A speed predictive control method of a permanent magnet synchronous motor based on a safe current limit

Abstract

The application relates to a permanent magnet synchronous motor speed prediction control method based on safe current limitation, comprising the following steps: obtaining relevant data of a permanent magnet synchronous motor, and constructing a system state space model containing matching and non-matching interference; based on the system state space model, constructing a system equivalent error model containing interference compensation; based on the system equivalent error model, constructing an optimal nominal speed prediction controller; considering the state constraint of the system, introducing a control barrier function based on current limitation; based on the current-limited control barrier function, constructing a robust current-limited control barrier function; fusing the nominal speed prediction controller and the robust current-limited control barrier function to construct a safe current-limited speed prediction controller, and realizing speed prediction control. Compared with the prior art, the application has the advantages of improving the current safety function of the permanent magnet synchronous motor, enhancing the anti-interference ability, dynamic response performance and tracking precision and the like.

Description

Technical Field

[0001] This invention relates to the field of permanent magnet synchronous motor control, and in particular to a speed prediction control method for permanent magnet synchronous motors based on a safe current limit. Background Technology

[0002] Permanent magnet synchronous motors (PMSMs) are widely used in various industries. Their high efficiency, small size, and precise control make them a popular choice for high-precision speed regulation control applications such as electric vehicles, robots, and industrial automation. Traditional cascaded dual-loop control strategies based on the field-oriented control (FOC) framework are widely adopted in PMSM speed control due to their reliable performance and low torque ripple. Meanwhile, to overcome the limited control bandwidth caused by cascaded structures, many non-cascaded control methods have been extensively studied. Although this method achieves high dynamic speed regulation, it also introduces two key problems. First, in non-cascaded structures, disturbances in the current loop are mismatched, meaning they cannot be canceled by the control input, which adversely affects the performance of the control system. Therefore, suppressing mismatched disturbances becomes crucial. Second, in non-cascaded structures, the current is no longer limited by a current reference value as in cascaded structures; therefore, current safety issues cannot be ignored.

[0003] To address the aforementioned issues, a current-constrained control method based on a penalty mechanism has been extensively studied. However, while this method achieves current constraint, the control input tends towards infinity as the system current reaches the constraint boundary value. This can lead to controller crashes or system instability in practical engineering, and it also significantly sacrifices system dynamic performance while pursuing current constraint. Meanwhile, to improve system dynamic performance and handle current constraint issues, the application of single-loop predictive control methods in permanent magnet synchronous motor speed control has been widely studied. In the paper "Single-loop robust model predictive speed regulation of PMSM based on exogenous signal preview," matched and unmatched disturbances in the system are uniformly treated as matched lumped disturbances for estimation and compensation. However, this approach struggles to achieve accurate speed tracking and rapid disturbance suppression. In the literature on the active disturbance rejection explicit model predictive direct speed control method for permanent magnet synchronous motors, matched and unmatched disturbances are estimated and compensated separately, which improves the system's disturbance rejection capability. However, this method involves a higher-order prediction model and requires solving the optimal problem in each control cycle, resulting in a large amount of computation and hindering the application of this method in engineering.

[0004] In summary, designing a single-loop speed predictive controller that comprehensively considers safe current limits, dynamic performance, and disturbance rejection performance, while also having a low application threshold, remains a rather challenging problem in engineering applications. Summary of the Invention

[0005] The purpose of this invention is to provide a speed prediction control method for permanent magnet synchronous motors based on a safe current limit to improve current safety.

[0006] The objective of this invention can be achieved through the following technical solutions:

[0007] A speed predictive control method for a permanent magnet synchronous motor based on a safety current limit includes the following steps:

[0008] Acquire relevant data of the permanent magnet synchronous motor and construct a system state-space model that includes matched and unmatched disturbances;

[0009] Based on the system state-space model, construct a system equivalent error model that includes disturbance compensation;

[0010] Based on the system equivalent error model, an optimal nominal speed predictive controller is constructed.

