A time series disturbance compensation high frequency motor drive multi-rate control method

Abstract

The application discloses a time series disturbance compensation high-frequency motor drive multi-rate control method, and the core principle is to transform low-rate disturbance compensation and reference tracking trajectory into a time-sequenced processing process matched with high-rate control instructions. s It can be known that a single disturbance compensation value is expanded into a time series prediction of disturbance changes on N future control points; meanwhile, a low-frequency single reference instruction is constructed into a smooth high-frequency reference trajectory covering N control points through a Lagrange interpolation method, and since the high-frequency control points T c The above all realize closed-loop correction of disturbance compensation and reference tracking, fundamentally solve the problem of open-loop prediction error accumulation and propagation caused by unmatched information update rates in traditional multi-rate control, and therefore realize high-precision robust control of a high-frequency motor drive system.

Description

Technical Field

[0001] This invention belongs to the field of high-frequency motor drive control technology, and more specifically, relates to a high-frequency motor drive multi-rate control method with time series disturbance compensation. Background Technology

[0002] In recent years, the rapid development of wide-bandgap semiconductor device technology, represented by silicon carbide (SiC), has enabled power converters to operate at extremely high switching frequencies. This has brought revolutionary breakthroughs in improving the power density, reducing system size, and enhancing energy efficiency of motor drive systems. However, high switching frequencies pose a significant challenge to digital controllers, as extremely short interrupt cycles limit the execution of complex control algorithms. To address this challenge, multi-rate model predictive control (MMC) schemes have emerged. This scheme decouples the sampling and control frequencies, allowing the controller to operate within a longer sampling period. Internal calculation A control sequence, and at a higher control frequency Execute sequentially. This solution alleviates the computational burden on the processor while achieving high switching frequency control.

[0003] However, although the above scheme successfully achieved high-frequency control, its control performance is highly dependent on the accuracy of the system model. Due to its multi-step prediction mechanism, when there is a mismatch between the model parameters and the actual system, the prediction error can exceed the accuracy of the system model within one sampling period. The disturbance propagates and accumulates continuously at control points with time-series characteristics. To address this issue, a common approach is to introduce a disturbance observer, which treats model mismatch and external disturbances as a single lumped disturbance for compensation. However, traditional disturbance observers can only perform disturbance compensation once per sampling time. They cannot provide real-time disturbance suppression for time series composed of multiple control inputs within a sampling period. This results in the cumulative effect of errors persisting within a sampling period, limiting the system's control accuracy and robustness. Therefore, a novel control method capable of accurately compensating for time-series disturbances within a sampling period is urgently needed. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of the prior art and propose a high-frequency motor drive multi-rate control method with time series disturbance compensation. In the case of inaccurate system model, it effectively compensates for and suppresses time series disturbances within the sampling period, thereby achieving high-performance control of the high-frequency motor drive system.

[0005] To achieve the above-mentioned objective, the present invention provides a high-frequency motor drive multi-rate control method with time-series disturbance compensation, characterized by comprising the following steps:

[0006] (1) Obtain motor system status and low-frequency reference commands: through The coordinate transformation module converts the three-phase current of the motor , , Convert to Actual current in stationary coordinate system , ; through speed loop PI controller and The coordinate transformation module obtains the motor's position. Reference current in stationary coordinate system ; Put the motor in Actual current in stationary coordinate system , As the current State vector at time step ;

[0007] (2) Estimation of time series disturbances and model compensation: The generalized proportional-integral observer is used to estimate the disturbances in the time series. and The time series perturbation between time points is then estimated and embedded into the current prediction model of the motor system.

[0008] (3) Generate high-frequency reference trajectory: based on the current Low-frequency reference current at time electrical angle and electric angular velocity ,predict The reference current at time t; then Lagrange interpolation is used to... and Interpolation calculations are performed on the reference current between time points to generate a smooth, high-dimensional reference trajectory vector. ;

[0009] (4) Solving for the optimal control sequence: Predicting the output vector of the motor system at the next moment based on the current prediction model. Combined with the high-dimensional reference trajectory vector generated in step (3) By minimizing the cost function An optimal control sequence is obtained by solving the problem. ;

[0010] (5) Drive execution: The optimal control sequence The space vector pulse width modulation module input to the motor system, in Each sampling period Inside, a high-frequency switching signal is generated through a space vector pulse width modulation module to control the motor.

