A robust predictive control method and system for permanent magnet synchronous fan
By using a multi-layer expanded state observer structure in a permanent magnet synchronous generator for series and parallel connection, the problem of MPC insufficient parameter mismatch and noise robustness is solved, and stronger parameter robustness and noise suppression are achieved, improving The dynamic and steady-state performance of wind power systems is achieved.
Patent Information
- Application Number
- CN202410134314.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-01-30
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2044-01-30
AI Technical Summary
Model Predictive Control (MPC) is not robust to parameter mismatches and sensor noise in permanent magnet synchronous generators, resulting in problems with steady-state errors and current ripple.
Multiple expansion state observers (ESOs) are used for serial connection, combining system disturbances and observation weighting to form a multi-layer structure, which is used to estimate system disturbances and optimize control targets, and improve parameter robustness and noise suppression capabilities.
Without increasing the computing burden, the dynamic and steady-state performance is significantly improved, high-frequency noise interference is effectively reduced, and the robustness and efficiency of wind power generation systems are improved.
Smart Images

Figure CN117997189B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of wind turbine predictive control, and in particular relates to a method and system for robust predictive control of a permanent magnet synchronous wind turbine. Background Art
[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.
[0003] Permanent magnet synchronous generator (PMSG) uses permanent magnet excitation instead of excitation winding excitation, making the motor structure simpler. It has the advantages of high operating reliability, wide working wind speed range, high energy conversion efficiency, and simple maintenance. It has become the mainstream of offshore wind power. MPC can compensate for the uncertainty caused by model mismatch, distortion, interference, etc. in time through continuous rolling local optimization. It has good dynamic response ability, multi-objective optimization ability and adaptability to various working conditions. In recent years, it has made rapid progress in theoretical derivation and practical application, and has been widely used in the control of PMSG back-to-back converters. However, model predictive control (MPC) has the disadvantage of being sensitive to noise and parameter changes. In actual PMSG, there are deviations in parameters such as permanent magnet flux, stator inductance and filter inductance, and parameter measurement will generate noise, which will lead to the deterioration of MPC control effect and cause steady-state error and current ripple. Therefore, improving the robustness of MPC to parameter mismatch and sensor noise has become an important research goal.
[0004] The model predictive control (MPC-ESO) framework based on the extended state observer adds the extended state observer (ESO) to the traditional MPC. Its implementation scheme: In MPC control, the continuous time model on the grid side is:
[0005]
[0006] Where x is the state variable, u is the control variable, F is the total disturbance of the system, and y is the output variable. If there is noise N in the sensor detecting y, the robustness of MPC needs to consider the error caused by parameter mismatch. And sensor noise pollution causes errors If the predicted value of the state variable in the normal undisturbed state of MPC is x p , then the total prediction error Δx p becomes:
[0007]
[0008] In order to make the system robust to parameter mismatch and noise, it is necessary to make Δx p is 0.
[0009] After adding ESO, the system control block diagram is as follows: Figure 1 As shown. At this time, the variable estimation model of the ESO system robust control is:
[0010]
[0011] in, y is the output variable of the system under test, N is the measurement noise of the sensor, and α is the non-zero input gain. is the estimated value of {x,F}, {γ1,γ2} is the ESO error gain of the estimated and measured state variables, and its characteristic polynomial is s 2 +γ1s+γ2. The parameters of the ESO error gain are selected according to the characteristic polynomial. The relationship between the controller bandwidth and the parameters γ1, γ2 is γ1=2ω0, Where -ω0 is the pole of ESO. The control framework of the generator side and grid side of this control scheme are as follows: Figure 2 , Figure 3 shown.
[0012] The above MPC-ESO method ensures that accurate power tracking can be achieved well in the presence of current measurement noise and parameter mismatch. However, ESO cannot balance noise suppression and parameter mismatch. This is because ESO requires a larger α to improve its robustness to parameter mismatch, but it will generate high-frequency noise interference, resulting in high-order harmonics and deteriorating the steady-state performance of the system; however, if α is reduced to attenuate noise, the transient performance will be poor and the robustness to parameter mismatch will be reduced. This contradiction causes MPC-ESO to have current ripple, torque offset, power deviation and other problems in the case of parameter mismatch, and its robustness needs to be improved. Summary of the invention
[0013] In order to solve the above problems, the present invention proposes a robust predictive control method and system for a permanent magnet synchronous wind turbine. The present invention connects a plurality of extended state observers (ESOs) in series and parallel, and weights the system disturbance and the observation value. The method has stronger parameter robustness, can more effectively balance parameter mismatch, interference suppression and high-frequency noise suppression, has good dynamic and steady-state performance, and has great application significance for improving the performance of wind power generation systems under actual working conditions.
