BLDC driving parameter self-tuning embedded system and method for multiple device types

Abstract

The invention discloses a BLDC (brushless direct current) driving parameter self-tuning embedded system and method for multiple equipment types, and the method comprises the steps: constructing a BLDC unified mathematical model containing an electromagnetic torque equation, a back electromotive force equation and a mechanical motion equation, and building a nonlinear mapping relation library of motor type identification codes and a key electrical parameter set; a predefined multi-mode excitation signal sequence is injected into the to-be-driven BLDC motor, dynamic response data are synchronously collected, and the dynamic response data are used for representing the characteristics of the key electrical parameter set; based on collected dynamic response data, equivalent inductance and equivalent resistance in a key electrical parameter set are decoupled and extracted through a frequency domain impedance analysis method, and meanwhile, a counter electromotive force coefficient and rotational inertia in the key electrical parameter set are identified by combining time domain counter electromotive force zero crossing point detection with rotating speed differential operation. Based on a motor fingerprint feature library and a fuzzy adaptive architecture, multi-model debugging-free plug-and-play is realized, and stable control performance is maintained under a dynamic working condition.

Description

Technical Field

[0001] The present application relates to the field of embedded drive control technology, and in particular to an embedded system and method for self-tuning BLDC drive parameters for multiple device types. Background Art

[0002] As brushless DC motors (BLDCs) penetrate advanced manufacturing and intelligent equipment, their drive systems face increasingly complex and diverse application scenarios. From precision medical instruments to industrial robot joints, motors of different specifications and manufacturers exhibit widely varying electrical characteristics. Current drive technologies rely on pre-stored fixed parameters or manual parameter adjustment, exposing fundamental limitations when addressing the need for rapid adaptation across multiple models.

[0003] First, due to differences in electromagnetic design and production processes, different motor models of the same power rating exhibit vastly different dynamic response characteristics during actual operation. Existing solutions require time-consuming, specialized debugging for each motor type, significantly extending equipment deployment cycles. Furthermore, manual tuning makes it difficult to accurately capture the nonlinear coupling relationship between winding parameters and mechanical characteristics. This leads to frequent torque oscillations, commutation loss, and other uncontrolled phenomena when switching motor models within the same series of drives, severely restricting the flexibility of multi-device collaborative operation systems.

[0004] Secondly, under complex operating conditions such as temperature fluctuations and sudden load changes, the motor's equivalent parameters exhibit time-varying drift. Traditional fixed-parameter control strategies are unable to track these changes in real time, leading to a series of problems such as increased speed response overshoot and degraded steady-state accuracy. Especially in long-term continuous operation scenarios, the accumulated errors caused by parameter mismatch often force system shutdown and readjustment, reducing equipment utilization and creating operational safety risks, creating an insurmountable technical gap in high-reliability applications. Summary of the Invention

[0005] In order to solve the above problems, an embodiment of the present invention provides an embedded method for self-tuning BLDC drive parameters for multiple device types, the method comprising:

[0006] Construct a unified BLDC mathematical model that includes the electromagnetic torque equation, back-electromotive force equation, and mechanical motion equation, and establish a nonlinear mapping relationship library between motor type identification codes and key electrical parameter sets;

[0007] injecting a predefined multi-mode excitation signal sequence into the BLDC motor to be driven and simultaneously collecting dynamic response data, wherein the dynamic response data is used to characterize a set of key electrical parameters;

[0008] Based on the collected dynamic response data, the equivalent inductance and equivalent resistance of the key electrical parameters are decoupled and extracted through frequency domain impedance analysis. At the same time, the back EMF coefficient and moment of inertia of the key electrical parameters are identified by combining time domain back EMF zero-crossing detection with speed differential operation.

[0009] The identified key electrical parameter set is input into a preset fuzzy adaptive tuner to generate an initial PID control parameter set. The control performance is verified through a closed-loop speed response test, and the PID control parameter set is dynamically optimized based on the speed overshoot and steady-state fluctuation rate.

[0010] The optimized PID control parameter group is bound to the motor type identification code, stored in the device parameter knowledge base of the embedded FLASH memory, and attached with a credibility label.

[0011] Furthermore, the nonlinear mapping relationship library construction method includes:

[0012] The motor no-load starting current waveform is input into a waveform encoder to generate a feature vector; the feature vector is processed by a lightweight convolutional neural network; and the output of the lightweight convolutional neural network is used as an iterative initial value for key electrical parameter identification.

[0013] Furthermore, the method for generating the multi-mode excitation signal sequence includes:

[0014] A basic excitation template is loaded according to the motor type identification code; a swarm intelligence algorithm is used to optimize the gradient parameters of the basic excitation template; and the optimized gradient parameters are injected into the dynamic response data collection link.

