Torque control method and system for surface-mounted permanent magnet synchronous motor for electric propulsion

Abstract

The application provides a torque control method and system of a surface-mounted permanent magnet synchronous motor for electric propulsion, and relates to the technical field of motor drive control. The method comprises the following steps: obtaining a torque reference value; obtaining stator flux linkage observation values by using a nonlinear flux linkage observer based on iteration learning based on the current, voltage and torque reference value of the permanent magnet synchronous motor; obtaining torque prediction values and stator flux linkage prediction values by using a hyper-local prediction model based on a composite discrete extended state observer based on the current, voltage, stator flux linkage observation values and candidate voltage vectors of the inverter of the permanent magnet synchronous motor; determining an optimal voltage vector from the candidate voltage vectors according to the torque reference value, a preset stator flux linkage reference value, the torque prediction value and the stator flux linkage prediction value, and controlling the torque of the permanent magnet synchronous motor according to the optimal voltage vector. The application solves the problem of large torque tracking error under parameter perturbation, and improves the system robustness and dynamic response capability.

Description

Technical Field

[0001] This invention relates to the field of motor drive control technology, and in particular to a torque control method and system for a surface-mounted permanent magnet synchronous motor for electric propulsion. Background Technology

[0002] In the field of motor drive control, permanent magnet synchronous motors for electric propulsion are widely used in marine electric propulsion systems due to their high torque-to-current ratio and high power-to-weight ratio. Ships operate in harsh sea conditions such as humid heat, salt spray, and deep-water high pressure, which can easily lead to a non-uniform temperature field distribution inside the motor, causing local demagnetization of the rotor's permanent magnets. This results in perturbations in inductance parameters and permanent magnet flux linkage parameters. Simultaneously, typical operating conditions such as variable frequency speed regulation and surge loads place high dynamic response requirements on the drive system.

[0003] Currently, torque control for permanent magnet synchronous motors used in electric propulsion mainly employs two schemes: First, the traditional proportional-integral (PI) controller, which offers good steady-state performance but suffers from slow dynamic response due to motor dq-axis current coupling and a fixed operating point, and is prone to system instability under parameter perturbations. Second, a robust model predictive control strategy based on hyperlocal predictive models, which acquires real-time motor parameters through an observer and compensates them for the predictive model to improve robustness. However, existing hyperlocal predictive models do not consider discretization truncation errors, resulting in insufficient model accuracy. Furthermore, the nonlinear flux observer, limited by its compensation mechanism, exhibits slow convergence under transient conditions, preventing the control system from quickly reducing torque tracking errors under parameter perturbations.

[0004] Therefore, how to effectively reduce torque tracking error and improve the dynamic response capability of the system under the condition of perturbation of inductance parameters and permanent magnet flux linkage parameters has become a technical problem that urgently needs to be solved in this field. Summary of the Invention

[0005] This invention provides a torque control method for a surface-mounted permanent magnet synchronous motor for electric propulsion, which solves the problems of large torque tracking error and insufficient dynamic response capability in the prior art when the inductance parameters and permanent magnet flux linkage parameters are perturbed.

[0006] This invention provides a torque control method for a surface-mounted permanent magnet synchronous motor (PMSM) for electric propulsion. The method includes: acquiring a torque reference value; obtaining a stator flux observation value using an iterative learning-based nonlinear flux observer based on the current, voltage, and torque reference values ​​of the PMSM; obtaining a torque prediction value and a stator flux prediction value using a hyperlocal prediction model based on a composite discrete extended state observer based on the current, voltage, and stator flux observation values ​​of the PMSM and the candidate voltage vector of the inverter; determining an optimal voltage vector from the candidate voltage vectors based on the torque reference value, the preset stator flux reference value, the torque prediction value, and the stator flux prediction value, and controlling the torque of the PMSM according to the optimal voltage vector.

[0007] According to the present invention, a torque control method for a surface-mounted permanent magnet synchronous motor for electric propulsion is provided, wherein obtaining a torque reference value includes: obtaining the rotor angular velocity of the permanent magnet synchronous motor; inputting the difference between a preset rotor angular velocity reference value and the rotor angular velocity into a PI controller to obtain the torque reference value.

[0008] The torque control method for a surface-mounted permanent magnet synchronous motor for electric propulsion provided by the present invention further includes, before obtaining the stator flux observation value using a nonlinear flux observer based on iterative learning: acquiring the three-phase stator current and rotor electrical angle of the permanent magnet synchronous motor; performing Park transformation on the three-phase stator current to obtain the α-axis current and β-axis current in the two-phase stationary coordinate system; and obtaining the optimal voltage vector determined in the previous control cycle to obtain the α-axis voltage and β-axis voltage in the two-phase stationary coordinate system.

[0009] The present invention provides a torque control method for a surface-mounted permanent magnet synchronous motor for electric propulsion, which uses a nonlinear flux linkage observer based on iterative learning to obtain stator flux linkage observation values, including: inputting α-axis current, β-axis current, α-axis voltage, β-axis voltage and torque reference values ​​into the nonlinear flux linkage observer to obtain α-axis stator flux linkage observation values ​​and β-axis stator flux linkage observation values.

[0010] According to the present invention, a torque control method for a surface-mounted permanent magnet synchronous motor for electric propulsion, comprising obtaining α-axis stator flux linkage observations and β-axis stator flux linkage observations, including: acquiring the α-axis stator flux linkage observations and β-axis stator flux linkage observations determined in the previous control cycle; calculating a torque estimate based on the α-axis stator flux linkage observations and β-axis stator flux linkage observations of the previous control cycle, the α-axis current and β-axis current of the current control cycle, and the number of pole pairs of the permanent magnet synchronous motor; calculating the torque tracking error between the torque reference value and the torque estimate; generating a flux linkage error compensation term related to the rotor electrical angle and motor inductance parameters using the torque tracking error; and superimposing the flux linkage error compensation term onto the basic flux linkage estimate calculated based on the α-axis voltage and β-axis voltage of the current control cycle to generate the α-axis stator flux linkage observations and β-axis stator flux linkage observations of the current control cycle.

[0011] According to the torque control method of a surface-mounted permanent magnet synchronous motor for electric propulsion provided by the present invention, the flux linkage error compensation term is generated by inputting the torque tracking error into a learning function matrix related to the rotor electrical angle and motor inductance parameters.

[0012] The torque control method for a surface-mounted permanent magnet synchronous motor for electric propulsion provided by the present invention further includes, before obtaining the torque prediction value and stator flux prediction value using a hyperlocal prediction model based on a composite discrete extended state observer, enumerating all switching state combinations of the two-level inverter connected to the permanent magnet synchronous motor to obtain the candidate voltage vector.

[0013] According to the present invention, a torque control method for a surface-mounted permanent magnet synchronous motor for electric propulsion is provided, wherein the selectable voltage vector is determined by eight switching state combinations of a two-level inverter, including six effective voltage vectors and two zero voltage vectors.

[0014] According to the present invention, a torque control method for a surface-mounted permanent magnet synchronous motor for electric propulsion is provided. The hyperlocal prediction model based on a composite discrete extended state observer is constructed through the following steps: A composite discretization method integrating second-order Taylor expansion, explicit Euler method, and implicit Euler method is used to discretize the continuous domain state equations of the permanent magnet synchronous motor, obtaining a composite discretized prediction model; the α-axis current, β-axis current, α-axis voltage, β-axis voltage, α-axis stator flux linkage observations, and β-axis stator flux linkage observations are input into the extended state observer to estimate the current disturbance and flux linkage disturbance estimates; the current disturbance estimates and flux linkage disturbance estimates are then compensated into the composite discretized prediction model to form a hyperlocal prediction model.

[0015] According to the present invention, a torque control method for a surface-mounted permanent magnet synchronous motor for electric propulsion estimates current disturbance estimates and flux linkage disturbance estimates are obtained by: comparing the α-axis current and β-axis current with the current prediction values ​​output by the composite discretized prediction model to obtain the current prediction error; comparing the observed values ​​of the α-axis stator flux linkage and β-axis stator flux linkage with the stator flux linkage prediction values ​​output by the composite discretized prediction model to obtain the flux linkage prediction error; and updating the current disturbance estimates online based on the current prediction error and updating the flux linkage disturbance estimates online based on the flux linkage prediction error.

