A harmonic current decoupling control method based on electromagnetic vibration noise suppression

Abstract

This invention discloses a harmonic current decoupling control method based on electromagnetic vibration and noise suppression, belonging to the field of motor control technology. It aims to address the shortcomings of existing harmonic current injection control strategies in terms of multi-harmonic coupling compensation, parameter uncertainty suppression, and dynamic condition adaptability. The method includes: constructing a harmonic voltage-current model in a multi-synchronous rotating coordinate system based on stator current and stator voltage equations; constructing state equations based on the harmonic voltage-current model, setting an uncertainty disturbance estimator, estimating the uncertainty disturbance, and obtaining the uncertainty disturbance matrix; and using an internal model control structure to perform closed-loop tracking control of the target harmonic current command, using the estimated value of the uncertainty disturbance matrix as feedforward compensation to offset system uncertainties. This invention can achieve accurate tracking and injection of the target harmonic current under dynamic conditions, thereby effectively suppressing electromagnetic vibration and noise caused by electromagnetic force harmonics.

Description

Technical Field

[0001] This invention relates to the field of motor control technology, and specifically to a harmonic current decoupling control method based on electromagnetic vibration noise suppression. Background Technology

[0002] Permanent magnet synchronous motors (PMSMs) are widely used in electric vehicles due to their high power density, wide speed range, and high efficiency. However, compared to the low-frequency ignition, mechanical, and combustion noise generated in traditional internal combustion engines, the main noise of electric drive systems becomes a sharp whistling sound caused by electromagnetic harmonics. This high-frequency monotonic noise is significant under many operating conditions, severely affecting the passenger's driving experience. For automotive PMSMs, the main component of sharp electromagnetic noise is order noise with a frequency proportional to the speed, caused by slotting, back EMF harmonics, and harmonic currents. This includes torsional vibration noise generated by tangential electromagnetic harmonics exciting the rotor shaft system and breathing mode noise generated by radial electromagnetic harmonics. This noise can be suppressed by actively injecting current harmonics. The core principle is to use the interaction between the armature magnetic field generated by the injected compensation current and the permanent magnet magnetic field to produce an electromagnetic force with the same spatial order and frequency as the original electromagnetic harmonics but with an opposite phase angle, thereby canceling out the main excitation electromagnetic harmonics that cause the whistling sound.

[0003] Key challenges in noise reduction based on active harmonic current injection (HCI) include determining the optimal reference value of harmonic currents and designing controllers for fast and accurate harmonic current tracking, such as CN102201770A. Determining the optimal reference value of harmonic currents primarily involves using finite element simulation, analytical modeling calculations, and bench experiments to obtain the harmonic current combination that minimizes the amplitude of the target harmonic electromagnetic force or the decibel value of the order noise. For controller design to achieve fast and accurate harmonic current tracking, existing harmonic control methods mainly fall into three categories: harmonic control based on multi-synchronous rotating coordinate transformation; harmonic control based on periodic controllers; and harmonic control based on self-learning algorithms. The shortcomings of existing methods include: 1. They neglect the coupling relationship between the 5th and 7th harmonic current variables, and the controller bandwidth is limited, so the harmonic current command tracking performance drops sharply as the speed increases; 2. Factors such as magnetic saturation, strong coupling effects between harmonic variables, and strong time-varying nonlinearity of the transfer model parameters make it difficult to achieve fast and accurate response to harmonic commands, especially when facing parameter uncertainties and external interference, the system's response accuracy further decreases; 3. In automotive scenarios, the motor speed changes are random, and the parameters of the harmonic detector and periodic controller need to adapt quickly to the vehicle speed changes, which greatly increases the control difficulty.