[0014] According to some embodiments, the present invention adopts the following technical solutions:
[0015] A method for robust predictive control of a permanent magnet synchronous fan comprises the following steps:
[0016] In the control link on the motor side and / or the grid side, a model predictive control method based on an extended state observer is used for control, and in the control process, a multi-layer structure with multiple stages in series, each stage containing several extended state observers in parallel, is used to predict the current and voltage and estimate the system disturbance.
[0017] As an optional implementation, a PI speed outer loop is used on the motor side, and a model predictive control inner loop based on an extended state observer is used. The control objectives of the inner loop include stator current control and machine-side converter switching frequency control.
[0018] As an optional implementation, on the grid side, a DC voltage outer loop is used, and a model predictive control inner loop based on an extended state observer is used. For the predictive control of the inner loop, the control objectives include grid current control, neutral point voltage difference control and grid-side converter switching frequency control.
[0019] As an optional implementation, the multi-layer structure contains M sub-expanded state observers, each sub-expanded state observer has the same sub-frequency, each stage is a cascade of two sub-expanded state observers, each sub-expanded state observer estimates at the sub-frequency based on the state variable, and the system disturbance is the sum of the estimated values of all sub-expanded state observers.
[0020] As an optional implementation, in the control process on the motor side, the penalty term of the current control in the control target is the stator current i in the dq coordinate system. d and i q ,i q The reference is obtained by the torque error, and the control i q , control the torque, and then control the speed to reach the reference speed, i d According to the maximum torque current control setting;
[0021] A penalty term for switching actions is also added to the cost function to reduce the switching frequency as much as possible while ensuring system performance.
[0022] As a further feature, the current control has a higher priority than the switching operation frequency.
[0023] As an optional implementation, in the control process on the motor side, the switching vector with the smallest cost function is used as the optimal switching vector, and the motor side converter is controlled with the optimal switching vector in the next control cycle.
[0024] As an optional implementation, in the control process on the grid side, the current control penalty term is the grid side current i in the dq coordinate system d and i q ,i d The reference is obtained by the DC bus voltage control outer loop, by controlling i d, to realize the control of DC bus voltage, i q The reference value is the set value;
[0025] The penalty term for switching frequency control is the switching frequency of the grid-side converter;
[0026] The penalty term for capacitor neutral point voltage control is the capacitor neutral point voltage unbalance of the back-to-back converter.
[0027] As an optional implementation, the switching vector with the smallest cost function is selected as the optimal switching vector, and the selected switching vector is output in the next control cycle to control the grid-side converter.
[0028] A robust predictive control system for a permanent magnet synchronous fan, comprising:
[0029] The hybrid multi-level model prediction module is configured to use a model prediction control method based on an extended state observer to perform control in the control link on the motor side or / and the grid side, and in the control process, a multi-layer structure with multiple stages in series, each stage containing several extended state observers in parallel, is used to predict based on current and voltage and estimate system disturbances.
[0030] Compared with the prior art, the present invention has the following beneficial effects:
[0031] The present invention takes the problem that MPC is difficult to be robust to parameter mismatch and measurement noise in PMSG control as the starting point, proposes a hybrid cascade parallel extended state observer model-free predictive control method, and solves the problem that only the ESO series structure has poor robustness to parameter mismatch by connecting multiple extended state observers (ESO) in series and parallel. On the basis of the ESO parallel structure, series connection is performed, and system disturbances and observation values are weighted, which has stronger parameter robustness, can more effectively balance parameter mismatch, interference suppression and high-frequency noise suppression, has good dynamic and steady-state performance, and has great application significance for improving the performance of wind power generation systems under actual working conditions.
[0032] The present invention has better parameter and noise robustness without sacrificing computational burden or efficiency. It has excellent dynamic and steady-state performance in the case of stator inductance and grid-side inductance parameter mismatch, can effectively attenuate high-frequency noise, and has improved performance compared with MPC and MPC-ESO.
[0033] The present invention has a wide range of applications and can be extended to scenarios such as double-fed asynchronous wind power generation systems and four-quadrant motor drives.
[0034] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS
[0035] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.