[0016] The present invention provides a torque control method for a surface-mounted permanent magnet synchronous motor for electric propulsion, which uses a hyperlocal prediction model based on a composite discrete extended state observer to obtain torque prediction values ​​and stator flux prediction values. The method includes: for each candidate voltage vector, using the hyperlocal prediction model to output current prediction values ​​and stator flux prediction values ​​after two control cycles relative to the current time; and calculating torque prediction values ​​based on current prediction values ​​and stator flux prediction values.

[0017] According to the present invention, a torque control method for a surface-mounted permanent magnet synchronous motor for electric propulsion includes a current prediction value including an α-axis current prediction value and a β-axis current prediction value; a stator flux prediction value including an α-axis stator flux prediction value and a β-axis stator flux prediction value; and a torque prediction value calculated based on the α-axis current prediction value, the β-axis current prediction value, the α-axis stator flux prediction value, the β-axis stator flux prediction value, and the number of pole pairs of the permanent magnet synchronous motor.

[0018] The torque control method for a surface-mounted permanent magnet synchronous motor for electric propulsion provided by the present invention determines the optimal voltage vector from candidate voltage vectors, including: calculating the torque deviation between a torque reference value and a torque prediction value, and the flux deviation between a preset stator flux reference value and a stator flux prediction value; using preset weighting coefficients to perform a weighted summation of the torque deviation and flux deviation to obtain the cost function value corresponding to each candidate voltage vector; and determining the candidate voltage vector corresponding to the minimum cost function value as the optimal voltage vector.

[0019] According to the present invention, a torque control method for a surface-mounted permanent magnet synchronous motor for electric propulsion includes a stator flux prediction value comprising an α-axis stator flux prediction value and a β-axis stator flux prediction value. The calculation step for the flux deviation between a preset stator flux reference value and the predicted stator flux value includes: calculating a predicted stator flux amplitude value based on the α-axis and β-axis stator flux prediction values; and calculating the difference between the preset stator flux reference value and the predicted stator flux amplitude value to obtain the flux deviation.

[0020] This invention also provides a torque control system for a surface-mounted permanent magnet synchronous motor for electric propulsion. The system includes: an acquisition module for acquiring torque reference values; a flux linkage observation module for obtaining stator flux linkage observation values ​​based on the permanent magnet synchronous motor current, voltage, and torque reference values ​​using a nonlinear flux linkage observer based on iterative learning; a state prediction module for obtaining torque prediction values ​​and stator flux linkage prediction values ​​based on the permanent magnet synchronous motor current, voltage, stator flux linkage observation values, and candidate voltage vectors of the inverter using a hyperlocal prediction model based on a composite discrete extended state observer; and a voltage vector determination module for determining the optimal voltage vector from the candidate voltage vectors based on the torque reference values, preset stator flux linkage reference values, torque prediction values, and stator flux linkage prediction values, and controlling the torque of the permanent magnet synchronous motor according to the optimal voltage vector.

[0021] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the torque control method for a surface-mounted permanent magnet synchronous motor for electric propulsion as described in any of the preceding claims.

[0022] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the torque control method for a surface-mounted permanent magnet synchronous motor for electric propulsion as described in any of the preceding claims.

[0023] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the torque control method for a surface-mounted permanent magnet synchronous motor for electric propulsion as described in any of the preceding claims.

[0024] The torque control method and system for a surface-mounted permanent magnet synchronous motor for electric propulsion provided by this invention obtains a torque reference value to provide target commands for subsequent control; by obtaining stator flux observation values ​​based on motor current, voltage, and torque reference values ​​and utilizing a nonlinear flux observation based on iterative learning, the system can leverage torque tracking error to drive the observation to converge quickly, and can still accurately obtain stator flux information during parameter perturbations; and by using a hyperlocal prediction model based on a composite discrete extended state observation, the system utilizes motor current, voltage, stator flux observation values, and inverter candidate voltage vectors. By obtaining the predicted torque and stator flux linkage values, a composite discretization method is used to reduce the truncation error of the prediction model, and an extended state observer is used to estimate and compensate for parameter perturbations and unmodeled dynamics in real time, thereby improving the prediction accuracy. By determining the optimal voltage vector from the candidate voltage vectors based on the torque reference value, the preset stator flux linkage reference value, the predicted torque value, and the predicted stator flux linkage value, and controlling the torque of the permanent magnet synchronous motor accordingly, high-precision tracking of torque is achieved under the perturbation conditions of inductance parameters and permanent magnet flux linkage parameters, while taking into account the strong robustness and high dynamic response capability of the system. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0026] Figure 1 This is a control system diagram of a torque control method for a surface-mounted permanent magnet synchronous motor for electric propulsion provided in an embodiment of the present invention.

[0027] Figure 2This is a schematic flowchart of a torque control method for a surface-mounted permanent magnet synchronous motor for electric propulsion provided in an embodiment of the present invention.

[0028] Figure 3 In this embodiment of the invention, the motor parameters are... , Simulation results of torque tracking performance of traditional model predictive torque control method.

[0029] Figure 4 In this embodiment of the invention, the motor parameters are... , The simulation results of torque tracking performance of the torque predictive control method using the traditional nonlinear flux observer and the traditional hyperlocal predictive model are shown in the figure.

[0030] Figure 5 In this embodiment of the invention, the motor parameters are... , The simulation results of torque tracking performance of the torque predictive control method using a nonlinear flux observer based on iterative control and a hyperlocal predictive model based on a composite discrete extended state observer are shown in the figure.

[0031] Figure 6 In this embodiment of the invention, the motor parameter is 1.2. 0.8 Simulation results of torque tracking performance of traditional model predictive torque control method.

[0032] Figure 7 In this embodiment of the invention, the motor parameter is 1.2. 0.8 The simulation results of torque tracking performance of the torque predictive control method using the traditional nonlinear flux observer and the traditional hyperlocal predictive model are shown in the figure.

[0033] Figure 8 In this embodiment of the invention, the motor parameter is 1.2. 0.8 The simulation results of torque tracking performance of the torque predictive control method using a nonlinear flux observer based on iterative control and a hyperlocal predictive model based on a composite discrete extended state observer are shown in the figure.

[0034] Figure 9 This is a schematic diagram of the torque control system of a surface-mounted permanent magnet synchronous motor for electric propulsion provided in an embodiment of the present invention.

[0035] Figure 10 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0036] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0037] Figure 1 This is a control system diagram of a torque control method for a surface-mounted permanent magnet synchronous motor for electric propulsion provided in an embodiment of the present invention.

[0038] like Figure 1 As shown, the torque control method described in this embodiment realizes complete speed-torque dual closed-loop predictive control based on the control system. The system includes: a speed reference value setting module 101, a speed loop PI controller 102, a nonlinear flux linkage observer based on iterative control 103, a hyperlocal predictive model based on a composite discrete extended state observer 104, a cost function optimization module 105, a three-phase full-bridge two-level inverter 106, a surface-mounted permanent magnet synchronous motor (PMSM) of the controlled object 107, and a coordinate transformation module 108.

[0039] For example, the overall closed-loop control flow of this system is as follows: After the system is powered on and started, it first acquires the rotor position information of the controlled surface-mounted permanent magnet synchronous motor 107 through the position sensor, obtains the rotor mechanical angular velocity through differential calculation, and then calculates the actual rotor angular velocity based on the number of motor pole pairs. Preset rotor angular velocity reference value (Output from speed reference value given module 101) and actual rotor angular velocity Simultaneously, the input speed deviation comparison circuit calculates the difference between the two values ​​and outputs the speed deviation signal.

[0040] The speed deviation signal is input to the speed loop PI controller 102. The controller performs proportional-integral regulation and outputs an electromagnetic torque reference value. ， As the target command for subsequent torque closed-loop control 。 At the same time, the preset stator flux reference value With electromagnetic torque reference value Jointly transmit to Cost function optimization module 105.