[0004] It is evident that existing harmonic current injection control strategies still have shortcomings in terms of multi-harmonic coupling compensation, parameter uncertainty suppression, and dynamic operating condition adaptability, making it difficult to simultaneously guarantee the tracking performance of harmonic current and the system's anti-disturbance performance. Summary of the Invention

[0005] To address the shortcomings of existing harmonic current injection control strategies in terms of multi-harmonic coupling compensation, parameter uncertainty suppression, and dynamic operating condition adaptability, a harmonic current decoupling control method based on electromagnetic vibration noise suppression is proposed. This method enables precise tracking and injection of the target harmonic current under dynamic operating conditions, thereby effectively suppressing electromagnetic vibration noise caused by electromagnetic force harmonics.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a harmonic current decoupling control method based on electromagnetic vibration noise suppression, comprising the following steps: S1. Construct a harmonic voltage-current model in a multi-synchronous rotating coordinate system based on the stator current and stator voltage equations; S2, construct the state equation based on the harmonic voltage-current model, set up the uncertain disturbance estimator, estimate the uncertain disturbance, and obtain the uncertain disturbance matrix; S3 employs an internal model control structure to perform closed-loop tracking control of the target harmonic current command, using the estimated value of the uncertainty disturbance matrix as feedforward compensation to offset system uncertainties.

[0007] In the technical solution of this invention, a harmonic voltage-current model is first established, which can describe the coupling relationship between harmonic voltages and currents of each order. Then, based on the above model, a state equation is established, and an observational model of the uncertainty disturbance matrix is ​​performed. Finally, a closed-loop tracking control of the target harmonic current command is realized according to the internal model control structure, and the estimated value of the uncertainty disturbance matrix in step S2 is used as feedforward compensation to offset the system uncertainty. The method of this invention can improve the robustness and response capability of the system.

[0008] The present invention is further configured such that step S1 includes: S11, performs multi-synchronous rotating coordinate transformation on the stator current of the motor, and converts the harmonic current into DC form in its corresponding rotating coordinate system; S12, Substitute the transformed harmonic current into the stator voltage equation of the permanent magnet synchronous motor to obtain the harmonic voltage equation; S13, apply a multi-synchronous rotating coordinate transformation to the harmonic voltage equation to obtain two synchronous rotating voltage vectors expressed in DC form.

[0009] In this technical solution, the construction of the harmonic voltage-current model can be achieved through the above three sub-steps.

[0010] The present invention is further configured such that: the left side of the state equation is the first derivative of the harmonic current vector, and the right side of the state equation is the nominal system dynamic A0x, the control input Bu, and the uncertainty disturbance matrix δ. UDE The sum; x is the harmonic current vector; u is the harmonic voltage vector; A0 is the simplified dynamic matrix after neglecting the angular velocity coupling term; B is the input matrix; δ UDE This is a composite uncertainty interference matrix.

[0011] In this technical solution, the above-mentioned state equations are established based on the harmonic voltage-current model constructed in step S1.

[0012] The present invention is further configured such that step S2 includes: S21, the harmonic voltage-current model is rewritten in the form of state equations; S22, Select a filter to obtain an approximate disturbance estimate and estimate the disturbance in the system; S23, after estimating the uncertainty disturbance, the uncertainty is offset by designing a feedforward path; S24, integrate steps S21 to S23 to obtain the feedforward compensation voltage.

[0013] The present invention is further configured such that the uncertainty disturbance includes cross-coupling terms containing rotational speed information between harmonic variables, uncertainty of system parameters, and external disturbances.

[0014] By designing an uncertain disturbance estimator, an observational model of the uncertain disturbance matrix is ​​performed to estimate online the cross-coupling terms between harmonic variables containing rotational speed information, the uncertainty of system parameters, and external disturbances.

[0015] The present invention is further configured such that: the filter is specifically selected as a first-order low-pass filter to estimate the disturbance in the system.

[0016] Using a first-order low-pass filter to estimate disturbances in the system can simplify the design process of an uncertain disturbance estimator.

[0017] The present invention is further configured such that: in step S2, the feedforward compensation voltage V generated by the control law of the uncertain disturbance estimator is finally obtained. UDE (s), the feedforward compensation voltage V UDE (s) is the inverse of the input matrix B multiplied by the first intermediate value and then multiplied by the first-order low-pass filter L. c (s), where the first intermediate quantity is the s-fold state vector X(s) in the s domain minus the product of matrix A0 and state vector X(s) minus the product of input matrix B and control input U(s).

[0018] The present invention is further configured such that: in step S3, the output of the internal model control structure is a harmonic voltage vector V. IMC (s), the harmonic voltage vector V IMC (s) represents the harmonic current reference command value X. * The difference between state vector X(s) and state vector F is multiplied by the internal model controller F. IMC (s) Equivalent PI control rate F of the main diagonal elements IMC-PI (s) and F IMC (s) is the sum of the residual coupling compensation terms D(s) constructed from off-diagonal elements.

