A method and device for maximum torque per ampere control of a synchronous reluctance motor
By tracking the local electromagnetic model online, the problem of insufficient control accuracy of synchronous reluctance motors under magnetic circuit saturation and cross-coupling was solved, achieving higher control accuracy and lower stator current consumption, and adapting to changes in different operating points.
Patent Information
- Application Number
- CN202611276291.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-08-21
- Publication Date
- 2026-11-10
AI Technical Summary
Existing maximum torque-to-current ratio control methods for synchronous reluctance motors suffer from insufficient control accuracy and poor adaptability when faced with magnetic circuit saturation and cross-coupling phenomena. In particular, traditional methods struggle to accurately reflect nonlinear electromagnetic characteristics online.
A control method based on an online tracking local electromagnetic model is adopted. By constructing a local electromagnetic model that includes cubic self-axis nonlinear terms and cross-coupling terms on the d-axis and q-axis, and using the least squares method with linear regression equations and forgetting factors to update the model parameters online, the optimal current angle can be directly obtained, reducing the amount of online calculation and improving control accuracy and adaptability.
It achieves higher control accuracy and lower stator current at different operating points, reduces the optimal current angle deviation caused by magnetic circuit nonlinearity, and improves the maximum torque-to-current ratio control effect of synchronous reluctance motor.
Smart Images

Figure CN123074355A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of motor drive control technology, and more specifically, relates to a method and device for controlling the maximum torque-to-current ratio of a synchronous reluctance motor based on an online tracking local electromagnetic model. Background Technology
[0002] Synchronous reluctance motors (SRRMs) offer advantages such as the absence of permanent magnets, low cost, high efficiency, and suitability for high-speed operation, making them promising for industrial servo and high-efficiency drive applications. However, SRRMs exhibit significant magnetic circuit saturation and d-axis / q-axis cross-coupling, causing their d-axis and q-axis electromagnetic parameters to vary considerably with the operating point. Under these conditions, if a fixed inductance model is still used to derive the maximum torque-to-current ratio control law, the actual optimal current angle will deviate from the theoretical value, leading to increased current, increased copper losses, and decreased efficiency.
[0003] In existing technologies, there are three main technical approaches for nonlinear MTPA control of synchronous reluctance motors. The first approach involves offline experimental modeling and generating lookup tables, then outputting the optimal current angle during runtime through table lookup or interpolation. While this method can account for nonlinearity to some extent, it requires significant offline testing workload and storage space, and recalibration is necessary after parameter drift. The second approach involves online identification of inductor parameters and calculation of the optimal current angle under traditional MTPA conditions. However, most methods only update the apparent inductance or a portion of the incremental inductance, making it difficult to systematically reflect the combined effects of saturation and cross-coupling. The third approach is the online perturbation search method, which directly approximates the optimal angle through micro-perturbations. However, this method suffers from additional ripple, limited convergence speed, and complex real-time implementation.
[0004] Existing research has paid relatively little attention to using local models to describe the nonlinear electromagnetic characteristics of a synchronous reluctance motor in the neighborhood of its current operating point, especially for schemes that simultaneously satisfy the conditions of "preservation of local nonlinear information" and "direct solution of the optimal current angle." Most existing methods still rely on offline table lookup, fixed parameter approximation, or online numerical search based on general high-order nonlinear models, making it difficult to simultaneously balance model accuracy, online computational load, and engineering real-time performance. Summary of the Invention
[0005] To address the shortcomings of related technologies, the present invention aims to provide a method and apparatus for controlling the maximum torque-to-current ratio of a synchronous reluctance motor based on an online tracking local electromagnetic model. The present invention aims to solve the problems of insufficient accuracy and poor adaptability of the maximum torque-to-current ratio control based on the fixed inductance model or offline lookup table model caused by significant magnetic circuit saturation and cross-coupling of the synchronous reluctance motor.
[0006] To achieve the above objectives, in a first aspect, the present invention provides a method for controlling the maximum torque-to-current ratio of a synchronous reluctance motor, comprising: For each time window, the motor operation data corresponding to that time window is substituted into the preset linear regression equation to obtain the linear regression equation for that time window. The preset linear regression equation is obtained by substituting the local electromagnetic model into the dq-axis voltage equation of the motor and integrating both sides of the dq-axis voltage equation; the local electromagnetic model of the synchronous reluctance motor is: ψ d , ψ q For the d-axis and q-axis flux linkages of a synchronous reluctance motor, i d , i q For d-axis and q-axis currents, l d , l dd , l q , l qq , l c These are the parameters of the local electromagnetic model. l d and l q d and q are the linear self-axis flux linkages of the axis, respectively. l dd and l qq These are the cubic self-axis nonlinearity coefficients of the d and q axes, respectively, used to characterize self-axis saturation. l c The local electromagnetic model is used to characterize the d-axis and q-axis cross-coupling effect; it is used to characterize the magnetic circuit self-saturation effect and cross-coupling effect in the neighborhood of the current operating point; a time window includes N control cycles, where N is a positive integer greater than 20; the motor operation data includes the d-axis voltage, q-axis voltage, d-axis current, q-axis current, and electric angular velocity during the operation of the synchronous reluctance motor; The linear regression equations corresponding to all historical time windows are combined to obtain a set of linear regression equations. Solving the set of linear regression equations yields the target local electromagnetic model parameters that reflect the local electromagnetic characteristics of the current working point. Substituting the parameters of the target local electromagnetic model into the expression of the local electromagnetic model, the d-axis and q-axis magnetic flux linkages are obtained. ψ d and ψ q ;Will ψ d and ψ qSubstituting the values into the electromagnetic torque equation of the synchronous reluctance motor, the optimal current angle corresponding to the maximum torque-to-current ratio is obtained, and then the d-axis and q-axis current commands for each control cycle are calculated. The d-axis and q-axis current commands are input into the synchronous reluctance motor current closed-loop controller to achieve maximum torque-to-current ratio control.
