A Harmonic Injection Control System and Method for Electric Drive Systems Based on Progressive Fitting Prediction
By employing a progressive fitting and prediction-based harmonic injection control method in the electric drive system, a current amplitude and phase prediction function is constructed, solving the problems of long time consumption and high cost in the existing technology. This achieves efficient and stable noise suppression across the entire torque range, improving the NVH performance of the electric drive system.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- ZHIXIN TECH CO LTD
- Filing Date
- 2026-04-28
- Publication Date
- 2026-07-10
AI Technical Summary
Existing harmonic injection methods require calibration testing for each torque requirement, which is time-consuming and costly. Furthermore, the parameters are inconsistent under different operating conditions, resulting in high test bench and labor costs.
A harmonic injection control method for electric drive systems based on progressive fitting and prediction is adopted. By progressively optimizing the harmonic current at a selected reference torque point, a current amplitude and phase prediction function is constructed to achieve optimal control across the entire torque range. Performance is optimized by combining parameter fitting and closed-loop verification.
It achieves efficient and low-cost harmonic injection calibration, reduces test time and resource consumption, ensures stable and smooth noise suppression across the entire torque range, and improves the NVH performance of the electric drive system.
Smart Images
Figure CN122371797A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of motor control and noise and vibration optimization, specifically to a harmonic injection control system and method for electric drive systems based on progressive fitting prediction. Background Technology
[0002] With the rapid growth of the electric vehicle market, the NVH performance of electric drive systems (motors, reducers, and motor controllers) has gradually become a focus of end-user attention. Traditional noise suppression methods typically rely on structural adjustments. Existing harmonic injection methods all employ a sweeping injection approach, requiring calibration tests for each required torque to select the optimal current amplitude and phase. Since higher torque leads to faster motor temperature rise, this process is time-consuming. The time-consuming harmonic injection calibration in existing technologies, coupled with the varying harmonic control parameters required under different operating conditions, results in high test bench costs and high labor costs. Summary of the Invention
[0003] The purpose of this invention is to provide, on the one hand, a harmonic injection control system for an electric drive system based on progressive fitting prediction, and on the other hand, a harmonic injection control method for an electric drive system based on progressive fitting prediction. This system and method can predict the optimal harmonic current amplitude and phase across the entire torque range using a small amount of data, enabling more efficient and lower-cost harmonic injection calibration, thereby meeting the NVH performance requirements of the electric drive.
[0004] To achieve this objective, the present invention provides a harmonic injection control system for an electric drive system based on progressive fitting prediction, comprising: The progressive optimization module is used to progressively optimize the harmonic current at a selected reference torque point to obtain the optimal harmonic current amplitude and optimal harmonic current phase corresponding to the selected reference torque point. The parameter fitting module is used to perform parameter fitting on the optimal harmonic current amplitude and the optimal harmonic current phase, respectively, to obtain the corresponding current amplitude prediction function and current phase prediction function. The parameter prediction and correction module is used to predict the optimal harmonic current amplitude and optimal harmonic current phase corresponding to the real-time torque of the motor based on the real-time torque of the motor, the current amplitude prediction function and the current phase prediction function after the current amplitude prediction function and the current phase prediction function have passed the performance optimization verification, and output the corresponding harmonic current command to drive the motor.
[0005] Furthermore, the method for progressively optimizing at the selected reference torque point to obtain the optimal harmonic current amplitude and optimal harmonic current phase corresponding to the selected reference torque point includes: at a selected motor speed, selecting multiple discrete torque values as reference torque points; based on the time order corresponding to the radial electromagnetic force wave of the motor at each reference torque point, injecting harmonic currents of different amplitudes and phases of the corresponding order into each reference torque point; and selecting the amplitude and phase of the harmonic current that minimizes the motor operating noise as the optimal harmonic current amplitude and optimal harmonic current phase.