[0041] In the feedback loop, the current sampling circuit acquires the three-phase instantaneous current of the stator winding of the controlled motor in real time. , , The position sensor synchronously collects the rotor's electrical angle. The three-phase stator current and rotor electrical angle are input into the coordinate transformation module 108. The module performs Park transformation to convert the stator current in the three-phase stationary abc coordinate system into the stator current in the two-phase stationary αβ coordinate system. The current signal is then transmitted to the nonlinear flux observer 103 based on iterative control and the hyperlocal prediction model 104 based on the composite discrete extended state observer, respectively.

[0042] The nonlinear flux linkage observer 103 based on iterative control receives the αβ-axis stator current. Rotor electrical angle and the electromagnetic torque reference value output by the speed loop PI controller 102 Internally, an iterative observation algorithm with torque tracking error compensation term outputs the αβ axis stator flux observation values ​​for the current control cycle. .

[0043] Stator flux linkage observations The input is further fed into a hyperlocal prediction model 104 based on a composite discrete extended state observer, which simultaneously receives the αβ-axis stator current. Rotor electrical angle The model enumerates all candidate voltage vectors for the three-phase full-bridge two-level inverter 106. Internally, it employs a composite discretization prediction algorithm that integrates second-order Taylor expansion, explicit Euler method, and implicit Euler method. Combined with real-time estimates of current disturbance and flux disturbance from the extended state observer, it outputs the torque prediction value after two control cycles for each candidate voltage vector. Compared with the predicted value of stator flux .

[0044] Torque prediction value Compared with the predicted value of stator flux Input cost function optimization module 105, and electromagnetic torque reference value 、 Preset stator flux reference value Both participate in the cost function calculation. The cost function optimization module 105 traverses all candidate voltage vectors, calculates the cost function value for each candidate voltage vector, selects the candidate voltage vector that minimizes the cost function value as the optimal voltage vector for the current control cycle, and converts it into the switching transistor drive signal for the three-phase full-bridge two-level inverter 106. .

[0045] drive signal The on / off states of the switching transistors of the three-phase bridge arms (a, b, and c) in the three-phase full-bridge two-level inverter 106 are controlled respectively. The inverter converts the DC voltage U supplied by the DC bus into DC voltage. dc It is converted into a three-phase AC voltage with corresponding amplitude and frequency, and output to the controlled surface-mount permanent magnet synchronous motor 107 to drive the motor to rotate and output mechanical torque.

[0046] It should be noted that during the operation of the motor, its rotor position, speed and stator current and other operating status signals are collected in real time, processed by the coordinate transformation module 108 and other feedback links and then sent back to the control front end, where they are compared with the preset reference values ​​to form a new round of control cycle calculation.

[0047] Thus, the above process is executed cyclically in each control cycle, forming a complete speed-torque dual closed-loop model predictive torque control loop.

[0048] The following is combined with Figure 2 This invention describes a torque control method for a surface-mounted permanent magnet synchronous motor for electric propulsion. For consistency, the entity executing this method will be uniformly referred to as the "system," and will not be described further thereafter.

[0049] It should be noted that the surface-mounted permanent magnet synchronous motor in this embodiment is a permanent magnet synchronous motor in which the rotor permanent magnet is attached to the outer surface of the rotor core and is used in marine electric propulsion systems.

[0050] In this embodiment of the invention, the ship electric propulsion system uses a battery to power the electric motor. The electric motor rotates and drives the propeller through the transmission shaft to achieve ship propulsion. It is widely used in aquaculture, marine transportation, leisure and recreation, flood control and disaster relief and other fields.

[0051] Figure 2 This is a schematic flowchart of the torque control method for a surface-mounted permanent magnet synchronous motor for electric propulsion provided in an embodiment of the present invention. Figure 2 As shown, the method includes the following steps: S201. Obtain the torque reference value.

[0052] In some embodiments, the rotor angular velocity of a permanent magnet synchronous motor can be obtained.

[0053] For example, the rotor mechanical angle of a permanent magnet synchronous motor can be collected by a position sensor, the rotor mechanical angular velocity can be obtained by differential calculation of the rotor mechanical angle, and then the rotor electrical angular velocity can be obtained by conversion based on the number of pole pairs of the motor, which is the rotor angular velocity in this embodiment.

[0054] Furthermore, the difference between the preset rotor angular velocity reference value and the rotor angular velocity is input into the PI controller to obtain the torque reference value.

[0055] For example, this embodiment uses a speed loop PI controller to achieve closed-loop speed control and outputs an electromagnetic torque reference value, the corresponding expression of which is: .

[0056] in, Permanent magnet synchronous motor Reference value of electromagnetic torque at any given time; for The rotor angular velocity at time t; This is a preset rotor angular velocity reference value; The proportional parameter of the PI control algorithm; These are the integral parameters for the PI control algorithm; This is the integral operation of the PI control algorithm.

[0057] It should be noted that the proportional and integral parameters of the PI controller can be tuned according to the dynamic response requirements of the ship's electric propulsion system to adapt to the speed control requirements under different operating conditions.

[0058] Thus, this invention generates an electromagnetic torque reference value through a closed-loop PI controller for rotational speed, providing precise target commands for subsequent torque tracking control and ensuring the stability of rotational speed control of the ship's electric propulsion system under typical operating conditions such as variable frequency speed regulation and surge impact loads.

[0059] S202. Based on the reference values ​​of current, voltage and torque of the permanent magnet synchronous motor, the stator flux observation value is obtained by using a nonlinear flux observer based on iterative learning.

[0060] Optionally, before obtaining the stator flux linkage observations, the three-phase stator current and rotor electrical angle of the permanent magnet synchronous motor can be collected first.

[0061] For example, the instantaneous current of the three-phase stator windings of a permanent magnet synchronous motor, i.e., the a-phase current, can be acquired through a current sampling circuit. b-phase current c-phase current The rotor electrical angle of a permanent magnet synchronous motor can be collected using a position sensor. Position sensors include, but are not limited to, rotary transformers and photoelectric encoders.

[0062] Specifically, the sampling period is consistent with the control period of this embodiment to ensure the real-time and synchronous nature of the sampling data.

[0063] Furthermore, the three-phase stator currents are subjected to Park transformation to obtain the α-axis current and β-axis current in the two-phase stationary coordinate system.

[0064] For example, the expression for coordinate transformation in this embodiment is as follows: .

[0065] in, for α-axis stator current in a two-phase stationary coordinate system at any given moment; for β-axis stator current in the two-phase stationary coordinate system at any given moment; , , They are respectively Time in three-phase stationary coordinate system Mutually, Mutually, Phase stator current.

[0066] It should be noted that, through the above coordinate transformation, the AC current in the three-phase stationary coordinate system can be converted into the orthogonal current in the two-phase stationary coordinate system, which simplifies the subsequent flux linkage observation and state prediction calculations.

[0067] Furthermore, the optimal voltage vector determined in the previous control cycle is obtained, resulting in the α-axis voltage and β-axis voltage in the two-phase stationary coordinate system.

[0068] For example, the stator voltage of the current control cycle is obtained by mapping the optimal voltage vector output from the previous control cycle, and the corresponding expression is: .

[0069] in, for α-axis stator voltage in the two-phase stationary coordinate system at any given moment; for β-axis stator voltage in the two-phase stationary coordinate system at a given moment; for The optimal α-axis voltage vector selected at any given time; for The optimal β-axis voltage vector selected at time t.

[0070] Thus, by synchronously acquiring three-phase current and rotor electrical angle, transforming coordinates, and obtaining stator voltage, this invention provides accurate basic input data for subsequent flux linkage observation and state prediction, ensuring the computational accuracy of the control algorithm.

[0071] Optionally, after obtaining the α-axis current, β-axis current, α-axis voltage, and β-axis voltage, the α-axis current, β-axis current, α-axis voltage, β-axis voltage, and torque reference values ​​can be input into the nonlinear flux linkage observer to obtain the α-axis stator flux linkage observation values ​​and the β-axis stator flux linkage observation values.

[0072] It should be noted that the nonlinear flux linkage observer based on iterative control in this embodiment is a closed-loop flux linkage observer that uses the stator flux linkage observation error as the basic correction term and the torque tracking error between the electromagnetic torque reference value and the actual torque estimation value as the iterative learning compensation term. It is used to solve the problem that traditional nonlinear flux linkage observers are limited by their own compensation mechanism and have slow convergence under transient conditions.