[0007] Optionally, the process of constructing the preset linear regression equation includes: In the k-th time window [ t k 1, t k The corresponding linear regression equation is: ; in, The local electromagnetic model parameter vector and the window observation vector are shown. b k and window regression matrix A k The expression is:
[0008]
[0009] Among them, among them, R s This refers to the stator resistance of the synchronous reluctance motor. , The current quantity in the k-th time window includes i d , i q , , , , Current f The average values of the multiple sampling points at the beginning and end of the window are respectively and .
[0010] Optionally, the linear regression equations are solved using the least squares method with a forgetting factor, including: Regarding the first The window regression matrix formed by the time windows and window observation vector ,according to and Calculate the cumulative matrix corresponding to all historical time windows. and cumulative vector ;in , Forgetting factor; According to the cumulative matrix and cumulative vector By solving Obtain the parameter vector Thus, the local electromagnetic model parameters of the target are obtained; where , The regularization coefficient is . For the cumulative matrix Identity matrices of the same order.
[0011] Optionally, the step of substituting the target local electromagnetic model parameters into the expression of the local electromagnetic model will... ψ d and ψ q Substituting into the electromagnetic torque equation of the synchronous reluctance motor, the optimal current angle corresponding to the maximum torque-to-current ratio is obtained, including: Let the local electromagnetic model in i d = I s cos , i q = I s sin ;in, I s Given a current amplitude, β The current angle is the current vector relative to the d-axis. Substituting the parameters of the target local electromagnetic model into the expression of the local electromagnetic model, the d-axis and q-axis magnetic flux linkages are obtained. ψ d and ψ q and will ψ d and ψ q Substituting the electromagnetic torque equation of the synchronous reluctance motor into the equation, the optimal current angle is obtained by solving the equation with the objective of satisfying the maximum torque-to-current ratio condition; wherein, the maximum torque-to-current ratio condition is to satisfy the electromagnetic torque of the synchronous reluctance motor with respect to the current angle. The result after taking the partial derivative is zero.
[0012] Optionally, the step of solving the equation to obtain the optimal current angle with the objective of satisfying the maximum torque-current ratio condition includes: Convert the maximum torque-current ratio condition to a condition related to the current angle. β The low-order algebraic equations are solved to obtain candidate solutions; from the candidate solutions, the candidate solution that is within the preset current angle range and has the largest corresponding electromagnetic torque is selected as the optimal current angle.
[0013] Optionally, it also includes: acquiring motor operation data for each time window in real time and updating the target local electromagnetic model parameters in real time.
[0014] Optionally, the process of updating the target's local electromagnetic model parameters in real time also includes: The partial derivative of the electromagnetic torque of the synchronous reluctance motor with respect to the current optimal current angle is calculated based on the updated target local electromagnetic model parameters and filtered. When the absolute value of the calculation result is less than the first preset threshold, it is determined that the synchronous reluctance motor has run to the local optimal region, the update of the local electromagnetic model parameters is paused, and the amplitude of the current sampled current vector is latched. When the deviation between the sampled current vector magnitude in subsequent control cycles and the latched sampled current vector magnitude exceeds the second preset threshold, the update of the local electromagnetic model parameters is restarted.
[0015] Secondly, the present invention also provides a maximum torque-to-current ratio control device for a synchronous reluctance motor, comprising: The linear regression equation construction module is used to take the motor operation data corresponding to each time window, substitute it into the preset linear regression equation, and obtain the linear regression equation for the corresponding time window. The preset linear regression equation is obtained by substituting the local electromagnetic model into the dq-axis voltage equation of the motor and integrating both sides of the dq-axis voltage equation; the local electromagnetic model of the synchronous reluctance motor is: ψ d , ψ q For the d-axis and q-axis flux linkages of a synchronous reluctance motor, i d , i q For d-axis and q-axis currents, l d , l dd , l q , l qq , l c These are the parameters of the local electromagnetic model. l d and l q d and q are the linear self-axis flux linkages of the axis, respectively. l dd and l qq These are the cubic self-axis nonlinearity coefficients of the d and q axes, respectively, used to characterize self-axis saturation. l cThe local electromagnetic model is used to characterize the d-axis and q-axis cross-coupling effect; it is used to characterize the magnetic circuit self-saturation effect and cross-coupling effect in the neighborhood of the current operating point; a time window includes N control cycles, where N is a positive integer greater than 20; the motor operation data includes the d-axis voltage, q-axis voltage, d-axis current, q-axis current, and electric angular velocity during the operation of the synchronous reluctance motor; The parameter update module is used to obtain a set of linear regression equations by combining the linear regression equations corresponding to all historical time windows, and to solve the set of linear regression equations to obtain the target local electromagnetic model parameters that reflect the local electromagnetic characteristics of the current working point. The optimal current angle direct solution control module is used to substitute the target local electromagnetic model parameters into the expression of the local electromagnetic model to obtain the d-axis and q-axis flux linkages. ψ d and ψ q ;Will ψ d and ψ q Substituting the values into the electromagnetic torque equation of the synchronous reluctance motor, the optimal current angle corresponding to the maximum torque-to-current ratio is obtained, and then the d-axis and q-axis current commands for each control cycle are calculated. The d-axis and q-axis current commands are input into the synchronous reluctance motor current closed-loop controller to achieve maximum torque-to-current ratio control.
