A robust model-free predictive control method and system for permanent magnet synchronous motor

Abstract

The application discloses a robust permanent magnet synchronous motor model-free predictive control method and system, which establishes an ultralocal model on the basis of a motor mathematical model, observes the lumped disturbance in the model through an adaptive sliding mode observer with an adaptive exponential approach rate, and feeds forward compensates the model; then, the ultralocal model is used to predict a reference voltage vector in combination with a control thought without a dead time and a one-step delay compensation method; finally, a sector in a space voltage vector hexagon is determined according to the reference voltage vector, and then, based on a vertical optimization overmodulation strategy, a switching control signal is obtained by using the duty cycles of two adjacent voltage vectors and a zero vector of the sector, and is used for motor control. The application improves the traditional model-free predictive control method based on the ultralocal model from two aspects of model prediction and modulation output, further improves the system robustness, and has excellent dynamic performance under the overmodulation working condition.

Description

Technical Field

[0001] The present invention belongs to the technical field of permanent magnet synchronous motor current control, and in particular relates to a robust permanent magnet synchronous motor model-free predictive control method and system. Background Art

[0002] Surface-mounted permanent magnet synchronous motors (SPMSMs) are widely used in industrial transmissions due to their high power density and wide speed regulation range. The current loop within the control structure plays a crucial role in PMSM drive, directly impacting both the dynamic and steady-state performance of the motor drive system. Predictive control, a control method with rapid response, simple concepts, and easily implemented algorithms, is becoming the mainstream approach for PMSM control.

[0003] However, the dynamic and steady-state performance of predictive control is heavily dependent on the accuracy of the mathematical model of the permanent magnet synchronous motor. In the actual engineering application of permanent magnet synchronous motors, when motor parameters such as resistance and inductance change with temperature and electromagnetic fields, it will lead to motor model parameter mismatch, which will cause tracking static error in the current loop, degraded control performance, and even cause system instability.

[0004] Model-free predictive control (MPC) is a method that does not require motor parameters and can effectively improve system robustness. Currently, MPC primarily includes two types of methods: current difference and hyperlocal model. The current difference method uses stored current differences at each voltage for prediction, but this method suffers from current spikes and current update hysteresis. The hyperlocal model method replaces the traditional motor model with a hyperlocal model, using the resulting centralized disturbance to predict the next cycle. The control performance of this method is related to the accuracy of the calculation of the centralized disturbance observation value.

[0005] And with the development of model-free predictive control, due to the equivalent high gain of predictive control and the limitation of power converter voltage drive, the predicted reference voltage will enter the overmodulation region when responding to load impact. At this time, the SVM adopted by most studies has poor control effect in the overmodulation region. However, the problem of degraded control performance under overmodulation is rarely discussed in model-free predictive control based on hyperlocal models. Summary of the Invention

[0006] This invention addresses the shortcomings of existing technologies, namely the poor control performance of the SVM (Supporting Virtual Machine) employed in most existing research in the overmodulation region. By doing so, we provide a robust model-free predictive control method and system for permanent magnet synchronous motors. This method utilizes a sliding mode observer to rapidly and effectively observe disturbances with minimal jitter, significantly improving the robustness of the system. Furthermore, an improved modulation output scheme improves the system's dynamic performance under overmodulation without compromising steady-state performance.

[0007] The technical solution of the present invention provides a robust permanent magnet synchronous motor model-free predictive control method, wherein the model-free prediction part includes a super-local model and a sliding mode observer; the modulation output part is a modulation output method based on an overmodulation strategy of vertical optimization.

[0008] In a first aspect, the permanent magnet synchronous motor model-free predictive control method provided by the technical solution of the present invention includes:

[0009] Step 1: Sample the three-phase current and voltage of the permanent magnet synchronous motor separately during the sampling period, and perform coordinate transformation to obtain the sampling current i of the d-axis and q-axis in the two-phase dq coordinate system. d (k), i q (k) and the sampling voltage u d (k),u q (k);

[0010] Step 2: The sampling current i of the d-axis and q-axis d (k), i q (k) and the sampling voltage u of the d-axis and q-axis d (k),u q (k) Input the sliding mode observer built based on the hyperlocal model to observe the stator current value and concentrated disturbance in the dq coordinate system at the next moment;

[0011] Step 3: Substitute the stator current value and the concentrated disturbance into the hyperlocal model after Euler discretization to obtain the reference voltage values ​​of the d-axis and q-axis

[0012] Step 4: Use the reference voltage values ​​of the d-axis and q-axis The sector of the spatial voltage vector hexagon where the reference voltage vector is located is identified. Then, based on an overmodulation strategy with vertical optimization, the duty cycle of the two adjacent voltage vectors and the zero vector in the sector is used to obtain the switching control signal of the inverter to achieve control of the permanent magnet synchronous motor.