[0019] In this technical solution, the harmonic voltage vector V is obtained through calculation. IMC (s) is used for subsequent adjustments.

[0020] The present invention is further configured such that: the F IMC The residual coupling compensation term D(s) constructed from off-diagonal elements does not contain angular velocity coupling terms that have been compensated by the uncertain disturbance estimator.

[0021] The present invention is further configured to include step S4: optimizing harmonic current decoupling control by adjusting the closed-loop bandwidth and the estimator bandwidth respectively.

[0022] By adjusting the closed-loop bandwidth and estimator bandwidth as described above, the harmonic current decoupling control achieves an optimal balance between robustness and dynamic performance, thereby improving the system's robustness and response capability.

[0023] The harmonic current decoupling control method based on electromagnetic vibration noise suppression of the present invention can bring the following beneficial effects: 1. Existing harmonic voltage-current controlled models neglect the strong coupling relationship between harmonic variables, which causes the harmonic current command tracking performance to drop sharply as the rotational speed increases. In the modeling process of this invention, the coupling effect between harmonic currents of the same and different orders in the controlled model as well as the instantaneous value change of the harmonic current variables themselves are considered simultaneously, so as to achieve accurate decoupling compensation for the harmonic coupling effect. 2. Existing harmonic current injection control methods do not consider magnetic saturation and the strong time-varying nonlinearity of the transferred model parameters, making it difficult to achieve fast and accurate response to harmonic commands when faced with parameter uncertainties and external disturbances. This invention proposes a harmonic current decoupling method based on the combination of Uncertainty Disturbance Estimator (UDE) and Internal Model Control (IMC). The UDE calculates and compensates for harmonic variable cross-coupling, parameter perturbation, and load disturbances in real time, and the IMC is combined to construct a robust tracking closed loop for the target harmonic current. This effectively suppresses the impact of parameter time-varying and harmonic coupling effects on control accuracy, significantly improves the dynamic response accuracy and anti-interference capability of the system, and achieves accurate harmonic injection under dynamic operating conditions. Attached Figure Description

[0024] Figure 1 This is a block diagram of the internal mold control structure.

[0025] Figure 2 This is a block diagram of the harmonic current decoupling control system based on UDE combined with IMC of the present invention.

[0026] Figure 3 This is a block diagram of the parallel connection of the fundamental current controller and the harmonic current controller of the present invention.

[0027] Figure 4 This is a waveform diagram of the harmonic current response of the present invention.

[0028] Figure 5 The spectrum diagram shows the results of the prototype vibration steady-state optimization experiment.

[0029] Figure 6 The spectrum diagram shows the results of the prototype noise steady-state optimization experiment.

[0030] Figure 7 The spectrum of the experimental results for dynamic run-up optimization of prototype noise. Detailed Implementation

[0031] Example 1 To address the shortcomings of existing harmonic current injection control strategies in terms of multi-harmonic coupling compensation, parameter uncertainty suppression, and dynamic operating condition adaptability, this embodiment proposes a harmonic current decoupling control method based on electromagnetic vibration and noise suppression. (Refer to...) Figure 1 , Figure 2 and Figure 3 It includes the following steps.

[0032] Step S1: First, construct a harmonic voltage-current model in a multi-synchronous rotating coordinate system, specifically based on the stator current and stator voltage equations.

[0033] The specific construction process of the above harmonic voltage-current model mainly includes the following sub-steps.

[0034] Step S11: Perform multi-synchronous rotating coordinate transformation on the stator current of the permanent magnet synchronous motor, and convert the harmonic current into DC form in its corresponding rotating coordinate system.

[0035] Step S12: Substitute the harmonic current obtained after the above transformation into the stator voltage equation of the permanent magnet synchronous motor to obtain the harmonic voltage equation.

[0036] Step S13: Apply a multi-synchronous rotating coordinate transformation to the harmonic voltage to obtain two synchronous rotating voltage vectors represented by direct current.

[0037] Furthermore, the stator current is first subjected to multi-synchronous rotating coordinate transformation (MSRFT) to convert the (6k±1)th harmonic current into a DC form in its corresponding rotating coordinate system, so as to realize the extraction and independent control of the harmonic current components.

