Motor current sound optimization method

Abstract

The invention provides a motor current sound optimization method, which relates to the technical field of motors and comprises the following steps: acquiring an original request current signal output by an upper controller; performing intelligent self-adaptive average value filtering processing based on motor operation state sensing and real-time signal feature analysis on the original request current signal, and outputting a request current signal subjected to average value filtering through a variable depth sliding window mechanism and a dual-threshold abnormal data processing strategy; performing composite architecture low-pass filtering processing based on a parallel filtering architecture on the request current signal subjected to average value filtering, and outputting a smooth request current signal subjected to low-pass filtering through the synergistic effect of broadband self-adaptive low-pass filtering and specific sub-harmonic wave trapping suppression; and the smoothing request current signal after average value filtering and low-pass filtering processing is sent to a current loop controller to serve as a reference input value, and the current loop controller generates a PWM driving signal.

Description

Technical Field

[0001] This invention relates to the field of motor technology, specifically a method for optimizing motor current noise. Background Technology

[0002] In motor control applications, such as industrial drives, automotive electric drives, or smart home systems that use pulse width modulation (PWM) control, audible current noise, often manifested as a high-frequency "squeaking" or "hissing" sound, remains a key issue restricting product noise quality.

[0003] To address this problem, common technical methods can be broadly categorized into three types: One approach is to increase the PWM frequency to push harmonics into a frequency band that is not sensitive to the human ear. However, this will increase switching losses, reduce efficiency, and is limited by the power device's frequency limit. Secondly, strengthening hardware filtering, such as adding inductors or capacitors on the motor side, will lead to an increase in size, weight and cost. Third, while using modulation methods such as random PWM or harmonic injection can disperse harmonic energy, it can also easily cause torque fluctuations and affect control stability.

[0004] In summary, existing technologies cannot effectively suppress audible electrical noise without significantly increasing losses and costs.

[0005] Therefore, a method for optimizing motor current noise is provided to solve the problem in the prior art that cannot effectively suppress audible current noise without significantly increasing losses and costs. Summary of the Invention

[0006] In order to solve the above-mentioned technical problems, the purpose of this invention is to provide a method for optimizing motor current noise, which solves the problem that existing technologies cannot effectively suppress audible current noise without significantly increasing losses and costs.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a method for optimizing motor current noise, comprising: Obtain the raw request current signal output by the upper-level controller; The original request current signal is subjected to intelligent adaptive average value filtering based on motor operating status perception and real-time signal feature analysis. The request current signal after average value filtering is output through a variable depth sliding window mechanism and a dual threshold abnormal data processing strategy. The requested current signal after average value filtering is subjected to low-pass filtering based on a composite architecture with a parallel filtering architecture. Through the synergistic effect of wideband adaptive low-pass filtering and specific subharmonic notch suppression, a smoothed requested current signal after low-pass filtering is output. At the same time, the filtering parameters are optimized online based on a closed-loop performance sensing mechanism. The smoothed request current signal, after being processed by the average value filtering and low-pass filtering, is sent to the current loop controller as a reference input value, and the current loop controller generates a PWM drive signal.

[0008] Preferably, based on motor operating status perception and real-time signal feature analysis, the original requested current signal is subjected to intelligent adaptive averaging filtering. Through a variable-depth sliding window mechanism and a dual-threshold abnormal data processing strategy, the output averaging-filtered requested current signal includes: Based on the collected continuous raw request current signals, a sampling sequence is obtained, and a sliding window is initialized. The sliding window is dynamically adjusted between a preset maximum window threshold and a minimum window threshold. For the sampling sequence within the current sliding window, outliers are removed based on a dual threshold decision strategy to obtain the corresponding valid sampling sequence; Based on the acquired real-time operating status of the motor, the changing trend of the requested current signal is predicted, and based on the changing trend, the reference length of the sliding window is determined. Calculate the real-time signal change rate of the current effective sampling sequence, and correct the reference length based on the real-time signal change rate to obtain the final sliding window length; the final sliding window length is constrained between a preset maximum window threshold and a minimum window threshold. Based on the effective sampling sequence within the final sliding window length, a two-dimensional weighting strategy is used to perform average value filtering calculation to obtain the requested current signal after average value filtering.