[0011] Considering the state constraints of the system, a control barrier function based on current constraints is introduced;

[0012] Based on the current-constrained control barrier function, a robust current-constrained control barrier function is constructed.

[0013] The nominal speed predictive controller and the robust current-limited control barrier function are fused together to construct a safe current-limited speed predictive controller, thereby achieving predictive control of speed.

[0014] Furthermore, the specific steps for constructing the system state-space model containing matched and unmatched disturbances include:

[0015] The three-phase current signals are transformed by Clark and Park to obtain the q-axis and d-axis measured currents.

[0016] A mathematical model of the magnetic synchronous motor in the dq coordinate system is established based on the q-axis measured current and the d-axis measured current. The expression of the mathematical model is as follows:

[0017]

[0018] Based on the mathematical model, a system state-space model including matched and unmatched disturbances is constructed. The expression of the system state-space model is:

[0019]

[0020]

[0021]

[0022] In the formula, i d and i q Angular velocity ω and d-axis current i, respectively d and q-axis current i q The differential, n p ψ is the number of pole pairs of the motor. f For rotor flux linkage, i q For measuring current along the q-axis, B m ω is the coefficient of viscous friction, T is the measurement velocity, and ω is the viscous friction coefficient. L R is the load torque, J is the moment of inertia, and R is the moment of inertia. s L is the stator resistance. s For stator inductance, u d and u q These are the voltages along the d-axis and q-axis in the dq coordinate system, respectively. Let x1 be the differential of system states x1 and x2, and x1 = ω ref -ω, For system status, Let be the system state vector. y is the system input, d1 and d2 are the unmatched and matched lumped disturbances of the system, respectively, and Δψ is the system output. f ΔB m ,ΔJ,ΔR s ΔL s These represent the deviations between the corresponding nominal model parameters and the actual model parameters of the motor. and For the known system model dynamics, ω ref The desired speed.

[0023] Furthermore, the specific steps for constructing the system equivalent error model including interference compensation include:

[0024] The system state-space model is decomposed into two state subsystems, and a high-order nonsmooth estimator is constructed.

[0025] Based on the nonsmooth estimator, the matching and non-matching disturbances in the system are reconstructed, and the reconstructed matching and non-matching disturbances are as follows:

[0026]

[0027]

[0028] Based on the reconstructed matched and unmatched disturbances, a system steady-state vector is defined, and a system steady-state signal value with disturbance compensation is constructed. The system steady-state signal value is:

[0029]

[0030] Based on the system steady-state signal value, the system error and system error vector are defined, and an equivalent system error model including disturbance compensation is constructed. The equivalent system error model is as follows:

[0031]

[0032]

[0033]

[0034] v q =uu *

[0035]

[0036] In the formula, d1 and d2 are the unmatched and matched lumped disturbances of the system, respectively. For ω 1，0 The differential, x1 = w ref -ω, For the system state, ω ref Let ω be the desired velocity, ω be the measured velocity, J be the moment of inertia, and n be the moment of inertia. p ψ is the number of pole pairs of the motor. f For rotor flux linkage, i q For measuring current along the q-axis, B m The coefficient of viscous friction is... and For the known system model dynamics, w 1，1 w 1，2 w 2，1 They are d1, The estimated value of d2, x i,0 for, and It is an auxiliary variable, l 1，j , l 2，i-1 and ρ i These are the gain coefficients of the positive estimator, i = 1, 2, j = 0, 1, 2. Let x1 be the steady-state vector of the system. * x2 * Let x1 and x2 be the steady-state signal values ​​of the system states, respectively, and u be the system input. * The steady-state signal value input to the system. v q =uu * For systematic error, Let be the system state error vector. and η ω and η q The differential.