[0013] The technical solution of the present invention provides a model-free predictive control method for a permanent magnet synchronous motor, which uses a super-local model to replace the traditional mathematical model of the permanent magnet synchronous motor to realize predictive current control, and uses a sliding mode observer to estimate parameter disturbances and perform feedback compensation. Then, in the process of modulation output, considering the optimization problem of overmodulation, an overmodulation strategy with vertical optimization is proposed based on SVM, which realizes the optimized modulation output of the reference voltage through simple calculation, thereby improving the dynamic control performance in the overmodulation area.

[0014] Further optionally, the content of the vertical optimization overmodulation strategy includes:

[0015] Using the duty ratios of the two adjacent voltage vectors and the zero vector, a modulation region where a reference voltage vector is located is identified, wherein the space voltage vector region is divided into a linear modulation region, an overmodulation region I, and an overmodulation region II;

[0016] If the reference voltage vector is located in the linear modulation region, a switch control signal is generated by modulating the duty cycle of the two adjacent voltage vectors and the zero vector;

[0017] If the reference voltage vector is located in the overmodulation region I, a new reference voltage vector synthesized by two adjacent voltage vectors is determined on the boundary line of the space voltage vector using a vertical method, and then the duty cycle and modulation are updated to generate a switch control signal;

[0018] If the reference voltage vector is located in the overmodulation II region, the voltage vector closest to the two adjacent voltage vectors is used as a new reference voltage vector, and the duty cycle and modulation are updated to generate a switch control signal.

[0019] Further optionally, when the space voltage vector region is divided into a linear modulation region, an overmodulation region I, and an overmodulation region II, the corresponding division rule is:

[0020] If d1+d2≤1, the reference voltage vector is located in the linear modulation region;

[0021] If d1+d2>1 and |d1-d2|<1, the reference voltage vector is located in the overmodulation region I;

[0022] If d1+d2>1 and |d1-d2|>1, the reference voltage vector is located in the overmodulation II region;

[0023] Wherein, d1 and d2 represent the updated duty cycles of the two adjacent voltage vectors respectively.

[0024] Further optionally, if the reference voltage vector is located in the overmodulation region I, the duty cycle is updated according to the following formula:

[0025]

[0026] If the reference voltage vector is in the overmodulation region II, the duty cycle is updated according to the following formula:

[0027]

[0028] Where d1 and d2 represent the duty cycles of the two adjacent voltage vectors, represents the reference voltage vector, and u1 and u2 are the two adjacent voltage vectors.

[0029] Further optionally, the sliding mode observer is a sliding mode observer based on an adaptive exponential approach rate, and the control law in the sliding mode observer is:

[0030]

[0031] Where U dsmo and U qsmo represents the sliding mode control law, is the adaptive coefficient, k and δ are coefficients greater than 0, ε is a coefficient greater than 0 and less than 1, and λ is the set proportional factor; s is the sliding surface, including the sliding parameter s corresponding to the dq axis direction d 、s q , and is defined as the error between the stator current observation value and the stator current sampling value under the dq axis, specifically expressed as:

[0032]

[0033] in, is the stator current observation value of the dq axis, i d 、i q is the stator current sampling value of the dq axis, e d 、e q They are respectively expressed as the error between the stator current observation value and the stator current sampling value under the dq axis, and Z(e) is a saturation function that satisfies:

[0034]

[0035] Further optionally, the observation formula of the sliding mode observer is expressed as:

[0036]

[0037] Among them, T s represents the sampling period, U dsmo and U qsmo represents the sliding mode control law of the sliding mode observer, g d and g q Both are gain coefficients of the control law, which are positive real numbers;

[0038] is the stator current value of the d-axis and q-axis in the dq coordinate system at the next moment; It is the concentrated disturbance of the d-axis and q-axis in the dq coordinate system at the next moment.

[0039] The present invention selects the parameters k, λ and g to be designed based on the Lyapunov function, which is an existing theory of sliding mode observer parameter design and is recorded in many existing papers. Therefore, the present invention does not make any specific restrictions on this. g refers to g d and g qcoefficient.

[0040] Further optionally, the Euler discretization is performed on the hyperlocal model to obtain a reference voltage value corresponding to the discretized model. The formula is as follows:

[0041]

[0042] Among them, T s represents the sampling period, They represent the current reference values ​​of the q-axis and d-axis respectively; α is the model input gain.

[0043] In a second aspect, the present invention provides a control system based on the permanent magnet synchronous motor model-free predictive control method, which at least includes:

[0044] The sampling module is used to sample the three-phase current and voltage of the permanent magnet synchronous motor during the sampling period, and perform coordinate transformation to obtain the sampling current i of the d-axis and q-axis in the two-phase dq coordinate system. d (k), i q (k) and the sampling voltage u d (k),u q (k);

[0045] Sliding mode observer module, used to use the sampling current i of the d-axis and q-axis d (k), i q (k) and the sampling voltage u of the d-axis and q-axis d (k),u q (k) Observe the stator current value and concentrated disturbance in the dq coordinate system at the next moment;

[0046] The hyperlocal model prediction module is used to substitute the stator current value and the centralized disturbance into the hyperlocal model after Euler discretization to obtain the reference voltage values ​​of the d-axis and q-axis.