[0009] Preferably, for the sampling sequence within the current sliding window, outlier removal is performed based on a dual-threshold decision strategy to obtain the corresponding valid sampling sequence, including: Based on the sampling sequence within the current sliding window, determine whether each sampling point in the sampling sequence simultaneously exceeds both the static threshold interval and the dynamic threshold interval; the dynamic threshold interval is an interval obtained based on the statistical characteristics of the sampling sequence within the current sliding window; the static threshold interval is a fixed interval set based on the maximum allowable current of the motor. If not, then the sampling point is a valid point and is saved to the valid sampling sequence; Conversely, if the sampling point is not found, it is determined to be an outlier. The sampling point initially determined to be an outlier is designated as a pending point, and multiple subsequent sampling points are collected to form a confirmation window. Calculate the mean and standard deviation of the sampled data within the confirmation window, and determine the corresponding interval for the undetermined point; If the value of the undetermined point is outside the defined range of the undetermined point, and the data points in the confirmation window are in the same trend of change as the undetermined point, then the undetermined point is determined to be a valid transient signal, and the valid transient signal is retained in the valid sampling sequence; If the value of the undetermined point is within the defined range of the undetermined point, then the undetermined point is determined to be abnormal and is removed.

[0010] Preferably, based on the acquired real-time operating status of the motor, the prediction of the changing trend of the requested current signal, and the determination of the reference length of the sliding window based on the changing trend, includes: Establish a mapping library between motor operating conditions and requested current signal change patterns; Based on the obtained real-time operating status of the motor, the current change pattern with the highest matching degree is matched from the mapping relationship library; the real-time operating status includes the current motor speed, load torque, and uploaded control commands; Based on the matched current change pattern, the optimal window length corresponding to the current change pattern is queried, and the optimal window length is set as the reference length of the sliding window.

[0011] Preferably, the final sliding window length is obtained by correcting the reference length based on the real-time signal change rate, including: When the rate of change of the real-time signal is less than the preset rate of change threshold, the real-time signal is determined to be stable, and an offset is added to the reference length to obtain the final sliding window length. Conversely, if the real-time signal changes drastically, the offset is subtracted from the reference length to obtain the final sliding window length.

[0012] Preferably, the final sliding window length is obtained by correcting the reference length based on the real-time signal change rate, including: When the rate of change of the real-time signal is less than the preset rate of change threshold, the real-time signal is determined to be stable, and an offset is added to the reference length to obtain the final sliding window length. Conversely, if the real-time signal changes drastically, the offset is subtracted from the reference length to obtain the final sliding window length.

[0013] Preferably, the requested current signal filtered by the average value is subjected to a composite low-pass filter based on a parallel filtering architecture. Through the synergistic effect of wideband adaptive low-pass filtering and specific subharmonic notch suppression, the output smoothed requested current signal after low-pass filtering includes: Based on the spectrum of the original requested current signal and the real-time operating status of the motor, the spectrum of the requested current signal after average filtering is analyzed to identify one or more specific dominant harmonic frequencies under the current operating conditions. Based on one or more identified specific subdominant harmonic frequencies, a wideband adaptive low-pass filter and one or more specific subdominant notch filters are dynamically configured; the center frequency of the specific subdominant notch filter is locked at the identified specific subdominant harmonic frequency. The requested current signal, after being filtered by the average value, is synchronously input into a wideband adaptive low-pass filter and a specific notch filter. The corresponding low-pass filter output signal and notch filter output signal are acquired, and the notch filter output signal is removed from the low-pass filter output signal, thus forming a smoothed requested current signal after the average value filtering and low-pass filtering processing.

[0014] Preferably, online optimization of filter parameters based on a closed-loop performance-aware mechanism includes: The actual output phase current of the current loop controller is collected in real time, and based on the actual output phase current, the corresponding total harmonic distortion (THD) and the distortion rate of one or more specific harmonics that contribute the most to the noise and vibration of the motor are calculated synchronously. Based on the total harmonic distortion (THD) and distortion rate, and based on a preset multi-objective collaborative optimization strategy, the cutoff frequency of the broadband adaptive low-pass filter and the depth and bandwidth parameters of the specific sub-notch filter are optimized and adjusted to form optimized filter parameters.