[0036] Figure 1 It is the control block diagram of MPC-ESO;
[0037] Figure 2 It is a motor side control block diagram of MPC-ESO and the method of this embodiment;
[0038] Figure 3 : is a grid-side control block diagram of MPC-ESO and the method of this embodiment;
[0039] Figure 4 is a control block diagram of the MFPPC-CPESO of this embodiment;
[0040] Figure 5 This is a CPESO example diagram with M=3 in this embodiment. DETAILED DESCRIPTION
[0041] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.
[0042] It should be noted that the following detailed descriptions are all illustrative and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.
[0043] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.
[0044] In the absence of conflict, the embodiments in this application and the features in the embodiments may be combined with each other.
[0045] Embodiment 1
[0046] A method for robust predictive control of a permanent magnet synchronous wind turbine includes control on a motor side and a grid side.
[0047] The wind turbine side uses the PI speed outer loop and predictive control inner loop structure. For the predictive control of the inner loop, the control targets include stator current control and machine-side converter switching frequency control. The grid side uses the DC voltage outer loop and predictive control inner loop structure. For the predictive control of the inner loop, the control targets include grid current control, neutral point voltage difference control and grid-side converter switching frequency control. The control methods of the motor side and the grid side are described below:
[0048] The control target on the motor side is the stator current error and the switching frequency of the machine-side converter. The control block diagram is as follows: Figure 2 As shown (this embodiment adopts the CP-ESO control method), the control steps are described as follows:
[0049] Step 1: Sensor samples motor side current The motor stator flux angle θ and the motor speed ω m .
[0050] Step 2: Convert the motor side current into the stator current dq axis component through park transformation
[0051] Step 3: Set the electric speed ω e and reference speed Compare and send to PI controller to get q-axis current References and will Set to 0.
[0052] Step 4: ESO predicts the current and voltage at time k to obtain the value of the dq-axis component of the motor-side current at time k+1 and the value of the dq-axis component of the system disturbance at time k+1:
[0053]
[0054] Step 5: Based on the value of the motor side current at time k+1 and the dq axis component of the system disturbance, the value of the motor side current dq axis component at time k+2 is predicted:
[0055]
[0056] Step 6: Introduce CPESO and use multiple sub-ESOs to perturb the system Estimate and sum, also use multiple sub-ESOs to estimate Its structure is as Figure 4As shown. Compared with a single ESO, the cascade extended state observer (CESO) with a series structure has the advantage that it is no longer a nonlinear filter, so it can effectively consider noise suppression and dynamic performance. However, the robustness of CESO to parameter mismatch is poor, so it is considered to add a parallel structure and mix the cascade and parallel ESO to solve the problem of poor robustness to parameter mismatch, which has certain prospects in the robust control of power electronics.
[0057] The structure contains M sub-ESOs, each with a sub-frequency of ω 0j , divided into (M+1) / 2 levels of cascade, each level is two sub-ESO cascades. Its time domain model is:
[0058]
[0059] In this formula, the expansion state is the state variable x j At sub-frequency ω 0j The estimated value of the j-th sub-ESO of ESO gain is γ 1,j =2ω 0j and With pole -ω 0j The state variable is And the estimated total disturbance is
[0060] Taking M=3 as an example, the control block diagram is as follows Figure 5 At this time, and variables for:
[0061]
[0062] Step 7: Obtain stator current through CPESO with M=3 and system disturbances The observed value at time k+1 gives the two-step predicted stator current:
[0063]
[0064] Step 8: The penalty term for the control target is as follows:
[0065] The current control penalty term is the stator current i in the dq coordinate system d and The reference is obtained by the torque error, and the control i q , the torque can be controlled, and then the speed can be controlled to reach the reference speed. d The maximum torque current control is set to 0. These two control objectives are combined into one item with the highest priority, and its cost function is:
[0066]
[0067] The switching frequency of the motor-side converter. In high-power wind power systems, low switching frequency can reduce power loss, thereby reducing the heat dissipation requirements of the equipment. Therefore, by adding a penalty term for the switching action in the cost function, the switching frequency can be reduced as much as possible while ensuring system performance. For this purpose, the priority control target is set, and the cost function is:
[0068]
[0069] Step 9: Select the switching vector with the smallest controller cost function as the optimal switching vector, and output the selected optimal switching vector in the next control cycle to control the motor-side converter.
[0070] Embodiment 2
[0071] This embodiment describes the specific steps of the grid-side control. The control targets are the grid-side current, neutral point voltage deviation and switching frequency of the grid-side converter. The control block diagram is as follows: Figure 3 As shown (this embodiment adopts the CP-ESO control method). The control steps are as follows:
[0072] Step 1: Sensors collect grid-side voltage and current And the DC bus voltage V dc and its reference value and the voltage value of the two capacitors V C1 、V C2 .
