Permanent magnet synchronous motor multi-source disturbance suppression method based on adaptive observer

Abstract

The invention discloses a permanent magnet synchronous motor multi-source disturbance suppression method based on an adaptive observer. Firstly, a multi-source disturbance model is reconstructed by designing non-singular coordinate transformation, and coupling between an unknown disturbance frequency and a to-be-estimated state is eliminated. On the basis, a self-adaptive observer is designed in combination with a virtual invariant manifold technology, and constant disturbance and multi-frequency unknown period disturbance components in the system can be estimated online. And finally, injecting a composite disturbance signal estimated by the observer in real time into a controller of the system as a feed-forward compensation item, thereby realizing suppression of multi-frequency disturbance. According to the method, accurate frequency and amplitude information of disturbance does not need to be known in advance, multiple periodic disturbances and constant disturbances can be effectively processed at the same time, and the application range of a permanent magnet synchronous motor driving system in a multi-source disturbance environment is effectively expanded.

Description

Technical Field

[0001] This invention belongs to the field of permanent magnet synchronous motor control, specifically a method for suppressing multi-source disturbances in permanent magnet synchronous motors based on an adaptive observer. Background Technology

[0002] Permanent magnet synchronous motors (PMSMs) have been widely used in modern industrial fields such as new energy vehicles, high-end equipment manufacturing, and industrial robots due to their high power density, high efficiency, and excellent control performance. However, in actual operation, motor drive systems often operate under complex conditions, facing various uncertainties such as load torque fluctuations, parameter perturbations, and inverter nonlinearities. These disturbances include both constant components and time-varying periodic components with unknown frequencies, severely degrading the system's tracking accuracy and dynamic performance, and even affecting the stable operation of the equipment. Therefore, effectively suppressing complex multi-source disturbances has become a key technical challenge for improving the performance and robustness of PMSM drive systems.

[0003] Existing disturbance suppression methods, such as repetitive control, resonant controllers, and various disturbance observers, have achieved some success, but most still have inherent limitations. On the one hand, methods such as resonant controllers based on the internal model principle can handle multi-frequency disturbances, but this requires the frequency of the disturbance to be precisely known; performance deteriorates sharply when the frequency changes or is unknown. On the other hand, while some advanced observer techniques based on adaptive or phase-locked loops can handle situations with unknown frequencies, their algorithmic structures typically only provide effective estimation for a single periodic disturbance of unknown frequency. When multiple periodic disturbances of unknown frequency exist simultaneously in the system, the aforementioned existing techniques are difficult to apply, resulting in poor suppression effects or even failure. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention proposes a multi-source disturbance suppression method for permanent magnet synchronous motors based on an adaptive observer. This method first reconstructs the multi-source disturbance model by designing a non-singular coordinate transformation, eliminating the coupling between the unknown disturbance frequency and the state to be estimated. Based on this, an adaptive observer is designed using virtual invariant manifold technology, capable of estimating constant disturbances and multi-frequency unknown periodic disturbance components in the system online. Finally, the composite disturbance signal estimated in real time by the observer is injected into the system controller as a feedforward compensation term, thereby suppressing multi-frequency disturbances. This invention achieves real-time and accurate estimation and compensation of constant disturbances and multi-frequency unknown periodic disturbances without prior knowledge of the disturbance frequency, thus significantly broadening the system's vibration suppression capability and improving its control performance and adaptability under complex operating conditions.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0006] The method for suppressing multi-source disturbances in permanent magnet synchronous motors based on adaptive observers includes the following steps:

[0007] S1. Reconstruct the multi-source perturbation model through non-singular coordinate transformation to eliminate the coupling between the unknown perturbation frequency and the state to be estimated.

[0008] S2. Based on the multi-source disturbance model constructed in step S1, an adaptive observer is designed using virtual invariant manifold technology to estimate constant disturbances and multi-frequency unknown periodic disturbances online.

[0009] S3. The composite disturbance signal estimated in real time by the adaptive observer in step S2 is introduced into the vector controller as a feedforward compensation quantity to suppress multi-source disturbances.