[0006] Furthermore, methods for obtaining corresponding current amplitude prediction functions and current phase prediction functions by performing parameter fitting on the optimal harmonic current amplitude and optimal harmonic current phase respectively include: Current amplitude prediction function: I = a1T + b1 or I = a2T 2 +b2T+c; Current phase prediction function: θ = d1T + e; Where I is the predicted optimal harmonic current amplitude, a1 is the amplitude prediction first-order coefficient determined by parameter fitting, T is the real-time torque of the motor, b1 is the first amplitude prediction constant determined by parameter fitting, a2 is the amplitude prediction second-order coefficient determined by parameter fitting, b2 is the amplitude prediction first-order coefficient determined by parameter fitting, c is the second amplitude prediction constant determined by parameter fitting, θ is the predicted optimal harmonic current phase, d1 is the phase prediction first-order coefficient determined by fitting, and e is the natural constant.
[0007] Furthermore, it also includes: selecting a torque drive motor outside the reference torque point, optimizing and verifying the current amplitude prediction function and the current phase prediction function, and determining whether the current amplitude prediction function and the current phase prediction function need to be corrected based on the performance verification results.
[0008] Furthermore, the method for optimizing and verifying the performance of the current amplitude prediction function and the current phase prediction function includes: selecting a torque-driven motor outside the reference torque point, recording the original noise of the motor when no harmonic current command corresponding to the current amplitude prediction function and the current phase prediction function is input, and the optimized noise of the motor when the harmonic current command corresponding to the current amplitude prediction function and the current phase prediction function is input. The difference between the original noise and the optimized noise is taken as the noise optimization value. When the noise optimization value generated after the motor is driven is less than the preset value, the optimization performance verification of the current amplitude prediction function and the current phase prediction function fails, and the current amplitude prediction function and the current phase prediction function need to be corrected.
[0009] Furthermore, the method for correcting the current amplitude prediction function and the current phase prediction function includes: when the noise optimization value generated after the motor is driven is less than the preset value, the predicted optimal harmonic current amplitude and optimal harmonic current phase are increased or decreased by a set step size under the current torque of the motor, and the harmonic current progressive optimization operation is re-executed to obtain the corrected optimal harmonic current amplitude and optimal harmonic current corresponding to the current torque of the motor; at the same time, a reference torque point is added on the basis of the original reference torque point and progressive optimization is performed, and the parameters of the corrected optimal harmonic current amplitude and optimal harmonic current phase under the current torque, the newly added reference torque point, and the optimal harmonic current amplitude and optimal harmonic current phase corresponding to the original reference torque point are respectively fitted to obtain the corresponding corrected current amplitude prediction function and current phase prediction function.
[0010] Furthermore, the method for predicting the optimal harmonic current amplitude and optimal harmonic current phase corresponding to the real-time torque of the motor based on the motor real-time torque, current amplitude prediction function and current phase prediction function includes: inputting the real-time torque of the motor into the current amplitude prediction function and the current phase prediction function to obtain the optimal harmonic current amplitude and optimal harmonic current phase corresponding to the real-time torque of the motor.
[0011] Furthermore, the method of driving the motor according to the optimal harmonic current amplitude and optimal harmonic current phase corresponding to the real-time torque of the motor includes: transforming the optimal harmonic current amplitude and optimal harmonic current phase corresponding to the real-time torque of the motor to the direct axis and quadrature axis coordinate system to obtain the harmonic current command, and superimposing the harmonic current command with the fundamental current command and inputting it to the current loop controller to drive the motor.
[0012] Furthermore, a harmonic injection control method for an electric drive system based on progressive fitting prediction, comprising: At the selected reference torque point, the harmonic current is progressively optimized to obtain the optimal harmonic current amplitude and optimal harmonic current phase corresponding to the selected reference torque point. By performing parameter fitting on the optimal harmonic current amplitude and the optimal harmonic current phase respectively, the corresponding current amplitude prediction function and current phase prediction function are obtained. The optimal harmonic current amplitude and optimal harmonic current phase corresponding to the real-time torque of the motor are predicted based on the real-time torque, current amplitude prediction function and current phase prediction function. The corresponding harmonic current command is output to drive the motor based on the optimal harmonic current amplitude and optimal harmonic current phase corresponding to the real-time torque of the motor. When the noise optimization value generated after the motor is driven is less than the preset value, the current amplitude prediction function and current phase prediction function are corrected.