[0038] Subsequently, the harmonic current transformed by MSRFT is substituted into the stator voltage equation of the permanent magnet synchronous motor (PMSM), and the instantaneous variation characteristics of the harmonic current are considered in the modeling process to obtain the harmonic voltage equation in the synchronous rotating coordinate system (SRF).

[0039] Finally, the harmonic voltage equation under the obtained SRF is subjected to a multi-synchronous rotating coordinate transformation again to eliminate the AC coupling terms and transform it into a DC expression in the corresponding (6k±1)th order synchronous rotating coordinate system.

[0040] Through the above transformation process, the harmonic voltage-current relationship that originally contained complex AC quantities in the SRF can be simplified to a synchronous rotating voltage vector represented by two DC quantities in the (6k±1)th order synchronous rotating coordinate system.

[0041] Step S2: Based on the harmonic voltage-current model obtained in Step S1, the state equation is transformed. An uncertain disturbance estimator is designed to estimate the uncertain disturbance, resulting in the uncertain disturbance matrix. Uncertain disturbances include cross-coupling terms containing speed information among harmonic variables, uncertainties in system parameters, and external disturbances. The uncertain disturbance estimator is designed to perform observational modeling of the uncertain disturbance matrix, online estimation of cross-coupling terms containing speed information among harmonic variables, uncertainties in system parameters, and external disturbances, real-time compensation for the negative impacts of disturbances, and improvement of the robustness of the control system to parameter changes and external disturbances.

[0042] For the above expression of the state equation, the left side of the equation is the first derivative of the harmonic current vector, and the right side of the equation is the nominal system dynamics A0x, the control input Bu, and the uncertainty disturbance matrix δ. UDE The sum of three terms, where x represents the harmonic current vector; u represents the harmonic voltage vector; A0 represents the simplified dynamic matrix after neglecting the angular velocity coupling term; B represents the input matrix; δ UDE This represents the composite uncertainty disturbance matrix; the nominal system dynamic A0x is the product of the simplified dynamic matrix (ignoring angular velocity coupling terms) and the harmonic current vector x. The control input Bu is the product of the input matrix B and the harmonic voltage vector u.

[0043] In this technical solution, the above-mentioned state equations are established based on the harmonic voltage-current model constructed in step S1.

[0044] The above step S2 mainly includes the following sub-steps.

[0045] Step S21: Rewrite the above harmonic voltage-current model into state equation form. For a more detailed explanation of the state equation expression, please refer to the above content.

[0046] Step S22: Select an appropriate filter to obtain an approximate disturbance estimate and estimate the system disturbance. Considering that the above model is a four-input and four-output system, in order to simplify the UDE control law design process, a first-order low-pass filter is selected to estimate the disturbance in the system; more specifically, it mainly includes the following process: By selecting a suitable filter, the disturbance estimate can approximately represent δ. UDE With filter l c The convolution of . Where l c s-domain expression L c To estimate the bandwidth ω c For the numerator, the Laplace variable s and the estimator bandwidth ω c The sum of these is the scalar in the denominator multiplied by the identity matrix I.

[0047] Step S23: After estimating the uncertainty disturbance, the uncertainty is counteracted by cleverly designing a feedforward path. More specifically, the harmonic voltage vector u is set equal to the harmonic voltage vector v output by the internal model controller. IMC The feedforward compensation voltage v generated by the control law of the uncertain disturbance estimator UDE The difference. The feedforward compensation voltage v generated by the control law of the uncertain disturbance estimator. UDE Equal to the inverse of the input matrix B and the perturbation estimation matrix The product of and . Taking the Laplace transform of the above expression for the harmonic voltage vector u, we obtain the s-domain voltage equation, where the s-domain voltage U(s) equals V. IMC (s) minus V UDE (s), where U(s) and V IMC (s), V UDE (s) correspond to the matrix vectors u and v respectively. IMC v UDE The s-domain expression. And the harmonic voltage vector V IMC (s) represents the harmonic current reference command value X. * The difference between state vector X(s) and state vector F is multiplied by the internal model controller F. IMC (s) Equivalent PI control rate F of the main diagonal elements IMC-PI (s) and F IMC (s) is the sum of the residual coupling compensation terms D(s) constructed from off-diagonal elements. Where F IMC The residual coupling compensation term D(s) constructed from off-diagonal elements does not contain angular velocity coupling terms that have been compensated by the uncertain disturbance estimator.