[0073] In some embodiments, the α-axis stator flux linkage observation value and the β-axis stator flux linkage observation value determined in the previous control cycle can be obtained.

[0074] For example, the α-axis stator flux linkage observation value and the β-axis stator flux linkage observation value of the previous control cycle are stored in the system's storage unit and are directly retrieved during the current control cycle operation as the initial values ​​for iterative operation.

[0075] Specifically, the initial value of the stator flux linkage observation during the initial control cycle can be set to 0, or initialized according to the rated value of the motor permanent magnet flux linkage.

[0076] Furthermore, based on the observed values ​​of the α-axis stator flux linkage and β-axis stator flux linkage from the previous control cycle, the α-axis current and β-axis current from the current control cycle, and the number of pole pairs of the permanent magnet synchronous motor, the torque estimate is calculated.

[0077] For example, the expression for calculating the torque estimate is as follows: .

[0078] in, for The torque estimate at time t; for The α-axis stator flux linkage observation at time t; for The observed value of the stator flux linkage along the β-axis at time t; for The α-axis stator current at time t; for The β-axis stator current at time t; This represents the number of pole pairs of the motor.

[0079] Furthermore, the torque tracking error between the torque reference value and the torque estimate value is calculated.

[0080] For example, the expression for calculating torque tracking error is: .

[0081] in, Let k be the torque tracking error. The reference value for the electromagnetic torque at time k is... This is the torque estimate at time k.

[0082] Furthermore, the torque tracking error is used to generate flux linkage error compensation terms related to the rotor electrical angle and motor inductance parameters.

[0083] Furthermore, the flux linkage error compensation term is superimposed on the basic flux linkage estimate calculated based on the α-axis voltage and β-axis voltage of the current control cycle to generate the α-axis stator flux linkage observation and β-axis stator flux linkage observation for the current control cycle.

[0084] For example, the expression of the nonlinear flux linkage observer model based on iterative control constructed in this embodiment is as follows: .

[0085] in, State matrix at time step ; State matrix at time step ; coefficient matrix ; matrix ; Input variable matrix ; Learning function matrix ; Magnetic flux error value .

[0086] in, for The α-axis stator flux linkage observation at time t; for The observed value of the stator flux linkage along the β-axis at time t; for The α-axis stator flux linkage observation at time t; The value of the stator flux linkage along the β axis at time k+1; This is the observer parameter, which is usually set to 1; For the algorithm execution cycle parameters; Let be the flux linkage error value at time k; for The rotor electrical angle at any given moment; for The α-axis stator voltage at time t; for The β-axis stator voltage at time t; This refers to the stator resistance of the motor. for The α-axis stator current at time t; for The β-axis stator current at time t; These are the stator inductance parameters of the motor; The second observer parameter has a value range of 0 < <30; This represents the number of pole pairs of the motor. These are the flux linkage parameters of the permanent magnet in the motor; for Reference value of electromagnetic torque at any given time; for The torque estimate at time t.

[0087] It should be noted that in the above observer model, This is the flux linkage error compensation term generated based on torque tracking error. This compensation term can effectively accelerate the convergence speed of the flux linkage observer under transient conditions and solve the problem of slow convergence of traditional nonlinear flux linkage observers.

[0088] In this embodiment of the invention, the flux linkage error compensation term is generated by inputting the torque tracking error into a learning function matrix related to the rotor electrical angle and motor inductance parameters.

[0089] Specifically, learning function matrices With rotor electrical angle Motor stator inductance parameters It is directly related and can be adjusted in real time according to the rotor position and motor parameters to ensure the accuracy and adaptability of the compensation item.

[0090] Thus, this invention, through a nonlinear flux linkage observer based on iterative control, introduces torque tracking error as an iterative learning compensation term, which effectively improves the convergence speed of the flux linkage observer under transient conditions. Even under conditions where the motor inductance parameters and permanent magnet flux linkage parameters are perturbed, it can still accurately obtain the stator flux linkage observation value, providing reliable basic data for subsequent state prediction.

[0091] S203. Based on the observed values ​​of current, voltage, and stator flux linkage of the permanent magnet synchronous motor and the candidate voltage vector of the inverter, the torque prediction value and stator flux linkage prediction value are obtained using a hyperlocal prediction model based on a composite discrete extended state observer.

[0092] Optionally, before obtaining the torque prediction value and the stator flux prediction value, all switching state combinations of the two-level inverter connected to the permanent magnet synchronous motor can be enumerated to obtain the candidate voltage vector.

[0093] In this embodiment of the invention, the candidate voltage vector is determined by a combination of eight switching states of a two-level inverter.

[0094] The eight switching state combinations include six effective voltage vectors and two zero voltage vectors. Among the eight switching state combinations, six combinations correspond to effective voltage vectors, and two combinations correspond to zero voltage vectors (the zero voltage vectors output by these two combinations are the same), thus corresponding to a total of seven different output voltage vectors.

[0095] For example, the calculation expression for the candidate voltage vector is: .

[0096] in, For the first One candidate voltage vector; for Phase instantaneous voltage, when the switch state is 1, When the switch state is 0, ; for Phase instantaneous voltage, when the switch state is 1, When the switch state is 0, ; This is the instantaneous voltage of phase c when the switch state is 1. When the switch state is 0 ; This is the DC bus voltage; It is the imaginary unit.

[0097] Specifically, based on the above expression, all candidate voltage vectors are enumerated: the first candidate voltage vector is... The second candidate voltage vector is The third candidate voltage vector is The fourth candidate voltage vector is The fifth candidate voltage vector is The sixth candidate voltage vector is The seventh candidate voltage vector is .

[0098] in, to For the effective voltage vector, It is a zero voltage vector.

[0099] Thus, by enumerating all switching state combinations of the two-level inverter, the present invention obtains all candidate voltage vectors, providing a complete set of control input candidates for subsequent model prediction rolling optimization, and ensuring that the selection range of the optimal voltage vector covers all feasible control outputs.

[0100] Optionally, a hyperlocal prediction model based on a composite discrete extended state observer can be constructed first.

[0101] It should be noted that the hyperlocal prediction model based on the composite discrete extended state observer in this embodiment is a motor state prediction model that uses a composite discretization method to reduce the discretization truncation error and uses an extended state observer to estimate and compensate for system disturbances caused by motor parameter perturbations in real time. It is used to solve the problem that traditional hyperlocal prediction models do not consider discretization truncation errors and have insufficient model accuracy.

[0102] In some embodiments, a composite discretization method combining second-order Taylor expansion, explicit Euler method, and implicit Euler method can be used to discretize the continuous domain state equations of the permanent magnet synchronous motor, thereby obtaining a composite discretized prediction model.

[0103] For example, firstly, a traditional continuous domain hyperlocal model of a permanent magnet synchronous motor is constructed, with the expression: .

[0104] in, = + This is the stator flux linkage vector in a two-phase stationary coordinate system; = + The stator current vector in a two-phase stationary coordinate system; = + The stator voltage vector in a two-phase stationary coordinate system; These are the stator inductance parameters of the motor; For magnetic flux disturbance term, This is the current disturbance term, used to characterize the total disturbance caused by motor parameter perturbations, unmodeled dynamics, etc.

[0105] Specifically, the state equations of the continuous domain are discretized using a second-order Taylor expansion. The expression of the second-order Taylor expansion is as follows: .

[0106] in, for The system state variables at any given time; To control the cycle; Here are the system state equations.

[0107] Meanwhile, the expression for the explicit Euler method is: .

[0108] The expression for the implicit Euler method: .

[0109] By combining the above second-order Taylor expansion, explicit Euler method, and implicit Euler method, a composite discretized prediction model is obtained, expressed as: .

[0110] in, for The stator flux linkage vector at time t; for The stator current vector at time t; for The stator voltage vector at time t; for The stator voltage vector at time t; for The flux perturbation term at time; for The flux perturbation term at time; for The current disturbance term at any given moment; for The current disturbance term at any given time.

[0111] In other embodiments, the α-axis current, β-axis current, α-axis voltage, β-axis voltage, α-axis stator flux linkage observation, and β-axis stator flux linkage observation can be input into the extended state observer to estimate the current disturbance and flux linkage disturbance.