[0011] The objective of this invention is achieved as follows:

[0012] The high-frequency motor drive multi-rate control method based on time-series disturbance compensation proposed in this invention is based on the principle of transforming low-rate disturbance compensation and reference tracking trajectory into a time-series processing process that matches high-rate control commands. Specifically, it uses a generalized proportional-integral observer to transform the disturbance compensation and reference tracking trajectory only at the sampling time... The known compensation value for a single disturbance is extended to the future. Time series prediction of disturbance changes at each control point; simultaneously, constructing a coverage of low-frequency individual reference commands using Lagrange interpolation. A smooth high-frequency reference trajectory at each control point, due to the fact that at each high-frequency control point... Both disturbance compensation and closed-loop correction for reference tracking are achieved, fundamentally solving the problem of continuous accumulation and propagation of open-loop prediction error caused by mismatch in information update rate in traditional multi-rate control. Therefore, high-precision robust control of high-frequency motor drive systems is achieved.

[0013] Meanwhile, the high-frequency motor drive multi-rate control method based on time series disturbance compensation of the present invention also has the following beneficial effects:

[0014] (1) In this invention, by using a generalized proportional-integral observer, the timing disturbance of each control point within the sampling period is estimated and compensated in real time. This method fundamentally solves the problem of propagation and accumulation of prediction error caused by model mismatch in the multi-step prediction process in traditional multi-rate predictive control methods, and significantly improves the accuracy of control.

[0015] (2) To address the data dimension mismatch problem of high-frequency reference trajectory in multi-rate control, this invention uses Lagrange interpolation to generate smooth high-frequency reference trajectory, which enables the system to have faster dynamic response speed and higher steady-state tracking accuracy.

[0016] (3) The present invention provides effective time-series compensation for disturbances caused by uncertainties such as parameter mismatch, so that the control system can still maintain stable and high-quality control performance when the key parameters of the motor change significantly.

[0017] (4) By decoupling the sampling and control frequency, the controller is allowed to complete the complex calculation of multiple future control sequences within a longer sampling period. This method alleviates the computational pressure on the digital processor under high switching frequency and provides sufficient time margin for executing the precise disturbance compensation algorithm proposed in this invention, ensuring the reliable implementation of the control strategy. Attached Figure Description

[0018] Figure 1 This is an overall structural diagram of a multi-rate control system for time-series disturbance compensation according to the present invention;

[0019] Figure 2 This is a detailed flowchart of the time-series disturbance compensation control algorithm;

[0020] Figure 3 It is a digital implementation method based on the time series disturbance compensation control algorithm;

[0021] Figure 4 This is a waveform comparison diagram of the present invention under the condition of motor inductance parameter mismatch. Detailed Implementation

[0022] The specific embodiments of the present invention will now be described with reference to the accompanying drawings to enable those skilled in the art to better understand the invention. It should be particularly noted that in the following description, detailed descriptions of known functions and designs that might obscure the main content of the invention will be omitted here.

[0023] Example

[0024] Figure 1 This is an overall structural diagram of a multi-rate control system for time-series disturbance compensation according to the present invention.

[0025] In this embodiment, we will first give a brief introduction to the overall structure of the multi-rate control system for time-series disturbance compensation, such as... Figure 1 As shown, it includes: a silicon carbide (SiC) based high-frequency motor driver 1, a three-phase permanent magnet synchronous motor (PMSM) 2, a DC power supply 3, a current sampling module 4, a position encoding module 5, and a disturbance compensation multi-rate controller 6.

[0026] The high-frequency motor driver 1 is a three-phase voltage source inverter (VSI), whose DC side is powered by a DC power supply 3, and whose AC side is connected to a three-phase permanent magnet synchronous motor 2.