[0048] Step S24: Obtain the feedforward compensation voltage based on the results of steps S21 to S23. Further, by integrating the above multiple expressions, the online compensation amount of UDE, i.e., the feedforward compensation voltage V, can be obtained. UDE (s), feedforward compensation voltage V UDE (s) is equal to the inverse of the input matrix B multiplied by the first intermediate quantity and then multiplied by the first-order low-pass filter L. c (s), the first intermediate quantity is the s times state vector X(s) in the s domain minus the product of matrix A0 and state vector X(s) minus the product of input matrix B and control input U(s).

[0049] In this technical solution, the feedforward compensation voltage generated by the control law of the uncertain disturbance estimator can be calculated.

[0050] Step S3: The target harmonic current command is tracked in a closed loop using an internal model control structure, and the estimated value of the uncertainty disturbance matrix obtained in step S2 is used as feedforward compensation to offset the system uncertainty.

[0051] This embodiment also includes step S4: optimizing harmonic current decoupling control by adjusting the closed-loop bandwidth and estimator bandwidth respectively.

[0052] Through the above step S4, the harmonic current decoupling control can achieve an optimal balance between robustness and dynamic performance, thereby improving the robustness and response capability of the system.

[0053] refer to Figure 2 This is a block diagram of the harmonic current decoupling control system combining UDE and IMC. The system mainly consists of three parts: a PI controller based on internal model decoupling design, an IMC feedforward decoupling compensation part, and an online compensation part for uncertainty interference error by UDE.

[0054] The method in this embodiment uses UDE to estimate online the cross-coupling terms between harmonic variables containing speed information, the uncertainty of system parameters, and external disturbances, thereby compensating for the negative impact of disturbances in real time and improving the robustness of the control system to parameter changes and external disturbances. The IMC structure is used to achieve closed-loop tracking control of the target harmonic current, ensuring improved system performance.

[0055] Example 2 This embodiment proposes a harmonic current decoupling control method based on electromagnetic vibration noise suppression, which includes the following steps.

[0056] Step S1: First, construct a harmonic voltage-current model in a multi-synchronous rotating coordinate system, specifically based on the stator current and stator voltage equations.

[0057] The specific construction process of the above harmonic voltage-current model mainly includes the following sub-steps.

[0058] Step S11: Perform multi-synchronous rotating coordinate transformation on the stator current of the permanent magnet synchronous motor, and convert the harmonic current into DC form in its corresponding rotating coordinate system.

[0059] Step S12: Substitute the harmonic current obtained after the above transformation into the stator voltage equation of the permanent magnet synchronous motor to obtain the harmonic voltage equation.

[0060] Step S13: Apply a multi-synchronous rotating coordinate transformation to the harmonic voltage to obtain two synchronous rotating voltage vectors represented by direct current.

[0061] Furthermore, the stator current is first subjected to multi-synchronous rotating coordinate transformation (MSRFT) to convert the (6k±1)th harmonic current into a DC form in its corresponding rotating coordinate system, so as to realize the extraction and independent control of the harmonic current components.

[0062] Subsequently, the harmonic current transformed by MSRFT is substituted into the stator voltage equation of the permanent magnet synchronous motor (PMSM), and the instantaneous variation characteristics of the harmonic current are considered in the modeling process to obtain the harmonic voltage equation in the synchronous rotating coordinate system (SRF).

[0063] Finally, the harmonic voltage equation under the obtained SRF is subjected to a multi-synchronous rotating coordinate transformation again to eliminate the AC coupling terms and transform it into a DC expression in the corresponding (6k±1)th order synchronous rotating coordinate system.

[0064] Through the above transformation process, the harmonic voltage-current relationship that originally contained complex AC quantities in the SRF can be simplified to a synchronous rotating voltage vector represented by two DC quantities in the (6k±1)th order synchronous rotating coordinate system.