[0015] Compared with existing technologies, the advantages of this invention are as follows: This invention innovatively combines intelligent adaptive average filtering and composite low-pass filtering in a two-stage preprocessing scheme, forming a progressive signal optimization link. The first-stage average filtering, through a variable-depth sliding window and dual-threshold anomaly removal, first filters out sudden abnormal data and high-frequency glitches in the original current signal, laying a stable signal foundation for subsequent processing. The second-stage composite low-pass filtering, through the synergy of wideband adaptive filtering and specific notch suppression, further suppresses the remaining harmonic components. The two-stage processing is progressive, avoiding the problem of incomplete suppression of complex interference by a single filter, while effectively preserving the effective characteristics of the current signal, significantly reducing the distortion of the smooth current signal output, and reducing the source of harmonic noise in the PWM drive stage from the source. That is, without significantly increasing losses and costs, by using average filtering and low-pass filtering in series, it can smooth the step changes of the waveform and filter out high-frequency glitches, achieving a combination of "fine-tuning" and "coarse-tuning" effects, thereby effectively suppressing audible current noise. Dual smoothing filtering at the command source significantly reduces high-frequency harmonics and glitches in the current that could generate noise, resulting in direct and significant noise reduction. Since the filtering process is performed on the open-loop feedforward path, it does not affect the feedback loop of the closed-loop control, thus enhancing the system's robustness.

[0016] Compared to traditional fixed-parameter filtering schemes, this invention incorporates adaptive logic in both stages of filtering. In the average value filtering stage, the current variation pattern is matched based on the motor's real-time speed, load torque, and other operating conditions. The sliding window reference length is dynamically adjusted, and the window size is corrected by combining the signal change rate, ensuring sufficient filtering depth when the current is stable and timely response when the signal changes abruptly. In the low-pass filtering stage, specific dominant harmonics under the current operating conditions are identified in real-time through spectrum analysis. The center frequency of the notch filter and the cutoff frequency of the low-pass filter are dynamically configured, accurately addressing harmonic characteristic changes under different motor operating conditions and overcoming the shortcomings of traditional fixed filtering parameters that are difficult to adapt to complex dynamic scenarios. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0018] Figure 1 This is a schematic diagram of a method for optimizing motor current noise. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be described in detail below. Obviously, the described embodiments are merely some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other implementation methods obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0020] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0021] Example 1 like Figure 1 As shown, this embodiment discloses a method for optimizing motor current noise, including: S1. Obtain the raw request current signal output by the upper-level controller; S2. Perform intelligent adaptive average value filtering on the original request current signal based on motor operating status perception and real-time signal feature analysis. Output the average value filtered request current signal through a variable depth sliding window mechanism and a dual threshold abnormal data processing strategy. S2 includes: S20. Based on the collected continuous raw request current signals, obtain the sampling sequence and initialize the sliding window, wherein the sliding window is dynamically adjusted between a preset maximum window threshold and a minimum window threshold. S21. For the sampling sequence within the current sliding window, outliers are removed based on a dual threshold decision strategy to obtain the corresponding valid sampling sequence. It should be noted that the sampling sequence within the current sliding window, based on a dual-threshold decision strategy, is used to remove outliers to obtain the corresponding valid sampling sequence, including: Based on the sampling sequence within the current sliding window, it is determined whether each sampling point in the sampling sequence simultaneously exceeds both the static threshold interval and the dynamic threshold interval. The dynamic threshold interval is an interval obtained based on the statistical characteristics of the sampling sequence within the current sliding window. The static threshold interval is a fixed interval set based on the maximum allowable current of the motor. In this embodiment, the key hardware parameters of the motor, the maximum operating current of the motor, and the rated operating current of the motor are obtained. When setting the static threshold interval, the corresponding expression is: ;in, This is directly related to the motor's maximum allowable current, which is typically taken as 90%–100% of the motor's maximum allowable current. In this embodiment, 95% of the motor's maximum allowable current is used. The minimum operating requirement is set based on the rated operating current of the motor, which is usually 10% to 20% of the rated operating current of the motor. In this embodiment, 15% of the rated operating current of the motor is used.