[0112] For example, the composite discrete extended state observer constructed in this embodiment observes flux disturbances and current disturbances as extended state variables in real time. The adjustable gain parameter of the observer is set as follows: First adjustable gain parameter ; Second adjustable gain parameter ; in, The bandwidth of the composite discrete extended state observer can be tuned according to the system's dynamic response requirements.

[0113] Specifically, the composite discrete extended state observer updates the flux disturbance estimate and current disturbance estimate online by using the error between the predicted value and the actual observed value of the system state, thereby achieving real-time tracking and compensation of the total system disturbance.

[0114] In one alternative implementation, the α-axis current and β-axis current can be compared with the current prediction values ​​output by the composite discretized prediction model to obtain the current prediction error.

[0115] For example, the expression for calculating the current prediction error is: = .

[0116] in, for Current prediction error at time; The output of the composite discretization prediction model The predicted current value at time; for The stator current vector acquired at any time.

[0117] Furthermore, the current disturbance estimate is updated online based on the current prediction error.

[0118] Specifically, the update logic for the current disturbance estimate is as follows: taking the current prediction error as input, the current disturbance estimate is corrected online by adjusting the observer gain parameter through proportional-derivative adjustment, so that the current prediction error converges to 0, thereby realizing real-time tracking of the current disturbance.

[0119] In another alternative implementation, the stator flux linkage observations along the α-axis and β-axis can be compared with the stator flux linkage predictions output by the composite discretized prediction model to obtain the flux linkage prediction error.

[0120] For example, the expression for calculating the flux linkage prediction error is as follows: = .

[0121] in, for The flux linkage prediction error at time t; The output of the composite discretization prediction model The predicted flux linkage value at time 1; For the output of a nonlinear flux linkage observer based on iterative control The stator flux linkage observation at time t.

[0122] Furthermore, the flux disturbance estimate is updated online based on the flux prediction error.

[0123] Specifically, the update logic for the flux linkage disturbance estimate is as follows: taking the flux linkage prediction error as input, the gain parameter of the observer is adjusted proportionally and derivatively to correct the flux linkage disturbance estimate online, so that the flux linkage prediction error converges to 0, thereby realizing real-time tracking of the flux linkage disturbance.

[0124] Thus, this invention uses a composite discrete extended state observer to observe and update the flux disturbance and current disturbance caused by motor parameter perturbation as a total disturbance in real time, providing accurate disturbance compensation for the prediction model and effectively improving the accuracy of the prediction model under parameter perturbation conditions.

[0125] Furthermore, after obtaining the current disturbance estimate and the flux disturbance estimate, the current disturbance estimate and the flux disturbance estimate are then compensated into the composite discretized prediction model to form a hyperlocal prediction model.

[0126] For example, considering one-step delay compensation in digital control, a hyperlocal prediction model based on a composite discrete extended state observer is constructed for... The motor state at time t is predicted, and the corresponding expression is: .

[0127] in, for Stator flux linkage prediction at time 10:00 for Predicted value of stator flux linkage along the α-axis at time t. for Moment Predicted values ​​of stator flux linkage; for Predicted current value at time 10:00 for Predicted α-axis current value at time t. for Predicted β-axis current value at time t; for The predicted torque value at time 1; for The estimated flux linkage perturbation at time t; for The estimated value of the current disturbance at time; for The candidate voltage vector input at any given time, ∈[ ]; for The αβ-axis stator voltage vector at time t; for The αβ-axis stator voltage vector at time t; for Observations of the αβ-axis stator flux linkage at time t; for Observations of the αβ-axis stator flux linkage at time t; for The αβ axis stator current vector at time t; for The αβ axis stator current vector at time t; is the first adjustable gain parameter of the composite discrete extended state observer; This is the second adjustable gain parameter of the composite discrete extended state observer; The bandwidth of the composite discrete extended state observer; To control the cycle; These are the stator inductance parameters of the motor; It is the imaginary unit.

[0128] Thus, by feeding forward the disturbance estimate into the composite discretized prediction model, the present invention forms a hyperlocal prediction model based on the composite discrete extended state observer, which effectively reduces the discretization truncation error of the prediction model and at the same time offsets the influence of motor parameter perturbation on prediction accuracy, thereby significantly improving the accuracy of state prediction.

[0129] In some embodiments, for each candidate voltage vector, a hyperlocal prediction model is used to output the predicted current and stator flux linkage values ​​two control cycles after the current time.

[0130] It should be noted that the two-step prediction (prediction at time k+2) is used to compensate for the one-step control delay in the digital control system, avoid the dynamic response lag caused by the control delay, and ensure the high dynamic response capability of the system.

[0131] The predicted current values ​​include α-axis current prediction and β-axis current prediction; the predicted stator flux linkage values ​​include α-axis stator flux linkage prediction and β-axis stator flux linkage prediction.

[0132] Furthermore, the torque prediction value is calculated based on the predicted current value and the predicted stator flux linkage value.

[0133] In this embodiment of the invention, the torque prediction value is calculated based on the α-axis current prediction value, the β-axis current prediction value, the α-axis stator flux prediction value, the β-axis stator flux prediction value, and the number of pole pairs of the permanent magnet synchronous motor.

[0134] For example, The expression for calculating the torque prediction value at time t is: .

[0135] in, for The predicted torque value at time 1; for Predicted α-axis stator flux linkage at time t; The predicted value of the stator flux linkage along the β axis at time k+2; for Predicted α-axis current value at time t; for Predicted β-axis current value at time t; This represents the number of pole pairs of the motor.

[0136] Thus, this invention uses a hyperlocal prediction model based on a composite discrete extended state observer to accurately predict the future motor state corresponding to each candidate voltage vector, obtaining the torque prediction value and the stator flux prediction value, providing accurate prediction data for subsequent optimal voltage vector selection.

[0137] S204. Based on the torque reference value, the preset stator flux reference value, the torque prediction value, and the stator flux prediction value, determine the optimal voltage vector from the candidate voltage vectors, and control the torque of the permanent magnet synchronous motor according to the optimal voltage vector.

[0138] Optionally, the torque deviation between the torque reference value and the torque prediction value, as well as the flux deviation between the preset stator flux reference value and the stator flux prediction value, can be calculated.

[0139] The stator flux linkage prediction values ​​include the α-axis stator flux linkage prediction values ​​and the β-axis stator flux linkage prediction values.

[0140] For example, the expression for calculating torque deviation is: .

[0141] in, For torque deviation, This is the reference value for electromagnetic torque. for The predicted torque value at time 1.

[0142] In some embodiments, the predicted value of stator flux linkage amplitude can be calculated based on the predicted values ​​of α-axis stator flux linkage and β-axis stator flux linkage.

[0143] Furthermore, the difference between the preset stator flux reference value and the predicted stator flux amplitude value is calculated to obtain the flux deviation.

[0144] For example, the expression for calculating flux linkage deviation is: .

[0145] in, For magnetic flux deviation, The preset stator flux linkage reference value, This is the predicted value of the stator flux linkage amplitude.

[0146] It should be noted that the preset stator flux reference value in this embodiment is set as the rated amplitude of the permanent magnet flux of the surface-mounted permanent magnet synchronous motor.

[0147] Furthermore, the torque deviation and flux deviation are weighted and summed using preset weighting coefficients to obtain the cost function value corresponding to each candidate voltage vector.

[0148] For example, the cost function expression constructed in this embodiment is as follows: .

[0149] in, This is the cost function value; This is a preset electromagnetic torque reference value; for The predicted torque value at time 1; This is a weighting coefficient used to adjust the priority of torque tracking control and stator flux linkage amplitude tracking control; This is a preset stator flux linkage reference value; for Predicted value of stator flux linkage amplitude at time t.

[0150] Furthermore, the candidate voltage vector corresponding to the minimum cost function value is determined as the optimal voltage vector.

[0151] Specifically, all seven candidate voltage vectors are iterated through, and the cost function value corresponding to each candidate voltage vector is calculated. The candidate voltage vector with the smallest cost function value is selected through rolling optimization, which is the optimal voltage vector for the current control cycle.

[0152] Thus, by constructing a cost function with torque tracking error and flux tracking error as optimization targets, and selecting the optimal voltage vector through rolling optimization, this invention achieves precise closed-loop control of motor torque, ensuring the steady-state control accuracy and dynamic response speed of the system.