[0015] Furthermore, the frequency domain impedance analysis method includes:

[0016] Wavelet packet decomposition is performed on the collected swept frequency response current; a neural differential equation solver is constructed based on the decomposition result; and the neural differential equation solver outputs equivalent circuit parameters to the fuzzy adaptive tuner.

[0017] Furthermore, the real-time temperature compensation method includes:

[0018] The power module temperature is collected as the input for parameter decoupling; the resistance drift compensation is generated through the heat transfer state space model; and the compensated resistance value is input into the optimization phase of the PID control parameter group.

[0019] Furthermore, the credibility label generation method includes:

[0020] Receive performance indicator data output by a closed-loop speed response test verification; process the performance indicator through a multi-criteria evaluation tree; set a tag state machine based on the evaluation results and store them in a device parameter knowledge base.

[0021] Furthermore, the fault fusing mechanism includes:

[0022] Monitor overcurrent and stall events in S4 closed-loop verification; trigger safety parameter group switching and generate key parameter feature hashes; update stored credibility tags based on event types.

[0023] Furthermore, the parameter iterative optimization function includes:

[0024] When the stored credibility tag meets the conditions, the historical parameter set is read; based on the historical parameter set, an incremental excitation is generated and injected into the dynamic response data acquisition link; the moment of inertia is updated through decoupling and the PID integral term is readjusted.

[0025] Furthermore, the method further comprises:

[0026] The optimized PID control parameter group is bound to the motor type identification code, stored in the device parameter knowledge base of the embedded FLASH memory, and attached with a credibility label. When the credibility label is high credibility, a dynamic load disturbance sequence is injected and dynamic performance indicators are collected. If the dynamic performance indicators reach the over-limit condition, the credibility label is downgraded to steady-state high credibility, and the applicable operating range is marked. The marked credibility label is associated with the applicable operating range and stored in the device parameter knowledge base.

[0027] An embedded system for automatic BLDC drive parameter tuning for multiple device types. The system includes:

[0028] A core model building module, which constructs a unified BLDC mathematical model including electromagnetic torque equations, back electromotive force equations, and mechanical motion equations, and establishes a nonlinear mapping relationship library between motor type identification codes and key electrical parameter sets;

[0029] an excitation injection acquisition module, which injects a predefined multi-mode excitation signal sequence into the BLDC motor to be driven and simultaneously acquires dynamic response data, wherein the dynamic response data is used to characterize the characteristics of the key electrical parameter set;

[0030] A parameter decoupling identification module, which extracts the equivalent inductance and equivalent resistance of key electrical parameters through frequency domain impedance analysis based on the collected dynamic response data. It also identifies the back electromotive force coefficient and moment of inertia of key electrical parameters using time domain back electromotive force zero-crossing detection combined with speed differential calculations.

[0031] A parameter self-tuning module, which inputs the identified key electrical parameter set into a preset fuzzy adaptive tuner to generate an initial PID control parameter set, verifies control performance through a closed-loop speed response test, and dynamically optimizes the PID control parameter set based on speed overshoot and steady-state fluctuation rate;

[0032] The knowledge base storage module binds the optimized PID control parameter group with the motor type identification code, stores the optimized PID control parameter group with the motor type identification code in the device parameter knowledge base of the embedded FLASH memory, and adds a credibility tag.

[0033] The technical effects and advantages of the embedded method for self-tuning BLDC drive parameters for multiple device types provided by the present invention are as follows:

[0034] The present invention is based on a motor fingerprint feature library and a fuzzy adaptive architecture to achieve plug-and-play debugging-free for multiple models and maintain stable control performance under dynamic working conditions. The present invention automatically matches the nonlinear parameter mapping library through the motor's unique identification code, and can adapt to motors with different electromagnetic characteristics without manual intervention; combining wavelet packet frequency domain analysis and time domain inertia identification, accurately constructs a motor dynamic model to avoid torque fluctuations caused by parameter mismatch in traditional solutions; adopts a fuzzy rule-driven PID parameter real-time tuning mechanism to autonomously adjust control parameters in load mutation and temperature drift scenarios; pre-compensates winding resistance drift based on a heat transfer state space model to significantly suppress speed overshoot and steady-state fluctuations; the parameter credibility multi-criteria verification mechanism automatically eliminates abnormal identification results and triggers safety parameter fuse switching; lightweight neural network deployment achieves microsecond dynamic response, ensuring efficient execution of complex algorithms in resource-constrained controllers. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 This is a flow chart of the embedded method for self-tuning BLDC drive parameters for multiple device types in Example 1;

[0036] Figure 2 This is a flow chart of the embedded method for self-tuning BLDC drive parameters for multiple device types in Example 2;

[0037] Figure 3 This is a connection diagram of the BLDC drive parameter self-tuning embedded system for multiple device types in Example 3. DETAILED DESCRIPTION