[0022] In another embodiment, the real-time temperature of the motor corresponding to the motor is also obtained. and current load torque ,in, ;in, This refers to the normal temperature threshold of the motor. This is the maximum allowable temperature for the motor. For temperature coefficient, This is the load torque coefficient. It is calibrated according to the motor model.

[0023] In this embodiment, the expression for the dynamic threshold interval is: ; in The standard deviation of the sampled sequence, The mean of the sampled sequence. For adjustment coefficients; ; ; In the formula, n is the current sliding window length, and the sampling sequence within the current sliding window is: ; This refers to the sampled value of the i-th sampling point in the sampling sequence. In this embodiment, the adjustment coefficient of the dynamic threshold interval... It is dynamically adjusted. Specifically, before the motor leaves the factory, the correspondence between the standard deviation of current fluctuation and the optimal adjustment coefficient under different operating conditions is collected through experiments to form an initial training library.

[0024] Within the current sliding window, calculate the fluctuation characteristic value of the sampled sequence within the current sliding window. ; If the fluctuation characteristic value Then, the corresponding k value for the low-fluctuation scenario is retrieved from the initial training library; If the fluctuation characteristic value If so, the default k value will be used; If the fluctuation characteristic value Then, the corresponding k value for the high volatility scenario is retrieved from the initial training library. The k value for the low volatility scenario is less than the default k value, and the default k value is less than the k value for the high volatility scenario.

[0025] If not, then the sampling point is a valid point and is saved to the valid sampling sequence; Conversely, if the sampling point is not found, it is determined to be an outlier. The sampling point initially determined to be an outlier is designated as a pending point, and multiple subsequent sampling points are collected to form a confirmation window. Calculate the mean and standard deviation of the sampled data within the confirmation window, and determine the corresponding interval for the undetermined point; If the value of the undetermined point is outside the defined interval of the undetermined point, and the data points within the confirmation window are in the same trend as the undetermined point, then the undetermined point is determined to be a valid transient signal, and the valid transient signal is retained in the valid sampling sequence; in this embodiment, the strategy for determining whether they are in the same trend is: Perform linear regression analysis on the sampled data within the confirmation window to obtain the slope of the regression line; Calculate the instantaneous rate of change of the point to be determined relative to its previous sampling point; If the sign of the slope of the regression line is the same as the sign of the instantaneous rate of change, and the absolute value of the slope of the regression line is greater than a preset significance threshold, then it is determined that they are in the same trend of change.

[0026] If the value of the undetermined point is within the defined range of the undetermined point, then the undetermined point is determined to be abnormal and is removed.

[0027] S22. Based on the acquired real-time operating status of the motor, predict the changing trend of the requested current signal, and determine the reference length of the sliding window based on the changing trend. It should be noted that the process of predicting the changing trend of the requested current signal based on the acquired real-time operating status of the motor, and determining the reference length of the sliding window based on the changing trend, includes: Establish a mapping library between motor operating conditions and requested current signal change patterns; Based on the obtained real-time operating status of the motor, the current change pattern with the highest matching degree is matched from the mapping relationship library; the real-time operating status includes the current motor speed, load torque, and uploaded control commands; Based on the matched current change pattern, the optimal window length corresponding to the current change pattern is queried, and the optimal window length is set as the reference length of the sliding window.

[0028] S23. Calculate the real-time signal change rate of the current effective sampling sequence, and correct the reference length based on the real-time signal change rate to obtain the final sliding window length; the final sliding window length is constrained between a preset maximum window threshold and a minimum window threshold. It should be noted that the step of correcting the reference length based on the real-time signal change rate to obtain the final sliding window length includes: When the rate of change of the real-time signal is less than the preset rate of change threshold, the real-time signal is determined to be stable, and the final sliding window length is increased by an offset based on the baseline length. Conversely, if the real-time signal changes drastically, the final sliding window length is reduced by the offset from the baseline length. In this embodiment, the value of the offset for adding or subtracting is dynamically changed, and the specific change logic is as follows: ; In the formula, For the corresponding increase or decrease in offset, and These are the corresponding weight coefficients. .