[0047] A space vector modulation module is used to utilize the reference voltage values ​​of the d-axis and q-axis The sector of the spatial voltage vector hexagon where the reference voltage vector is located is identified. Then, based on an overmodulation strategy with vertical optimization, the duty cycle of the two adjacent voltage vectors and the zero vector in the sector is used to obtain the switching control signal of the inverter to achieve control of the permanent magnet synchronous motor.

[0048] In a third aspect, the present invention provides a system based on the permanent magnet synchronous motor model-free predictive control method, which includes a permanent magnet synchronous motor and a control subsystem. The control subsystem uses the permanent magnet synchronous motor model-free predictive control method to generate a switching control signal of the inverter, thereby controlling the permanent magnet synchronous motor.

[0049] In a fourth aspect, the present invention provides a computer-readable storage medium storing a computer program, wherein the computer program is called by a processor to implement:

[0050] The three-phase current and voltage of the permanent magnet synchronous motor are sampled separately during the sampling period, and the coordinate transformation is performed to obtain the sampling current i of the d-axis and q-axis in the two-phase dq coordinate system. d (k), i q (k) and the sampling voltage u d (k),u q (k);

[0051] The sampling d-axis and q-axis current i d (k), i q (k) and the sampling voltage u of the d-axis and q-axis d (k),u q (k) Input the sliding mode observer built based on the hyperlocal model to observe the stator current value and concentrated disturbance in the dq coordinate system at the next moment;

[0052] Substitute the stator current value and the concentrated disturbance into the hyperlocal model after Euler discretization to obtain the reference voltage values ​​of the d-axis and q-axis

[0053] Using the reference voltage values ​​of the d-axis and q-axis The sector of the spatial voltage vector hexagon where the reference voltage vector is located is identified. Then, based on an overmodulation strategy with vertical optimization, the duty ratios of the two adjacent voltage vectors and the zero vector in the sector are used to obtain the switching control signal of the inverter for controlling the permanent magnet synchronous motor.

[0054] Beneficial effects

[0055] The present invention discloses a robust model-free predictive control method for a permanent magnet synchronous motor. This method uses a hyperlocal model instead of the traditional mathematical model of the permanent magnet synchronous motor to implement predictive current control, and utilizes a sliding mode observer to estimate parameter disturbances and perform feedback compensation. Finally, during the modulation output process, considering the optimization problem of overmodulation, an overmodulation strategy with vertical optimization is proposed based on the Support Vector Machine (SVM), achieving optimized modulation output of the reference voltage through simple calculations. This method, compared to the prior art, has at least the following advantages:

[0056] (1) The model-free predictive control method for the permanent magnet synchronous motor adopts a super-local model that does not use any motor parameters to replace the mathematical model of the permanent magnet synchronous motor, thereby improving the robustness of the system, continuing the advantages of the model-free predictive control method, and providing another technical idea to realize the model-free predictive control method.

[0057] (2) The model-free predictive control method for the permanent magnet synchronous motor proposes an overmodulation strategy with vertical optimization based on the SVM principle, which divides the spatial voltage vector area into a linear modulation area, an overmodulation area I, and an overmodulation area II. If the reference voltage vector is located in the overmodulation area I, a new reference voltage vector synthesized by two adjacent voltage vectors is determined on the boundary line using the vertical method for modulation output. If the reference voltage vector is located in the overmodulation area II, the voltage vector closest to the two adjacent voltage vectors is used as the new reference voltage vector for modulation output. Through simple calculations, the technical solution of the present invention improves the dynamic control performance in the overmodulation area and solves the problem that the existing SVM has poor control effect and reduced control performance in the overmodulation area.

[0058] (3) In a further preferred embodiment of the present invention, a sliding mode observer based on an adaptive exponential reaching rate is proposed to observe and compensate for parameter disturbances. Under the action of the adaptive exponential reaching rate, the disturbances of the sliding mode controller are suppressed and the tracking speed of the sliding mode controller is improved, thereby improving the accuracy of predictive control and the parameter disturbance suppression capability, and increasing the robustness of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 It is a system model block diagram corresponding to the method of the present invention;

[0060] Figure 2 The block diagram of the adaptive sliding mode observer model corresponding to the method of the present invention;

[0061] Figure 3 Schematic diagram of vertical overmodulation corresponding to the method of the present invention, (a), (b), and (c) correspond to the linear modulation area, overmodulation area I, and overmodulation area II, respectively;

[0062] Figure 4 The present invention provides a comparison simulation diagram of the traditional and improved model-free predictive control parameters mismatch, wherein (a) corresponds to the current i of the traditional model-free predictive current control model. d 、i q Waveform diagram, (b) corresponds to the current i of the improved model-free prediction current model of the present invention d 、i q Waveform diagram, (c) corresponds to the harmonic analysis diagram of the traditional model-free predictive current control model, (d) corresponds to the harmonic analysis diagram of the improved model-free predictive current model of the present invention;