[0038] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0039] Example 1

[0040] See also Figure 1 As shown, an embodiment of the present invention provides an embedded method for self-tuning BLDC drive parameters for multiple device types, the method comprising:

[0041] S1. Construct a unified BLDC mathematical model including the electromagnetic torque equation, back electromotive force equation, and mechanical motion equation, and establish a nonlinear mapping relationship library between the motor type identification code and the key electrical parameter set;

[0042] S2. Inject a predefined multi-mode excitation signal sequence into the BLDC motor to be driven, and simultaneously collect dynamic response data, which is used to characterize the characteristics of the key electrical parameter set;

[0043] S3. Based on the collected dynamic response data, the equivalent inductance and equivalent resistance of the key electrical parameter set are decoupled and extracted through frequency domain impedance analysis. At the same time, the back EMF coefficient and moment of inertia of the key electrical parameter set are identified by combining time domain back EMF zero-crossing detection with speed differential operation.

[0044] S4. Input the identified key electrical parameter set into a preset fuzzy adaptive tuner to generate an initial PID control parameter set. Verify the control performance through a closed-loop speed response test, and dynamically optimize the PID control parameter set based on the speed overshoot and steady-state fluctuation rate.

[0045] S5. Bind the optimized PID control parameter group to the motor type identification code, store it in the device parameter knowledge base of the embedded FLASH memory, and attach a credibility label.

[0046] S1-S5 are automatically completed in a single embedded microcontroller without the need for external debugging equipment.

[0047] The key electrical parameter set includes equivalent inductance, equivalent resistance, back electromotive force coefficient and moment of inertia.

[0048] The dynamic response data includes three-phase current waveform, rotor position signal, swept frequency response current and DC bus voltage.

[0049] The method for constructing the nonlinear mapping relationship library in S1 includes:

[0050] Input the motor no-load starting current waveform into the waveform encoder to generate a feature vector;

[0051] When a new motor is first connected to the system, the controller controls the motor to perform a no-load start operation (that is, no mechanical load is connected to the motor shaft end) and simultaneously collects the three-phase current waveform (a part of the dynamic response data). The current waveform during the startup phase is input into the waveform encoder for feature extraction. This includes: the waveform encoder generates a compact feature vector by analyzing the geometric characteristics of the current waveform (such as peak amplitude, rising edge slope, oscillation frequency, etc.). For example, a certain type of motor may exhibit a typical "double-peak oscillation" current waveform during no-load startup, which the encoder converts into a 128-dimensional floating-point vector.

[0052] Processing the feature vector through a lightweight convolutional neural network;

[0053] This feature vector is then processed by a pre-built lightweight convolutional neural network (CNN). This CNN model is optimized through offline training, with a network architecture designed to accommodate embedded resource constraints (e.g., containing only three convolutional layers and one fully connected layer), allowing real-time execution on a microcontroller. The network output directly correlates to estimated values ​​for a set of key electrical parameters (equivalent inductance, equivalent resistance, back-EMF coefficient, and moment of inertia). For example, in equivalent inductance identification, given the starting current characteristic of a flat motor, the CNN outputs an initial estimate of its equivalent inductance (e.g., 0.85mH), whereas traditional methods require a complex iterative process.

[0054] The lightweight convolutional neural network output is used as the iterative initial value for S3 parameter identification.

[0055] This estimated parameter set will serve as the iterative initial value for the frequency-domain impedance analysis method and time-domain back-electromotive force zero-crossing detection in step S3. Compared with random initialization, this method significantly improves the efficiency of parameter identification.

[0056] In the subsequent S3-S5 processes, the system performs refined parameter identification based on this initial value. The resulting optimized PID parameter set is stored in the device parameter knowledge base along with the motor type identification code (such as "EC_Flat_48V"). When the same model motor is connected again, the system can directly call the historical parameter set, achieving a rapid deployment effect of learning once and reusing multiple times.

[0057] The multi-mode excitation signal sequence in S2 is generated in the following way:

[0058] Load the basic excitation template according to the motor type identification code defined in S1;

[0059] Optimizing the gradient parameters of the basic excitation template using a swarm intelligence algorithm;

[0060] Inject the optimized sequence into the S2 response acquisition link.

[0061] In step S2, the generation of the multi-mode excitation signal sequence is achieved by:

[0062] When the system detects the connection of a new motor (by reading the QR code on the motor nameplate or manually entering it by the user), it first extracts its motor type identification code (such as "EC_Disk_24V"). ​​Based on this identification code, the controller loads the corresponding basic excitation template from the device parameter knowledge base (the FLASH database established by S5). This template is a set of predefined PWM waveform pattern combinations, including excitation units with different duty cycles, frequencies, and commutation timings. For example, for a disk motor, the basic excitation template may include two core modes: step-by-step boost starting and sinusoidal frequency sweep.