[0010] As a preferred technical solution of the present invention, the specific process of step S1 is as follows:

[0011] The mechanical equation of a permanent magnet synchronous motor is:

[0012] (1)

[0013] in Describes the differential operator. Indicates the mechanical rotation speed. Represents the moment of inertia. Indicates the load torque. The electromagnetic torque can be expressed as follows:

[0014] (2)

[0015] in Represents the extreme logarithm. Indicates magnetic flux. express Shaft stator current, express Shaft stator inductance;

[0016] Assuming the dynamics of the current loop are faster than those of the velocity loop, and that they satisfy the principle of multi-timescale separation, the mechanical dynamics equation (1) can be rewritten as:

[0017] (3) express Reference value for shaft current;

[0018] Considering periodic disturbances, the mechanical dynamics equation (1) of the permanent magnet synchronous motor can be further rewritten as:

[0019] (4)

[0020] in and This indicates the motor's rated flux linkage and rated moment of inertia. Indicates a constant load. The harmonic load, whose frequency and amplitude are both unknown, can be further modeled as follows:

[0021] (5)

[0022] in Indicates the first The amplitude of the periodic disturbance, Indicates the first The frequency of the periodic disturbance Indicates the first The phase of a periodic disturbance;

[0023] When considering two periodic disturbances, the mechanical dynamics equation (4) of the permanent magnet synchronous motor can be rewritten as:

[0024] (6)

[0025] in , Indicates system output , , , Indicates the equivalent constant perturbation. , This represents the pseudo-input of the velocity loop;

[0026] From (6), we can see the unknown frequency signal. and Respectively with the state to be estimated as well as Due to coupling, a non-singular coordinate transformation is defined to facilitate observer design:

[0027] (7)

[0028] Where the coordinate transformation matrix It can be represented as:

[0029] (8)

[0030] Substituting (7) into (6) yields the new state-space equation:

[0031] (9)

[0032] Substituting (8) into (7) yields:

[0033] (10)

[0034] When considering more than two periodic disturbances, the above method is extended.

[0035] As a preferred technical solution of the present invention, the specific process of step S2 is as follows:

[0036] Rewrite (9) in the standard form of the Brunovsky observer:

[0037] (11)

[0038] in , , , , Model (11) guarantees that the parameters to be estimated are... and regression vector Depends only on known signals;

[0039] For the system Design an adaptive observer:

[0040] (12)

[0041] superscript Represents the estimated value of the signal; This represents the observer gain, which needs to guarantee the matrix. It is a Hurwitz matrix; This represents the auxiliary variable introduced to construct the virtual pop culture; its specific expression is:

[0042] (13)

[0043] Frequency to be estimated The dynamic equations can be designed as follows:

[0044] (14)

[0045] in This represents the positive definite observer gain matrix;

[0046] Combining (12)-(14), we can obtain the complete expression for the observer and guarantee the state to be estimated. and the parameters to be estimated All converge to the true value;

[0047] Based on (10) and the designed adaptive observers (12)-(14), the estimated values ​​of the equivalent constant perturbation and the perturbations of each period can be obtained:

[0048] (15).

[0049] As a preferred technical solution of the present invention, the specific process of step S3 is as follows:

[0050] Based on vector control, the final designed speed loop controller is as follows:

[0051] (16)

[0052] in This represents a reference value for the machine's rotational speed. The PI controller is represented;

[0053] The current loop controller can be designed as follows:

[0054] (17)

[0055] in and They represent Shaft stator voltage, and They represent Reference and actual values ​​of shaft current and These represent the resonant angular frequencies as follows: and Quasi-resonant controller;

[0056] Effective suppression of multi-source disturbances can be achieved by using a speed loop controller (16) and a current loop controller (17).

[0057] Compared with the prior art, the present invention has the following advantages:

[0058] This invention fundamentally overcomes the limitations of traditional disturbance suppression methods, such as their prior dependence on disturbance frequencies and their ability to handle only single frequencies, by introducing non-singular coordinate transformation and an adaptive observer based on invariant manifold technology. This invention requires no prior knowledge of the frequency information of periodic disturbances and can accurately estimate constant disturbances and multiple periodic disturbances with unknown frequencies in the system online in real time. These disturbances are then used as feedforwards to compensate the controller, thereby achieving coordinated suppression of complex multi-source disturbances. This not only significantly broadens the system's effective vibration suppression capability but also enables it to maintain excellent dynamic tracking performance and steady-state control accuracy under complex operating conditions with time-varying frequencies, greatly improving the robustness and engineering applicability of permanent magnet synchronous motor drive systems. Attached Figure Description