[0063] Figure 5 The present invention provides a comparative simulation diagram of the traditional and improved model-free predictive control under overmodulation, wherein (a) corresponds to the u of the traditional model-free predictive current control model. sDynamic response diagram, (b) corresponds to the u of the improved model-free prediction current model of the present invention s Dynamic response diagram, (c) corresponds to the current i of the traditional model-free predictive current control model d 、i q Waveform diagram, (d) corresponds to the current i of the improved model-free prediction current model of the present invention d 、i q Waveform graph. DETAILED DESCRIPTION

[0064] The technical solution of the present invention uses a new technical idea to solve some of the existing technical problems existing in the existing model-free predictive control method. Figure 1 This is a system model block diagram corresponding to the method described in the present invention, such as Figure 1 As shown, the system mainly includes a sliding mode observer module, a hyperlocal model prediction module, and a space vector modulation module with an overmodulation strategy with vertical optimization. Based on the above three core modules, the technical solution of the present invention provides a model-free predictive control method for a permanent magnet synchronous motor. The sampled current and voltage after coordinate transformation are input into the adaptive sliding mode observer module to obtain the current value at the next moment and the concentrated disturbance term of the hyperlocal model, and the above-mentioned observations and the reference current value obtained under the speed loop PI are input into the hyperlocal model prediction module. Based on the control idea of ​​zero beat, the reference voltage value is predicted by the hyperlocal model; finally, after the reference voltage value is inversely Park transformed, the space vector modulation module generates the switching control signal of the two-level inverter through the overmodulation strategy with vertical optimization to realize the control of the permanent magnet synchronous motor. The present invention further optimizes the sliding mode observer and proposes a sliding mode observer based on an adaptive exponential convergence rate. The above technical solution achieves compensation for concentrated disturbances through a sliding mode observer with an adaptive exponential convergence rate, further increasing the robustness of the system. Furthermore, an overmodulation strategy with vertical optimization based on the SVM principle is adopted. Due to the inverter voltage limitation, the reference voltage entering the overmodulation region under sudden load changes can be optimized, improving the system's dynamic performance in the overmodulation region. The present invention will be further described below with reference to an embodiment, using a surface-mounted permanent magnet synchronous motor as an example.

[0065] Example 1:

[0066] This embodiment provides a robust model-free predictive control method for a permanent magnet synchronous motor, comprising the following steps:

[0067] Step 1: Sample the three-phase current and voltage of the surface-mounted permanent magnet synchronous motor separately during the sampling period, and perform coordinate transformation to obtain the sampling current i of the d-axis and q-axis in the two-phase dq coordinate system. d (k), i q (k) and the sampling voltage u d(k),u q Specifically, after sampling the three-phase current and voltage of the surface-mounted permanent magnet synchronous motor, the sampled current i of the d-axis and q-axis in the two-phase dq coordinate system is obtained through Clack and Park coordinate transformation. d (k), i q (k) and the sampling voltage u d (k),u q (k).

[0068] Step 2: The sampling current i d (k), i q (k) and the sampling voltage u d (k),u q (k) Input the sliding mode observer built based on the super-local model to observe the stator current values ​​of the d-axis and q-axis in the dq coordinate system at the next moment and concentrated disturbances

[0069] Among them, the construction principle of the sliding mode observer based on the hyperlocal model is as follows:

[0070] First, a mathematical model of the surface-mounted permanent magnet synchronous motor in the dq axis coordinate system is established. Based on the mathematical model, the stator voltage is used as the control variable and the stator current change rate is used as the output variable to establish a first-order superlocal model of the permanent magnet synchronous motor.

[0071] The mathematical model of the surface-mounted permanent magnet synchronous motor in the dq axis coordinate system is:

[0072]

[0073] Where u d 、u q are the d-axis and q-axis voltages respectively; i d 、i q is the stator current of the d-axis and q-axis, which is regarded as the sampling current; w e is the electrical angular velocity of the motor; R s , L s and ψ f They are stator resistance, stator inductance and rotor flux respectively.

[0074] A first-order hyperlocal model is established in the dq-axis coordinate system, which can be expressed as:

[0075]

[0076] Where, parameter α = 1 / L s is the model input gain; parameter f d 、f qis the unknown part of the model (centralized disturbance) and is also the desired quantity of the first-order hyperlocal model.

[0077] Then, based on the above-mentioned first-order hyperlocal model, a sliding mode observer with a novel adaptive approach rate is established. d (k), i q (k) and the sampling voltage u d (k),u q (k) After inputting the sliding mode observer and performing first-order forward discretization on the mathematical model, the current value at the next moment can be observed. and concentrated disturbances

[0078] The mathematical model of the adaptive sliding mode observer in the dq-axis coordinate system is as follows:

[0079]

[0080]

[0081] Where, is the parameter perturbation f d 、f q The estimated value (observed value) of is the estimated value (observed value) of the stator current of the d-axis and q-axis, U dsmo and U qsmo represents the sliding mode control law, g d and g q is the gain coefficient of the control law.