[0016] Thirdly, the present invention also provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the method as described in any one of the first aspects.
[0017] Fourthly, the present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method as described in any one of the first aspects.
[0018] Compared with the prior art, the above-described technical solutions conceived in this invention can achieve the following beneficial effects: 1. This invention provides a method for controlling the maximum torque-to-current ratio of a synchronous reluctance motor. By establishing a local electromagnetic model including cubic self-axis nonlinear terms and cross-coupling terms on the d-axis and q-axis, and updating the model parameters based on real-time operating data of the synchronous reluctance motor, the model can characterize the magnetic circuit saturation and cross-coupling characteristics in the neighborhood of the current operating point. This reduces the deviation of the optimal current angle caused by magnetic circuit nonlinearity when using fixed inductance parameters, thereby improving the accuracy of maximum torque-to-current ratio control at different operating points and reducing the stator current required to generate the same torque.
[0019] 2. The present invention provides a maximum torque-to-current ratio (MTPA) control method for a synchronous reluctance motor. It constructs a regression matrix and observation vector for a continuous time window, and reduces the impact of noise and sampling fluctuations on individual boundary sampling points by averaging multiple sampling points at the beginning and end of the window. Simultaneously, it attenuates historical regression information window by window using a forgetting factor, giving recent sampling data higher weight in model parameter identification. This allows the local electromagnetic model to be continuously updated as the motor's operating point changes, improving the adaptability of MTPA control to changes in load and operating state.
[0020] 3. The present invention provides a maximum torque-to-current ratio control method for a synchronous reluctance motor. When the torque deflection meets the preset convergence condition, the local electromagnetic model parameters are frozen to reduce repeated parameter fluctuations near the local optimum caused by measurement noise. When the sampled current amplitude changes more than the latched current amplitude at the time of freezing than the preset threshold, the model parameter update is restarted, thereby taking into account both the control stability at the stable operating point and the model tracking capability after significant changes in the operating point.
[0021] 4. The present invention provides a maximum torque-to-current ratio control method for a synchronous reluctance motor. By utilizing the expression of the constructed local electromagnetic model, the maximum torque-to-current ratio condition is transformed into a low-order algebraic equation with respect to the current angle variable. The MTPA current angle is obtained based on the online updated model parameters, avoiding global numerical search within the control cycle, which helps to reduce the amount of online calculation and meets the real-time operation requirements of the embedded motor controller. Attached Figure Description
[0022] Figure 1 A flowchart illustrating the overall process of a method for controlling the maximum torque-to-current ratio of a synchronous reluctance motor, as provided in an embodiment of the present invention. Figure 2 This is a schematic diagram of online identification of local electromagnetic models based on time windows in an embodiment of the present invention; Figure 3 The figure shows a comparison between a synchronous reluctance motor maximum torque-to-current ratio control method provided in this embodiment of the invention and the traditional MTPA method that does not consider magnetic circuit nonlinearity. Detailed Implementation
[0023] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.
[0024] The following description, in conjunction with a preferred embodiment, illustrates the content involved in the above embodiments.
[0025] To address the existing technical problems, how to construct a local electromagnetic model that can characterize the saturation and cross-coupling characteristics of the magnetic circuit in the vicinity of the current operating point, and directly express the optimal current angle corresponding to the maximum torque-current ratio, and further realize the online tracking of the model parameters, is a key issue that has been relatively little studied but has significant research value in the current control of synchronous reluctance motors (MTPA).
[0026] This invention provides a method for controlling the maximum torque-to-current ratio of a synchronous reluctance motor. A specially structured local electromagnetic model is constructed around the current operating point of the synchronous reluctance motor. This model characterizes both local self-axis saturation and cross-coupling characteristics, and allows the optimal current angle corresponding to the maximum torque-to-current ratio to be directly expressed as an explicit function, semi-explicit function, or solution of a low-order algebraic equation relating model parameters and the current amplitude. During operation, the local model parameters are tracked online using linear regression based on a time window and a least-squares identification method with a forgetting factor. The optimal current angle and d-axis and q-axis current commands are then directly obtained from the updated local model to achieve maximum torque-to-current ratio control of the synchronous reluctance motor.
[0027] The core of this scheme is to construct a local electromagnetic model with a special structure around the current operating point of the synchronous reluctance motor, identify the parameters of the local model online through voltage, current and electric angular velocity data within a time window, and directly obtain the maximum torque-to-current ratio and optimal current angle considering the nonlinearity of the magnetic circuit based on the local model, thereby generating d-axis and q-axis current commands to achieve online tracking control of the actual MTPA operating point.