[0065] Step S2: Based on the harmonic voltage-current model obtained in Step S1, the state equation is transformed. An uncertain disturbance estimator is designed to estimate the uncertain disturbance, resulting in the uncertain disturbance matrix. Uncertain disturbances include cross-coupling terms containing speed information among harmonic variables, uncertainties in system parameters, and external disturbances. The uncertain disturbance estimator is designed to perform observational modeling of the uncertain disturbance matrix, online estimation of cross-coupling terms containing speed information among harmonic variables, uncertainties in system parameters, and external disturbances, real-time compensation for the negative impacts of disturbances, and improvement of the robustness of the control system to parameter changes and external disturbances.

[0066] For the above expression of the state equation, the left side of the equation is the first derivative of the harmonic current vector, and the right side of the equation is the nominal system dynamics A0x, the control input Bu, and the uncertainty disturbance matrix δ. UDE The sum of three terms, where x represents the harmonic current vector; u represents the harmonic voltage vector; A0 represents the simplified dynamic matrix after neglecting the angular velocity coupling term; B represents the input matrix; δ UDE This represents the composite uncertainty disturbance matrix; the nominal system dynamic A0x is the product of the simplified dynamic matrix (ignoring angular velocity coupling terms) and the harmonic current vector x. The control input Bu is the product of the input matrix B and the harmonic voltage vector u.

[0067] In this technical solution, the above-mentioned state equations are established based on the harmonic voltage-current model constructed in step S1.

[0068] The above step S2 mainly includes the following sub-steps.

[0069] Step S21: Rewrite the above harmonic voltage-current model into state equation form. For a more detailed explanation of the state equation expression, please refer to the above content.

[0070] Step S22: Select an appropriate filter to obtain an approximate disturbance estimate and estimate the system disturbance. Considering that the above model is a four-input and four-output system, in order to simplify the UDE control law design process, a first-order low-pass filter is selected to estimate the disturbance in the system; more specifically, it mainly includes the following process: By selecting a suitable filter, the disturbance estimate can approximately represent δ. UDE With filter l c The convolution of . Where l c s-domain expression L c To estimate the bandwidth ω c For the numerator, the Laplace variable s and the estimator bandwidth ω c The sum of these is the scalar in the denominator multiplied by the identity matrix I.

[0071] Step S23: After estimating the uncertainty disturbance, the uncertainty is counteracted by cleverly designing a feedforward path. More specifically, the harmonic voltage vector u is set equal to the harmonic voltage vector v output by the internal model controller. IMC The feedforward compensation voltage v generated by the control law of the uncertain disturbance estimator UDE The difference. The feedforward compensation voltage v generated by the control law of the uncertain disturbance estimator. UDE Equal to the inverse of the input matrix B and the perturbation estimation matrix The product of and . Taking the Laplace transform of the above harmonic voltage vector u, we obtain the s-domain voltage equation, where the s-domain voltage U(s) is equal to V. IMC(s) minus V UDE (s), where U(s) and V IMC (s), V UDE (s) correspond to the matrix vectors u and v respectively. IMC v UDE The s-domain expression. And the harmonic voltage vector V IMC (s) represents the harmonic current reference command value X. * The difference between state vector X(s) and state vector F is multiplied by the internal model controller F. IMC (s) Equivalent PI control rate F of the main diagonal elements IMC-PI (s) and F IMC (s) is the sum of the residual coupling compensation terms D(s) constructed from off-diagonal elements. Where F IMC The residual coupling compensation term D(s) constructed from off-diagonal elements does not contain angular velocity coupling terms that have been compensated by the uncertain disturbance estimator.

[0072] Step S24: Obtain the feedforward compensation voltage based on the results of steps S21 to S23. Further, by integrating the above multiple expressions, the online compensation amount of UDE, i.e., the feedforward compensation voltage V, can be obtained. UDE (s), feedforward compensation voltage V UDE (s) is equal to the inverse of the input matrix B multiplied by the first intermediate quantity and then multiplied by the first-order low-pass filter L. c (s), the first intermediate quantity is the s times state vector X(s) in the s domain minus the product of matrix A0 and state vector X(s) minus the product of input matrix B and control input U(s).

[0073] In this technical solution, the feedforward compensation voltage generated by the control law of the uncertain disturbance estimator can be calculated.

[0074] Step S3: The target harmonic current command is tracked in a closed loop using an internal model control structure, and the estimated value of the uncertainty disturbance matrix obtained in step S2 is used as feedforward compensation to offset the system uncertainty.