[0082] The sliding mode observer mathematical model is discretized into the first-order Euler form to predict the stator current value of the dq axis at the next moment. and concentrated disturbance compensation The corresponding formula is as follows:

[0083]

[0084]

[0085] Step 3: Set the stator current value and the concentrated disturbance Substitute the hyperlocal model after Euler discretization to obtain the reference voltage values ​​of the d-axis and q-axis Among them, in order to accurately track the reference current of the d-axis and q-axis The deadbeat control concept is used to predict and calculate the reference voltage values ​​of the d-axis and q-axis.

[0086] It should be noted that the given speed n* The deviation between the actual speed n measured by the encoder is obtained by the PI controller of the speed loop to obtain the q-axis current reference value. And give the reference value of d-axis current

[0087] The formula for the reference voltage value is as follows:

[0088]

[0089] Step 4: Use the reference voltage values ​​of the d-axis and q-axis Identify the sector of the spatial voltage vector hexagon where the reference voltage vector is located, and then, based on an overmodulation strategy with vertical optimization, use the duty cycle of the two adjacent voltage vectors and the zero vector in the sector to obtain the switch control signal to achieve control of the permanent magnet synchronous motor. The details are as follows:

[0090] The calculated reference voltage value Perform inverse Park transform to obtain the reference voltage in the two-phase stationary αβ coordinate system Then calculate the reference voltage vector and the angle with the α axis Then, the sector of the hexagonal space voltage vector is determined based on the angle. It is the reference voltage value of the d-axis and q-axis in the dq coordinate system Determine the voltage vector, that is

[0091] Then, let the two adjacent voltage vectors in the sector be u1 and u2, and the duty cycle of the three voltage vectors can be calculated as:

[0092]

[0093] Wherein, d1, d2, and d0 represent the duty ratios of the two adjacent voltage vectors and the zero vector, respectively; E1, E2, E3, x2, and x1 are all intermediate variables; and the subscript αβ represents a two-phase stationary αβ coordinate system.

[0094] If d1+d2≤1, the reference voltage is in the linear modulation region and is output according to the original reference voltage.

[0095] If d1+d2>1 and |d1-d2|<1, the reference voltage is in the overmodulation region I. The new reference voltage vector is determined on the boundary line of the space voltage vector by the vertical line (a vertical line is drawn on the boundary line, and the new reference voltage vector is determined based on the intersection of the boundary line and the vertical line). The new duty cycle is calculated as follows:

[0096] make

[0097] Using the Pythagorean theorem we have:

[0098]

[0099] By changing, it is easy to obtain:

[0100]

[0101] Then the new voltage vector duty cycle can be expressed as:

[0102]

[0103] If d1+d2>1 and |d1-d2|>1, the reference voltage is in the overmodulation region II. The closest adjacent voltage vector is determined as the new voltage vector, and the duty cycle is:

[0104]

[0105] Figure 3 The vertical overmodulation principle diagram, where (a), (b), and (c) correspond to the linear modulation area, overmodulation area I, and overmodulation area II, respectively. The value of is used to judge the space voltage vector sector, and the duty ratio of the two optimal adjacent voltage vectors (i.e. the two adjacent voltage vectors in the sector) and the zero vector are calculated based on the basic principle of SVM (space vector modulation) and the duty ratio of the zero vector is used to make regional judgments. When in the linear modulation area, the reference voltage vector is directly compared with the adjacent voltage vector and zero vector of the sector. Synthesis; when the reference voltage vector When it is in the overmodulation I region, a vertical line is drawn on the boundary line of the space voltage vector to obtain a new reference voltage vector synthesized by two adjacent voltage vectors on the boundary line of the space voltage vector; when the reference voltage In the overmodulation II region, the closest of the two adjacent voltage vectors is used for output. It should be understood that, based on the above technical ideas, the technical solution of the present invention fully considers the optimization problem of overmodulation and proposes an overmodulation strategy with vertical optimization based on SVM. This achieves optimized modulation output of the reference voltage through simple calculations, improving dynamic control performance in the overmodulation region.

[0106] In addition, this embodiment also preferably adopts a sliding mode observer based on an adaptive exponential approach rate, such as Figure 2The figure shows the block diagram of the adaptive sliding mode observer model. Based on the hyperlocal model, a sliding mode observer mathematical model based on an adaptive exponential approach rate is established. The use of a new adaptive exponential approach rate not only enables the sliding mode observer to converge quickly but also effectively suppresses the sliding mode chattering problem. The control law in the sliding mode observer is:

[0107]

[0108] The error between the observed current value and the actual current value under the dq axis is set as the sliding surface:

[0109]

[0110] The new adaptive approach rate is designed as follows:

[0111]

[0112] in is the saturation function; λ is the designed proportional factor; is the adaptive coefficient, k, δ is a coefficient greater than 0, and ε is a coefficient greater than 0 and less than 1. As |s| increases, the adaptive coefficient approaches This means that the new approximation rate converges faster than the traditional approximation rate; when |s| decreases, the adaptive coefficient converges to Where |s| gradually decreases to 0 as the sliding mode control function decreases, which means that when the system trajectory approaches the sliding mode surface, the adaptive coefficient gradually decreases to 0 to suppress the sliding mode chattering. Therefore, using this adaptive reaching law can be achieved by making the variable between 0 and The saturation function is used to replace the traditional sign function, which can further reduce chattering while ensuring accuracy.