[0028] like Figure 1 As shown, the present invention provides a method for controlling the maximum torque-to-current ratio of a synchronous reluctance motor, comprising: S1. For each time window, substitute the motor operation data corresponding to that time window into the preset linear regression equation to obtain the linear regression equation for the corresponding time window. The preset linear regression equation is obtained by substituting the local electromagnetic model into the dq-axis voltage equation of the motor and integrating both sides of the dq-axis voltage equation; the local electromagnetic model of the synchronous reluctance motor is: ψ d , ψ q For the d-axis and q-axis flux linkages of a synchronous reluctance motor, i d , i q For d-axis and q-axis currents, l d , l dd ,l q , l qq , l c These are the parameters of the local electromagnetic model. l d and l q d and q are the linear self-axis flux linkages of the axis, respectively. l dd and l qq These are the cubic self-axis nonlinearity coefficients of the d and q axes, respectively, used to characterize self-axis saturation. l c The local electromagnetic model is used to characterize the d-axis and q-axis cross-coupling effect; it is used to characterize the magnetic circuit self-saturation effect and cross-coupling effect in the neighborhood of the current operating point; a time window includes N control cycles, where N is a positive integer greater than 20; the motor operation data includes the d-axis voltage, q-axis voltage, d-axis current, q-axis current, and electric angular velocity during the operation of the synchronous reluctance motor; The local electromagnetic model satisfies reciprocity and axisymmetric structural constraints; wherein, the condition for satisfying reciprocity is:
[0029] at the same time, right For odd functions, It is an even function. right For odd functions, It is an even function, reflecting the axially symmetric form used.
[0030] S2. Based on the linear regression equations corresponding to all historical time windows, a set of linear regression equations is obtained. The set of linear regression equations is solved to obtain the target local electromagnetic model parameters that reflect the local electromagnetic characteristics of the current working point. S3. Substitute the parameters of the target local electromagnetic model into the expression of the local electromagnetic model to obtain the d-axis and q-axis magnetic flux linkages. ψ d and ψ q ;Will ψ d and ψ q Substituting the values into the electromagnetic torque equation of the synchronous reluctance motor, the optimal current angle corresponding to the maximum torque-to-current ratio is obtained, and then the d-axis and q-axis current commands for each control cycle are calculated. The d-axis and q-axis current commands are input into the synchronous reluctance motor current closed-loop controller to achieve maximum torque-to-current ratio control.
[0031] In this scheme, instead of directly using long-time integration to obtain instantaneous flux linkage, a window observation vector is constructed using integral form within the time window. The flux linkage increment term is constructed by averaging multiple sampling points at the beginning and end of the window, thus forming the window regression matrix and window observation vector for a single time window. The window regression matrix and window observation vector formed by different time windows are not vertically concatenated. Instead, the matrix contribution and vector contribution of each time window to the least squares equation are recursively accumulated into the cumulative matrix and cumulative vector to reduce the impact of integral drift, sampling noise, and current ripple on the parameter identification results, and to avoid storing all historical window data for a long time.
[0032] In this way, the current time window data is given a greater weight than the historical time window data, thereby improving the model parameters' ability to track changes in the current operating conditions.
[0033] During operation, the controller collects the d-axis and q-axis voltages of the synchronous reluctance motor in real time. v d , v q d-axis and q-axis currents i d , i q Electric angular velocity oh e and stator resistance R s Information such as... The voltage equation of the synchronous reluctance motor in the dq coordinate system is:
[0034] To avoid integral drift and DC bias problems caused by direct long-term integration, this invention does not use instantaneous flux direct regression, but instead constructs a regression equation within a time window.
[0035] Specifically, in the k-th time window [ t k 1, t k Integrating the above dq-axis voltage equation, we obtain:
[0036] To further reduce sampling noise, this invention employs the averaging of multiple sampling points at the beginning and end of the window to construct the flux linkage increment term. Instead of directly using the single-point value at the window boundary, let the average value of the current f at multiple sampling points at the beginning and end of the window be... , Then we have:
[0037] Substituting the local electromagnetic model into the above integral equation, the k-th time window can be rearranged as a function of the parameter vector. The linear regression equation:
[0038] in, A k For the first k The window regression matrix is constructed using time windows, and each column corresponds to the regression coefficient of the parameter within that time window. b k For the first k The window observation vector, constructed using time windows, consists of elements derived from the integration of the d-axis and q-axis voltage equations within the window. In other words, A k This reflects the linear contribution of each parameter to be identified to the window integral equation within that time window. b k It reflects the actual observation results formed by voltage, current, and electric angular velocity data within this time window.
[0039] Specifically, in this embodiment, b k Indicates the first k The window observation vector obtained by integrating the dq-axis voltage equation within a time window is expressed as follows:
[0040]
[0041] in, A k It contains two rows and five columns, corresponding to the dq-axis integral voltage equation and the parameters to be identified, respectively. l d , l dd , l q , l qq , l c The linear coefficients. In the first... k A time window is formed: ,in, .
[0042] In this embodiment, each time the first... k Data processing for each time window is used to form the window regression matrix corresponding to that time window. A k and window observation vector b kTo combine the regression information from multiple windows and achieve least squares identification of model parameters with a forgetting factor, this embodiment does not vertically concatenate the window regression matrix and window observation vector formed by different time windows, nor does it store all historical window data for a long time. Instead, it recursively accumulates the contribution of each time window to the least squares equation into the cumulative matrix and cumulative vector.
[0043] like Figure 2 As shown, this invention uses multiple adjacent time windows to construct a regression matrix. Each window contains... v d , v q , i d , i q , oh e The data is averaged using multiple sampling points at the beginning and end of the window to suppress sampling noise; the regression information of the window is recursively accumulated after each time window is completed; the finite window length is used to constrain integral drift.
[0044] Figure 2 The window regression information recursive accumulation in the context refers to the accumulation of the first... k The window regression matrix formed by the time windows A k and window observation vector b k Convert to the matrix contribution of this time window to the least squares equation Sum vector contribution Then, it is updated by recursively combining the accumulated information from the previous time window.