[0013] The beneficial effects of this invention are as follows: Addressing the shortcomings of existing technologies that employ scanning methods, which require time-consuming calibration of numerous operating points within the target torque range and are exacerbated by rapid motor temperature rise under high torque, a more efficient calibration method is proposed. This method selects only a limited number of representative reference torque points at the target speed for harmonic parameter optimization, obtaining discrete optimal parameter samples. Then, using least squares fitting techniques, a predictive model describing the continuous functional relationship between the optimal amplitude, phase, and torque of the harmonic current is constructed, overcoming the enormous time and bench resource consumption caused by traditional methods that scan point-by-point across the entire torque range. Furthermore, the parameter continuity and smoothness achieved through mathematical modeling eliminate the risk of control command abrupt changes and secondary noise that may arise from interpolation in traditional table lookup methods. The closed-loop verification and correction mechanism of this scheme can increase the reference point and adjust the prediction function online by local optimization when the optimization effect of randomly selected torque points is not good. This ensures the generalization ability of the prediction model under all working conditions. With the lowest experimental cost, it achieves stable, smooth and superior active noise suppression of electric drive system in the full torque operating range, which significantly improves the efficiency and economy of harmonic injection calibration. Attached Figure Description
[0014] Figure 1 This is a schematic diagram of the process of the present invention; Figure 2 This is a schematic diagram of the structure of the present invention. Detailed Implementation
[0015] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, not all of them. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to represent selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.
[0016] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments: Example 1 like Figure 2 As shown, a harmonic injection control system for an electric drive system based on progressive fitting prediction includes: The progressive optimization module is used to progressively optimize the harmonic current at a selected reference torque point to obtain the optimal harmonic current amplitude and optimal harmonic current phase corresponding to the selected reference torque point. The parameter fitting module is used to perform parameter fitting on the optimal harmonic current amplitude and the optimal harmonic current phase, respectively, to obtain the corresponding current amplitude prediction function and current phase prediction function. The parameter prediction and correction module is used to predict the optimal harmonic current amplitude and optimal harmonic current phase corresponding to the real-time torque of the motor based on the real-time torque of the motor, the current amplitude prediction function and the current phase prediction function after the current amplitude prediction function and the current phase prediction function have passed the performance optimization verification, and output the corresponding harmonic current command to drive the motor.
[0017] In some technical solutions, the method of progressively optimizing at a selected reference torque point to obtain the optimal harmonic current amplitude and optimal harmonic current phase corresponding to the selected reference torque point includes: at a selected motor speed, selecting multiple discrete torque values as reference torque points; based on the time order corresponding to the radial electromagnetic force wave of the motor at each reference torque point, injecting harmonic currents of different amplitudes and phases of the corresponding order into each reference torque point; and selecting the amplitude and phase of the harmonic current that minimizes the motor operating noise as the optimal harmonic current amplitude and optimal harmonic current phase.
[0018] In some embodiments, the number of reference torque points is selected based on a non-uniform sampling strategy for characteristic operating conditions. For example, dense sampling is performed in regions with drastic torque changes and magnetic circuit saturation, while sparse sampling is performed in regions with smooth torque. Based on the time order corresponding to the radial electromagnetic force wave of the motor at each reference torque point, harmonic currents of different amplitudes and phases of the corresponding order are injected into each reference torque point. For example, reference torque point T... i The corresponding radial electromagnetic force wave time order is 6f (i.e., the vibration noise frequency is 6 times the fundamental electrical frequency f of the motor), requiring the injection of 5th harmonic current, meaning the frequency of the current harmonics is 5 times the fundamental frequency (the fundamental electrical frequency f of the motor); if the reference torque point T i The corresponding radial electromagnetic force wave has a time order of 12f, and 11th harmonic current needs to be injected. The amplitude and phase of the harmonic current that minimizes the motor's operating noise are selected as the optimal harmonic current amplitude and optimal harmonic current phase.
[0019] Radial electromagnetic force waves refer to the periodically changing electromagnetic force distribution generated in the radial direction of the stator and rotor air gap due to the interaction between the stator and rotor magnetic fields within the motor. They are the root cause of motor structural vibration and vibration noise. Harmonic current refers to the current component injected into the stator windings of the motor through the motor controller, with a frequency that is an integer multiple of the fundamental electrical frequency f (e.g., the 5th or 7th order). By adjusting the phase to be out of phase with the original harmonic noise, the energy of the sound waves can be mutually canceled.
