there is a micro-delay in their sampling time, a unified sampling period is used for time alignment so that the data participating in the disturbance analysis all correspond to the same drive condition, reducing misjudgments caused by different sampling delays.

[0084] The drive control signal described in this disclosure can be a PWM drive signal, a DAC analog drive signal, a bridge arm gate drive signal, a multiphase drive control quantity, or a digital drive control word, as long as it can control the output of the power drive module.

[0085] It is important to note that this disclosure does not establish a fixed compensation model for a single disturbance. Instead, within the same driving cycle, it constructs a continuously interconnected closed-loop control process, including load dynamic characteristic extraction, drive output prediction, comprehensive disturbance construction, disturbance source analysis, and compensation control generation. Load characteristic description information determines the drive output prediction result, which, together with the actual drive output result, forms comprehensive disturbance information. This comprehensive disturbance information further determines the disturbance component information, with each component corresponding to a different drive adjustment object. The drive adjustment result then participates in updating the disturbance characterization information and compensation control relationship in the next driving cycle, thus forming an adaptive closed-loop compensation mechanism for nonlinear load driving processes. Compared to existing technologies that only perform unified compensation based on output error or closed-loop control based on fixed parameters, this disclosure achieves full-process coordination between load dynamic characteristic identification, disturbance source analysis, and drive compensation control. This allows the compensation strategy to evolve in real time with the driving conditions, improving the stability of the drive output, disturbance suppression capability, and long-term operational reliability under complex nonlinear load conditions.

[0086] The above provides a detailed description of an anti-interference compensation method for an intelligent drive circuit with nonlinear load characteristics provided in this application. Those skilled in the art will recognize that the algorithm steps of the various examples disclosed herein can be implemented using electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this disclosure. The above description of the embodiments is only for the purpose of helping to understand the method and core ideas of this application. It should be noted that those skilled in the art can make several improvements and modifications to this application without departing from the principles of this application, and these improvements and modifications also fall within the protection scope of the claims of this application.

Claims

1. A method for anti-interference compensation of intelligent drive circuits with nonlinear load characteristics, characterized in that, include: The system acquires operational data during the process of an intelligent drive circuit driving a nonlinear load. The operational data includes at least the drive control signal, input voltage, output voltage, output current, power device operating status, and load status information. Based on the operating data, load characteristic description information describing the dynamic electrical behavior of nonlinear loads under the current driving condition is constructed, and disturbance characterization information corresponding to the current driving condition is established based on the load characteristic description information. The disturbance characterization information includes the driving output prediction result, the actual driving output result, and the comprehensive disturbance information formed by the driving output prediction result and the actual driving output result. Based on the disturbance characterization information, the disturbance source is analyzed for the comprehensive disturbance information to generate disturbance composition information including load-intrinsic disturbances, power supply disturbances, switching disturbances, and externally coupled disturbances; Based on the disturbance component information, a compensation control relationship is established between each disturbance component and the drive adjustment object, and compensation control information corresponding to the current drive condition is generated based on the compensation control relationship. Based on the compensation control information, the drive control signal is corrected, a compensated drive control signal is generated, and the nonlinear load is driven. The disturbance characterization information is updated based on the compensated drive output, and the compensation control relationship is adjusted based on the updated disturbance characterization information.

2. The method according to claim 1, characterized in that: The step of constructing load characteristic description information based on the operating data to describe the dynamic electrical behavior of the nonlinear load under the current driving condition includes: Based on the input voltage, output voltage, and output current, a first set of characteristic parameters is generated to characterize the electrical response of the load. Based on the operating status of the power devices and the drive control signals, a second set of characteristic parameters is generated to characterize the drive switching process. Based on the load status information, a third set of characteristic parameters representing the load operating status is generated; The first, second, and third feature parameter groups are time-aligned according to a uniform sampling period to obtain a set of feature parameters corresponding to the same driving condition, which serves as load feature description information.