[0063] To improve the accuracy of parameter identification, the system uses a swarm intelligence algorithm (in this embodiment, the particle swarm optimization (PSO) algorithm) to dynamically optimize the gradient parameters of the basic excitation template, specifically including:

[0064] Define the key change characteristics of the stimulus signal, including:

[0065] Voltage ramp rate (e.g., from 5% to 95% rated voltage in 0.1s); frequency switching step size (e.g., incremental interval from 50Hz to 200Hz); and transition time between modes.

[0066] The optimization process includes: constructing the objective function based on the parameter estimates output by the CNN in S1, running the PSO online in the embedded system, generating 20 sets of gradient parameter combinations, and evaluating the signal-to-noise ratio and spectrum richness of the three-phase current waveform.

[0067] Real-time adjustment: When current sampling saturation is detected, the voltage ramp rate parameter is automatically reduced.

[0068] The optimized excitation sequence is injected into the motor through a three-phase inverter, simultaneously triggering the acquisition of dynamic response data (three-phase current, rotor position, and bus voltage). This process is completely executed in a closed loop within the microcontroller. The resulting optimized excitation sequence is stored in conjunction with the motor type identification code. When the same type of motor is connected again, the historical optimal sequence can be directly called to avoid repeated optimization calculations.

[0069] Frequency domain impedance analysis methods in S3 include:

[0070] Perform wavelet packet decomposition on the swept frequency response current collected by S2;

[0071] Build a neural differential equation solver based on the decomposition results;

[0072] The neural differential equation solver outputs equivalent circuit parameters to the fuzzy adaptive tuner.

[0073] When executing the frequency domain impedance analysis in step S3, the system first preprocesses the three-phase current waveform under the swept-frequency excitation collected in S2. Taking a 12-pole BLDC motor as an example, when the microcontroller injects a linear swept-frequency voltage signal of 20Hz-2kHz, the synchronously collected stator current response data is input into the wavelet packet decomposition module. Here, the db4 wavelet basis function is used for three-layer decomposition, dividing the current signal into eight characteristic frequency bands in the frequency domain, effectively separating the winding fundamental response, switching noise and harmonic interference components.

[0074] The separated frequency band energy feature matrix is ​​fed into a neural differential equation solver. This neural differential equation solver uses a lightweight fully connected network structure (for example, an input layer with 8 nodes, a hidden layer with 5 nodes, and an output layer with 2 nodes). Its core function is to establish a mapping model between frequency domain response and equivalent circuit parameters by learning the differential constraints of the electromagnetic system. Specifically, it includes:

[0075] Input: normalized energy value of each frequency band, corresponding frequency point and bus voltage;

[0076] Output: equivalent inductance L eq (unit: mH) and equivalent resistance R eq (unit: Ω);

[0077] Training mechanism: Network training is completed based on a synthetic dataset generated offline (covering frequency response simulation data of various inductor and resistor combinations).

[0078] The equivalent circuit parameters output by the solver are then passed to the fuzzy adaptive tuner of S4. For example, when R eq =0.8Ω、L eq =2.1mH (example value), the fuzzy adaptive tuner automatically generates an initial value range for the proportional gain Kp based on the moment of inertia identification results, providing key electrical constraints for subsequent PID parameter optimization. The entire process runs in a closed loop within the microcontroller, eliminating the need to export intermediate data for frequency domain analysis.

[0079] The fuzzy adaptive tuner is an intelligent controller that combines fuzzy logic control and adaptive control theory. It is specifically used to automatically adjust and optimize the parameters of the PID controller, namely the proportional gain, integral gain, and differential gain, when the system model is inaccurate, nonlinear, or the parameters change. Its working principle includes:

[0080] S101. Map precise input parameters (equivalent inductance L, equivalent resistance R, back-electromotive force coefficient ke, moment of inertia J) to pre-defined fuzzy linguistic variables, such as "small inductance", "medium back-electromotive force coefficient", and "large moment of inertia". Each linguistic variable is defined by a membership function, which indicates the degree to which the precise value belongs to a certain fuzzy concept.

[0081] S102. Set the fuzzy rule base. The fuzzy rule base is the core knowledge base of the fuzzy adaptive tuner, which contains a series of IF-THEN rules based on expert experience or system knowledge. The IF-THEN rules describe what kind of PID parameter combination (THEN part) should correspond to different motor parameter combinations (IF part).

[0082] Example Rule 1:

[0083] IF (inductance L is large) AND (moment of inertia J is medium) THEN (proportional gain K p Should be medium to high), (integral gain K i should be low), (differential gain K d should be middle).