[0020] It should be noted that the correspondence between the order of the radial electromagnetic force wave and the order of the current harmonics is determined by the fundamental physical laws governing the interaction between the electromagnetic characteristics and magnetic fields of the motor windings. A stator current time harmonic of a specific order generates a spatial harmonic magnetic field of a specific order. This magnetic field interacts with the rotor main magnetic field, exciting electromagnetic force waves with a specific frequency, thereby inducing vibration noise at the corresponding frequency. For a three-phase integer-slot winding, the time order corresponding to the radial electromagnetic force wave of the motor is... : , The k-th harmonic current generates the magnetic field harmonic order. When a k-th harmonic current is injected, the dominant magnetic field harmonic order is generated. =k, This represents the number of pole pairs of the motor rotor. This is the fundamental frequency of the motor. The number of pole pairs in a 6-pole motor. =3, when the time order of the radial electromagnetic force wave is 3. When it is 6f, Solving for =1 or =3; According to u=6n±k (n is an integer), the 5th harmonic current (k=5) can be obtained. When n=-1, the following occurs: =6×(-1)+5=-1st order magnetic field (i.e., 1st order reverse magnetic field), or 7th harmonic current (k=7) when n=1, generates =6×1-7=-1st order magnetic field, the -1st order magnetic field cancels out The 6f noise generated when =1.
[0021] In some technical solutions, methods for obtaining corresponding current amplitude prediction functions and current phase prediction functions by parameter fitting of the optimal harmonic current amplitude and optimal harmonic current phase include: Current amplitude prediction function: I = a1T + b1 or I = a2T 2 +b2T+c; Current phase prediction function: θ = d1T + e; Where I is the predicted optimal harmonic current amplitude, a1 is the amplitude prediction first-order coefficient determined by parameter fitting, T is the real-time torque of the motor, b1 is the first amplitude prediction constant determined by parameter fitting, a2 is the amplitude prediction second-order coefficient determined by parameter fitting, b2 is the amplitude prediction first-order coefficient determined by parameter fitting, c is the second amplitude prediction constant determined by parameter fitting, θ is the predicted optimal harmonic current phase, d1 is the phase prediction first-order coefficient determined by fitting, and e is the natural constant.
[0022] In some embodiments, the reference torque point and its corresponding optimal harmonic current amplitude or optimal harmonic current phase are input into spreadsheet software Excel (such as Microsoft Excel) to obtain the corresponding scatter plot. A linear trend line is added to the data in the scatter plot and the formula of the linear trend line is displayed as the corresponding current amplitude prediction function and current phase prediction function.
[0023] By transforming discrete and limited experimental calibration data into a general solution that covers continuous full torque conditions, this method requires only experiments at a few reference torque points. Through fitting a simple polynomial function (such as a linear or quadratic function), it can predict the optimal harmonic current amplitude and phase for any operating torque in real time, continuously, and smoothly. This completely avoids the extremely high time and cost associated with calibrating a massive number of torque points one by one using traditional scanning methods. Simultaneously, it ensures the continuity of control parameters as torque changes, eliminates the risk of parameter abrupt changes and noise mutations that may be caused by lookup table methods, and ultimately achieves stable and reliable noise optimization capabilities across all operating conditions with minimal calibration costs.
[0024] Some technical solutions also include: selecting a torque drive motor outside the reference torque point, optimizing and verifying the performance of the current amplitude prediction function and the current phase prediction function, and determining whether the current amplitude prediction function and the current phase prediction function need to be corrected based on the performance verification results.
[0025] In some technical solutions, the method for optimizing and verifying the performance of the current amplitude prediction function and the current phase prediction function includes: selecting a torque-driven motor outside the reference torque point, recording the original noise of the motor when no harmonic current command corresponding to the current amplitude prediction function and the current phase prediction function is input, and the optimized noise of the motor when the harmonic current command corresponding to the current amplitude prediction function and the current phase prediction function is input, taking the difference between the original noise and the optimized noise as the noise optimization value, and when the noise optimization value generated after the motor is driven is less than the preset value, the optimization performance verification of the current amplitude prediction function and the current phase prediction function fails, and the current amplitude prediction function and the current phase prediction function need to be corrected.
[0026] In some embodiments, the preset value for the noise optimization value is A-weighted 20 dB(A).