[0113] In summary, the method provided in this embodiment optimizes the sliding mode controller and also proposes modulation based on an overmodulation strategy with vertical optimization. Figure 4 The conventional model-free predictive current control based on the common sliding mode observer and the improved model-free predictive current control based on the present invention are shown in Figure 2. The results show that the improved model-free predictive current control based on the conventional sliding mode observer and the improved model-free predictive current control based on the present invention can be used to calculate the current under the conditions of 1200 r / min speed, 3 N.m torque and double inductance mismatch (L′ s =2L s ), the current i d 、i q Comparison of waveforms and harmonic analysis, where (a) corresponds to the current i of the traditional model-free predictive current control model. d 、i q Waveform diagram, (b) corresponds to the current i of the improved model-free prediction current model of the present inventiond 、i q Waveform diagram, (c) corresponds to the harmonic analysis diagram of the traditional model-free predictive current control model, and (d) corresponds to the harmonic analysis diagram of the improved model-free predictive current model of the present invention. It can be seen from the images that the current i under the improved model-free predictive current control is d 、i q The ripple is smaller and the harmonic content is lower, which means that the robustness is higher under parameter mismatch.

[0114] Figure 5 The traditional model-free predictive current control based on SVM and the improved model-free predictive current control based on the present invention, at a speed of 2500r / min and a torque load that suddenly changes from 0N.m to 7N.m, the current i d 、i q and u s Dynamic response comparison diagram. Among them, (a) corresponds to the u of the traditional model-free predictive current control model. s Dynamic response diagram, (b) corresponds to the u of the improved model-free prediction current model of the present invention s Dynamic response diagram, (c) corresponds to the current i of the traditional model-free predictive current control model d 、i q Waveform diagram, (d) corresponds to the current i of the improved model-free prediction current model of the present invention d 、i q Waveform. It can be seen from the image that the improved model-free predicted current control is in the overmodulation area, the current i d 、i q Less overshoot, u s The time to return to the modulation area is shorter, and the dynamic performance in the overmodulation area is better.

[0115] Example 2:

[0116] This embodiment provides a control system based on the permanent magnet synchronous motor model-free predictive control method, which at least includes: a sampling module, a sliding mode observer module, a super-local model prediction module and a space vector modulation module.

[0117] The sampling module is used to sample the three-phase current and voltage of the permanent magnet synchronous motor in the sampling period, and perform coordinate transformation to obtain the sampling current i of the d-axis and q-axis in the two-phase dq coordinate system. d (k), i q (k) and the sampling voltage u d (k),u q (k).

[0118] The sliding mode observer module is used to utilize the sampling current i of the d-axis and q-axis d (k), i q(k) and the sampling voltage u of the d-axis and q-axis d (k),u q (k) Observe the stator current values ​​of the d-axis and q-axis in the dq coordinate system at the next moment and concentrated disturbances

[0119] The super local model prediction module is used to convert the stator current value and concentrated disturbances Substitute the hyperlocal model after Euler discretization to obtain the reference voltage values ​​of the d-axis and q-axis

[0120] The space vector modulation module is used to use the reference voltage values ​​of the d-axis and q-axis The sector of the spatial voltage vector hexagon where the reference voltage vector is located is identified, and then based on the overmodulation strategy with vertical optimization, the duty cycle of the two adjacent voltage vectors and the zero vector in the sector is used to obtain the switching control signal to achieve control of the permanent magnet synchronous motor.

[0121] Please refer to the above-mentioned methods for the specific implementation process of each module, and will not be elaborated here. It should be understood that the above-mentioned division of functional modules is merely a division of logical functions. In actual implementation, other division methods can be used. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not implemented. At the same time, the above-mentioned integrated units can be implemented in the form of hardware or software functional units.

[0122] Example 3:

[0123] This embodiment provides a system based on the permanent magnet synchronous motor model-free predictive control method, which includes a permanent magnet synchronous motor and a control subsystem. The control subsystem uses the permanent magnet synchronous motor model-free predictive control method to generate a switching control signal for a two-level inverter, thereby controlling the permanent magnet synchronous motor.

[0124] The formation of the control subsystem may be achieved by referring to the method described in Example 2, or a control terminal device storing a program corresponding to the permanent magnet synchronous motor model-free predictive control method may be used to generate a switch control signal.