[0084] Example Rule 2:

[0085] IF (resistance R is large) AND (back electromotive force Ke is small) THEN (K p should be high), (K p should be high), (K i should be medium to high), (K d should be small).

[0086] S103. Evaluate all rules related to the current input state (fuzzified parameters). For each rule, calculate the activation strength of the IF part, usually using AND operation (minimum value min) or OR operation (maximum value max). For example, for example rule 1, the activation strength = min(L large, J medium); based on the activation strength, crop or scale the output fuzzy set corresponding to the THEN part (K p Medium-high, K i Low, K d The common method is minimum operation or product operation; all the output fuzzy sets of the activated rules (i.e. the K p , K i , K d fuzzy sets) are combined to form each output variable (K p , K i , K d ) is the total output fuzzy set, and the maximum value operation is often used for aggregation.

[0087] S104: Convert the total output fuzzy set (a fuzzy, irregularly shaped area) obtained in step S103 back into an accurate numerical value as the initial recommended value of the PID parameter. Common methods include: center of gravity method, maximum average method and maximum median method.

[0088] S104. The initial recommended value is combined with the moment of inertia J and other factors (such as the desired dynamic response speed and anti-disturbance requirements). The fuzzy adaptive tuner defines a reasonable range around the initial recommended value. This range is passed to the subsequent PID parameter optimization algorithm as a key electrical constraint.

[0089] S3's real-time temperature compensation methods include:

[0090] Collect the power module temperature as the input for S3 parameter decoupling;

[0091] Generate resistance drift compensation through heat transfer state space model;

[0092] The compensated resistance value is input into the S4 control parameter optimization link.

[0093] During the parameter identification process in step S3, the system also performs real-time temperature compensation to eliminate the effect of power device heating on electrical parameters. The specific operations include:

[0094] The power module's temperature signal (e.g., IGBT substrate temperature) is synchronously acquired through the microcontroller's analog input channels. This temperature signal is acquired on the same time basis as the dynamic response data (three-phase current, rotor position, and DC bus voltage). The acquired temperature data is input into a heat transfer state-space model. This model describes the process of heat transfer from the power module to the motor winding, taking into account the effects of thermal capacitance and thermal resistance (e.g., the thermal resistance characteristics from junction to housing). Its output is an estimated value of the winding temperature rise. The state-space model is essentially a mathematical abstraction of the thermal circuit network. The temperature of each node is dynamically updated through state equations. The dynamic equations use a third-order model (junction → substrate → winding), including: ;

[0095] Where, is the state vector, is the updated state vector, is the junction temperature of the power device, is the baseplate or case temperature, is the motor winding temperature, is the total power device loss, is the thermal resistance from junction to substrate, is the thermal resistance from substrate to winding, 、 and is the heat capacity of the junction, substrate and winding.

[0096] Extract from the state vector , calculate the difference from the ambient temperature, which is the estimated value of the winding temperature rise.

[0097] Based on this estimated value of the winding temperature rise, the system calculates the resistance drift compensation amount. For example, when the IGBT substrate temperature is detected to rise from room temperature 25°C to 70°C during operation, the heat transfer state space model may estimate that the winding temperature rise is 45°C. Then, based on the resistance temperature coefficient of the conductor material (for example, copper wire is approximately 0.00393 / °C), the resistance compensation coefficient α is calculated (α≈1+0.00393*45≈1.177). At this time, if the equivalent resistance R calculated by the original solution of the frequency domain impedance analysis method is 0.82Ω, the compensated resistance value R comp That is: R comp =α*R≈0.82Ω*1.177≈0.96Ω.

[0098] The equivalent resistance value after compensation is R comp , together with the equivalent inductance L obtained from frequency domain analysis, and the back EMF coefficient Ke and moment of inertia J identified through time domain back EMF zero-crossing point detection and speed differential resolution, constitute a temperature-corrected key electrical parameter set. This complete parameter set is then passed to the fuzzy adaptive tuner in step S4 to generate the initial PID control parameter group. The accurate correction of the resistance value ensures the rationality of the current loop gain calculation and avoids the resistance increase caused by temperature rise being misidentified as other electrical characteristic changes, thereby improving the accuracy of subsequent PID parameter tuning and the stability of the control system. Finally, when the optimized PID parameter group is stored in the device parameter knowledge base together with the motor type identification code, its corresponding credibility label is also improved due to the introduction of temperature compensation.

[0099] The credibility label generation method in S5 includes:

[0100] Receive performance indicator data output by S4 closed-loop verification;

[0101] Processing the performance indicators through a multi-criteria evaluation tree;

[0102] The tag state machine is set according to the evaluation results and stored in the S5 device parameter knowledge base.