[0037] Furthermore, the specific steps for constructing the optimal nominal speed prediction controller include:

[0038] Based on the system equivalent error model, the output prediction expression is obtained by Taylor continuous expansion, and the output prediction expression is:

[0039]

[0040] Construct a value function, and further expand the value function based on the output prediction expression. The expanded value function is as follows:

[0041]

[0042] Based on the expanded value function, the optimal control law of the equivalent error model of the system is calculated, and the optimal nominal speed predictive controller is constructed. The expression for calculating the optimal control law is as follows:

[0043]

[0044] In the formula, For the system's predicted output, η ω and η q These are the speed and current errors, respectively. v q The optimal control law for the equivalent error system. For future moments, J is the moment of inertia, T > 0 is the prediction time domain, and k ω * k q * To control the gain.

[0045] Furthermore, the expression for the optimal nominal speed prediction controller is:

[0046]

[0047] In the formula, u op For the output of the speed prediction controller, k ω * k q * For control gain, T > 0 is the prediction time domain, η ω For the velocity deviation, η q For current deviation, u * This is the steady-state input value for the system.

[0048] Furthermore, the current-limited control barrier function is:

[0049]

[0050] In the formula, Let x1 be the current-limited control barrier function, x2 be the system state containing current information, and x be the system state vector.

[0051] Furthermore, the specific steps for constructing the robust current-constrained control barrier function include:

[0052] Based on the control barrier function constrained by the current, an initial arbitrary local Lipschitz continuous feedback controller is constructed.

[0053] A higher-order nonsmooth estimator is used to compensate for the disturbances in the initially constructed feedback controller to obtain the final feedback controller.

[0054] The robust current-limited control barrier function is obtained by replacing the estimation error in the feedback controller.

[0055] Furthermore, the robust current-limited control barrier function is:

[0056]

[0057] In the formula, It is a robust safety set, where u is the system control input. Let be the system state vector. Let Γ be a sufficiently large positive constant. It is the Lie derivative, a type of extension. The function, where λ is the gain coefficient. For system disturbance estimation information, w 1，1 and w 2，1 These are the estimates for systems d1 and d2, respectively.

[0058] Furthermore, the specific steps for constructing the safe current-limited speed prediction controller include:

[0059] The nominal speed predictive controller and the robust current-limited control barrier function are fused to construct an optimization problem to obtain the Lipschitz continuous safe current-limited speed predictive controller.

[0060] The Lipschitz continuous safety current-limited speed prediction controller is simplified;

[0061] Based on the simplified results, an explicit form of a safe current-limited speed prediction controller is obtained.

[0062] Furthermore, the expression for the safe current-limited speed prediction controller is:

[0063]

[0064] In the formula, u q * To meet the desired q-axis output voltage for current safety, J is the moment of inertia, and L... s For stator inductance, n p Let ψ be the extreme logarithm. f For rotor flux linkage, To meet the upper limit of the q-axis voltage for system current safety, To meet the lower voltage limit of the input u for system current safety, It is Li Daoshu, For system disturbance estimation information, w 1，1 and w 2，1 Let be the estimates of system d1 and d2, respectively, and λ be the gain parameter. and For the selected current-limited control barrier function, u nom For a nominal speed prediction controller u that does not consider current safety op .

[0065] Compared with the prior art, the present invention has the following beneficial effects:

[0066] (1) The safety current-limited speed prediction controller provided by the present invention introduces a current-constrained control barrier function, and on this basis, a robust current-limited control barrier function based on disturbance compensation is designed to achieve strict current safety guarantee of the system, even when the system is disturbed.

[0067] (2) The safety current-limited speed prediction controller provided by the present invention integrates the designed nominal speed prediction controller and the current-limited robust control barrier function by constructing an optimization problem, thereby realizing a comprehensive consideration of current constraints and speed tracking performance. That is, while ensuring current safety, the optimal tracking performance of the system is guaranteed to the greatest extent.