[0029] S24. Based on the effective sampling sequence within the final sliding window length, a dual-dimensional weighting strategy is used to calculate the average value filtering, resulting in the requested current signal after average value filtering. The dual-dimensional weighting strategy primarily employs a dual weighting of the time dimension and the data quality dimension: the time weight makes new data more important, while the quality weight automatically reduces the proportion of outliers. The two are multiplied to form a comprehensive weight, which is then used for weighted averaging. Furthermore, the weight parameters can be adaptively adjusted according to the signal characteristics.

[0030] S3. Perform low-pass filtering on the average value filtered request current signal using a composite architecture based on a parallel filtering architecture. Through the synergistic effect of wideband adaptive low-pass filtering and specific subharmonic notch suppression, output a smooth request current signal after low-pass filtering. At the same time, optimize the filtering parameters online based on a closed-loop performance sensing mechanism. The S3 includes: S30. Based on the spectrum of the original requested current signal and the real-time operating status of the motor, the spectrum of the requested current signal after average value filtering is analyzed to identify one or more specific dominant harmonic frequencies under the current operating condition. In this embodiment, the selection strategy for the dominant harmonic frequency is as follows: Perform a Fast Fourier Transform on the original request current signal to obtain the original spectrum containing complete harmonic components. ; Perform a Fast Fourier Transform on the average-filtered request current signal to obtain the smoothed current spectrum. ; From the original spectrum Potential harmonic components were identified, including the frequencies of each harmonic. and initial amplitude Establish the corresponding harmonic candidate set; The characteristic harmonic frequency under the current operating condition is calculated based on the real-time operating status of the motor. The calculation formula is as follows: ; In the formula, This represents the number of pole pairs corresponding to the motor. This represents the motor speed.

[0031] Verify the correlation between measured harmonics and theoretical harmonics, and screen out harmonic components that are physically related to the current operating state.

[0032] Based on the current spectrum Accurately locate candidate harmonics and obtain the current actual amplitude. ; Based on a pre-defined formula for calculating the overall importance of harmonics, the overall importance of each harmonic is calculated. The harmonics are then ranked according to their importance index, and the top K most important harmonics are selected as the dominant harmonics, outputting a set of dominant harmonic frequencies. .

[0033] S31. Based on the identified one or more specific subdominant harmonic frequencies, dynamically configure a wideband adaptive low-pass filter and one or more specific subdominant notch filters; the center frequency of the specific subdominant notch filter is locked at the identified specific subdominant harmonic frequency; in this embodiment, the dynamic configuration strategy is as follows: Based on the identified set of dominant harmonic frequencies Initialize configuration parameters: Cutoff frequency of wideband adaptive low-pass filter ,in This corresponds to the safety factor; For each dominant harmonic frequency Independently configured notch filter, notch filter execution frequency Notch filter bandwidth It is determined by the following formula: ; , This is the bandwidth adjustment factor. Dominant harmonic frequency The short-term standard deviation of the fluctuation.

[0034] Based on the cutoff frequency of the broadband adaptive low-pass filter, the operating frequency of each notch filter, and the corresponding bandwidth values, the performance of the entire system is evaluated, and the harmonic distortion reduction effect under the current filter parameters is calculated. Phase lag and computational complexity In this embodiment, the harmonic distortion reduction effect is evaluated by comparing the total harmonic distortion of the signal before and after filtering. Phase lag is evaluated by analyzing the phase shift of a signal at a specific frequency. Specifically, a standard sinusoidal test signal with the same fundamental frequency as the current motor is input into the filtering system, and the phase change of the signal after passing through the filter is measured. This phase difference is the phase lag of the system under the current configuration. Computational complexity is evaluated by statistically analyzing the amount of computation required to execute the filter. For a parallel wideband low-pass filter and multiple notch filters, the total computational complexity is the sum of the computational complexities of each filter.