[0103] During the parameter knowledge base storage process in step S5, the system dynamically generates credibility labels through a multi-criteria evaluation tree. The specific process includes: when the fuzzy adaptive tuner in S4 outputs the optimized PID parameter set, the system immediately performs a verification run in closed-loop control mode (for example: the motor accelerates to the rated speed of 3000 rpm without load and then suddenly increases the load by 50%). The performance indicator data collected during this process (including the customized torque fluctuation coefficient, speed overshoot, and steady-state tracking error) is transmitted to the evaluation module in real time.

[0104] The evaluation module uses a tree-like decision structure to process performance indicator data, including:

[0105] Primary node verification of basic stability (e.g. speed overshoot <5%);

[0106] Secondary node analysis dynamic response (e.g. torque fluctuation coefficient ≤ 0.03 N·m);

[0107] The final node comprehensively evaluates the steady-state accuracy (such as position tracking error < 0.1°).

[0108] Based on the above evaluation results, the tag state machine is updated, including:

[0109] If all indicators meet the standards (for example: overshoot 2.3% / fluctuation coefficient 0.028N·m / error 0.08°), the state machine switches to the high confidence state;

[0110] If a single indicator slightly exceeds the limit (such as a fluctuation coefficient of 0.035N·m), it will be downgraded to a pending verification state;

[0111] When key indicators are abnormal (such as overshoot > 10%), a failure state is triggered and parameter backtracking is started.

[0112] Finally, the motor type identification code, optimized PID parameter group and corresponding credibility label (such as BLDC-12P-48V: K p =0.85, K i =12,K d =0.02|Confidence: High) is associated and stored in the device parameter knowledge base. When the system subsequently calls this parameter group, the confidence label will be used as the priority sorting basis. For example, when matching the same motor model, the parameter set with high confidence is loaded first to improve control robustness.

[0113] S4's fault fusing mechanisms include:

[0114] Monitor overcurrent and stall events during S4 closed-loop verification;

[0115] Trigger security parameter group switching and generate key parameter feature hash;

[0116] Update the credibility label stored in S5 based on the event type.

[0117] When performing closed-loop verification in step S4, the system simultaneously activates the fault fuse mechanism, which continuously monitors abnormal events in the control loop, including:

[0118] By real-time analysis of defined dynamic response data (three-phase current, rotor position), overcurrent (e.g., instantaneous phase current value > 30A) or stall (e.g., speed < 5rpm within 500ms) events can be identified.

[0119] If an overcurrent event is detected (such as a U-phase current peak of 35A), the currently verified PID parameter group is immediately disconnected and switched to a pre-stored safety parameter group (such as a conservative parameter with a 60% reduction in the current loop proportional gain).

[0120] When the switch is triggered, the key parameter feature hash under the current working condition is extracted, that is, the equivalent resistance, inductance and back electromotive force coefficient at this moment are combined and encoded. For example, equivalent resistance = 0.95Ω | inductance = 1.2mH | back electromotive force coefficient = 0.05V / rpm, that is, the hash value is 0x8E3A.

[0121] Event processing results will be fed back to the S5 device parameter knowledge base, including:

[0122] If the fuse is triggered by overcurrent and stall (e.g. stall event lasts for 800ms), the system will automatically downgrade the credibility tag of the corresponding PID parameter group to invalid;

[0123] The feature hash value is associated with the event type (overcurrent, stall, and voltage abnormality) and timestamp and stored;

[0124] When the parameter group is called again for the same motor model, the system will first compare the key parameter feature hash (such as the current working condition hash value 0x8E3A matches the historical failure record) and directly skip the parameter group to prevent repeated failures.

[0125] Parameter iterative optimization functions include:

[0126] When the credibility tag stored in S5 meets the conditions, the historical parameter set is read;

[0127] Generate incremental incentives based on historical parameter sets and inject them into the S2 acquisition phase;

[0128] Update the moment of inertia through S3 decoupling and readjust the PID integral term of S4.

[0129] During the continuous operation of the parameter knowledge base, when the system detects multiple credibility tags in a pending verification state for a motor model, it automatically triggers a parameter iterative optimization loop. Take a brushless DC motor (identification code BLDC-48V-200W) as an example:

[0130] The system reads the latest three sets of historical parameter sets of the model from S5 and analyzes their difference characteristics (such as the integral term K i Fluctuates within the range of 12 to 18);

[0131] Based on the historical parameter discreteness, a controlled incremental excitation is injected during the S2 dynamic response acquisition phase. This means a step disturbance of ±4% of the rated torque is superimposed on the torque command (the safety margin is constrained by the fuse mechanism to ensure that the current does not exceed the safety threshold).