[0068] (3) The safety current-limited speed prediction controller provided by the present invention has a very simple structure, which can effectively meet the requirements of real-time performance and limited computing resources. It can achieve fast real-time calculation in ordinary embedded processors, which can effectively reduce costs and implementation complexity, and therefore has considerable engineering application value. Attached Figure Description

[0069] Figure 1 This is a schematic diagram of the method flow of the present invention;

[0070] Figure 2 This is an overall diagram of the motor speed control system according to an embodiment of the present invention;

[0071] Figure 3 This is a block diagram of a safety current-limited speed prediction controller according to an embodiment of the present invention;

[0072] Figure 4 This is a system response curve for condition 1 (t=8s) in an embodiment of the present invention.

[0073] Figure 5 This is a system response curve for condition 2 (t=8s) in an embodiment of the present invention.

[0074] Figure 6 This is a system response curve for condition 3 (t=8s) in an embodiment of the present invention.

[0075] Figure 7 The figure shows the system response curves under three operating conditions at t=10s in the embodiments of the present invention. Detailed Implementation

[0076] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. These embodiments are based on the technical solution of the present invention and provide detailed implementation methods and specific operating procedures. However, the scope of protection of the present invention is not limited to the following embodiments.

[0077] This embodiment provides a speed prediction control method for permanent magnet synchronous motors based on a safe current limit, such as... Figure 3 As shown, this method, based on the speed of the permanent magnet synchronous motor and the d-axis and q-axis current feedback, first constructs a high-order non-smooth disturbance estimator to accurately estimate and compensate for matched and unmatched disturbances in the system. Then, based on the established equivalent error model of the system with disturbance compensation, an explicit nominal speed predictive controller is designed. Simultaneously, this embodiment introduces a control barrier function (CBF) to handle the current constraint problem. To provide a strict current safety guarantee for the motor speed control system, this embodiment designs a robust current-limited control barrier function (RCBF). Finally, by constructing an optimization problem, the nominal speed predictive controller and the robust current-limited control barrier function designed above are fused to synthesize a safe current-limited speed predictive controller. The overall block diagram of the system in this embodiment is shown below. Figure 2 As shown. First, the d-axis current loop is still designed using a PI controller, and the d-current reference is set to... Secondly, under the action of the proposed permanent magnet synchronous motor speed predictive control controller based on safe current limitation, the collected sensor feedback data, including speed ω and q-axis current i, are used to... q and d-axis current i d This allows for optimal speed tracking while ensuring current safety. Specifically, for example... Figure 1 As shown, the method includes the following steps:

[0078] Step 1: Obtain relevant data of the permanent magnet synchronous motor and construct a system state-space model that includes matched and unmatched disturbances.

[0079] In this specific example, the electrical parameters of the permanent magnet synchronous motor selected here are shown in Table 1:

[0080] Table 1 Electrical parameters of permanent magnet synchronous motor

[0081]

[0082] An angle sensor on the rotor of a permanent magnet synchronous motor is used to obtain the measured angle θ and measured speed ω. Then, a hardware ADC is used to sample the three-phase current signals in a three-phase stationary coordinate system, and the q-axis measured current i is obtained after Clark and Park transforms. q and d-axis measurement current i d Establish a mathematical model of PMSM in the dq coordinate system:

[0083]

[0084] Where, np ψ is the number of pole pairs of the motor. f For rotor flux linkage, B m T is the coefficient of viscous friction. L R is the load torque, J is the moment of inertia, and R is the moment of inertia. s L is the stator resistance. s For stator inductance, u d and u q These represent the voltages along the d-axis and q-axis in the dq coordinate system, respectively. The d-axis current loop still uses a PI controller, which is assumed to be functioning correctly.

[0085] Based on the obtained mathematical model of the motor (1), x1 = w is selected. ref -w, As the system state, w ref为 To determine the desired speed, the state-space model of the direct speed regulation system is as follows:

[0086]

[0087] in, The system state vector; y is the system input; y is the system output; d1 and d2 are the unmatched and matched lumped disturbances of the system, respectively, and their specific forms are as follows:

[0088]

[0089] Where, Δψ f ΔB m ,ΔJ,ΔR s ΔL s These represent the deviations between the corresponding nominal model parameters and the actual model parameters of the motor; and For a known system model dynamic, its specific form is:

[0090]

[0091] Step 2: Based on the system state-space model, construct the system equivalent error model including disturbance compensation.