[0075] This embodiment also includes step S4: optimizing harmonic current decoupling control by adjusting the closed-loop bandwidth and estimator bandwidth respectively.

[0076] Through the above step S4, the harmonic current decoupling control can achieve an optimal balance between robustness and dynamic performance, thereby improving the robustness and response capability of the system.

[0077] refer to Figure 2This is a block diagram of the harmonic current decoupling control system combining UDE and IMC. The system mainly consists of three parts: a PI controller based on internal model decoupling design, an IMC feedforward decoupling compensation part, and an online compensation part for uncertainty interference error by UDE.

[0078] The method in this embodiment uses UDE to estimate online the cross-coupling terms between harmonic variables containing speed information, the uncertainty of system parameters, and external disturbances, thereby compensating for the negative impact of disturbances in real time and improving the robustness of the control system to parameter changes and external disturbances. The IMC structure is used to achieve closed-loop tracking control of the target harmonic current, ensuring improved system performance.

[0079] Based on the above technical solution, this embodiment also verifies the harmonic current decoupling control method based on electromagnetic vibration noise suppression. Specifically, the algorithm performance is verified on a vehicle drive motor with a peak power of 250kW and a peak torque of 350Nm. The harmonic current response waveform can be found by referring to... Figure 4 Furthermore, vibration noise steady-state optimization experiments and dynamic run-up optimization experiments were conducted in a semi-anechoic chamber (compliant with GB50800-2012 standard) with background noise below 15 dB(A). The optimization results can be referenced. Figure 5 , Figure 6 as well as Figure 7 .

[0080] This invention utilizes a harmonic voltage-current dynamic model based on a multi-synchronous rotating coordinate system. It combines an Uncertainty Estimator (UDE) with Internal Model Control (IMC). The UDE is used to calculate and compensate for harmonic variable cross-coupling, parameter perturbations, and load disturbances in real time. Combined with IMC, a robust tracking closed-loop control for the target harmonic current is constructed. This invention enables precise harmonic current injection under dynamic operating conditions, effectively suppressing the impact of parameter time-varying characteristics and harmonic coupling effects on control accuracy. It improves the system's dynamic response accuracy and anti-interference capability, achieving functional decoupling of tracking performance and anti-interference performance. It also boasts advantages such as simple parameter tuning and strong robustness. Bench tests verify that this invention effectively reduces PMSM order noise and improves the acoustic and vibration performance of electric vehicles.

[0081] Compared with the prior art, this embodiment can bring the following technical effects.

[0082] 1. Existing harmonic voltage-current controlled models neglect the strong coupling relationship between harmonic variables, which causes the harmonic current command tracking performance to drop sharply as the rotational speed increases. In the modeling process of this invention, the coupling effect between harmonic currents of the same and different orders in the controlled model as well as the instantaneous value change of the harmonic current variable itself are considered simultaneously, so as to achieve accurate decoupling compensation for the harmonic coupling effect.

[0083] 2. Existing harmonic current injection control methods do not consider factors such as magnetic saturation and strong time-varying nonlinearity of the transferred model parameters, making it difficult to achieve fast and accurate response to harmonic commands when facing parameter uncertainties and external disturbances. This invention proposes a harmonic current decoupling method based on the combination of Uncertainty Disturbance Estimator (UDE) and Internal Model Control (IMC). The UDE calculates and compensates for harmonic variable cross-coupling, parameter perturbation, and load disturbance in real time, and the IMC is combined to construct a robust tracking closed loop for the target harmonic current. This effectively suppresses the impact of parameter time-varying and harmonic coupling effects on control accuracy, significantly improves the dynamic response accuracy and anti-interference capability of the system, and achieves accurate harmonic injection under dynamic operating conditions.

[0084] 3. This invention can effectively improve the robustness and responsiveness of the system by simply adjusting the closed-loop bandwidth and the estimator bandwidth.

[0085] 4. This invention effectively suppresses the acoustic and vibration energy of the motor characteristic order under a wide range of operating conditions by dynamically adjusting the harmonic injection reference command. The noise reduction effect is particularly significant under medium and high speed load conditions, with a maximum reduction in vibration acceleration of 79.2% and a maximum reduction in noise of 18.2 dBA.

[0086] 5. This invention effectively suppresses acoustic vibration energy while maintaining the stability of key performance indicators of the motor, verifying its good adaptability and engineering application potential under complex working conditions.