[0073] Step 2: Transform the grid-side current and voltage into the dq-axis coordinate system through park transformation to obtain and
[0074] Step 3: Compare the DC bus voltage with the reference and send it to the PI controller to get the reference d-axis current reference Will Set to 0.
[0075] Step 4: ESO predicts the current based on the existing dq-axis components of the current and voltage, and obtains the value of the dq-axis component of the grid-side current at time k+1, and the value of the dq-axis component of the system disturbance at time k+1:
[0076]
[0077] Step 5: Based on the values of the grid-side current and the dq-axis components of the system disturbance at time k+1, the value of the dq-axis components of the grid-side current at time k+2 is predicted:
[0078]
[0079] Step 6: Obtain grid current through CPESO with M=3 and system disturbances The observed value at time k+1, the two-step predicted grid current is:
[0080]
[0081] Step 7: Predict various penalty item parameters under different switching vector states of the converter in the next control cycle. The penalty items of the control target are as follows.
[0082] The current controller penalty term is the grid-side current i in the dq coordinate system d and i q .i d The reference is obtained by the DC bus voltage control outer loop, by controlling i d , the DC bus voltage can be controlled. q The reference value is set to 0 to ensure that the power factor on the grid side is 1. These two control objectives are combined into one item with the highest priority, and its cost function is:
[0083]
[0084] The penalty term of the switching frequency controller is the switching frequency of the grid-side converter. In high-power wind power systems, low switching frequency can reduce power loss, thereby reducing the heat dissipation requirements of the equipment. Therefore, by adding a penalty term for the switching action to the cost function, the switching frequency can be reduced as much as possible while ensuring system performance. For this purpose, the priority control target is set, and the cost function of the second-stage control target is:
[0085]
[0086] The penalty term of the capacitor neutral point voltage controller is the capacitor neutral point voltage imbalance of the back-to-back converter. For the neutral point clamped three-level converter, under normal working conditions, the voltages of the upper and lower voltage-equalizing capacitors C1 and C2 should be equal, and the maximum voltage borne by the switch tubes of the upper and lower bridge arms is the DC bus voltage V dc However, under certain working conditions, the neutral point voltage will deviate, resulting in output voltage waveform distortion. Severe deviation may cause the switch tube to break down. Therefore, a control algorithm is needed to ensure capacitor voltage balance. The cost function is:
[0087] J V =(V c1 -V c2 ) 2 (16)
[0088] Step 8: Select the switching vector with the smallest controller cost function as the optimal switching vector, and output the selected switching vector in the next control cycle to control the grid-side converter.
[0089] Embodiment 2
[0090] A robust predictive control system for a permanent magnet synchronous fan, comprising:
[0091] The hybrid multi-level model prediction module is configured to use a model prediction control method based on an extended state observer to perform control in the control link on the motor side or / and the grid side, and in the control process, a multi-layer structure with multiple stages in series, each stage containing several extended state observers in parallel, is used to predict based on current and voltage and estimate system disturbances.
[0092] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0093] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.
[0094] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.
[0095] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.
[0096] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made by those skilled in the art within the spirit and principle of the present invention without creative labor shall be included in the protection scope of the present invention.