[0092] The obtained system state-space model is decomposed into two state subsystems, and a higher-order nonsmooth estimator is constructed to reconstruct the matched and unmatched disturbances in the system, as follows:

[0093]

[0094] Among them, w i，0 It is x i The estimated value; w 1，1 w 1，2 w 2，1They are d1, The estimated value of d2; and It is an auxiliary variable; l 1，j , l 2，i-1 and ρ i These are the positive gain coefficients of the estimator (i = 1, 2, j = 0, 1, 2).

[0095] Based on the obtained system disturbance and its derivative estimate, define Given the system's steady-state vector, the system's steady-state signal value with disturbance compensation is constructed as follows:

[0096]

[0097] Based on the obtained steady-state signal, define υ q =uu * For systematic error, Let be the system error vector. By ignoring the estimation error, the linear equivalent error model of system (2) is established as follows:

[0098]

[0099] Step 3: Based on the equivalent error model of the system, construct the optimal nominal speed predictive controller.

[0100] Based on the obtained error system model, through Taylor continuous expansion, the output of system (7) at a future time τ is... It can be predicted as follows:

[0101]

[0102] in, To achieve rapid convergence of the tracking error to zero, the value function is designed as follows:

[0103]

[0104] Where T > 0 is the prediction time domain.

[0105] Based on the obtained output prediction expression (8), the value function (9) can be further expanded as follows:

[0106]

[0107] By order The optimal control law of system (7) can be derived as follows:

[0108]

[0109] Therefore, the optimal controller for system (2) can be designed as follows:

[0110]

[0111] in, and It is the controller gain.

[0112] Step 4: Consider the state constraints of the system and introduce a control barrier function based on current constraints.

[0113] To address the current constraint problem by introducing a control barrier function (CBF), the resulting system model is transformed into the following form:

[0114]

[0115] in, g2=I 2×2 , In practical applications, the stator current of a PMSM is subject to the following physical constraints:

[0116]

[0117] Among them, I max This is the maximum permissible stator current. Combining step 1, and also considering... The current-constrained problem (14) can be transformed into the following state-constrained problem:

[0118]

[0119] in,

[0120] Based on the derived system state constraints, the current-limited CBF is selected as follows:

[0121]

[0122] According to the given The corresponding security set described therein can be obtained. for:

[0123]

[0124] It is not difficult to verify. The relative order is 1, that is And satisfy hour, therefore It is a valid CBF.

[0125] Step 5: Based on the current-limited control barrier function, construct a robust current-limited control barrier function.

[0126] Based on the designed current-limited CBF, if the initial state of the system satisfies Then, for any local Lipschitz continuous feedback controller, it is:

[0127]

[0128] This ensures that the system state x2 is always within the safe set. In, that is This means that the current safety constraint is guaranteed, in which... It is the Lie derivative, and α is a class of extensions. function.

[0129] Considering that (18) above requires the system disturbance d to be precisely known, a higher-order nonsmooth estimator is used to compensate for the disturbance in (18) to facilitate engineering implementation. Then, by letting (18) can be rewritten as:

[0130]

[0131] in, However, due to the estimation error inherent in the estimator during the initial stage, a strict current constraint cannot still be guaranteed. To address this issue, a sufficiently large constant Γ is defined and ensured that: sup t＞0 {|ε 1，1 (t)|,ε 2，1 (t)|}≤Γ,ε 1，1 =w 1，1 -d1, ε 2，1 =w 2，1 -d2, then the estimation error in (19) Replace with A robust current-limited control barrier function (RCBF) can be obtained as follows:

[0132]

[0133] in, It is a robust safety set. It's not hard to see that... yes Therefore, compared to (18), (20) imposes a relatively conservative restriction on the control input u, thus ensuring a strict current constraint even if there is a disturbance estimation error.