[0027] By randomly selecting an operating point not involved in the fitting process, the actual noise attenuation before and after applying the prediction function is directly compared. This transforms the abstract goodness of fit into an intuitive noise reduction decibel value. When the optimization effect fails to meet the target, iterative correction of the current amplitude prediction function and the current phase prediction function is immediately triggered, driving the system to upgrade from a one-time offline fitting to a continuously self-optimizing learning system. This ensures the reliability of optimization under all operating conditions with minimal verification cost.
[0028] In some technical solutions, when the noise optimization value generated after the motor is driven is greater than or equal to the preset value, the current amplitude prediction function and the current phase prediction function are effective and do not need to be corrected.
[0029] When the noise optimization value at the verification point meets the standard, the prediction function is deemed valid and requires no correction. This proves that the current amplitude prediction function and current phase prediction function constructed from the initial limited reference point data have high accuracy and generalization ability, and can reliably predict the optimal parameters for untested operating conditions.
[0030] In some technical solutions, the methods for correcting the current amplitude prediction function and the current phase prediction function include: when the noise optimization value generated after the motor is driven is less than the preset value, the predicted optimal harmonic current amplitude and optimal harmonic current phase are increased or decreased by a set step size under the current torque of the motor, and the harmonic current progressive optimization operation is re-executed to obtain the corrected optimal harmonic current amplitude and optimal harmonic current corresponding to the current torque of the motor; at the same time, a reference torque point is added based on the original reference torque point and progressive optimization is performed, and the parameters of the corrected optimal harmonic current amplitude and optimal harmonic current phase under the current torque, the newly added reference torque point, and the optimal harmonic current amplitude and optimal harmonic current phase corresponding to the original reference torque point are respectively fitted to obtain the corresponding corrected current amplitude prediction function and current phase prediction function.
[0031] When the performance at randomly verified torque points fails to meet the requirements, a fine-tuning process is performed near these points to quickly locate a better parameter combination and incorporate it as a new benchmark. Subsequently, the coefficients of the predicted function are fine-tuned online using these new benchmark torque points. This approach can identify and correct complex nonlinear regions that the initial fit could not accurately describe, thereby continuously approximating the true optimal parameter surface. Ultimately, through a dynamic and progressive method, it ensures that the harmonic injection control strategy achieves stable and superior noise suppression across the entire operating range.
[0032] In some technical solutions, the method for predicting the optimal harmonic current amplitude and optimal harmonic current phase corresponding to the real-time torque of the motor based on the real-time torque, current amplitude prediction function and current phase prediction function includes: inputting the real-time torque of the motor into the current amplitude prediction function and the current phase prediction function to obtain the optimal harmonic current amplitude and optimal harmonic current phase corresponding to the real-time torque of the motor.
[0033] By directly substituting the real-time torque into the pre-fitted mathematical function for calculation, the absolute smoothness of control commands as torque changes is ensured, fundamentally eliminating the risks of command abrupt changes and noise surges that may arise from looking up discrete data. Simultaneously, this calculation method has extremely low computational load, meeting the high-speed real-time control requirements of electric drive systems, enabling efficient and stable execution in actual operation, ultimately achieving NVH performance optimization.
[0034] In some technical solutions, the method of driving the motor by outputting the corresponding harmonic current command based on the optimal harmonic current amplitude and optimal harmonic current phase corresponding to the real-time torque of the motor includes: transforming the optimal harmonic current amplitude and optimal harmonic current phase corresponding to the real-time torque of the motor to the direct axis and quadrature axis coordinate system to obtain the harmonic current command, and superimposing the harmonic current command with the fundamental current command and inputting it to the current loop controller to drive the motor.
[0035] In some embodiments, the conversion formulas for transforming the amplitude and phase of the optimal harmonic current corresponding to the real-time torque of the motor to the direct-axis and quadrature-axis coordinate systems are Id = I•cosθ; Iq = I•sinθ, where Id is the direct-axis current in the direct-axis and quadrature-axis coordinate systems, Iq is the quadrature-axis current in the direct-axis and quadrature-axis coordinate systems, I is the amplitude of the optimal harmonic current, and θ is the phase of the harmonic current. The fundamental current command is a sinusoidal current command with the same frequency as the motor's electrical frequency, output by the motor controller according to the target torque requirement.