[0059] Figure 1 Overall block diagram of the multi-source disturbance suppression method for permanent magnet synchronous motors based on adaptive observers according to embodiments of the present invention;

[0060] Figure 2Adaptive observer control block diagram of an embodiment of the present invention;

[0061] Figure 3 Simulation results of single-cycle load disturbance suppression of the proposed scheme in this embodiment of the invention;

[0062] Figure 4 Simulation results of the dual-cycle load disturbance suppression scheme proposed in the embodiments of the present invention;

[0063] Figure 5 Simulation results of dual-cycle load disturbance suppression in existing solutions according to embodiments of the present invention.

[0064] List of reference numerals in the attached diagram:

[0065] 1. Three-phase inverter; 2. Permanent magnet synchronous motor; 3. Speed ​​loop controller; 4. Current loop controller; 5. Inverse Park transform; 6. Inverse Clark transform; 7. Space vector pulse width modulation; 8. Clark transform; 9. Park transform; 10. Adaptive observer; 11. Encoder. Detailed Implementation

[0066] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments:

[0067] This invention provides a method for suppressing multi-source disturbances in permanent magnet synchronous motors based on an adaptive observer. It does not require prior knowledge of any frequency information of the periodic disturbances, and can estimate the constant disturbances and multiple periodic disturbances of unknown frequency in the system online in real time and accurately. It then uses these disturbances as feedforward quantities to compensate the controller, thereby achieving the coordinated suppression of complex multi-source disturbances. This significantly broadens the vibration suppression capability of the permanent magnet synchronous motor drive system and improves its control performance and adaptability under complex operating conditions.

[0068] The method for suppressing multi-source disturbances in permanent magnet synchronous motors based on adaptive observers includes the following steps:

[0069] S1. Reconstruct the multi-source perturbation model through non-singular coordinate transformation to eliminate the coupling between the unknown perturbation frequency and the state to be estimated.

[0070] The specific process is as follows:

[0071] The mechanical equation of a permanent magnet synchronous motor is:

[0072] (1)

[0073] in Describes the differential operator. Indicates the mechanical rotation speed. Represents the moment of inertia. Indicates the load torque. The electromagnetic torque can be expressed as follows:

[0074] (2)

[0075] in Represents the extreme logarithm. Indicates magnetic flux. express Shaft stator current, express Shaft stator inductance;

[0076] Assuming the dynamics of the current loop are faster than those of the velocity loop, and that they satisfy the principle of multi-timescale separation, the mechanical dynamics equation (1) can be rewritten as:

[0077] (3) express Reference value for shaft current;

[0078] Considering periodic disturbances, the mechanical dynamics equation (1) of the permanent magnet synchronous motor can be further rewritten as:

[0079] (4)

[0080] in and This indicates the motor's rated flux linkage and rated moment of inertia. Indicates a constant load. The harmonic load, whose frequency and amplitude are both unknown, can be further modeled as follows:

[0081] (5)

[0082] in Indicates the first The amplitude of the periodic disturbance, Indicates the first The frequency of the periodic disturbance Indicates the first The phase of a periodic disturbance;

[0083] When considering two periodic disturbances, the mechanical dynamics equation (4) of the permanent magnet synchronous motor can be rewritten as:

[0084] (6)

[0085] in , Indicates system output , , , Indicates the equivalent constant perturbation. , This represents the pseudo-input of the velocity loop;

[0086] From (6), we can see that the unknown frequency signal and Respectively with the state to be estimated as well as Due to coupling, a non-singular coordinate transformation is defined to facilitate observer design:

[0087] (7)

[0088] Where the coordinate transformation matrix It can be represented as:

[0089] (8)

[0090] Substituting (7) into (6) yields the new state-space equation:

[0091] (9);

[0092] Substituting (8) into (7) yields:

[0093] (10)

[0094] When considering more than two periodic disturbances, the above method is extended.

[0095] S2. Based on the multi-source disturbance model constructed in step S1, an adaptive observer is designed using virtual invariant manifold technology to estimate constant disturbances and multi-frequency unknown periodic disturbances online.