[0045] In this embodiment, when the cumulative time window reaches a preset number and the preset parameter solution cycle is reached, the Cholesky decomposition is used to solve the above linear equation system. When the linear equation system fails to be solved due to non-positive definite matrices or numerical anomalies, the effective local electromagnetic model parameters of the previous solution cycle remain unchanged. After successful solution, the obtained model parameters can be low-pass smoothed, and then the smoothed model parameters can be used for subsequent MTPA current angle calculations.
[0046] The above recursive update process is equivalent to solving the following weighted least squares problem with forgetting factors and regularization terms:
[0047] in, l It is the forgetting factor, and 0 < l ≤1, is the regularization coefficient, and By introducing a forgetting factor, the influence of earlier time windows on parameter identification results gradually decays over time, giving data from recent time windows a relatively high weight. This allows the identified parameters to preferentially reflect local electromagnetic characteristic changes in the neighborhood of the current operating point, enabling online tracking of the operating status.
[0048] To implement the weighted least squares identification described above online, a cumulative matrix is set. and cumulative vector and initialized to: ; ; Regarding the first The window regression matrix formed by the time windows and window observation vector Update the cumulative matrix according to the following formula and cumulative vector : ; ; in, , Forgetting factor; According to the cumulative matrix and cumulative vector By solving the following system of linear equations:
[0049] Get the first k Estimated values of local electromagnetic model parameters corresponding to each time window: ; in, , The regularization coefficient is . For the cumulative matrix A 5th-order identity matrix of the same order.
[0050] Optionally, the step of obtaining the optimal current angle corresponding to the maximum torque-current ratio based on the updated local electromagnetic model parameters includes: With a fixed current amplitude I s Under the condition, let i d = I s cos , i q = I s sin ;in, βLet be the current angle relative to the d-axis. The expression of the local electromagnetic model is updated based on the target local electromagnetic model parameters, and this expression is substituted into the electromagnetic torque equation of the synchronous reluctance motor. The optimal current angle is obtained by solving the equation with the maximum torque-to-current ratio condition as the objective; where β The current angle is the current vector relative to the d-axis, and the maximum torque-to-current ratio condition is the electromagnetic torque of the synchronous reluctance motor with respect to the current angle. β Find the partial derivative and set it to zero.
[0051] Because the local electromagnetic model employs a special structure, the condition equations can be further rearranged into low-order algebraic equations about β, or into equations about cos(β). 2β The low-order algebraic equation of ) is used to directly obtain the optimal current angle. Furthermore, to simplify the expression, intermediate variables K and M are constructed from the local electromagnetic model parameters and current amplitude, and the maximum torque-current ratio condition equation is rearranged to be about cos(2 β The quadratic algebraic equation of ) is used to directly obtain the optimal current angle. β .
[0052] The maximum torque-to-current ratio condition is the ratio of the electromagnetic torque of the synchronous reluctance motor to the current angle. Find the partial derivative and set it to zero; The process of finding the optimal current angle includes: solving the current angle as a function of the maximum torque-current ratio condition transformation. β The low-order algebraic equations are used to obtain candidate solutions; from the candidate solutions, the candidate solution that is located within the preset current angle range and has the largest corresponding electromagnetic torque is selected as the optimal current angle.
[0053] Specifically, substituting the d-axis and q-axis current parameters into the local electromagnetic model, we have:
[0054] The expression for the electromagnetic torque of a synchronous reluctance motor is:
[0055] in, p Let be the number of pole pairs of the synchronous reluctance motor. Substituting the parameterized expression and rearranging, we get:
[0056] To simplify the expression, an intermediate variable is introduced:
[0057] right Taking the partial derivative and setting it to zero, we can obtain the condition equation for the maximum torque-current ratio:
[0058] Again The MTPA condition can then be summarized as follows:
[0059] Further calculation yields:
[0060] Thus, the optimal current angle is obtained: .
[0061] When the absolute value of M is greater than the preset numerical threshold, the value is calculated according to the above radical formula. x Only retain those that meet the requirements. x Candidate roots ∈ [-1,1] whose corresponding current angles are within the preset current angle range. When the absolute value of M is less than or equal to the preset numerical threshold, π / 4 is used as the candidate current angle; when there are no candidate roots that meet the feasible conditions, the preset safe current angle or the effective current angle of the previous control cycle is used. When multiple candidate roots meet the feasible conditions, their corresponding torque values can be further compared, or continuous root selection can be achieved by combining the angle of the previous cycle to obtain the final optimal current angle. Then, according to: Reference values for the d-axis and q-axis currents of the synchronous reluctance motor are generated and input into the current closed-loop controller for control, thereby tracking the true maximum torque-to-current ratio operating point considering the nonlinearity of the magnetic circuit online.
[0062] To suppress sudden changes in the optimal current angle caused by parameter variations or operating point changes, continuity constraints, step limits, or low-pass filtering can be applied to the obtained optimal current angle. Alternatively, the final output d-axis and q-axis current commands can be smoothed to improve the stability of inverter drive and current closed-loop control, and enhance the system control robustness.
[0063] Furthermore, based on the obtained optimal current angle, the d-axis and q-axis current commands of the synchronous reluctance motor are generated and input into the current closed-loop controller, thereby tracking the actual MTPA operating point considering the nonlinearity of the magnetic circuit online.
[0064] The process of updating the local electromagnetic model parameters also includes parameter update freezing and re-triggering steps.