[0125] Example 4:

[0126] The present invention provides a computer-readable storage medium storing a computer program, wherein the computer program is called by a processor to implement:

[0127] The three-phase current and voltage of the permanent magnet synchronous motor are sampled separately during the sampling period, and the coordinate transformation is performed to obtain the sampling current i of the d-axis and q-axis in the two-phase dq coordinate system. d (k), i q (k) and the sampling voltage u d (k),u q (k);

[0128] The sampling current i of the d-axis and q-axis d (k), i q (k) and the sampling voltage u of the d-axis and q-axis d (k),u q (k) Input the sliding mode observer built based on the super local model to observe the stator current value at the next moment and concentrated disturbances

[0129] The stator current value and concentrated disturbances Substitute the hyperlocal model after Euler discretization to obtain the reference voltage values ​​of the d-axis and q-axis

[0130] Using the reference voltage values ​​of the d-axis and q-axis The sector of the spatial voltage vector hexagon where the reference voltage vector is located is identified, and then based on an overmodulation strategy with vertical optimization, the duty ratios of the two adjacent voltage vectors and the zero vector in the sector are used to obtain the switching control signal for controlling the permanent magnet synchronous motor.

[0131] It should be understood that the implementation process of some steps and whether to execute some steps and the execution order can refer to the implementation process of the above-mentioned embodiment.

[0132] The readable storage medium is a computer-readable storage medium, which can be the internal storage unit of the controller described in any of the aforementioned embodiments, such as the hard disk or memory of the controller. For example, the terrain element model constructed in the present invention exists in the hard disk, and then the computer program that performs the fusion step is stored in the memory, so that the fusion process is implemented based on the memory. The readable storage medium can also be an external storage device of the controller, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. equipped on the controller. Furthermore, the readable storage medium can also include both the internal storage unit of the controller and an external storage device. The readable storage medium is used to store the computer program and other programs and data required by the controller. The readable storage medium can also be used to temporarily store data that has been output or is to be output.

[0133] Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present invention. The aforementioned readable storage medium includes various media capable of storing program code, such as a USB flash drive, a mobile hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0134] It should be emphasized that the examples described in the present invention are illustrative rather than restrictive. Therefore, the present invention is not limited to the examples described in the specific embodiments. Any other embodiments derived by those skilled in the art based on the technical solution of the present invention that do not depart from the purpose and scope of the present invention, whether modified or replaced, also fall within the scope of protection of the present invention.

Claims

1. A robust model-free predictive control method for a permanent magnet synchronous motor, characterized by: The following steps are involved: Step 1: Sample the three-phase current and voltage of the permanent magnet synchronous motor separately during the sampling period, and perform coordinate transformation to obtain the sampling current i of the d-axis and q-axis in the two-phase dq coordinate system. d (k), i q (k) and the sampling voltage u d (k),u q (k); Step 2: The sampling current i of the d-axis and q-axis d (k), i q (k) and the sampling voltage u of the d-axis and q-axis d (k),u q (k) Input the sliding mode observer built based on the hyperlocal model to observe the stator current value and concentrated disturbance in the dq coordinate system at the next moment; Step 3: Substitute the stator current value and the concentrated disturbance into the hyperlocal model after Euler discretization to obtain the reference voltage values ​​of the d-axis and q-axis Step 4: Use the reference voltage values ​​of the d-axis and q-axis Identify the sector of the spatial voltage vector hexagon where the reference voltage vector is located. Then, based on an overmodulation strategy with vertical optimization, use the duty cycle of the two adjacent voltage vectors and the zero vector in the sector to obtain the switching control signal of the inverter to achieve control of the permanent magnet synchronous motor. The content of the vertical optimization overmodulation strategy includes: Using the duty ratios of the two adjacent voltage vectors and the zero vector, a modulation region where a reference voltage vector is located is identified, wherein the space voltage vector region is divided into a linear modulation region, an overmodulation region I, and an overmodulation region II; If the reference voltage vector is located in the linear modulation region, a switch control signal is generated by modulating the duty cycle of the two adjacent voltage vectors and the zero vector; If the reference voltage vector is located in the overmodulation region I, a new reference voltage vector synthesized by the two adjacent voltage vectors is determined on the boundary line of the space voltage vector using a vertical method, and then the duty cycle and modulation are updated to generate a switch control signal; If the reference voltage vector is located in the overmodulation zone II, the voltage vector closest to the two adjacent voltage vectors is used as a new reference voltage vector, and the duty cycle and modulation are updated to generate a switch control signal; If the reference voltage vector is in the overmodulation region I, the duty cycle is updated according to the following formula: If the reference voltage vector is in the overmodulation region II, the duty cycle is updated according to the following formula: Where d1, d2, and d0 represent the updated duty ratios of the two adjacent voltage vectors and the zero vector, respectively. represents the reference voltage vector, and u1 and u2 are the two adjacent voltage vectors.