[0132] Analyze the step response waveform through S3 time-frequency domain hybrid decoupling and update the moment of inertia identification value (for example, from the initial 0.0015kg·m² to 0.0018kg·m²);

[0133] Automatically readjust the S4 PID integral gain based on the change in the moment of inertia, including: Where, is the new integral gain, is the original integral gain, is the updated moment of inertia, is the original moment of inertia, the original parameter group β has Ki=15 → new value=15× ≈16.4, the optimized parameter set immediately enters the S4 closed-loop verification phase. If the torque fluctuation coefficient decreases from 0.028 N·m to 0.025 N·m (a decrease greater than 10%), the new parameter set's credibility is upgraded to high credibility. Otherwise, the system reverts to the historically optimal parameter set α and suspends iterative optimization for this model. The valid parameter set generated during this process is continuously accumulated in the device parameter knowledge base, promoting the autonomous evolution of control parameters.

[0134] Example 2

[0135] like Figure 2 As shown, this embodiment further improves the design based on the first embodiment. The difference is that in the actual operation of the first embodiment, it is found that the credibility label is not adaptable enough to transient working conditions, resulting in the high credibility parameter group still having the risk of instability during dynamic load switching. Based on this, the embedded method for BLDC drive parameter self-tuning for multiple device types also includes:

[0136] S5: Bind the optimized PID control parameter group with the motor type identification code, store it in the device parameter knowledge base of the embedded FLASH memory, and attach a credibility tag. When the credibility tag is high credibility, inject a dynamic load disturbance sequence and collect dynamic performance indicators. If the dynamic performance indicators reach the over-limit condition, downgrade the credibility tag to steady-state high credibility, mark the applicable operating range, and associate the marked credibility tag with the applicable operating range and store it in the device parameter knowledge base.

[0137] Dynamic performance indicators include speed fluctuation rate and torque response delay time; speed fluctuation rate is the maximum relative deviation of the speed from the steady-state value when the load suddenly changes, and torque response delay is the lag time from the torque command change to the actual torque reaching the standard; over-limit conditions include speed fluctuation rate exceeding the first threshold (preset) and torque response delay exceeding the second threshold (preset). If either of the above two over-limit conditions is met, the credibility label will be downgraded.

[0138] Example 3

[0139] like Figure 3 As shown, based on the same inventive concept as the embedded method for self-tuning BLDC drive parameters for multiple device types in the aforementioned embodiment, this application provides an embedded system for self-tuning BLDC drive parameters for multiple device types. The system and method embodiments in the embodiments of this application are based on the same inventive concept. The system includes:

[0140] The core model building module builds a unified BLDC mathematical model that includes the electromagnetic torque equation, back electromotive force equation, and mechanical motion equation, and establishes a nonlinear mapping relationship library between motor type identification codes and key electrical parameter sets;

[0141] An excitation injection acquisition module injects a predefined multi-mode excitation signal sequence into the BLDC motor to be driven and simultaneously acquires dynamic response data, wherein the dynamic response data is used to characterize a set of key electrical parameters;

[0142] Parameter decoupling identification module: Based on the collected dynamic response data, the parameter decoupling identification module uses frequency domain impedance analysis to decouple and extract the equivalent inductance and equivalent resistance of the key electrical parameter set. At the same time, it uses time domain back electromotive force zero-crossing detection combined with speed differential operation to identify the back electromotive force coefficient and moment of inertia of the key electrical parameter set;

[0143] The parameter self-tuning module inputs the key electrical parameter set obtained by identification into a preset fuzzy adaptive tuner to generate an initial PID control parameter set. The control performance is verified through a closed-loop speed response test, and the PID control parameter set is dynamically optimized based on the speed overshoot and steady-state fluctuation rate.

[0144] The knowledge base storage module binds the optimized PID control parameter group with the motor type identification code, stores it in the device parameter knowledge base of the embedded FLASH memory, and adds a credibility label.

[0145] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if such changes and modifications fall within the scope of the claims and their equivalents, the present invention is intended to include such changes and modifications.

[0146] The above is only a preferred specific implementation method of the embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes based on the technical solution and concept of the present application within the technical scope disclosed in the present application, and they should be covered by the scope of protection of the present application.

Claims

1. An embedded method for self-tuning BLDC drive parameters for multiple device types, characterized by: Methods include: Construct a unified BLDC mathematical model that includes the electromagnetic torque equation, back-electromotive force equation, and mechanical motion equation, and establish a nonlinear mapping relationship library between motor type identification codes and key electrical parameter sets; injecting a predefined multi-mode excitation signal sequence into the BLDC motor to be driven and simultaneously collecting dynamic response data, wherein the dynamic response data is used to characterize a set of key electrical parameters; Based on the collected dynamic response data, the equivalent inductance and equivalent resistance of the key electrical parameters are decoupled and extracted through frequency domain impedance analysis. At the same time, the back EMF coefficient and moment of inertia of the key electrical parameters are identified by combining time domain back EMF zero-crossing detection with speed differential operation. The identified key electrical parameter set is input into a preset fuzzy adaptive tuner to generate an initial PID control parameter set. The control performance is verified through a closed-loop speed response test, and the PID control parameter set is dynamically optimized based on the speed overshoot and steady-state fluctuation rate. The optimized PID control parameter group is bound to the motor type identification code, stored in the device parameter knowledge base of the embedded FLASH memory, and attached with a credibility label.