[0027] The three-phase permanent magnet synchronous motor 2 is the final controlled object of this control system; in this embodiment, it is specifically a surface-mounted permanent magnet synchronous motor; its stator three-phase windings receive three-phase AC power from the high-frequency motor driver 1.

[0028] The DC power supply 3 is the energy source for the entire motor drive system, and its function is to provide a stable and continuous DC voltage for the high-frequency motor driver 1.

[0029] The current sampling module 4 is located between the driver 1 and the permanent magnet synchronous motor 2, and is used to collect three-phase AC current in real time. And feed it back to the multi-rate controller 6;

[0030] The position encoding module 5 is installed on the shaft of the motor 2 and is used to detect the electrical angle of the motor rotor in real time. And feed it back to the multi-rate controller 6;

[0031] The disturbance-compensated multi-rate controller 6 is the core of this invention. It receives signals from the current sampling module 4 and the position encoding module 5, processes them through a series of internal algorithms, and outputs a space vector pulse width modulation (SVPWM) signal. The high-frequency motor driver 1; the disturbance compensation multi-rate controller 6 internally includes: a speed calculation module 7, a difference calculation unit 8, and a speed loop PI controller 9. / The module includes: coordinate transformation module 10, Lagrange interpolation module 11, time series disturbance observation module 12, prediction model module 13, control law solving module 14, and space vector pulse width modulation module 15.

[0032] The input terminal of the speed calculation module 7 is connected to the position encoding module 5, and is used to calculate the speed based on the input electrical angle. Calculate the current actual angular velocity of the motor. ;

[0033] The input terminal of the difference operation unit 8 is connected to an externally given reference angular velocity. and the actual angular velocity from velocity calculation module 7 This is used to calculate the difference between the two, thus obtaining the speed error signal;

[0034] The input terminal of the speed loop PI controller 9 is connected to the output terminal of the difference operation unit 8, receives the speed error signal, and controls the output through proportional-integral (PI) control. Shaft reference current In addition, Shaft reference current In this embodiment, it is set to zero;

[0035] The / The coordinate transformation module 10 is responsible for converting the actual three-phase currents. Convert to Actual current in stationary coordinate system general shaft reference current converted to Reference current in stationary coordinate system ;

[0036] The input terminal of the Lagrange interpolation module 11 and The coordinate transformation module 10 is connected to receive a low-frequency reference current. Fitting low-frequency reference commands to generate a model that is similar to future... A high-dimensional reference trajectory matching each high-frequency control point And output it to the prediction model module 13;

[0037] The input terminal of the time series disturbance observation module 12 and The coordinate transformation module 10 is connected to receive the actual current. This module is used for online estimation of time series disturbances within the sampling period caused by uncertainties such as model parameter mismatch. ;

[0038] The prediction model module 13 receives data from... Actual current of coordinate transformation module 10 Time series disturbance information and time series disturbance observation module 12 This module embeds the estimated time-series disturbances into the system's discretized state-space model, forming a more accurate prediction model that has been compensated in real time.

[0039] The input terminal of the control law solving module 14 is connected to the Lagrange interpolation module 11 and the prediction model module 13; this module is based on the compensated prediction model and the high-dimensional reference trajectory. By minimizing the cost function The online solution yields a result containing the future. Optimal control sequence for each control input ;

[0040] The input terminal of the space vector pulse width modulation module 15 is connected to the control law solving module 14; this module generates driving signals sequentially at a higher control frequency within one sampling period. , used to control the high-frequency motor driver 1.