[0153] In this embodiment of the invention, after obtaining the optimal voltage vector, the switching state of the three-phase full-bridge inverter switching transistor can be changed according to the optimal voltage vector, and the corresponding voltage can be output to drive the surface-mount permanent magnet synchronous motor, completing the torque closed-loop control of one control cycle, and entering the cyclic calculation of the next control cycle.

[0154] The torque control method for a surface-mounted permanent magnet synchronous motor for electric propulsion provided by this invention obtains a torque reference value to provide target commands for subsequent control; by obtaining stator flux observation values ​​based on motor current, voltage, and torque reference values ​​and utilizing a nonlinear flux observation based on iterative learning, the method can leverage torque tracking error to drive the observation to converge quickly, and can still accurately obtain stator flux information during parameter perturbations; and by obtaining stator flux observation values ​​based on motor current, voltage, stator flux observation values, and inverter candidate voltage vectors using a hyperlocal prediction model based on a composite discrete extended state observation, the method can obtain... The predicted torque and stator flux linkage values ​​are improved by using a composite discretization method to reduce the truncation error of the prediction model and by using an extended state observer to estimate and compensate for parameter perturbations and unmodeled dynamics in real time. The optimal voltage vector is determined from the candidate voltage vectors based on the torque reference value, the preset stator flux linkage reference value, the predicted torque value, and the predicted stator flux linkage value, and the torque of the permanent magnet synchronous motor is controlled accordingly. This achieves high-precision tracking of torque under the perturbation of inductance parameters and permanent magnet flux linkage parameters, while taking into account the system's strong robustness and high dynamic response capability.

[0155] To verify the effectiveness of the control method in this embodiment, Matlab / Simulink software was used for simulation verification. The simulation period was set to 1e-5, and a fixed discrete step size solver was used for the simulation.

[0156] The specific parameters of the surface-mounted permanent magnet synchronous motor used in the ship's electric propulsion system in this embodiment are shown in Table 1 below. Table 1 The simulation conditions were set as follows: the reference speed was set to 3000 rpm, and a step load torque of 1 N·m was applied at 0.1 s; the reference value of the stator flux linkage was set to the amplitude of the permanent magnet flux linkage.

[0157] To highlight the performance advantages of the control method in this embodiment, simulation comparisons are performed on the following three control methods: 1) Traditional model predictive torque control method; 2) Torque prediction control method using traditional nonlinear flux observer and traditional hyperlocal prediction model; 3) The torque prediction control method in this embodiment uses a nonlinear flux observer based on iterative control and a hyperlocal prediction model based on a composite discrete extended state observer.

[0158] First, assuming the motor's electrical parameters remain unchanged, the inductance and permanent magnet flux linkage parameters are at their rated values. and magnetic chain Under the following working conditions: Figure 3 In this embodiment of the invention, the motor parameters are... , The figure shows the simulation results of torque tracking performance of the traditional model predictive torque control method. The horizontal axis of the figure represents the simulation time in seconds (s), ranging from 0 to 0.2 s; the vertical axis represents the motor output torque in Newton-meters (N·m). The figure contains two curves: the curve for the given torque reference value (solid line) and the curve for the actual feedback torque of the motor (dashed line).

[0159] from Figure 3 It can be seen that after applying a 1 N·m step load torque at 0.1 s, the actual feedback torque of the traditional model predictive torque control method shows a large overshoot, the torque tracking error is large under dynamic conditions, and the torque curve fluctuates significantly after entering steady state. This verifies that the prediction model of this method has low accuracy and cannot achieve high-precision torque tracking.

[0160] Figure 4 In this embodiment of the invention, the motor parameters are... , The figure shows the simulation results of torque tracking performance using a conventional nonlinear flux observer and a conventional hyperlocal predictive model for torque predictive control. The horizontal axis represents the simulation time in seconds (s), ranging from 0 to 0.2 s; the vertical axis represents the motor output torque in Newton-meters (N·m). The figure contains two curves: the curve for the given torque reference value (solid line) and the curve for the actual feedback torque of the motor (dashed line).

[0161] from Figure 4 It can be seen that after applying a step load torque at 0.1s, the dynamic torque tracking error of this method is significantly smaller than that of the traditional method, and the torque fluctuation under steady-state conditions is also significantly reduced. However, there is still a fixed steady-state error between the actual feedback torque and the given torque reference value during steady-state operation, which verifies that the performance of the traditional nonlinear flux observer and the traditional hyperlocal prediction model is still insufficient.

[0162] Figure 5 In this embodiment of the invention, the motor parameters are... , The figure shows the simulation results of torque tracking performance of the torque predictive control method using a nonlinear flux observer based on iterative control and a hyperlocal predictive model based on a composite discrete extended state observer. The horizontal axis of the figure represents the simulation time in seconds (s), ranging from 0 to 0.2 s; the vertical axis represents the motor output torque in Newton-meters (N·m). The figure contains two curves: the curve for the given torque reference value (solid line) and the curve for the actual feedback torque of the motor (dashed line).

[0163] from Figure 5It can be seen that after applying a step load torque in 0.1s, the actual feedback torque of the method of the present invention can track the given torque reference value quickly without overshoot, with fast dynamic response speed and almost zero dynamic tracking error. After entering steady state, the torque curve is smooth without obvious fluctuations and completely coincides with the given torque reference value, which verifies that the method of the present invention has high dynamic torque tracking capability under rated parameter conditions, the prediction model has the highest accuracy, and the torque tracking error is the smallest.

[0164] Secondly, when the motor parameters are perturbed, the inductance parameter becomes 1.2. The flux linkage parameter of the permanent magnet becomes 0.8. Under the following working conditions: Figure 6 In this embodiment of the invention, the motor parameter is 1.2. 0.8 The figure shows the simulation results of torque tracking performance of the traditional model predictive torque control method. The horizontal axis of the figure represents the simulation time in seconds (s), ranging from 0 to 0.2 s; the vertical axis represents the motor output torque in Newton-meters (N·m). The figure contains two curves: the curve for the given torque reference value (solid line) and the curve for the actual feedback torque of the motor (dashed line).

[0165] from Figure 6 It can be seen that under parameter perturbation conditions, due to the decrease in the flux linkage amplitude of the permanent magnet and the increase in the inductance parameter, the torque tracking performance of the traditional model predictive torque control method is severely deteriorated. After a 0.1s step load, the torque overshoot increases significantly, and there is a significant steady-state tracking error between the actual feedback torque and the given torque reference value. Moreover, the steady-state torque fluctuates violently, which verifies that the prediction model of this method has extremely low accuracy and poor robustness under parameter perturbation conditions.

[0166] Figure 7 In this embodiment of the invention, the motor parameter is 1.2. 0.8 The figure shows the simulation results of torque tracking performance using a conventional nonlinear flux observer and a conventional hyperlocal predictive model for torque predictive control. The horizontal axis represents the simulation time in seconds (s), ranging from 0 to 0.2 s; the vertical axis represents the motor output torque in Newton-meters (N·m). The figure contains two curves: the curve for the given torque reference value (solid line) and the curve for the actual feedback torque of the motor (dashed line).

[0167] from Figure 7It can be seen that under parameter perturbation conditions, the torque tracking error of this method gradually decreases as the simulation time progresses, but the convergence speed is slow. After a 0.1s step load, it takes a long time to approach the given torque reference value, and there is still a small torque tracking error under steady-state conditions. This verifies that the traditional nonlinear flux observer has a slow convergence speed under parameter perturbation conditions and cannot meet the requirements of high dynamic response of the system.

[0168] Figure 8 In this embodiment of the invention, the motor parameter is 1.2. 0.8 The figure shows the simulation results of torque tracking performance of the torque predictive control method using a nonlinear flux observer based on iterative control and a hyperlocal predictive model based on a composite discrete extended state observer. The horizontal axis of the figure represents the simulation time in seconds (s), ranging from 0 to 0.2 s; the vertical axis represents the motor output torque in Newton-meters (N·m). The figure contains two curves: the curve for the given torque reference value (solid line) and the curve for the actual feedback torque of the motor (dashed line).