[0036] By converting the harmonic current command to the orthogonal axis (dq) coordinate system, it can be synthesized with the fundamental current command of the motor vector control within a unified control framework. This not only ensures that the injected harmonic components can be quickly and accurately tracked and output by the current loop controller, but also avoids interference from harmonic injection on the motor's main torque and field control, thus guaranteeing the core drive performance of the system. Ultimately, this method enables the optimized parameters based on the predictive model to be directly converted into control signals that actually act on the motor, completing a full closed loop from parameter optimization to real-time active noise reduction, significantly improving the NVH performance of the electric drive system under complex operating conditions.
[0037] like Figure 1As shown, the implementation process of this invention is as follows: First, the target speed point and the torque operating range to be optimized are identified. Within this range, a finite number of discrete reference torque points are selected. Simultaneously, the specific order of the current harmonics to be injected is determined based on the electromagnetic characteristics of the motor. Then, progressive optimization is performed at the selected reference torque points to obtain the optimal harmonic current amplitude and phase parameters corresponding to each reference torque point. Using the optimal harmonic current amplitude and phase parameters, a current amplitude prediction function and a phase prediction function covering the entire torque range are constructed through mathematical fitting. After importing this prediction function model into the current controller, the effect of harmonic injection is verified under real operating conditions. If the verification effect is good, the process ends; if the effect is poor, the process returns to the optimization step, and the prediction function is closed-loop corrected and iteratively optimized by adding reference points, etc., until a satisfactory noise suppression effect is obtained.
[0038] A preferred embodiment of the present invention is as follows: using a 6-pole, 54-slot motor (pole pairs) Taking (e.g., 3) as an example, its 18th-order time-order noise exceeds the target threshold (e.g., 40dB(A)) at a peak of 660rpm, and needs to be optimized through harmonic injection.
[0039] At a motor speed of 660 rpm, the optimal torque range of the motor (i.e., the motor operating torque range corresponding to a speed of 660 rpm) is 140~360 Nm. Based on the optimal torque range of the motor, three torque points of 140 Nm, 240 Nm and 360 Nm are selected as the reference torque points.
[0040] The 18th-order time harmonic is a high-order spatial force wave, typically generated by the interaction of two stator magnetic field harmonics of different orders. Its spatial order... Calculation formula: , The time order of the spatial harmonics of the first stator magnetic field. The time order of the spatial harmonics of the second stator magnetic field can be obtained. , To generate an 18th-order spatial force wave, it is necessary to find two stator spatial harmonic magnetic fields with different time orders. The stator space harmonic magnetic field is generated by the time harmonics of the current injected into the stator winding. For a three-phase integer slot winding, the harmonic order μ of the space magnetic field generated by the k-th current harmonic is determined by the winding theory formula: μ = 6n ± k (n = 0, ±1, ±2, ...). The main harmonic orders μ of the space magnetic field generated by the 5th harmonic current (k=5) are as follows: When n=0, μ=5; when n=1, μ=6+5=11; when n=-1, μ=-6+5=-1 (i.e., the first-order reversed magnetic field), and μ=7, 13, 17, 19... etc. The main harmonic orders μ of the space magnetic field generated by the 7th harmonic current (k=7) are as follows: When n=0, μ=7; when n=1, μ=6+7=13; when n=-1, μ=-6+7=1, and μ=5, 11, 19, etc. By iterating through the 5th harmonic current (k=5) and the 7th harmonic current (k=7), we can obtain... ,like (Can be generated by the 5th harmonic, n=1), (Can be generated by the 5th or 7th harmonic, n=0). Both the 5th and 7th current harmonics affect the 18th time-order noise performance. In this embodiment, the 5th current harmonic is selected for harmonic injection. The noise within 0.1m around the motor controller is used as the noise evaluation point. The current harmonic amplitude and phase with the minimum 18th time-order noise are selected during the calibration process. The optimal 5th harmonic current amplitude and phase for 140Nm, 240Nm, and 360Nm are shown in Table 1.