[0041] Figure 2 This is a detailed flowchart of the time-series disturbance compensation control algorithm of the present invention, characterized by including the following steps:

[0042] (1) Obtain motor status and low-frequency reference trajectory;

[0043] (1.1) Obtain the three-phase current of the motor through the current sampling module. , , ;

[0044] (1.2) Obtain the real-time electrical angle of the motor through the position encoding module. ;

[0045] (1.3) Through The coordinate transformation module converts the actual three-phase current Convert to Actual current in stationary coordinate system , The coordinate transformation formula is expressed as follows:

[0046] ;

[0047] (1.4) Using the velocity calculation module to calculate the derivative, the electrical angle is... Converted to electric angular velocity ;

[0048] (1.5) Given a reference speed The electrical angular velocity error is calculated using the difference operation unit and then input into the speed loop PI controller to obtain... Shaft reference current To achieve maximum torque-to-current ratio control, Shaft reference current Set to zero;

[0049] (1.6) To obtain the low-frequency current reference trajectory, by... The coordinate transformation module will The reference current of the rotating coordinate system is transformed into The reference current in the stationary coordinate system. The coordinate transformation formula is expressed as follows:

[0050] ;

[0051] (2) Time series disturbance observation and compensation;

[0052] (2.1) The errors caused by all uncertainties such as inaccurate model parameters and unmodeled dynamics are uniformly equivalent to a single time series disturbance term. The discretized state equations of the constructed system can be expressed as follows:

[0053] ;

[0054] in, For the system in The time series disturbance term at time t can be specifically represented as:

[0055] ;

[0056] in, , , Representing the system matrix respectively Input matrix Perturbation matrix The uncertain part at the current moment;

[0057] definition ,

[0058] ;

[0059] in, , , , They are respectively Current and voltage under the shaft, , and Separate the resistance, flux linkage, and inductance of the motor;

[0060] (2.2) Estimate each sampling period using a reduced-order generalized proportional-integral observer. Time series disturbances within;

[0061] (2.2.1) First, a high-order difference model of the discrete-time perturbation is constructed, as follows:

[0062] ;

[0063] in, Represented as time series disturbance Higher-order difference terms;

[0064] (2.2.2) Design a discretized reduced-order generalized proportional-integral observer:

[0065] ;

[0066] in, ,here , and The matrix is ​​defined in the continuous domain of the system. For the gain of the observer, Here, is the state vector of the observer. ;

[0067] (2.2.3) Current State vector at time step The input is fed into the generalized proportional-integral observer to estimate the... and Time series perturbations between moments :

[0068] ;

[0069] In this embodiment, the designed lifting time series perturbation matrix This will be incorporated into the subsequent calculation of the control rate, and the uncertainty components in the model will be compensated.

[0070] (2.3) Model compensation;

[0071] Perturb the time series Time series perturbation information Embedded into the current state equation, the compensated current state prediction equation is obtained:

[0072] ;

[0073] in, for The output vector at time t, It is a second-order identity matrix. This is the factor by which the frequency is increased.

[0074] (3) Lagrange interpolation high-frequency reference trajectory;

[0075] To obtain the lifting reference trajectory vector In this embodiment, the Lagrange interpolation method is used to obtain high-frequency reference trajectory data within the sampling period. The specific process is as follows:

[0076] First, the construction The interpolating polynomial of order 1 can be expressed as follows:

[0077] ;

[0078] in, For interpolation polynomials, Let be the basis functions for Lagrange interpolation, and satisfy:

[0079] ;

[0080] in, These are the interpolation points within the sampling period. , , , Includes reference current at each sampling period. ;

[0081] Next, regarding and Interpolation calculations are performed on the reference current between time points to generate a high-dimensional reference trajectory vector. It is expressed as follows:

[0082] ;

[0083] (4) Obtaining the optimal control sequence;

[0084] (4.1) To obtain the high-dimensional control sequence, the following cost function is constructed:

[0085] ;

[0086] in, For tracking trajectory error, i.e. , and They are respectively and increase control vector Weighting coefficients;

[0087] (4.2) By making the cost function right The partial derivatives are equal to zero, and the lift control vector is obtained as follows:

[0088] ;

[0089] Among them, the lifting matrix and They are represented as follows:

[0090] ;

[0091] ;

[0092] Boosting Matrix Represented as:

[0093] ;

[0094] Among them, the lifting matrix express:

[0095] ;

[0096] In the formula, To improve the time series perturbation matrix;

[0097] (5) Output the optimal control sequence in an ordered manner;

[0098] Based on the extended rolling optimization strategy, the control vector is improved. The first N sequence is applied to the system, and the optimal control sequence is represented as:

[0099] ;

[0100] in, , dimension ;

[0101] Finally, the optimal control sequence will be obtained. Space vector pulse modulation is performed sequentially to generate a switching signal that drives a high-frequency silicon carbide motor driver.