[0096] The specific process is as follows:

[0097] Rewrite (9) in the standard form of the Brunovsky observer:

[0098] (11)

[0099] in , , , , Model (11) guarantees that the parameters to be estimated are... and regression vector Depends only on known signals;

[0100] Design an adaptive observer for system (9):

[0101] (12)

[0102] superscript Represents the estimated value of the signal; This represents the observer gain, which needs to guarantee the matrix. It is a Hurwitz matrix; This represents the auxiliary variable introduced to construct the virtual pop culture; its specific expression is:

[0103] (13)

[0104] Frequency to be estimated The dynamic equations can be designed as follows:

[0105] (14)

[0106] in This represents the positive definite observer gain matrix;

[0107] Combining (12)-(14), we can obtain the complete expression for the observer and guarantee the state to be estimated. and the parameters to be estimated All converge to the true value;

[0108] Based on (10) and the designed adaptive observers (12)-(14), the estimated values ​​of the equivalent constant perturbation and the perturbations of each period can be obtained:

[0109] (15)

[0110] S3. The composite disturbance signal estimated in real time by the adaptive observer in step S2 is introduced into the vector controller as a feedforward compensation quantity to achieve suppression of multi-source disturbances.

[0111] The specific process is as follows:

[0112] Based on vector control, the final designed speed loop controller is as follows:

[0113] (16)

[0114] in This represents a reference value for the machine's rotational speed. The PI controller is represented;

[0115] The current loop controller can be designed as follows:

[0116] (17)

[0117] in and They represent Shaft stator voltage, and They represent Reference and actual values ​​of shaft current and These represent the resonant angular frequencies as follows: and Quasi-resonant controller;

[0118] Effective suppression of multi-source disturbances can be achieved by using a speed loop controller (16) and a current loop controller (17).

[0119] like Figure 1 As shown, the overall framework of this invention includes a three-phase inverter 1, a permanent magnet synchronous motor 2, a speed loop controller 3, a current loop controller 4, an inverse Park transform 5, an inverse Clark transform 6, a space vector pulse width modulation 7, a Clark transform 8, a Park transform 9, an adaptive observer 10, and an encoder 11; by acquiring two-phase current... and The dq-axis stator current is obtained through Clark transform 8 and Park transform 9. The rotor position and speed information are obtained using encoder 11, and the frequencies of multi-source disturbances and periodic disturbances in the system are estimated using adaptive observer 10. The q-axis reference current is then obtained using speed loop controller 3. ; Utilizing dq-axis stator current , In addition to the period frequency information estimated by the adaptive observer, the voltage reference in the rotating coordinate system is obtained by combining it with the current loop controller 4. and Finally, the switching signal is obtained through inverse Park transform (5), inverse Clark transform (6), and space vector pulse width modulation (7). Used to control three-phase inverter 1.

[0120] Figure 2 This is a control block diagram of the adaptive observer described in this invention. The specific implementation is as follows:

[0121] By collecting , , as well as The state in the new coordinate system is estimated using the designed adaptive observer. :

[0122] (8);

[0123] The estimated constant perturbation and multiple periodic perturbations of unknown frequency are obtained by using coordinate transformation:

[0124] (9);

[0125] Figure 3The simulation results show the single-cycle load disturbance suppression of the proposed scheme. The results indicate that before applying adaptive observer compensation, the motor speed fluctuates significantly due to constant and unknown periodic disturbances. Once the observer starts working and uses the estimated disturbance value for controller feedforward compensation, the speed fluctuations are quickly suppressed, and the motor rapidly recovers to the reference value for stable operation.

[0126] Figure 4 The simulation results show the proposed scheme for suppressing dual-cycle load disturbances. The results demonstrate that even under complex conditions with multiple frequency disturbances present simultaneously, the proposed adaptive observer can accurately estimate the disturbance amplitude and frequency. After compensation, the system speed fluctuation is significantly reduced, and the control performance is effectively restored, verifying the cooperative suppression capability of this method against multi-source disturbances.

[0127] Figure 5 The simulation results show the effectiveness and advancement of the existing dual-cycle load disturbance suppression method. The results demonstrate that the existing method has limited ability to suppress multi-frequency disturbances, and significant fluctuations in speed still exist after compensation, resulting in poor control performance improvement. This comparison highlights the effectiveness and advancement of the method proposed in this invention.