3. The method according to claim 2, characterized in that: The first set of characteristic parameters includes at least one of the following: equivalent impedance characteristic parameters, dynamic incremental impedance characteristic parameters, equivalent inductance characteristic parameters, and equivalent capacitance characteristic parameters, used to characterize the relationship between the voltage and current response of a nonlinear load; the second set of characteristic parameters includes at least one of the following: switching state characteristic parameters, output ripple characteristic parameters, switching transient characteristic parameters, and harmonic characteristic parameters, used to characterize the power device drive switching process; the third set of characteristic parameters includes at least one of the following: load temperature characteristic parameters, current rate of change characteristic parameters, voltage rate of change characteristic parameters, and output response delay characteristic parameters, used to characterize the operating state of a nonlinear load.

4. The method according to claim 1, characterized in that: The establishment of disturbance characterization information corresponding to the current driving condition includes establishing a predicted driving output, specifically including: The equivalent impedance characteristic parameters, dynamic incremental impedance characteristic parameters, equivalent inductance characteristic parameters, and equivalent capacitance characteristic parameters in the load characteristic description information are normalized and then concatenated in the order of their arrangement to obtain the load state vector. Based on the PWM duty cycle, switching frequency and dead time in the drive control signal, calculate the effective on time, effective off time and average output voltage of the bridge arm in one switching cycle. The ratio of the equivalent impedance characteristic parameter to the dynamic incremental impedance characteristic parameter and the ratio of the equivalent inductance characteristic parameter to the equivalent capacitance characteristic parameter are compared with preset segmentation thresholds to determine the working segment as one of resistive-dominated working segment, inductive-dominated working segment, or capacitive-dominated working segment. The state update coefficient matrix corresponding to each working segment is obtained offline based on the load state vector and historical sampling data. The state update coefficient matrix includes state transition coefficient, conduction correction coefficient, turn-off correction coefficient and voltage mapping coefficient. Based on this, the load state vector, effective conduction time, effective turn-off time and average output voltage of the bridge arm are linearly combined to obtain the predicted output voltage value and predicted output current value, which are used as the drive output prediction result.

5. The method according to claim 4, characterized in that: The establishment of disturbance characterization information corresponding to the current driving condition also includes establishing actual driving output and comprehensive disturbance information, including: Obtain the predicted output voltage and predicted output current values ​​corresponding to the drive output prediction results, as well as the actual output voltage and actual output current values ​​corresponding to the current drive condition; Calculate the voltage deviation between the predicted output voltage value and the actual output voltage value, and the current deviation between the predicted output current value and the actual output current value, respectively. A perturbation sampling sequence is generated based on the voltage deviation, current deviation, and their corresponding sampling times; The disturbance sampling sequence is combined in time according to a preset sampling window to generate comprehensive disturbance information, which includes disturbance amplitude information, disturbance change direction information, and disturbance duration information.

6. The method according to claim 1, characterized in that: The analysis of disturbance sources based on disturbance characterization information of the comprehensive disturbance information includes: The integrated disturbance information is divided into multiple disturbance analysis windows according to the sampling time sequence, and the disturbance amplitude sequence, disturbance frequency sequence and disturbance phase sequence are extracted in each disturbance analysis window respectively; The disturbance frequency sequence is frequency matched with the reference frequency sequence corresponding to the switching frequency of the power device and its harmonics, and the disturbance data that meets the preset frequency deviation range is marked as the switching disturbance component. The disturbance phase sequence is time-aligned with the state switching sequence corresponding to the load state switching time, and the disturbance data that meets the preset time deviation range is marked as the load disturbance component. The disturbance frequency sequence is frequency matched with the reference frequency sequence corresponding to the input power supply ripple frequency, and the disturbance data that meets the preset frequency deviation range is marked as power supply disturbance components. Disturbance data that have not been marked as described above are collected into externally coupled disturbance components, and the switching disturbance components, load disturbance components, power supply disturbance components and externally coupled disturbance components are combined in chronological order to generate disturbance composition information.