[0102] like Figure 3 This demonstrates the specific implementation of the control algorithm of this invention on a digital controller. Its core is the use of a dual-rate interrupt mechanism to balance computational load and control frequency; low-rate interrupt: this interrupt uses the sampling frequency... Execution, within a relatively long sampling period. Inside, the controller sequentially completes all complex computational tasks, including: obtaining the state through sampling. A high-dimensional reference is generated using the Lagrange interpolation module; the time-series perturbation observation module estimates the perturbation time series; and the control law solving module calculates the value containing the future... Optimal control sequence of instructions; high-rate interrupt: this interrupt uses... Control frequency times Execution; in each control cycle Initially, the pre-calculated control sequence is interrupted only at a low rate. In this process, a control command is retrieved sequentially; the modulation module compares the modulated wave with the triangular carrier wave to generate the switching signal for that control cycle. By separating the time-consuming computational task from the high-speed command execution, this invention can achieve high switching frequencies while ensuring sufficient computational redundancy in the digital controller.

[0103] This embodiment will be described with reference to examples, such as... Figure 4 The diagram shows experimental waveforms comparing different motor inductance parameters under mismatch conditions. The inductance value set in the controller is 50% of the actual value. From top to bottom, the waveforms represent the motor speeds. electromagnetic torque and phase current (a) is the traditional multi-rate control scheme and (b) is the control scheme proposed in this invention. It can be seen that the traditional multi-rate control scheme has a total harmonic distortion of current of up to 7.21% under parameter mismatch, and there are obvious fluctuations in speed and electromagnetic torque. In contrast, the control scheme proposed in this invention reduces the total harmonic distortion of current to 2.67%, has better current quality, and effectively suppresses disturbances within the sampling period, thereby improving the robustness of the system.

[0104] Although the illustrative specific embodiments of the present invention have been described above to enable those skilled in the art to understand the invention, it should be understood that the invention is not limited to the scope of the specific embodiments. For those skilled in the art, various changes are obvious as long as they are within the spirit and scope of the invention as defined and determined by the appended claims, and all inventions utilizing the concept of the present invention are protected.

Claims

1. A high-frequency motor drive multi-rate control method with time-series disturbance compensation, characterized in that, Includes the following steps: (1) Obtain motor system status and low-frequency reference commands: through The coordinate transformation module converts the three-phase current of the motor , , Convert to Actual current in stationary coordinate system , ; Through speed loop PI controller and The coordinate transformation module obtains the motor's position. Reference current in stationary coordinate system ; Put the motor in Actual current in stationary coordinate system , As the current State vector at time step ; (2) Estimation of time series disturbances and model compensation: The generalized proportional-integral observer is used to estimate the disturbances in the time series. and The time series perturbation between time points is then estimated and embedded into the current prediction model of the motor system. (3) Generate high-frequency reference trajectory: based on the current Low-frequency reference current at time electrical angle and electric angular velocity ,predict The reference current at time t; then Lagrange interpolation is used to... and Interpolation calculations are performed on the reference current between time points to generate a smooth, high-dimensional reference trajectory vector. ; (4) Solving for the optimal control sequence: Predicting the output vector of the motor system at the next moment based on the current prediction model. Combined with the high-dimensional reference trajectory vector generated in step (3) By minimizing the cost function An optimal control sequence is obtained by solving the problem. ; (5) Drive execution: The optimal control sequence The space vector pulse width modulation module input to the motor system, in Each sampling period Inside, a high-frequency switching signal is generated through a space vector pulse width modulation module to control the motor.