2. The model-free predictive control method for a permanent magnet synchronous motor according to claim 1, wherein: When the space voltage vector region is divided into the linear modulation region, the overmodulation region I and the overmodulation region II, the corresponding division rules are: If d1+d2≤1, the reference voltage vector is located in the linear modulation region; If d1+d2>1 and |d1-d2|<1, the reference voltage vector is located in the overmodulation region I; If d1+d2>1 and |d1-d2|>1, the reference voltage vector is located in the overmodulation II region; Wherein, d1 and d2 represent the duty ratios of the two adjacent voltage vectors respectively.

3. The model-free predictive control method for a permanent magnet synchronous motor according to claim 1, wherein: The sliding mode observer is a sliding mode observer based on an adaptive exponential approach rate, and the control law in the sliding mode observer is: Where U dsmo and U qsmo represents the sliding mode control law, is the adaptive coefficient, k and δ are coefficients greater than 0, ε is a coefficient greater than 0 and less than 1, and λ is the set proportional factor; s is the sliding surface, including the sliding parameter s corresponding to the dq axis direction d 、s q , and is defined as the error between the stator current observation value and the stator current sampling value under the dq axis, specifically expressed as: in, is the stator current observation value of the dq axis, i d 、i q is the stator current sampling value of the dq axis, e d 、e q They are respectively expressed as the error between the stator current observation value and the stator current sampling value under the dq axis, and Z(e) is a saturation function that satisfies:

4. The model-free predictive control method for a permanent magnet synchronous motor according to claim 1, wherein: The observation formula of the sliding mode observer is expressed as: Among them, T s represents the sampling period, U dsmo and U qsmo represents the sliding mode control law of the sliding mode observer, g d and g q Both are gain coefficients of the control law, which are positive real numbers; is the stator current value of the d-axis and q-axis in the dq coordinate system at the next moment; It is the concentrated disturbance of the d-axis and q-axis in the dq coordinate system at the next moment.

5. The model-free predictive control method for a permanent magnet synchronous motor according to claim 1, wherein: Perform Euler discretization on the hyperlocal model to obtain the reference voltage values ​​of the discretized model corresponding to the d-axis and q-axis The formula is as follows: Among them, T s represents the sampling period, They represent the current reference values ​​of the q-axis and d-axis respectively; α is the model input gain.

6. A control system based on the permanent magnet synchronous motor model-free predictive control method according to any one of claims 1 to 5, characterized in that: At least: The sampling module is used to sample the three-phase current and voltage of the permanent magnet synchronous motor during the sampling period, and perform coordinate transformation to obtain the sampling current i of the d-axis and q-axis in the two-phase dq coordinate system. d (k), i q (k) and the sampling voltage u d (k),u q (k); Sliding mode observer module, used to use the d-axis and q-axis sampling current i d (k), i q (k) and the sampling voltage u of the d-axis and q-axis d (k),u q (k) Observe the stator current value and concentrated disturbance in the dq coordinate system at the next moment; The hyperlocal model prediction module is used to substitute the stator current value and the centralized disturbance into the hyperlocal model after Euler discretization to obtain the reference voltage values ​​of the d-axis and q-axis. A space vector modulation module is used to utilize the reference voltage values ​​of the d-axis and q-axis The sector of the spatial voltage vector hexagon where the reference voltage vector is located is identified. Then, based on an overmodulation strategy with vertical optimization, the duty cycle of the two adjacent voltage vectors and the zero vector in the sector is used to obtain the switching control signal of the inverter to achieve control of the permanent magnet synchronous motor.

7. A system based on the permanent magnet synchronous motor model-free predictive control method according to any one of claims 1 to 5, characterized in that: The invention comprises a permanent magnet synchronous motor and a control subsystem. The control subsystem adopts the permanent magnet synchronous motor model-free predictive control method to generate a switching control signal of an inverter, thereby controlling the permanent magnet synchronous motor.

8. A computer-readable storage medium, characterized in that: A computer program is stored, and the computer program is called by a processor to implement the permanent magnet synchronous motor model-free predictive control method according to any one of claims 1 to 5, specifically performing: The three-phase current and voltage of the permanent magnet synchronous motor are sampled separately during the sampling period, and the coordinate transformation is performed to obtain the sampling current i of the d-axis and q-axis in the two-phase dq coordinate system. d (k), i q (k) and the sampling voltage u d (k),u q (k); The sampling current i of the d-axis and q-axis d (k), i q (k) and the sampling voltage u of the d-axis and q-axis d (k),u q (k) Input the sliding mode observer built based on the hyperlocal model to observe the stator current value and concentrated disturbance in the dq coordinate system at the next moment; Substitute the stator current value and the concentrated disturbance into the hyperlocal model after Euler discretization to obtain the reference voltage values ​​of the d-axis and q-axis Using the reference voltage values ​​of the d-axis and q-axis The sector of the spatial voltage vector hexagon where the reference voltage vector is located is identified. Then, based on an overmodulation strategy with vertical optimization, the duty ratios of the two adjacent voltage vectors and the zero vector in the sector are used to obtain the switching control signal of the inverter for controlling the permanent magnet synchronous motor.

Images