[0065] Specifically, under the current output current angle, based on the updated local electromagnetic model parameters and the current given current amplitude, the electromagnetic torque of the synchronous reluctance motor and the partial derivative of the electromagnetic torque with respect to the current angle are calculated, and the partial derivative is low-pass filtered to obtain the filtered torque partial derivative.
[0066] When the filtered torque deflection enters the threshold region from outside the threshold region defined by the first preset threshold, it is determined that the current operating point has entered the local optimal region near the operating point with the maximum torque-to-current ratio. The update of local electromagnetic model parameters is then paused, and the current sampled current vector magnitude is latched. Preferably, parameter update freezing is triggered when the filtered torque deflection enters the threshold region from the negative side. More preferably, when parameter update freezing is triggered, accumulated window identification information is cleared, and the currently valid local electromagnetic model parameters remain unchanged.
[0067] When the increase in the magnitude of the sampled current vector in subsequent control cycles relative to the magnitude of the latched sampled current vector exceeds the second preset threshold, the parameter update freeze is lifted, and the time window data accumulation and local electromagnetic model parameter update are restarted to allow the local electromagnetic model to adapt to the changed operating point.
[0068] By using the above parameter update freeze and re-trigger mechanism, parameter updates are paused when the current operating point is close to the maximum torque-to-current ratio operating point, which can reduce the repeated fluctuations of local electromagnetic model parameters caused by sampling noise and current ripple; parameter updates are restarted when the current amplitude changes significantly, which can balance the parameter stability at the stable operating point and the model tracking capability after the operating point changes.
[0069] The first preset threshold can be an absolute threshold corresponding to the absolute value of the torque partial derivative, or a normalized threshold corresponding to the normalized torque partial derivative. The normalized torque partial derivative can be expressed as: or =| g f | / max(| T e |, e ); in, g f This is the torque partial derivative after low-pass filtering. T e The electromagnetic torque is calculated based on the current output current angle. e A positive number set to prevent the denominator from being zero.
[0070] The second preset threshold can be an absolute deviation threshold for the current amplitude, or a proportional threshold relative to the rated current of the motor or the amplitude of the latched current vector. The first preset threshold should be greater than the torque partial derivative fluctuation range caused by steady-state sampling noise and current ripple, and sufficient to indicate that the current operating point has entered the local optimum region; the second preset threshold should be greater than the current amplitude fluctuation range caused by normal sampling noise and current ripple, and sufficient to indicate that the current operating point has left the applicable region of the original local electromagnetic model.
[0071] For example, when using normalized torque partial derivative as the freezing criterion, the first preset threshold can be set to 0.02; the second preset threshold can be set to 0.5 A. The above thresholds are only examples and can be set according to the motor's rated current, torque and flux linkage scale, sampling noise, current ripple, filter bandwidth, and desired re-identification sensitivity.
[0072] Figure 3 The diagram shows a comparison between the method of this invention and the traditional MTPA method that does not consider magnetic circuit nonlinearity. The embodiment was implemented on a 5.5 kW synchronous reluctance motor platform with a sampling frequency of 10 kHz. By comparing the current amplitudes of the two control methods under the same load, the current reduction of more than 10% under heavy load further illustrates that in the medium-to-high load region, the method of this invention can more accurately track the true optimal current angle, thus resulting in a significantly lower stator current than the traditional method. This verifies the effectiveness of the online tracking of the local electromagnetic model and the direct solution of the optimal angle method.
[0073] In this embodiment, a local electromagnetic model is established, including cubic self-axis nonlinear terms on the d and q axes and cross-coupling terms. The model parameters are updated based on real-time operating data of the synchronous reluctance motor, enabling the model to characterize the magnetic circuit saturation and cross-coupling characteristics in the neighborhood of the current operating point. This addresses the problem of determining the MTPA current angle considering magnetic circuit saturation and cross-coupling based on real-time motor operating data and adapting to changes in operating point without relying on high-density offline table lookup and online global numerical search across all operating conditions. This reduces the deviation of the optimal current angle caused by magnetic circuit nonlinearity when using fixed inductance parameters, thereby improving the accuracy of maximum torque-to-current ratio control at different operating points and reducing the stator current required to generate the same torque.
[0074] Example 2 The present invention also provides a maximum torque-to-current ratio control device for a synchronous reluctance motor, comprising: The linear regression equation construction module is used to take the motor operation data corresponding to each time window, substitute it into the preset linear regression equation, and obtain the linear regression equation for the corresponding time window. The preset linear regression equation is obtained by substituting the local electromagnetic model into the dq-axis voltage equation of the motor and integrating both sides of the dq-axis voltage equation; the local electromagnetic model of the synchronous reluctance motor is: ψ d , ψ q For the d-axis and q-axis flux linkages of a synchronous reluctance motor, i d , i q For d-axis and q-axis currents, ld , l dd , l q , l qq , l c These are the parameters of the local electromagnetic model. l d and l q d and q are the linear self-axis flux linkages of the axis, respectively. l dd and l qq These are the cubic self-axis nonlinearity coefficients of the d and q axes, respectively, used to characterize self-axis saturation. l c The local electromagnetic model is used to characterize the d-axis and q-axis cross-coupling effect; it is used to characterize the magnetic circuit self-saturation effect and cross-coupling effect in the neighborhood of the current operating point; a time window includes N control cycles, where N is a positive integer greater than 20; the motor operation data includes the d-axis voltage, q-axis voltage, d-axis current, q-axis current, and electric angular velocity during the operation of the synchronous reluctance motor; The parameter update module is used to obtain a set of linear regression equations by combining the linear regression equations corresponding to all historical time windows, solve the set of linear regression equations, and obtain the target local electromagnetic model parameters that reflect the local electromagnetic characteristics of the current working point. The optimal current angle direct solution control module is used to substitute the target local electromagnetic model parameters into the expression of the local electromagnetic model to obtain the d-axis and q-axis flux linkages. ψ d and ψ q ;Will ψ d and ψ q Substituting the values into the electromagnetic torque equation of the synchronous reluctance motor, the optimal current angle corresponding to the maximum torque-to-current ratio is obtained, and then the d-axis and q-axis current commands for each control cycle are calculated. The d-axis and q-axis current commands are input into the synchronous reluctance motor current closed-loop controller to achieve maximum torque-to-current ratio control.