2. The embedded method for self-tuning BLDC drive parameters for multiple device types according to claim 1 is characterized in that: The method for constructing a nonlinear mapping relationship library includes: The motor no-load starting current waveform is input into a waveform encoder to generate a feature vector; the feature vector is processed by a lightweight convolutional neural network; and the output of the lightweight convolutional neural network is used as an iterative initial value for key electrical parameter identification.

3. The embedded method for self-tuning BLDC drive parameters for multiple device types according to claim 1, characterized in that: The method for generating a multi-mode excitation signal sequence includes: A basic excitation template is loaded according to the motor type identification code; a swarm intelligence algorithm is used to optimize the gradient parameters of the basic excitation template; and the optimized gradient parameters are injected into the dynamic response data collection link.

4. The embedded method for self-tuning BLDC drive parameters for multiple device types according to claim 1, characterized in that: Frequency domain impedance analysis methods include: Wavelet packet decomposition is performed on the collected swept frequency response current; a neural differential equation solver is constructed based on the decomposition result; and the neural differential equation solver outputs equivalent circuit parameters to the fuzzy adaptive tuner.

5. The embedded method for self-tuning BLDC drive parameters for multiple device types according to claim 1, characterized in that: Real-time temperature compensation methods include: The power module temperature is collected as the input for parameter decoupling; the resistance drift compensation is generated through the heat transfer state space model; and the compensated resistance value is input into the optimization phase of the PID control parameter group.

6. The embedded method for self-tuning BLDC drive parameters for multiple device types according to claim 1, characterized in that: Credibility label generation methods include: Receive performance indicator data output by a closed-loop speed response test verification; process the performance indicator through a multi-criteria evaluation tree; set a tag state machine based on the evaluation results and store them in a device parameter knowledge base.

7. The embedded method for self-tuning BLDC drive parameters for multiple device types according to claim 1, characterized in that: Fault fuse mechanisms include: Monitor overcurrent and stall events in S4 closed-loop verification; trigger safety parameter group switching and generate key parameter feature hashes; update stored credibility tags based on event types.

8. The embedded method for BLDC drive parameter self-tuning for multiple device types according to claim 1, characterized in that: Parameter iterative optimization functions include: When the stored credibility tag meets the conditions, the historical parameter set is read; based on the historical parameter set, an incremental excitation is generated and injected into the dynamic response data acquisition link; the moment of inertia is updated through decoupling and the PID integral term is readjusted.

9. The embedded method for BLDC drive parameter self-tuning for multiple device types according to claim 1, characterized in that: The method also includes: The optimized PID control parameter group is bound to the motor type identification code, stored in the device parameter knowledge base of the embedded FLASH memory, and attached with a credibility label. When the credibility label is high credibility, a dynamic load disturbance sequence is injected and dynamic performance indicators are collected. If the dynamic performance indicators reach the over-limit condition, the credibility label is downgraded to steady-state high credibility, and the applicable operating range is marked. The marked credibility label is associated with the applicable operating range and stored in the device parameter knowledge base.

10. BLDC drive parameter self-tuning embedded system for multiple device types, characterized by: The system includes: A core model building module, which constructs a unified BLDC mathematical model including electromagnetic torque equations, back electromotive force equations, and mechanical motion equations, and establishes a nonlinear mapping relationship library between motor type identification codes and key electrical parameter sets; an excitation injection acquisition module, which injects a predefined multi-mode excitation signal sequence into the BLDC motor to be driven and simultaneously acquires dynamic response data, wherein the dynamic response data is used to characterize the characteristics of the key electrical parameter set; A parameter decoupling identification module, which extracts the equivalent inductance and equivalent resistance of key electrical parameters through frequency domain impedance analysis based on the collected dynamic response data. It also identifies the back electromotive force coefficient and moment of inertia of key electrical parameters using time domain back electromotive force zero-crossing detection combined with speed differential calculations. A parameter self-tuning module, which inputs the identified key electrical parameter set into a preset fuzzy adaptive tuner to generate an initial PID control parameter set, verifies control performance through a closed-loop speed response test, and dynamically optimizes the PID control parameter set based on speed overshoot and steady-state fluctuation rate; The knowledge base storage module binds the optimized PID control parameter group with the motor type identification code, stores the optimized PID control parameter group with the motor type identification code in the device parameter knowledge base of the embedded FLASH memory, and adds a credibility tag.

Images