[0035] Based on the calculated expected harmonic distortion reduction effect, phase hysteresis, and computational complexity, and using a pre-defined multi-objective collaborative optimization function, the cutoff frequency of the broadband adaptive low-pass filter and the bandwidth of each notch filter are optimized using existing genetic algorithms or gradient descent methods to minimize the objective function. The multi-objective collaborative optimization function is as follows: ; In the formula, This is an estimate of the maximum harmonic distortion reduction achievable under current operating conditions. The minimum acceptable phase lag for the system. This serves as the benchmark for the lowest computing cost of the processor. , These are the corresponding weighting coefficients.

[0036] The parameters are adaptively adjusted in real time based on minimizing the corresponding filter parameters according to the objective function.

[0037] S32. The request current signal after average filtering is synchronously input into a wideband adaptive low-pass filter and a specific notch filter. The corresponding low-pass filter output signal and notch filter output signal are collected, and the notch filter output signal is deleted from the low-pass filter output signal to finally form a smooth request current signal after the average filtering and low-pass filtering processing.

[0038] It should be noted that the online optimization of filter parameters based on the closed-loop performance awareness mechanism includes: S300. Real-time acquisition of the actual output phase current of the current loop controller, and based on the actual output phase current, synchronous calculation of the corresponding total harmonic distortion (THD), and the distortion rate of one or more specific harmonics that contribute the most to motor noise and vibration; in this embodiment, the calculation logic of the distortion rate is as follows: ; In the formula, , It is the effective value of the h-th harmonic. It is the effective value of the fundamental frequency.

[0039] S301. Based on the total harmonic distortion (THD) and distortion rate, and based on a preset multi-objective collaborative optimization strategy, the cutoff frequency of the broadband adaptive low-pass filter and the depth and bandwidth parameters of the specific sub-notch filter are optimized and adjusted to form optimized filter parameters; in this embodiment, the multi-objective collaborative optimization strategy adopts a two-level dynamic priority optimization mechanism, specifically: Level 1 If the distortion rate exceeds the preset threshold, then the focus shifts to core harmonic suppression, prioritizing a significant increase in the notch depth (D) of the corresponding 6k±1 harmonic notch filter. The notch depth D determines the degree of attenuation for harmonics at a specific frequency. Taking the i-th notch filter as an example, the corresponding adjustment formula is: ; To adjust the notch depth step size, The notch depth of the i-th notch filter before adjustment. This represents the adjusted notch depth of the i-th notch filter. Of course, if increasing the notch depth is ineffective, or if the notch depth has reached its limit, the bandwidth (BW) of the corresponding notch filter can be appropriately reduced.

[0040] Level 2 If the total harmonic distortion (THD) is less than or equal to a preset distortion rate threshold, and if the THD is greater than a preset THD threshold, the cutoff frequency of the broadband adaptive low-pass filter is slightly reduced. The adjustment formula is as follows: ; The adjustment step size for the low-pass cutoff frequency; If the total harmonic distortion (THD) is less than or equal to the preset THD threshold, the overall waveform quality is considered good. In this case, the cutoff frequency of the low-pass filter can be slightly increased, provided that both the total harmonic distortion (THD) and the threshold are met. The adjustment formula is as follows: ; It is a ratio Smaller, finer adjustment steps.

[0041] S302, The optimized filter parameters are fed back to S31 in real time.

[0042] S4. The smoothed request current signal after the average value filtering and low-pass filtering is sent to the current loop controller as its reference input value, and the current loop controller generates a PWM drive signal.

[0043] Example 2 Compared with Embodiment 1, this embodiment provides another method for optimizing motor current noise. The overall steps are basically the same as those in Embodiment 1, except for the step of correcting the reference length of the sliding window based on the real-time signal change rate. The specific modifications are as follows: The step of correcting the reference length based on the real-time signal change rate to obtain the final sliding window length includes: Based on the real-time signal change rate, extract the dynamic characteristic index of the real-time signal change rate; Based on the dynamic characteristic indicators, the current signal change pattern is identified by a pattern recognition algorithm. The signal change pattern includes a steady gradual change, a step change, a periodic oscillation, and a random disturbance. Based on the identified current signal change pattern, a multi-modal correction strategy is used to adaptively correct the reference length.