[0134] Step 6: Combine the nominal speed predictive controller and the robust current-limited control barrier function to construct a safe current-limited speed predictive controller, thereby achieving predictive control of speed.

[0135] Based on the nominal speed predictive controller (12) and the current-constrained robust control barrier function (20), in order to ensure the speed tracking performance of the system to the greatest extent while strictly ensuring current safety, the following optimization problem is designed to generate a Lipschitz continuous safe current-constrained speed predictive controller:

[0136]

[0137] Among them, u nom =u op This is the nominal speed predictive controller derived from the above steps. It's easy to understand that the RCBF here can be considered a safety filter, capable of filtering out current constraint violation control inputs from the nominal speed predictive controller. It is the safety control input that is closest to the nominal speed prediction controller after being filtered by RCBF, thereby achieving the goal of ensuring the current while ensuring the system tracking performance as much as possible.

[0138] Based on the aforementioned safe current-limited speed prediction controller (21), by... and Substituting into (21), it can be further simplified to the following form:

[0139]

[0140] in, Therefore, the explicit form of the safety current-limited speed prediction controller for the q-axis voltage can be obtained as follows:

[0141]

[0142] Furthermore, the control parameter adjustment rules involved in the predictive controller for safe current limiting of the present invention are as follows:

[0143] 1) Non-smooth higher-order estimator: For the parameter selection of estimator (5), firstly, select parameter l 1，j , l 2，i-1 For some small positive numbers, such that: l 1，0 >l 1，1 >l 1，2 and l 1，j >l 2，2 The conditions are, for example, 4>2>1 and 2>1. Next, the estimator is brought to convergence by adjusting the gain coefficients ρ1 and ρ2. This parameter is typically set to a large positive constant, its value being proportional to the magnitude of the system disturbance. Finally, the control parameter l is adjusted. 1，j and l 2，i-1 Make fine adjustments to meet actual control requirements.

[0144] 2) Speed ​​Prediction Controller: The selection of parameters for the designed speed prediction controller (12) is very simple. Among them, the control gain... and As calculated in step 3, the control performance of the speed predictive controller can be achieved simply by adjusting parameter T. Furthermore, the magnitude of parameter T is inversely proportional to the system response. In practical applications, a positive T can be gradually selected from large to small until the desired control performance is achieved.

[0145] 3) Current-constrained robust control barrier function: For the selection of parameters of the current-constrained robust control barrier function (20), only two parameters, λ and Γ, need to be adjusted. Among them, λ is used to adjust the robustness of the control barrier function. The larger λ is, the less the control barrier function is affected by system disturbances, and the faster the system response speed. However, it should be noted that an excessively large λ may cause the controller output to jitter, so it needs to be selected according to the requirements. In addition, the parameter T is used to compensate for the estimation error of the estimator. When λ is determined, Γ usually needs to be selected step by step in ascending order according to the actual current constraint, until the strict current constraint is achieved in the actual working condition.

[0146] Three operating conditions for permanent magnet synchronous motors (PMSMs) are set up here to illustrate the effectiveness of the invention:

[0147] In the test, the nominal velocity predictive controller (12) and the disturbance observer-based model predictive controller (MPC-DOB) were selected as the comparison method. The test scenarios are shown in Table 2.

[0148] To ensure fairness in the comparison, the selection criteria for controller parameters comprehensively considered the system's transient performance and disturbance rejection capability. The control parameter settings are shown in Table 3.