2. The high-frequency motor drive multi-rate control method with time series disturbance compensation according to claim 1, characterized in that, The actual current , With reference current The specific method for obtaining it is as follows: (2.1) Obtain the three-phase current of the motor through the current sampling module. , , ; (2.2) Obtain the real-time electrical angle of the motor through the position encoding module. ; (2.3) Through The coordinate transformation module converts the three-phase current , , Convert to Actual current in stationary coordinate system , , The coordinate transformation formula is expressed as follows: ； (2.4) Using the velocity calculation module to calculate the derivative, the electrical angle is... Converted to electric angular velocity ; (2.5) Given a reference speed The electrical angular velocity error is calculated using the difference operation unit and then input into the speed loop PI controller to obtain... Shaft reference current To achieve maximum torque-to-current ratio control, Shaft reference current Set to zero; (2.6) To obtain the low-frequency current reference trajectory, by... The coordinate transformation module will The reference current of the rotating coordinate system is transformed into Reference current in stationary coordinate system , The coordinate transformation formula is expressed as follows: 。 3. The high-frequency motor drive multi-rate control method with time series disturbance compensation according to claim 1, characterized in that, The estimation method for the time series perturbation is as follows: (3.1) In Construct the motor current state equation for one sampling period on a time scale: ； in, For the motor system in The time series disturbance term at time 1. Specifically, it is expressed as: ； in, , , Representing the system matrix respectively Input matrix Perturbation matrix The uncertain part at the current moment; definition , ； in, , , , They are respectively Current and voltage under the shaft, , and Separate the resistance, flux linkage, and inductance of the motor; (3.2) Estimation using a reduced-order generalized proportional integral observer Time series perturbations within each sampling period; (3.2.1) Construct a high-order difference model for discrete-time perturbations, as follows: ； in, Represented as time series disturbance Higher-order difference terms, Represented as order; (3.2.2) Design a discretized generalized proportional-integral observer, expressed as: ； Among them, variables , , and The matrix is ​​defined in the continuous domain of the motor system. For the gain of the observer, Let be the state vector of the observer. ; (3.2.3) Current State vector at time step The input is fed into the generalized proportional-integral observer to estimate the... and Time series perturbations between moments : 。 4. The high-frequency motor drive multi-rate control method with time series disturbance compensation according to claim 3, characterized in that, The model compensation method in step (2) is as follows: Perturb the time series Time series perturbation information Embedded into the current state equation, the compensated current state prediction equation is obtained: ； in, for The output vector at time step 1 It is a second-order identity matrix. This is the factor by which the frequency is increased.

5. The high-frequency motor drive multi-rate control method with time series disturbance compensation according to claim 1, characterized in that, The high-dimensional reference trajectory vector The generation method is as follows: (5.1) Construction The interpolating polynomial of order n is expressed as: ； in, For interpolation polynomials, Let be the basis functions for Lagrange interpolation, and satisfy: ； in, These are the interpolation points within the sampling period. , , , Includes reference current at each sampling period. ; (5.2) Regarding and Interpolation calculations are performed on the reference current between time points to generate a high-dimensional reference trajectory vector. It is expressed as follows: 。 6. A high-frequency motor drive multi-rate control method for time-series disturbance compensation according to any one of claims 1-5, characterized in that, The optimal control sequence The method for obtaining it is as follows: (6.1) Constructing the cost function: ； in, For tracking trajectory error, i.e. , and They are respectively and increase control vector Weighting coefficients; (6.2) By making the cost function right The partial derivatives are equal to zero, and the lift control vector is obtained as follows: ； Among them, the lifting matrix and They are represented as follows: ； ； Boosting Matrix Represented as: ； Among them, the lifting matrix express: ； In the formula, To improve the time series perturbation matrix; (6.3) Improve the control vector based on the extended rolling optimization strategy. The first N sequence is applied to the system, and the optimal control sequence is represented as: ； in, , dimension .

Images