[0044] The multimodal correction strategy is as follows: For the steady-state gradient type, a gradual correction strategy is adopted, and the window length is slowly increased according to the gradual saturation correction function; the gradual saturation correction function is: ; In the formula, This is the final sliding window length. This is the baseline length of the sliding window. Based on the offset, For real-time signal change rate, For gradual adjustment coefficients; For step-change mutations, an aggressive correction strategy is adopted, rapidly reducing the window length according to the aggressive correction function; the aggressive correction function is: ; In the formula, For the radical adjustment coefficient, To prevent small positive numbers from being divided by zero.

[0045] For periodic oscillations, a resonance suppression correction formula is used to synchronize the window length with the oscillation period; the resonance suppression correction formula is as follows: ; In the formula, For modulation amplitude, As the dominant oscillation frequency, Sampling frequency, This is the floor function.

[0046] For random disturbances, a robust preservation strategy is adopted to basically keep the current window length unchanged.

[0047] Optionally, in this embodiment, those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing the hardware related to the terminal device. The program can be stored in a computer-readable storage medium, which may include: flash drive, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0048] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0049] If the integrated units in the above embodiments are implemented as software functional units and sold or used as independent products, they can be stored in the aforementioned computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause one or more electronic devices to execute all or part of the steps of the methods described in the various embodiments of this application.

[0050] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0051] In the several embodiments provided in this application, it should be understood that the disclosed application can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection of units or modules may be electrical or other forms.

[0052] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0053] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0054] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for optimizing motor current noise, characterized in that, The method includes: Obtain the raw request current signal output by the upper-level controller; The original request current signal is subjected to intelligent adaptive average value filtering based on motor operating status perception and real-time signal feature analysis. The request current signal after average value filtering is output through a variable depth sliding window mechanism and a dual threshold abnormal data processing strategy. The requested current signal after average value filtering is subjected to low-pass filtering based on a composite architecture with a parallel filtering architecture. Through the synergistic effect of wideband adaptive low-pass filtering and specific subharmonic notch suppression, a smoothed requested current signal after low-pass filtering is output. At the same time, the filtering parameters are optimized online based on a closed-loop performance sensing mechanism. The smoothed request current signal, after being processed by the average value filtering and low-pass filtering, is sent to the current loop controller as a reference input value, and the current loop controller generates a PWM drive signal.

2. The method for optimizing motor current noise according to claim 1, characterized in that, Based on motor operating status perception and real-time signal feature analysis, the original requested current signal is intelligently and adaptively filtered by means of a variable-depth sliding window mechanism and a dual-threshold abnormal data processing strategy. The output requested current signal after being filtered by means of a variable-depth sliding window mechanism includes: Based on the collected continuous raw request current signals, a sampling sequence is obtained, and a sliding window is initialized. The sliding window is dynamically adjusted between a preset maximum window threshold and a minimum window threshold. For the sampling sequence within the current sliding window, outliers are removed based on a dual threshold decision strategy to obtain the corresponding valid sampling sequence; Based on the acquired real-time operating status of the motor, the changing trend of the requested current signal is predicted, and based on the changing trend, the reference length of the sliding window is determined. Calculate the real-time signal change rate of the current effective sampling sequence, and correct the reference length based on the real-time signal change rate to obtain the final sliding window length; The final sliding window length is constrained between a preset maximum window threshold and a minimum window threshold. Based on the effective sampling sequence within the final sliding window length, a two-dimensional weighting strategy is used to perform average value filtering calculation to obtain the requested current signal after average value filtering.