[0041] Table 1 Amplitude and Phase of 5th Harmonic Current Based on three sets of current harmonic amplitude and phase pairs (140, 240, and 360 Nm), the prediction functions for the fifth harmonic current amplitude and phase are constructed using the least squares method (i.e., formulas built using Excel): I = 0.06854 * T - 8.5742 θ = -0.25 * T + 135 Based on the amplitude and phase prediction function of the fifth harmonic current, the amplitude and phase of the fifth harmonic current at each remaining torque point are predicted with a step size of 20 Nm as an example, as shown in Table 2.
[0042] Table 2 Predicted current amplitude and phase for 5 cycles The optimal harmonic current amplitude and harmonic current phase corresponding to the real-time torque of the motor are transformed into the dq axis coordinate system, superimposed on the fundamental current command, and the motor is driven by the current loop controller.
[0043] 300 Nm was selected as the torque point for verifying the harmonic effect. Under the predicted current amplitude and phase, the optimized value of the 18th time order noise was 37 dB(A), indicating that the parameter fitting model has a good optimization effect.
[0044] Example 2 A harmonic injection control method for an electric drive system based on progressive fitting prediction, comprising: At the selected reference torque point, the harmonic current is progressively optimized to obtain the optimal harmonic current amplitude and optimal harmonic current phase corresponding to the selected reference torque point. By performing parameter fitting on the optimal harmonic current amplitude and the optimal harmonic current phase respectively, the corresponding current amplitude prediction function and current phase prediction function are obtained. The optimal harmonic current amplitude and optimal harmonic current phase corresponding to the real-time torque of the motor are predicted based on the real-time torque, current amplitude prediction function and current phase prediction function. The corresponding harmonic current command is output to drive the motor based on the optimal harmonic current amplitude and optimal harmonic current phase corresponding to the real-time torque of the motor. When the noise optimization value generated after the motor is driven is less than the preset value, the current amplitude prediction function and current phase prediction function are corrected.
[0045] Example 3 The present invention provides a computer-readable storage medium storing a computer program, which, when executed by a processor, performs the steps of the method described in Embodiment 2.
[0046] This invention can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented in whole or in part as a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., a solid-state drive (SSD)).
[0047] Those skilled in the art will readily understand that the above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, combinations, substitutions, improvements, etc., made under the spirit and principles of the present invention are included within the protection scope of the present invention.
[0048] The contents not described in detail in this specification are existing technologies known to those skilled in the art.
Claims
1. A harmonic injection control system for an electric drive system based on progressive fitting prediction, characterized in that, It includes: The progressive optimization module is used to progressively optimize the harmonic current at a selected reference torque point to obtain the optimal harmonic current amplitude and optimal harmonic current phase corresponding to the selected reference torque point. The parameter fitting module is used to perform parameter fitting on the optimal harmonic current amplitude and the optimal harmonic current phase, respectively, to obtain the corresponding current amplitude prediction function and current phase prediction function. The parameter prediction and correction module is used to predict the optimal harmonic current amplitude and optimal harmonic current phase corresponding to the real-time torque of the motor based on the motor real-time torque, current amplitude prediction function and current phase prediction function, and output the corresponding harmonic current command to drive the motor.
2. The harmonic injection control system for an electric drive system based on progressive fitting prediction according to claim 1, characterized in that: The method for progressively optimizing the selected reference torque point to obtain the optimal harmonic current amplitude and optimal harmonic current phase corresponding to the selected reference torque point includes: at a selected motor speed, selecting multiple discrete torque values as reference torque points; based on the time order corresponding to the radial electromagnetic force wave of the motor at each reference torque point, injecting harmonic currents of different amplitudes and phases of the corresponding order into each reference torque point; and selecting the amplitude and phase of the harmonic current that minimizes the motor operating noise as the optimal harmonic current amplitude and optimal harmonic current phase.
3. A harmonic injection control system for an electric drive system based on progressive fitting prediction according to claim 2, characterized in that: Methods for obtaining corresponding current amplitude prediction functions and current phase prediction functions by parameter fitting of the optimal harmonic current amplitude and optimal harmonic current phase include: Current amplitude prediction function: I = a1T + b1 or I = a2T 2 +b2T+c; Current phase prediction function: θ = d1T + e; Where I is the predicted optimal harmonic current amplitude, a1 is the amplitude prediction first-order coefficient determined by parameter fitting, T is the real-time torque of the motor, b1 is the first amplitude prediction constant determined by parameter fitting, a2 is the amplitude prediction second-order coefficient determined by parameter fitting, b2 is the amplitude prediction first-order coefficient determined by parameter fitting, c is the second amplitude prediction constant determined by parameter fitting, θ is the predicted optimal harmonic current phase, d1 is the phase prediction first-order coefficient determined by fitting, and e is the natural constant.