[0149] Table 2 Experimental Test Scenario Design

[0150]

[0151] Table 3 Selection of Control Parameters

[0152]

[0153] Figure 4-7 The response curves of speed, q-axis current, and q-axis voltage under different test conditions for the three methods are presented respectively. Figure 4 The test results are for t=8s under operating condition 1. Figure 5 The test results are for condition 2 at t=8s. Figure 6 The test results are for condition 3 at t=8s. Figure 7The experimental results are presented under three operating conditions at t=10s. It can be seen that both the current-limited velocity predictive controller and the model predictive controller with disturbance observation (MPC-DOB) provided by this invention achieve current constraint. However, the response speed of the current-limited velocity predictive controller of this invention is slightly faster than that of MPC-DOB, and when faced with sudden disturbances, the controller of this invention exhibits a smaller velocity drop, possessing disturbance suppression capability similar to that of nominal velocity predictive control. Therefore, the experimental results show that the safe current-limited velocity predictive controller design method provided by this invention not only achieves excellent dynamic response while ensuring current safety, but also has excellent anti-interference capability under load. This confirms the excellent performance and feasibility of this invention in practical applications.

[0154] Meanwhile, to illustrate the computational efficiency of the current-limited speed predictive control of this invention, the execution time of the three methods described above was tested using the Simulink Profiler tool in MATLAB. Specific data are shown in Table 4.

[0155] Table 4 Comparison of computational efficiency

[0156]

[0157] It can be seen that the single-step execution time of the disturbance observer-based model predictive controller (MPC-DOB) is significantly higher than that of the current-limited speed predictive controller and the nominal speed predictive controller provided in this invention. This difference indicates that the current-limited speed predictive controller proposed in this embodiment has greater practicality in industrial applications, especially in scenarios with microprocessors that have limited computing power. This means that in actual production environments, the controller of this invention can more effectively meet the requirements of real-time performance and limited computing resources, bringing significant advantages to industrial applications.

[0158] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

Claims

1. A speed prediction control method for a permanent magnet synchronous motor based on a safe current limit, characterized in that, Includes the following steps: Acquire relevant data of the permanent magnet synchronous motor and construct a system state-space model that includes matched and unmatched disturbances; Based on the system state-space model, an equivalent error model of the system including disturbance compensation is constructed. Based on the system equivalent error model, an optimal nominal speed predictive controller is constructed. Considering the system's state constraints, a current-limited control barrier function is introduced; Based on the current-limited control barrier function, a robust current-limited control barrier function is constructed, wherein the specific steps for constructing the robust current-limited control barrier function include: Based on the current-limited control barrier function, an initial arbitrary local Lipschitz continuous feedback controller is constructed. A higher-order nonsmooth estimator is used to compensate for the disturbances in the initially constructed feedback controller to obtain the final feedback controller. Based on the estimation error in the feedback controller, a robust current-limited control barrier function is obtained by substitution. The nominal speed predictive controller and the robust current-limited control barrier function are fused to construct a safe current-limited speed predictive controller, thereby achieving predictive speed control. The specific steps for constructing the safe current-limited speed predictive controller include: The nominal speed predictive controller and the robust current-limited control barrier function are fused to construct an optimization problem to obtain the Lipschitz continuous safe current-limited speed predictive controller. The Lipschitz continuous safety current-limited speed prediction controller is simplified; Based on the simplified results, an explicit form of a safe current-limited speed prediction controller is obtained.

2. The speed prediction control method for a permanent magnet synchronous motor based on a safety current limit according to claim 1, characterized in that, The specific steps for constructing a system state-space model that includes matched and unmatched disturbances include: The three-phase current signal is obtained by performing Clark transform and Park transform. Shaft measurement current and Shaft measuring current; Based on the above Shaft measurement current and Shaft measurement current establishes permanent magnet synchronous motor in The mathematical model in the coordinate system is expressed as follows: Based on the mathematical model, a system state-space model including matched and unmatched disturbances is constructed. The expression of the system state-space model is: In the formula, , and Angular velocity , shaft current and shaft current The differential, This represents the number of pole pairs of the motor. For rotor flux linkage, for Shaft measuring current, The coefficient of viscous friction is... To measure speed, For load torque, For rotational inertia, For stator resistance, For stator inductance, and They are respectively coordinate system shaft and Shaft voltage; , System status and The differential, , For system status, Let be the system state vector. For system input, For system output; and These are the system's non-matching and matched lumped disturbances, respectively. , , , , These represent the deviations between the corresponding nominal model parameters and the actual model parameters of the motor. and For the known system model dynamics, The desired speed.