[0075] like Figure 1 As shown, the sampling module acquires the d-axis voltage of the synchronous reluctance motor in real time. v d q-axis voltage v q d-axis current i d q-axis current i q and electric angular velocity oh e The system collects operational information and sends it to the local electromagnetic model construction and online identification module. Based on the current-magnetic flux relationship near the current operating point, the local electromagnetic model construction module establishes a local electromagnetic model that characterizes local magnetic circuit saturation and cross-coupling effects. Subsequently, the time window regression equation construction module uses voltage, current, and electric angular velocity data within a time window to construct a window regression matrix and observation vector through integral form, and then superimposes multiple time windows to form the overall regression equation. The least squares parameter update module with a forgetting factor tracks the local electromagnetic model parameters online according to the overall regression equation, enabling the model parameters to be corrected in real time as the operating point changes.
[0076] After obtaining the updated local electromagnetic model parameters, the optimal current angle direct solution module utilizes the special structure of the local model to transform the maximum torque-current ratio condition into a low-order algebraic equation about the current angle under a fixed current amplitude, thereby directly obtaining the optimal current angle corresponding to the current operating point. The current command generation module then converts the optimal current angle into d-axis and q-axis current reference values and inputs them into the current closed-loop controller and inverter for control, thereby driving the synchronous reluctance motor to operate near the true MTPA operating point considering magnetic circuit nonlinearity. Furthermore, the real-time operating status of the synchronous reluctance motor is returned to the sampling module via a feedback path for updating the local model parameters and correcting the optimal current angle at the next moment. Therefore, the entire system forms a dynamic closed loop of "sampling—identification—solving—control—re-sampling".
[0077] The synchronous reluctance motor maximum torque-current ratio control device provided in this embodiment of the invention is used to execute the synchronous reluctance motor maximum torque-current ratio control method provided in Embodiment 1, and has the same beneficial effects.
[0078] Example 3 The present invention also provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the method as described in any one of Embodiment 1.
[0079] The electronic device can be a desktop computer, laptop, handheld computer, or cloud server, etc. The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The memory can be used to store computer programs and / or modules. The processor performs various functions of the electronic device by running or executing the computer programs and / or modules stored in the memory, and by accessing data stored in the memory.
[0080] Example 4 The present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method as described in any one of Embodiment 1.
[0081] Specifically, the memory may include high-speed random access memory, as well as non-volatile memory, such as hard disks, RAM, plug-in hard disks, smart media cards (SMC), secure digital (SD) cards, flash cards, at least one disk storage device, flash memory device, or other volatile solid-state storage devices.
[0082] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.
Claims
1. A method for controlling the maximum torque-to-current ratio of a synchronous reluctance motor, characterized in that, include: For each time window, the motor operation data corresponding to that time window is substituted into the preset linear regression equation to obtain the linear regression equation for that time window. The preset linear regression equation is obtained by substituting the local electromagnetic model into the dq-axis voltage equation of the motor and integrating both sides of the dq-axis voltage equation; the local electromagnetic model of the synchronous reluctance motor is: ψ d , ψ q For the d-axis and q-axis flux linkages of a synchronous reluctance motor, i d , i q For d-axis and q-axis currents, l d , l dd , l q , l qq , l c These are the parameters of the local electromagnetic model. l d and l q d and q are the linear self-axis flux linkages of the axis, respectively. l dd and l qq These are the cubic self-axis nonlinearity coefficients of the d and q axes, respectively, used to characterize self-axis saturation. l c The local electromagnetic model is used to characterize the d-axis and q-axis cross-coupling effect; it is used to characterize the magnetic circuit self-saturation effect and cross-coupling effect in the neighborhood of the current operating point; a time window includes N control cycles, where N is a positive integer greater than 20; the motor operation data includes the d-axis voltage, q-axis voltage, d-axis current, q-axis current, and electric angular velocity during the operation of the synchronous reluctance motor; The linear regression equations corresponding to all historical time windows are combined to obtain a set of linear regression equations. Solving the set of linear regression equations yields the target local electromagnetic model parameters that reflect the local electromagnetic characteristics of the current working point. Substituting the parameters of the target local electromagnetic model into the expression of the local electromagnetic model, the d-axis and q-axis magnetic flux linkages are obtained. ψ d and ψ q ;Will ψ d and ψ q Substituting the values into the electromagnetic torque equation of the synchronous reluctance motor, the optimal current angle corresponding to the maximum torque-to-current ratio is obtained, and then the d-axis and q-axis current commands for each control cycle are calculated. The d-axis and q-axis current commands are input into the synchronous reluctance motor current closed-loop controller to achieve maximum torque-to-current ratio control.