4. A harmonic injection control system for an electric drive system based on progressive fitting prediction according to claim 3, characterized in that: Also includes: A torque-driven motor outside the reference torque point is selected, and the current amplitude prediction function and current phase prediction function are optimized and verified. Based on the performance verification results, it is determined whether the current amplitude prediction function and current phase prediction function need to be corrected.
5. A harmonic injection control system for an electric drive system based on progressive fitting prediction according to claim 4, characterized in that: The method for optimizing and verifying the performance of the current amplitude prediction function and the current phase prediction function includes: selecting a torque-driven motor outside the reference torque point, recording the original noise of the motor when no harmonic current command corresponding to the current amplitude prediction function and the current phase prediction function is input, and the optimized noise of the motor when the harmonic current command corresponding to the current amplitude prediction function and the current phase prediction function is input. The difference between the original noise and the optimized noise is taken as the noise optimization value. When the noise optimization value generated after the motor is driven is less than the preset value, the optimization performance verification of the current amplitude prediction function and the current phase prediction function fails, and the current amplitude prediction function and the current phase prediction function need to be corrected.
6. A harmonic injection control system for an electric drive system based on progressive fitting prediction according to claim 4, characterized in that: The method for correcting the current amplitude prediction function and the current phase prediction function includes: when the noise optimization value generated after the motor is driven is less than the preset value, the predicted optimal harmonic current amplitude and optimal harmonic current phase are increased or decreased by a set step size under the current torque of the motor, and the harmonic current progressive optimization operation is re-executed to obtain the corrected optimal harmonic current amplitude and optimal harmonic current corresponding to the current torque of the motor; at the same time, a reference torque point is added on the basis of the original reference torque point and progressive optimization is performed. The corrected optimal harmonic current amplitude and optimal harmonic current phase under the current torque, the newly added reference torque point, and the optimal harmonic current amplitude and optimal harmonic current phase corresponding to the original reference torque point are respectively fitted with parameters to obtain the corresponding corrected current amplitude prediction function and current phase prediction function.
7. A harmonic injection control system for an electric drive system based on progressive fitting prediction according to claim 1 or 3, characterized in that: The method for predicting the optimal harmonic current amplitude and optimal harmonic current phase corresponding to the real-time torque of a motor based on the motor real-time torque, current amplitude prediction function and current phase prediction function includes: inputting the real-time torque of the motor into the current amplitude prediction function and the current phase prediction function to obtain the optimal harmonic current amplitude and optimal harmonic current phase corresponding to the real-time torque of the motor.
8. A harmonic injection control system for an electric drive system based on progressive fitting prediction according to claim 7, characterized in that: The method for driving a motor based on the optimal harmonic current amplitude and optimal harmonic current phase corresponding to the real-time torque of the motor includes: transforming the optimal harmonic current amplitude and optimal harmonic current phase corresponding to the real-time torque of the motor to the direct axis and quadrature axis coordinate system to obtain the harmonic current command; and superimposing the harmonic current command with the fundamental current command and inputting it to the current loop controller to drive the motor.
9. A harmonic injection control method for an electric drive system based on progressive fitting prediction according to claim 1, characterized in that, include: At the selected reference torque point, the harmonic current is progressively optimized to obtain the optimal harmonic current amplitude and optimal harmonic current phase corresponding to the selected reference torque point. By performing parameter fitting on the optimal harmonic current amplitude and the optimal harmonic current phase respectively, the corresponding current amplitude prediction function and current phase prediction function are obtained. The optimal harmonic current amplitude and optimal harmonic current phase corresponding to the real-time torque of the motor are predicted based on the real-time torque, current amplitude prediction function and current phase prediction function. The corresponding harmonic current command is output to drive the motor based on the optimal harmonic current amplitude and optimal harmonic current phase corresponding to the real-time torque of the motor. When the noise optimization value generated after the motor is driven is less than the preset value, the current amplitude prediction function and current phase prediction function are corrected.
10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method of claim 9.