7. The method according to claim 6, characterized in that: Based on the disturbance composition information, a compensation control relationship is established corresponding to each disturbance composition, compensation control information is generated, and the drive control signal is corrected, including: The action time interval corresponding to the switching disturbance component is mapped to the drive cycle corresponding to the power drive module. The bridge arm turn-on control node and bridge arm turn-off control node are configured according to the mapped action time interval to obtain the bridge arm timing adjustment parameters. The load disturbance component is mapped to the drive output control cycle, and the drive output control quantity is configured according to the mapped time interval to obtain the drive output adjustment parameter. The drive output control quantity includes at least one of the drive duty cycle and the drive output reference value. The action time interval corresponding to the power disturbance component is mapped to the driving modulation period, and the driving modulation clock is configured according to the mapped action time interval to obtain the driving clock adjustment parameter. The driving modulation clock includes at least one of the carrier period, modulation period and driving update period. The action time interval corresponding to the external coupling disturbance component is mapped to the drive protection cycle. Based on the disturbance duration and disturbance amplitude corresponding to the mapped action time interval, at least one of the drive output slope, bridge arm dead time and drive current establishment time is increased or decreased according to the preset adjustment coefficient to obtain the drive constraint adjustment parameters. The timing adjustment parameters of the bridge arm, the drive output adjustment parameters, the drive cycle adjustment parameters, and the drive constraint adjustment parameters are associated and organized according to the chronological order of each action time interval to generate a set of drive adjustment parameters.

8. The method according to claim 7, characterized in that: Generate a compensated drive control signal based on the set of drive adjustment parameters, including: Read the bridge arm timing adjustment parameters, drive output adjustment parameters, drive cycle adjustment parameters, and drive constraint adjustment parameters from the drive adjustment parameter set; Based on the bridge arm timing adjustment parameters, establish the turn-on timing and turn-off timing corresponding to each power bridge arm to form the bridge arm driving timing. Configure the drive output control quantity corresponding to the bridge arm drive timing according to the drive output adjustment parameters to form the bridge arm drive parameters; Configure the drive modulation timing corresponding to the bridge arm drive parameters according to the drive timing adjustment parameters to form the bridge arm drive timing parameters; Configure the drive constraint control quantity corresponding to the bridge arm drive cycle parameter according to the drive constraint adjustment parameter to form the bridge arm drive constraint parameter; The drive control signal is generated by combining the drive timing, drive parameters, drive cycle parameters, and drive constraint parameters corresponding to each power arm to generate the drive control signal output to the power drive module.

9. The method according to claim 1, characterized in that: After generating the drive control signal, the method further includes: Based on the output voltage, output current and load temperature collected within the corresponding drive cycle of the drive control signal, at least two of the output ripple, harmonic distortion rate, steady-state error, dynamic response time and current overshoot corresponding to the current drive cycle are extracted to form a set of drive evaluation parameters. Each drive evaluation parameter in the drive evaluation parameter set is mapped one-to-one with each drive adjustment parameter in the drive adjustment parameter set corresponding to the current drive control signal generation according to parameter type and corresponding drive cycle to obtain the evaluation result corresponding to each drive adjustment parameter. For drive adjustment parameters whose evaluation results meet preset adjustment conditions, the corresponding drive evaluation parameters are used as constraints to recalculate the drive adjustment parameters, and the recalculated drive adjustment parameters are used to replace the corresponding parameters in the drive adjustment parameter set to generate the drive adjustment parameter set for the next drive cycle.

10. The method according to claim 1, characterized in that: When constructing the disturbance characterization information, corresponding load characteristic quantities are selected based on the electrical characteristics of the nonlinear load to participate in the construction of the disturbance characterization information, wherein: When the nonlinear load exhibits inductive dynamic response characteristics, the equivalent inductance corresponding to the rate of change of the output current is used as the load characteristic quantity. When a nonlinear load exhibits capacitive dynamic response characteristics, the equivalent capacitance corresponding to the rate of change of the output voltage is taken as the load characteristic quantity. When the nonlinear load is a motor load and there is a back electromotive force, the back electromotive force characteristic quantity is used as the load characteristic quantity. When the nonlinear load is a piezoelectric load and there are phase change characteristics of voltage and current, the change in capacitive reactance corresponding to the phase difference between voltage and current is taken as the load characteristic quantity. When the nonlinear load is a light-emitting diode load and has a nonlinear current-voltage response characteristic, the change in current-voltage characteristic corresponding to voltage and current is taken as the load characteristic quantity; The selected one or more load characteristics, together with the equivalent impedance, dynamic incremental impedance, output current rate of change, output voltage rate of change, and load temperature, constitute the disturbance characterization information corresponding to the current nonlinear load.

Images