2. The method as described in claim 1, characterized in that, The process of constructing the preset linear regression equation includes: In the k-th time window [ t k 1, t k The corresponding linear regression equation is: ; in, The local electromagnetic model parameter vector and the window observation vector are shown. b k and window regression matrix A k The expression is: Among them, among them, R s This refers to the stator resistance of the synchronous reluctance motor. , The current quantity in the k-th time window includes i d , i q , , , , Current f The average values of the multiple sampling points at the beginning and end of the window are respectively and .
3. The method as described in claim 2, characterized in that, The linear regression equations are solved using the least squares method with a forgetting factor, including: Regarding the first The window regression matrix formed by the time windows and window observation vector ,according to and Calculate the cumulative matrix corresponding to all historical time windows. and cumulative vector ;in , Forgetting factor; According to the cumulative matrix and cumulative vector By solving Obtain the parameter vector Thus, the local electromagnetic model parameters of the target are obtained; where , The regularization coefficient is . For the cumulative matrix Identity matrices of the same order.
4. The method as described in claim 1, characterized in that, The step of substituting the target local electromagnetic model parameters into the expression of the local electromagnetic model will... ψ d and ψ q Substituting into the electromagnetic torque equation of the synchronous reluctance motor, the optimal current angle corresponding to the maximum torque-to-current ratio is obtained, including: Let the local electromagnetic model in i d = I s cos , i q = I s sin ;in, I s Given a current amplitude, β The current angle is the current vector relative to the d-axis. Substituting the parameters of the target local electromagnetic model into the expression of the local electromagnetic model, the d-axis and q-axis magnetic flux linkages are obtained. ψ d and ψ q and will ψ d and ψ q Substituting the electromagnetic torque equation of the synchronous reluctance motor into the equation, the optimal current angle is obtained by solving the equation with the objective of satisfying the maximum torque-to-current ratio condition; wherein, the maximum torque-to-current ratio condition is to satisfy the electromagnetic torque of the synchronous reluctance motor with respect to the current angle. The result after taking the partial derivative is zero.
5. The method as described in claim 4, characterized in that, The process of solving the equation to obtain the optimal current angle with the objective of satisfying the maximum torque-to-current ratio condition includes: Convert the maximum torque-current ratio condition to a condition related to the current angle. β The low-order algebraic equations are solved to obtain candidate solutions; from the candidate solutions, the candidate solution that is within the preset current angle range and has the largest corresponding electromagnetic torque is selected as the optimal current angle.
6. The method as described in claim 1, characterized in that, Also includes: The system acquires motor operation data for each time window in real time and updates the target local electromagnetic model parameters in real time.
7. The method as described in claim 6, characterized in that, The process of updating the target's local electromagnetic model parameters in real time also includes: The partial derivative of the electromagnetic torque of the synchronous reluctance motor with respect to the current output current angle is calculated based on the updated target local electromagnetic model parameters and filtered. When the absolute value of the calculation result is less than the first preset threshold, it is determined that the synchronous reluctance motor has run to the local optimal region, the update of the local electromagnetic model parameters is paused, and the amplitude of the current sampled current vector is latched. When the absolute value of the deviation between the magnitude of the sampled current vector in the subsequent control cycle and the magnitude of the latched sampled current vector exceeds the second preset threshold, the update of the local electromagnetic model parameters is restarted.
8. A maximum torque-to-current ratio control device for a synchronous reluctance motor, characterized in that, include: The linear regression equation construction module is used to take the motor operation data corresponding to each time window, substitute it into the preset linear regression equation, and obtain the linear regression equation for the corresponding time window. The preset linear regression equation is obtained by substituting the local electromagnetic model into the dq-axis voltage equation of the motor and integrating both sides of the dq-axis voltage equation; the local electromagnetic model of the synchronous reluctance motor is: ψ d , ψ q For the d-axis and q-axis flux linkages of a synchronous reluctance motor, i d , i q For d-axis and q-axis currents, l d , l dd , l q , l qq , l c These are the parameters of the local electromagnetic model. l d and l q d and q are the linear self-axis flux linkages of the axis, respectively. l dd and l qq These are the cubic self-axis nonlinearity coefficients of the d and q axes, respectively, used to characterize self-axis saturation. l c The local electromagnetic model is used to characterize the d-axis and q-axis cross-coupling effect; it is used to characterize the magnetic circuit self-saturation effect and cross-coupling effect in the neighborhood of the current operating point; a time window includes N control cycles, where N is a positive integer greater than 20; the motor operation data includes the d-axis voltage, q-axis voltage, d-axis current, q-axis current, and electric angular velocity during the operation of the synchronous reluctance motor; The parameter update module is used to obtain a set of linear regression equations by combining the linear regression equations corresponding to all historical time windows, and to solve the set of linear regression equations to obtain the target local electromagnetic model parameters that reflect the local electromagnetic characteristics of the current working point. The optimal current angle direct solution control module is used to substitute the target local electromagnetic model parameters into the expression of the local electromagnetic model to obtain the d-axis and q-axis flux linkages. ψ d and ψ q ;Will ψ d and ψ q Substituting the values into the electromagnetic torque equation of the synchronous reluctance motor, the optimal current angle corresponding to the maximum torque-to-current ratio is obtained, and then the d-axis and q-axis current commands for each control cycle are calculated. The d-axis and q-axis current commands are input into the synchronous reluctance motor current closed-loop controller to achieve maximum torque-to-current ratio control.
9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the method according to any one of claims 1 to 7.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method described in any one of claims 1 to 7.