[0128] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any other way. Any modifications or equivalent changes made based on the technical essence of the present invention shall still fall within the scope of protection claimed by the present invention.

Claims

1. A method for suppressing multi-source disturbances in permanent magnet synchronous motors based on adaptive observers, characterized in that, Includes the following steps: S1. Reconstruct the multi-source perturbation model through non-singular coordinate transformation to eliminate the coupling between the unknown perturbation frequency and the state to be estimated. S2. Based on the multi-source disturbance model constructed in step S1, an adaptive observer is designed using virtual invariant manifold technology to estimate constant disturbances and multi-frequency unknown periodic disturbances online. S3. The composite disturbance signal estimated in real time by the adaptive observer in step S2 is introduced into the vector controller as a feedforward compensation quantity to suppress multi-source disturbances.

2. The method for suppressing multi-source disturbances in a permanent magnet synchronous motor based on an adaptive observer according to claim 1, characterized in that: The specific process of step S1 is as follows: The mechanical equation of a permanent magnet synchronous motor is: （1）； in Describes the differential operator. Indicates the mechanical rotation speed. Indicates the moment of inertia. Indicates the load torque. The electromagnetic torque can be expressed as follows: （2）； in Represents the extreme logarithm. Indicates magnetic flux. express Shaft stator current, express Shaft stator inductance; Assuming the dynamics of the current loop are faster than those of the velocity loop, and that they satisfy the principle of multi-timescale separation, the mechanical dynamics equation (1) can be rewritten as: (3); among which express Reference value for shaft current; Considering periodic disturbances, the mechanical dynamics equation (1) of the permanent magnet synchronous motor can be further rewritten as: （4）； in and This indicates the motor's rated flux linkage and rated moment of inertia. Indicates a constant load. The harmonic load, whose frequency and amplitude are both unknown, can be further modeled as follows: （5）； in Indicates the first The amplitude of the periodic disturbance, Indicates the first The frequency of the periodic disturbance Indicates the first The phase of a periodic disturbance; When considering two periodic disturbances, the mechanical dynamics equation (4) of the permanent magnet synchronous motor can be rewritten as: （6）； in , Indicates system output , , , Indicates the equivalent constant perturbation. , This represents the pseudo-input of the velocity loop; From (6), we can see the unknown frequency signal. and Respectively with the state to be estimated as well as Since coupling exists, a non-singular coordinate transformation is defined to facilitate observer design: （7）； Where the coordinate transformation matrix It can be represented as: （8）； Substituting (7) into (6) yields the new state-space equation: （9）； Substituting (8) into (7) yields: （10）； When considering more than two periodic disturbances, the above method is extended.

3. The method for suppressing multi-source disturbances in a permanent magnet synchronous motor based on an adaptive observer according to claim 1 or 2, characterized in that: The specific process of step S2 is as follows: Rewrite (9) in the standard form of the Brunovsky observer: （11）； in , , , , ; Model (11) guarantees the parameters to be estimated and regression vector Depends only on known signals; Design an adaptive observer for system (9): (12)； superscript Represents the estimated value of the signal; This represents the observer gain, which needs to guarantee the matrix. It is a Hurwitz matrix; This represents the auxiliary variable introduced to construct the virtual pop culture; its specific expression is: (13)； Frequency to be estimated The dynamic equations can be designed as follows: (14)； in This represents the positive definite observer gain matrix; Combining (12)-(14), we obtain the complete expression for the observer and ensure that the state to be estimated is... and the parameters to be estimated All converge to the true value; Based on (10) and the designed adaptive observers (12)-(14), the estimated values ​​of the equivalent constant perturbation and the perturbations of each period are obtained: (15)。 4. The method for suppressing multi-source disturbances in a permanent magnet synchronous motor based on an adaptive observer according to claim 1, characterized in that: The specific process of step S3 is as follows: Based on vector control, the final designed speed loop controller is as follows: (16)； in This represents a reference value for the machine's rotational speed. The PI controller is represented; The current loop controller can be designed as follows: (17)； in and They represent Shaft stator voltage, and They represent Reference and actual values ​​of shaft current and These represent the resonant angular frequencies as follows: and Quasi-resonant controller; Effective suppression of multi-source disturbances can be achieved by using a speed loop controller (16) and a current loop controller (17).

Images