3. The method for optimizing motor current noise according to claim 2, characterized in that, For the sampling sequence within the current sliding window, outlier removal is performed based on a dual-threshold decision strategy to obtain the corresponding valid sampling sequences, including: Based on the sampling sequence within the current sliding window, determine whether each sampling point in the sampling sequence simultaneously exceeds both the static threshold interval and the dynamic threshold interval; the dynamic threshold interval is an interval obtained based on the statistical characteristics of the sampling sequence within the current sliding window; the static threshold interval is a fixed interval set based on the maximum allowable current of the motor. If not, then the sampling point is a valid point and is saved to the valid sampling sequence; Conversely, if the sampling point is not found, it is determined to be an outlier. The sampling point initially determined to be an outlier is designated as a pending point, and multiple subsequent sampling points are collected to form a confirmation window. Calculate the mean and standard deviation of the sampled data within the confirmation window, and determine the corresponding interval for the undetermined point; If the value of the undetermined point is outside the defined range of the undetermined point, and the data points in the confirmation window are in the same trend of change as the undetermined point, then the undetermined point is determined to be a valid transient signal, and the valid transient signal is retained in the valid sampling sequence; If the value of the undetermined point is within the defined range of the undetermined point, then the undetermined point is determined to be abnormal and is removed.

4. The method for optimizing motor current noise according to claim 3, characterized in that, Based on the acquired real-time operating status of the motor, the changing trend of the requested current signal is predicted, and based on the changing trend, the reference length of the sliding window is determined, including: Establish a mapping library between motor operating conditions and requested current signal change patterns; Based on the obtained real-time operating status of the motor, the current change pattern with the highest matching degree is matched from the mapping relationship library; the real-time operating status includes the current motor speed, load torque, and uploaded control commands; Based on the matched current change pattern, the optimal window length corresponding to the current change pattern is queried, and the optimal window length is set as the reference length of the sliding window.

5. The method for optimizing motor current noise according to claim 4, characterized in that, The reference length is corrected based on the real-time signal change rate to obtain the final sliding window length, which includes: When the rate of change of the real-time signal is less than the preset rate of change threshold, the real-time signal is determined to be stable, and an offset is added to the reference length to obtain the final sliding window length. Conversely, if the real-time signal changes drastically, the offset is subtracted from the reference length to obtain the final sliding window length.

6. The method for optimizing motor current noise according to claim 4, characterized in that, The step of correcting the reference length based on the real-time signal change rate to obtain the final sliding window length also includes: Based on the real-time signal change rate, extract the dynamic characteristic index of the real-time signal change rate; Based on the dynamic characteristic indicators, the current signal change pattern is identified by a pattern recognition algorithm. The signal change pattern includes a steady gradual change, a step change, a periodic oscillation, and a random disturbance. Based on the identified current signal change pattern, a multimodal correction strategy is used to adaptively correct the reference length, thereby obtaining the final sliding window length.

7. The method for optimizing motor current noise according to claim 5, characterized in that, The average-value filtered request current signal is subjected to a composite low-pass filter based on a parallel filtering architecture. Through the synergistic effect of wideband adaptive low-pass filtering and specific harmonic notch suppression, the output smoothed request current signal after low-pass filtering includes: Based on the spectrum of the original requested current signal and the real-time operating status of the motor, the spectrum of the requested current signal after average filtering is analyzed to identify one or more specific dominant harmonic frequencies under the current operating conditions. Based on one or more identified specific subdominant harmonic frequencies, a wideband adaptive low-pass filter and one or more specific subdominant notch filters are dynamically configured; the center frequency of the specific subdominant notch filter is locked at the identified specific subdominant harmonic frequency. The requested current signal, after being filtered by the average value, is synchronously input into a wideband adaptive low-pass filter and a specific notch filter. The corresponding low-pass filter output signal and notch filter output signal are acquired, and the notch filter output signal is removed from the low-pass filter output signal, thus forming a smoothed requested current signal after the average value filtering and low-pass filtering processing.

8. The method for optimizing motor current noise according to claim 7, characterized in that, Online optimization of filter parameters based on a closed-loop performance-aware mechanism includes: The actual output phase current of the current loop controller is collected in real time, and based on the actual output phase current, the corresponding total harmonic distortion (THD) and the distortion rate of one or more specific harmonics that contribute the most to the noise and vibration of the motor are calculated synchronously. Based on the total harmonic distortion (THD) and distortion rate, and based on a preset multi-objective collaborative optimization strategy, the cutoff frequency of the broadband adaptive low-pass filter and the depth and bandwidth parameters of the specific sub-notch filter are optimized and adjusted to form optimized filter parameters.

Images