Claims

1. A harmonic current decoupling control method based on electromagnetic vibration noise suppression, characterized in that, Includes the following steps: S1. Construct a harmonic voltage-current model in a multi-synchronous rotating coordinate system based on the stator current and stator voltage equations; S2, construct the state equation based on the harmonic voltage-current model, set up the uncertain disturbance estimator, estimate the uncertain disturbance, and obtain the uncertain disturbance matrix; S3 employs an internal model control structure to perform closed-loop tracking control of the target harmonic current command, using the estimated value of the uncertainty disturbance matrix as feedforward compensation to offset system uncertainties.

2. The harmonic current decoupling control method based on electromagnetic vibration noise suppression according to claim 1, characterized in that, Step S1 includes: S11, performs multi-synchronous rotating coordinate transformation on the stator current of the motor, and converts the harmonic current into DC form in its corresponding rotating coordinate system; S12, Substitute the transformed harmonic current into the stator voltage equation of the permanent magnet synchronous motor to obtain the harmonic voltage equation; S13, apply a multi-synchronous rotating coordinate transformation to the harmonic voltage equation to obtain two synchronous rotating voltage vectors expressed in DC form.

3. A harmonic current decoupling control method based on electromagnetic vibration noise suppression according to claim 1 or 2, characterized in that, The left side of the state equation is the first derivative of the harmonic current vector, and the right side of the state equation is the nominal system dynamics A0x, the control input Bu, and the uncertainty interference matrix δ UDE ; x is the harmonic current vector; u is the harmonic voltage vector; A0is the simplified dynamic matrix after ignoring the angular velocity coupling term, B is the input matrix; δ UDE is the composite uncertainty interference matrix.

4. A harmonic current decoupling control method based on electromagnetic vibration noise suppression according to claim 1 or 2, characterized in that, Step S2 includes: S21, the harmonic voltage-current model is rewritten in the form of state equations; S22, Select a filter to obtain an approximate disturbance estimate and estimate the disturbance in the system; S23, after estimating the uncertainty disturbance, the uncertainty is offset by designing a feedforward path; S24, integrate steps S21 to S23 to obtain the feedforward compensation voltage.

5. The harmonic current decoupling control method based on electromagnetic vibration noise suppression according to claim 1, characterized in that, The uncertainty disturbances include cross-coupling terms containing rotational speed information between harmonic variables, uncertainties in system parameters, and external disturbances.

6. The harmonic current decoupling control method based on electromagnetic vibration noise suppression according to claim 4, characterized in that, Specifically, a first-order low-pass filter is selected to estimate disturbances in the system.

7. The harmonic current decoupling control method based on electromagnetic vibration noise suppression according to claim 6, characterized in that, In step S2, the feedforward compensation voltage V generated by the control law of the uncertain disturbance estimator is finally obtained. UDE (s), the feedforward compensation voltage V UDE (s) is the inverse of the input matrix B multiplied by the first intermediate value and then multiplied by the first-order low-pass filter L. c (s), where the first intermediate quantity is the s-fold state vector X(s) in the s domain minus the product of matrix A0 and state vector X(s) minus the product of input matrix B and control input U(s).

8. A harmonic current decoupling control method based on electromagnetic vibration noise suppression according to claim 1, 2, or 7, characterized in that, In step S3, the output of the internal model control structure is a harmonic voltage vector V. IMC (s), the harmonic voltage vector V IMC (s) represents the harmonic current reference command value X. * The difference between state vector X(s) and state vector F is multiplied by the internal model controller F. IMC (s) Equivalent PI control rate F of the main diagonal elements IMC-PI (s) and F IMC (s) is the sum of the residual coupling compensation terms D(s) constructed from off-diagonal elements.

9. The harmonic current decoupling control method based on electromagnetic vibration noise suppression according to claim 8, characterized in that, The F IMC The residual coupling compensation term D(s) constructed from off-diagonal elements does not contain angular velocity coupling terms that have been compensated by the uncertain disturbance estimator.

10. A harmonic current decoupling control method based on electromagnetic vibration noise suppression according to claim 1 or 2, characterized in that, It also includes step S4: optimizing harmonic current decoupling control by adjusting the closed-loop bandwidth and estimator bandwidth respectively.

Images