Claims
1. A method for robust predictive control of a permanent magnet synchronous fan, characterized in that: The following steps are involved: In the control link on the motor side or / and the grid side, a model predictive control method based on an extended state observer is used for control, and in the control process, a multi-layer structure of multiple stages of extended state observers connected in series, each stage including a plurality of parallel extended state observers, is used to predict and estimate system disturbances based on current and voltage; The control target on the motor side is the stator current error and the switching frequency of the machine-side converter. The control steps are: Step 11: Sensor sampling motor side current Motor stator flux angle θ And the motor speed on the motor side ωm ; Step 12: Convert the motor side current into the stator current dq axis component through park transformation Step 13: Set the electric speed ω e and reference speed Compare and send to PI controller to get q-axis current References and will Set to 0; Step 14: ESO predicts the current and voltage at time k to obtain the value of the dq-axis component of the motor side current at time k+1 and the value of the dq-axis component of the system disturbance at time k+1: in, is the value of the dq axis component of the motor side current, k is the k moment, k+1 is the k+1 moment, is the value of the dq-axis component of the system disturbance, γ1 and γ2 are the ESO error gains of the estimated and measured state variables; Step 15: Based on the value of the motor side current at time k+1 and the dq axis component of the system disturbance, the value of the motor side current dq axis component at time k+2 is predicted: Among them, k+2 is the k+2 moment; Step 16: Introduce CPESO and use multiple sub-ESOs to perturb the system Estimate and sum, use multiple sub-ESOs to estimate Step 17: Passing Stator Current and system disturbances The observed value at time k+1 gives the two-step predicted stator current; Step 18: The penalty term for the control target is as follows: The penalty term of current control is the stator current i in the dq coordinate system d and i q ,i q The reference is obtained by the torque error, and the control i q , control the torque, and then control the speed to reach the reference speed, i d According to the maximum torque current control setting; i d and i q The two control objectives are combined into one item with the highest priority, and its cost function is: in, is the cost function of the current control penalty term on the motor side; The penalty term for the switch action is also added to the cost function, and the cost function is: in, is the cost function of the penalty term of the switching action of the motor side converter; Step 19: Select the switch vector with the smallest controller cost function as the optimal switch vector, and use the selected optimal switch vector in the next control cycle to control the motor-side converter; The control targets on the grid side are the grid-side current, neutral point voltage deviation and switching frequency of the grid-side converter. The control steps are: Step 21: Sensors collect grid-side voltage and current And the DC bus voltage V dc and its reference value and the voltage value of the two capacitors V C1 、V C2 ; Step 22: Convert the grid-side current and voltage to the dq-axis coordinate system through park transformation to obtain and Step 23: Compare the DC bus voltage with the reference and send it to the PI controller to obtain the reference d-axis current reference Will Set to 0; Step 24: The ESO predicts the current based on the existing dq-axis components of the current and voltage, and obtains the value of the dq-axis component of the grid-side current at time k+1, and the value of the dq-axis component of the system disturbance at time k+1: in, is the value of the dq-axis component of the grid-side current, is the value of the dq axis component of the system disturbance; Step 25: Based on the grid-side current at time k+1 and the value of the dq-axis component of the system disturbance, the value of the dq-axis component of the grid-side current at time k+2 is predicted: Step 26: Passing Grid Current and system disturbances The observation value at time k+1 is used to obtain the two-step predicted grid current; Step 27: Predict various penalty item parameters of the converter under different switching vector states in the next control period; the penalty item of the control target is as follows: The current control penalty term is the grid-side current i in the dq coordinate system d and i q ,i d The reference is obtained by the DC bus voltage control outer loop, by controlling i d , to realize the control of DC bus voltage, i q The reference value is the set value; i d and i q The two control objectives are combined into one item with the highest priority, and its cost function is: in, is the cost function of the current control penalty term on the grid side; The penalty term of switching frequency control is the switching frequency of the grid-side converter, and the cost function is: in, is the cost function of the switching frequency of the grid-side converter; The penalty term of capacitor neutral point voltage control is the capacitor neutral point voltage imbalance of the back-to-back converter, and its cost function is: J V =(V c1 -V c2 ) 2 ; Among them, J V is the cost function of the penalty term of capacitor neutral point voltage control; Step 28: Select the switching vector with the smallest controller cost function as the optimal switching vector, and output the selected switching vector in the next control cycle to control the grid-side converter.
2. A method for robust predictive control of a permanent magnet synchronous wind turbine as claimed in claim 1, characterized in that: The PI speed outer loop is used on the motor side, and the model predictive control inner loop based on the extended state observer is used. The control objectives of the inner loop include stator current control and machine-side converter switching frequency control.
3. A method for robust predictive control of a permanent magnet synchronous fan according to claim 1, characterized in that On the grid side, a DC voltage outer loop is used, and a model predictive control inner loop based on an extended state observer is used. For the predictive control of the inner loop, the control objectives include grid current control, neutral point voltage difference control, and grid-side converter switching frequency control.
4. A method for robust predictive control of a permanent magnet synchronous wind turbine as claimed in claim 1, characterized in that: The multi-layer structure contains M sub-expanded state observers, each sub-expanded state observer has the same sub-frequency, each level is a cascade of two sub-expanded state observers, each sub-expanded state observer estimates at the sub-frequency according to the state variable, and the system disturbance is the sum of the estimated values of all sub-expanded state observers.