[0169] from Figure 8 As can be seen, under parameter perturbation conditions, the method of the present invention can still quickly and without overshoot track the given torque reference value after applying a step load of 0.1s. The dynamic response speed is basically consistent with that under rated parameter conditions, and the torque tracking error is extremely small. After entering steady state, the actual feedback torque completely coincides with the given torque reference value, with no steady-state error and obvious fluctuations. This verifies that the method of the present invention still has excellent dynamic response capability, strong robustness and high-precision torque tracking performance under perturbation conditions of inductance parameters and permanent magnet flux linkage parameters.

[0170] In summary, the simulation results show that, combined with Figures 3 to 8 Comparative analysis shows that the torque control method in this embodiment has higher prediction model accuracy and smaller torque tracking error under the rated parameters of the motor. It still has excellent dynamic response capability and strong robustness under the perturbation conditions of inductance parameters and permanent magnet flux linkage parameters, effectively solving the problems of large torque tracking error and insufficient dynamic response capability under parameter perturbation in the prior art.

[0171] The torque control system of the surface-mounted permanent magnet synchronous motor for electric propulsion provided by the present invention will be described below. The torque control system of the surface-mounted permanent magnet synchronous motor for electric propulsion described below can be referred to in correspondence with the torque control method of the surface-mounted permanent magnet synchronous motor for electric propulsion described above.

[0172] Figure 9This is a structural diagram of a torque control system for a surface-mounted permanent magnet synchronous motor for electric propulsion, provided in an embodiment of the present invention. The torque control system for this surface-mounted permanent magnet synchronous motor for electric propulsion includes: an acquisition module 901, a flux linkage observation module 902, a state prediction module 903, and a voltage vector determination module 904.

[0173] The system includes: an acquisition module 901 for acquiring torque reference values; a flux linkage observation module 902 for obtaining stator flux linkage observation values ​​based on the current, voltage, and torque reference values ​​of the permanent magnet synchronous motor using a nonlinear flux linkage observer based on iterative learning; a state prediction module 903 for obtaining torque prediction values ​​and stator flux linkage prediction values ​​based on the current, voltage, and stator flux linkage observation values ​​of the permanent magnet synchronous motor and the candidate voltage vectors of the inverter using a hyperlocal prediction model based on a composite discrete extended state observer; and a voltage vector determination module 904 for determining the optimal voltage vector from the candidate voltage vectors based on the torque reference values, preset stator flux linkage reference values, torque prediction values, and stator flux linkage prediction values, and controlling the torque of the permanent magnet synchronous motor based on the optimal voltage vector.

[0174] In some embodiments, the acquisition module 901 is specifically used to: acquire the rotor angular velocity of the permanent magnet synchronous motor; input the difference between the preset rotor angular velocity reference value and the rotor angular velocity into the PI controller to obtain the torque reference value.

[0175] In some embodiments, the acquisition module 901 is further configured to: acquire the three-phase stator current and rotor electrical angle of the permanent magnet synchronous motor before acquiring the stator flux observation value using the nonlinear flux observer based on iterative learning; perform Park transformation on the three-phase stator current to obtain the α-axis current and β-axis current in the two-phase stationary coordinate system; acquire the optimal voltage vector determined in the previous control cycle to obtain the α-axis voltage and β-axis voltage in the two-phase stationary coordinate system.

[0176] In some embodiments, the flux linkage observation module 902 is specifically used to: input the α-axis current, β-axis current, α-axis voltage, β-axis voltage and torque reference values ​​into the nonlinear flux linkage observer to obtain the α-axis stator flux linkage observation value and the β-axis stator flux linkage observation value.

[0177] In some embodiments, the flux linkage observation module 902 is specifically used for: acquiring the α-axis stator flux linkage observation value and the β-axis stator flux linkage observation value determined in the previous control cycle; calculating the torque estimate value based on the α-axis stator flux linkage observation value and the β-axis stator flux linkage observation value of the previous control cycle, the α-axis current and the β-axis current of the current control cycle, and the number of pole pairs of the permanent magnet synchronous motor; calculating the torque tracking error between the torque reference value and the torque estimate value; generating a flux linkage error compensation term related to the rotor electrical angle and motor inductance parameters using the torque tracking error; and superimposing the flux linkage error compensation term onto the basic flux linkage estimate value calculated based on the α-axis voltage and the β-axis voltage of the current control cycle to generate the α-axis stator flux linkage observation value and the β-axis stator flux linkage observation value of the current control cycle.

[0178] In some embodiments, the acquisition module 901 is further configured to: before obtaining the torque prediction value and the stator flux prediction value using the hyperlocal prediction model based on the composite discrete extended state observer, enumerate all switching state combinations of the two-level inverter connected to the permanent magnet synchronous motor to obtain the candidate voltage vector.

[0179] In some embodiments, the torque control system of the above-mentioned surface-mounted permanent magnet synchronous motor for electric propulsion further includes a construction module. The construction module is used to: discretize the continuous domain state equations of the permanent magnet synchronous motor using a composite discretization method that integrates second-order Taylor expansion, explicit Euler method, and implicit Euler method to obtain a composite discretized prediction model; input the α-axis current, β-axis current, α-axis voltage, β-axis voltage, α-axis stator flux linkage observations, and β-axis stator flux linkage observations into an extended state observer to estimate the current disturbance estimate and flux linkage disturbance estimate; and compensate the current disturbance estimate and flux linkage disturbance estimate into the composite discretized prediction model to form a hyperlocal prediction model.

[0180] In some embodiments, the above-described construction module is specifically used to: compare the α-axis current and β-axis current with the current prediction values ​​output by the composite discretized prediction model to obtain the current prediction error; compare the α-axis stator flux linkage observation values ​​and β-axis stator flux linkage observation values ​​with the stator flux linkage prediction values ​​output by the composite discretized prediction model to obtain the flux linkage prediction error; and update the current disturbance estimate online based on the current prediction error and the flux linkage disturbance estimate online based on the flux linkage prediction error.

[0181] In some embodiments, the state prediction module 903 is specifically used to: for each candidate voltage vector, use a hyperlocal prediction model to output the current prediction value and the stator flux prediction value after two control cycles relative to the current time; and calculate the torque prediction value based on the current prediction value and the stator flux prediction value.

[0182] In some embodiments, the voltage vector determination module 904 is specifically used to: calculate the torque deviation between the torque reference value and the torque prediction value, and the flux deviation between the preset stator flux reference value and the stator flux prediction value; use preset weighting coefficients to perform a weighted summation of the torque deviation and flux deviation to obtain the cost function value corresponding to each candidate voltage vector; and determine the candidate voltage vector corresponding to the minimum cost function value as the optimal voltage vector.

[0183] In some embodiments, the stator flux linkage prediction value includes the α-axis stator flux linkage prediction value and the β-axis stator flux linkage prediction value; the voltage vector determination module 904 is specifically used to: calculate the stator flux linkage amplitude prediction value based on the α-axis stator flux linkage prediction value and the β-axis stator flux linkage prediction value; calculate the difference between the preset stator flux linkage reference value and the stator flux linkage amplitude prediction value to obtain the flux linkage deviation.

[0184] In the torque control system of the surface-mounted permanent magnet synchronous motor for electric propulsion provided by this invention, target commands are provided for subsequent control by acquiring torque reference values; stator flux observation values ​​are obtained by using a nonlinear flux observer based on iterative learning, based on motor current, voltage, and torque reference values, and the torque tracking error drives the observer to converge quickly, and stator flux information can still be accurately acquired during parameter perturbations; and a hyperlocal prediction model based on a composite discrete extended state observer is used to obtain the stator flux observation values ​​based on motor current, voltage, stator flux observation values, and inverter candidate voltage vectors. The predicted torque and stator flux linkage values ​​are obtained. A composite discretization method is used to reduce the truncation error of the prediction model. An extended state observer is used to estimate and compensate for parameter perturbations and unmodeled dynamics in real time, thereby improving the prediction accuracy. The optimal voltage vector is determined from the candidate voltage vectors based on the torque reference value, the preset stator flux linkage reference value, the predicted torque value, and the predicted stator flux linkage value. The torque of the permanent magnet synchronous motor is controlled accordingly. This achieves high-precision tracking of torque under the perturbation of inductance parameters and permanent magnet flux linkage parameters, while taking into account the system's strong robustness and high dynamic response capability.

[0185] Figure 10 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 10 As shown, the electronic device may include: a processor 1010, a communications interface 1020, a memory 1030, and a communications bus 1040, wherein the processor 1010, the communications interface 1020, and the memory 1030 communicate with each other through the communications bus 1040.

[0186] The processor 1010 can call logic instructions in the memory 1030 to execute a torque control method for a surface-mounted permanent magnet synchronous motor for electric propulsion. The method includes: acquiring a torque reference value; obtaining a stator flux observation value using an iterative learning-based nonlinear flux observer based on the permanent magnet synchronous motor current, voltage, and torque reference values; obtaining a torque prediction value and a stator flux prediction value using a hyperlocal prediction model based on a composite discrete extended state observer based on the permanent magnet synchronous motor current, voltage, stator flux observation values, and candidate voltage vectors of the inverter; determining an optimal voltage vector from the candidate voltage vectors based on the torque reference value, the preset stator flux reference value, the torque prediction value, and the stator flux prediction value, and controlling the torque of the permanent magnet synchronous motor according to the optimal voltage vector.

[0187] Furthermore, the logical instructions in the aforementioned memory 1030 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, essentially, or the part that contributes to the prior art, or a 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 several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0188] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the torque control method for a surface-mounted permanent magnet synchronous motor for electric propulsion provided by the above methods. The method includes: obtaining a torque reference value; obtaining a stator flux observation value using an iterative learning-based nonlinear flux linkage observer based on the permanent magnet synchronous motor current, voltage, and torque reference values; obtaining a torque prediction value and a stator flux linkage prediction value using a hyperlocal prediction model based on a composite discrete extended state observer based on the permanent magnet synchronous motor current, voltage, stator flux linkage observation values, and candidate voltage vectors of the inverter; determining an optimal voltage vector from the candidate voltage vectors according to the torque reference value, the preset stator flux linkage reference value, the torque prediction value, and the stator flux linkage prediction value, and controlling the torque of the permanent magnet synchronous motor according to the optimal voltage vector.

[0189] In another aspect, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements a torque control method for a surface-mounted permanent magnet synchronous motor for electric propulsion provided by the methods described above. The method includes: acquiring a torque reference value; obtaining a stator flux linkage observation value using an iterative learning-based nonlinear flux linkage observer based on the permanent magnet synchronous motor current, voltage, and torque reference values; obtaining a torque prediction value and a stator flux linkage prediction value using a hyperlocal prediction model based on a composite discrete extended state observer based on the permanent magnet synchronous motor current, voltage, stator flux linkage observation values, and candidate voltage vectors of the inverter; determining an optimal voltage vector from the candidate voltage vectors according to the torque reference value, the preset stator flux linkage reference value, the torque prediction value, and the stator flux linkage prediction value, and controlling the torque of the permanent magnet synchronous motor according to the optimal voltage vector.

[0190] The system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0191] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0192] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A torque control method for a surface-mounted permanent magnet synchronous motor for electric propulsion, characterized in that, include: Obtain the torque reference value; Based on the current, voltage and torque reference values ​​of the permanent magnet synchronous motor, the stator flux observation value is obtained using a nonlinear flux observer based on iterative learning. Based on the current and voltage of the permanent magnet synchronous motor, the observed values ​​of the stator flux linkage, and the candidate voltage vector of the inverter, the torque prediction value and the stator flux linkage prediction value are obtained using a hyperlocal prediction model based on a composite discrete extended state observer. Based on the torque reference value, the preset stator flux reference value, the torque prediction value, and the stator flux prediction value, the optimal voltage vector is determined from the candidate voltage vectors, and the torque of the permanent magnet synchronous motor is controlled according to the optimal voltage vector.

2. The method according to claim 1, characterized in that, The acquisition of the torque reference value includes: Obtain the rotor angular velocity of the permanent magnet synchronous motor; The difference between the preset rotor angular velocity reference value and the rotor angular velocity is input into the proportional-integral controller to obtain the torque reference value.

3. The method according to claim 1, characterized in that, Before obtaining stator flux observations using an iterative learning-based nonlinear flux observer, the method further includes: The three-phase stator current and rotor electrical angle of the permanent magnet synchronous motor are collected. The three-phase stator currents are subjected to Park transformation to obtain the α-axis current and β-axis current in the two-phase stationary coordinate system; The optimal voltage vector determined in the previous control cycle is obtained to obtain the α-axis voltage and β-axis voltage in the two-phase stationary coordinate system.

4. The method according to claim 3, characterized in that, The method of obtaining stator flux observations using a nonlinear flux observer based on iterative learning includes: The α-axis current, β-axis current, α-axis voltage, β-axis voltage, and torque reference value are input into the nonlinear flux linkage observer to obtain the α-axis stator flux linkage observation value and the β-axis stator flux linkage observation value.

5. The method according to claim 1, characterized in that, Before obtaining the torque prediction value and stator flux prediction value using the hyperlocal prediction model based on the composite discrete extended state observer, the method further includes: The candidate voltage vector is obtained by enumerating all the switching state combinations of the two-level inverter connected to the permanent magnet synchronous motor.

6. The method according to claim 4, characterized in that, The hyperlocal prediction model based on the composite discrete extended state observer is constructed through the following steps: A composite discretization method combining second-order Taylor expansion, explicit Euler method, and implicit Euler method is used to discretize the continuous domain state equation of the permanent magnet synchronous motor, resulting in a composite discretization prediction model. The α-axis current, β-axis current, α-axis voltage, β-axis voltage, α-axis stator flux linkage observation value, and β-axis stator flux linkage observation value are input into the extended state observer to estimate the current disturbance estimate and flux linkage disturbance estimate. The current disturbance estimate and the flux disturbance estimate are compensated into the composite discretized prediction model to form the hyperlocal prediction model.

7. The method according to claim 1, characterized in that, The method of obtaining torque prediction values ​​and stator flux prediction values ​​using a hyperlocal prediction model based on a composite discrete extended state observer includes: For each of the candidate voltage vectors, the superlocal prediction model is used to output the predicted current value and the predicted stator flux linkage value after two control cycles relative to the current time. The torque prediction value is calculated based on the predicted current value and the predicted stator flux linkage value.

8. The method according to claim 1, characterized in that, The step of determining the optimal voltage vector from the candidate voltage vectors includes: Calculate the torque deviation between the torque reference value and the torque prediction value, and the flux deviation between the preset stator flux reference value and the stator flux prediction value; The torque deviation and the flux deviation are weighted and summed using preset weighting coefficients to obtain the cost function value corresponding to each candidate voltage vector; The candidate voltage vector corresponding to the minimum cost function value is determined as the optimal voltage vector.

9. The method according to claim 8, characterized in that, The stator flux linkage prediction values ​​include the α-axis stator flux linkage prediction value and the β-axis stator flux linkage prediction value; The calculation steps for the flux deviation between the preset stator flux reference value and the predicted stator flux value include: Calculate the predicted value of stator flux linkage amplitude based on the predicted value of stator flux linkage along the α-axis and the predicted value of stator flux linkage along the β-axis; The difference between the preset stator flux reference value and the predicted stator flux amplitude value is calculated to obtain the flux deviation.

10. A torque control system for a surface-mounted permanent magnet synchronous motor for electric propulsion, characterized in that, The system includes: The acquisition module is used to obtain torque reference values; The flux linkage observation module is used to obtain stator flux linkage observation values ​​based on the current, voltage and torque reference values ​​of the permanent magnet synchronous motor using a nonlinear flux linkage observer based on iterative learning. The state prediction module is used to obtain the torque prediction value and the stator flux prediction value based on the current and voltage of the permanent magnet synchronous motor, the stator flux observation value, and the candidate voltage vector of the inverter, using a hyperlocal prediction model based on a composite discrete extended state observer. The voltage vector determination module is used to determine the optimal voltage vector from the candidate voltage vectors based on the torque reference value, the preset stator flux linkage reference value, the torque prediction value, and the stator flux linkage prediction value, and to control the torque of the permanent magnet synchronous motor based on the optimal voltage vector.

Images