5. A robust predictive control system for a permanent magnet synchronous fan, characterized in that: include: The hybrid multi-level model prediction module is configured to use a model prediction control method based on an extended state observer to perform control in the control link of the motor side or / and the grid side, and in the control process, a multi-layer structure of multiple stages of extended state observers connected in series, each stage including a plurality of parallel extended state observers, is used to predict according to current and voltage and estimate system disturbances; The control target on the motor side is the stator current error and the switching frequency of the machine-side converter. The control steps are: Step 11: Sensor sampling motor side current Motor stator flux angle θ And the motor speed on the motor side ωm ; Step 12: Convert the motor side current into the stator current dq axis component through park transformation Step 13: Set the electric speed ω e and reference speed Compare and send to PI controller to get q-axis current References and will Set to 0; Step 14: ESO predicts the current and voltage at time k to obtain the value of the dq-axis component of the motor side current at time k+1 and the value of the dq-axis component of the system disturbance at time k+1: in, is the value of the dq axis component of the motor side current, k is the k moment, k+1 is the k+1 moment, is the value of the dq-axis component of the system disturbance, γ1 and γ2 are the ESO error gains of the estimated and measured state variables; Step 15: Based on the value of the motor side current at time k+1 and the dq axis component of the system disturbance, the value of the motor side current dq axis component at time k+2 is predicted: Among them, k+2 is the k+2 moment; Step 16: Introduce CPESO and use multiple sub-ESOs to perturb the system Estimate and sum, use multiple sub-ESOs to estimate Step 17: Passing Stator Current and system disturbances The observed value at time k+1 gives the two-step predicted stator current; Step 18: The penalty term for the control target is as follows: The penalty term of current control is the stator current i in the dq coordinate system d and i q ,i q The reference is obtained by the torque error, and the control i q , control the torque, and then control the speed to reach the reference speed, i d According to the maximum torque current control setting; i d and i q The two control objectives are combined into one item with the highest priority, and its cost function is: in, is the cost function of the current control penalty term on the motor side; The penalty term for the switch action is also added to the cost function, and the cost function is: in, is the cost function of the penalty term of the switching action of the motor side converter; Step 19: Select the switch vector with the smallest controller cost function as the optimal switch vector, and use the selected optimal switch vector in the next control cycle to control the motor-side converter; The control targets on the grid side are the grid-side current, neutral point voltage deviation and switching frequency of the grid-side converter. The control steps are: Step 21: Sensors collect grid-side voltage and current And the DC bus voltage V dc and its reference value and the voltage value of the two capacitors V C1 、V C2 ; Step 22: Convert the grid-side current and voltage to the dq-axis coordinate system through park transformation to obtain and Step 23: Compare the DC bus voltage with the reference and send it to the PI controller to obtain the reference d-axis current reference Will Set to 0; Step 24: The ESO predicts the current based on the existing dq-axis components of the current and voltage, and obtains the value of the dq-axis component of the grid-side current at time k+1, and the value of the dq-axis component of the system disturbance at time k+1: in, is the value of the dq-axis component of the grid-side current, is the value of the dq axis component of the system disturbance; Step 25: Based on the grid-side current at time k+1 and the value of the dq-axis component of the system disturbance, the value of the dq-axis component of the grid-side current at time k+2 is predicted: Step 26: Passing Grid Current and system disturbances The observed value at time k+1 is used to obtain the two-step predicted grid current; Step 27: Predict various penalty item parameters of the converter under different switching vector states in the next control period; the penalty item of the control target is as follows: The current control penalty term is the grid-side current i in the dq coordinate system d and i q ,i d The reference is obtained by the DC bus voltage control outer loop, by controlling i d , to realize the control of DC bus voltage, i q The reference value is the set value; i d and i q The two control objectives are combined into one item with the highest priority, and its cost function is: in, is the cost function of the current control penalty term on the grid side; The penalty term of switching frequency control is the switching frequency of the grid-side converter, and the cost function is: in, is the cost function of the switching frequency of the grid-side converter; The penalty term of capacitor neutral point voltage control is the capacitor neutral point voltage imbalance of the back-to-back converter, and its cost function is: J V =(V c1 -V c2 ) 2 ; Among them, J V is the cost function of the penalty term of capacitor neutral point voltage control; Step 28: Select the switching vector with the smallest controller cost function as the optimal switching vector, and output the selected switching vector in the next control cycle to control the grid-side converter.
Citation Information
Patent Citations
Permanent magnet synchronous motor high-reliability current predictive control method and system thereof
CN109660170A
Model prediction current control method of permanent magnet synchronous motor
CN111600523A
Converter cascade predictive control method and system based on precise discretization
CN113675888A
Grid-connected converter prediction control method and system based on structure self-adaptive ESO
CN115276094A
Model-free predictive control method and system for grid-connected converter
CN115425642A