3. The speed prediction control method for a permanent magnet synchronous motor based on a safety current limit according to claim 2, characterized in that, The specific steps for constructing the system equivalent error model including interference compensation include: The system state-space model is decomposed into two state subsystems, and a high-order nonsmooth estimator is constructed. Based on the nonsmooth estimator, the matching and non-matching disturbances in the system are reconstructed, and the reconstructed matching and non-matching disturbances are as follows: ； Based on the reconstructed matched and unmatched disturbances, a system steady-state vector is defined, and a system steady-state signal value with disturbance compensation is constructed. The system steady-state signal value is: ； Based on the system steady-state signal value, the system error and system error vector are defined, and an equivalent system error model including disturbance compensation is constructed. The equivalent system error model is as follows: ； In the formula, and These are the system's non-matching and matched lumped disturbances, respectively. for The differential, , For system status, For the desired speed, To measure speed, For rotational inertia, This represents the number of pole pairs of the motor. For rotor flux linkage, for Shaft measuring current, The coefficient of viscous friction is... and For the known system model dynamics, , , They are , , The estimated value, , and It is an auxiliary variable. , and It is the gain coefficient of the positive estimator. i =1,2, j =0,1,2 Let be the system steady-state vector. , System status and The steady-state signal value, u For system input, The steady-state signal value input to the system. , , For systematic error, Let be the system state error vector. and They are respectively and The differential.

4. The speed prediction control method for a permanent magnet synchronous motor based on a safety current limit according to claim 3, characterized in that, The specific steps for constructing the optimal nominal velocity prediction controller include: Based on the system equivalent error model, the output prediction expression is obtained by Taylor continuous expansion, and the output prediction expression is: Construct a value function, and further expand the value function based on the output prediction expression. The expanded value function is as follows: Based on the expanded value function, the optimal control law of the equivalent error model of the system is calculated, and the optimal nominal speed predictive controller is constructed. The expression for calculating the optimal control law is as follows: In the formula, For the system's predicted output, and These are the speed and current errors, respectively. , v q The optimal control law for the equivalent error system. , For the future moment, For rotational inertia, It predicts the time domain. , To control the gain.

5. The speed prediction control method for a permanent magnet synchronous motor based on a safety current limit according to claim 4, characterized in that, The expression for the optimal nominal speed prediction controller is: In the formula, For the speed prediction controller output, , To control the gain, It predicts the time domain. For speed deviation, For current deviation, This is the steady-state input value for the system.

6. The speed prediction control method for a permanent magnet synchronous motor based on a safety current limit according to claim 5, characterized in that, The current-limited control barrier function is: In the formula, , For current-limited control barrier functions, x 2 represents the system state, which includes current information. This is the system state vector.

7. The speed prediction control method for a permanent magnet synchronous motor based on a safety current limit according to claim 6, characterized in that, The robust current-limited control barrier function is: In the formula, It is a robust security set. u For system control input, Let be the system state vector. , For a sufficiently large positive number, , , , It is Li Daoshu. This is the gain coefficient. Information for system disturbance estimation. and For each system and The estimate.

8. The speed prediction control method for a permanent magnet synchronous motor based on a safety current limit according to claim 7, characterized in that, The expression for the safe current-limited speed prediction controller is: In the formula, To meet the requirements of current safety Desired output voltage of shaft For rotational inertia, L s For stator inductance, For extreme logarithms, For rotor flux linkage, To meet the system current safety requirements Upper limit of shaft voltage To meet the system current safety input requirements The lower limit of voltage, , It is Li Daoshu, Information for system disturbance estimation. and For each system and The estimate, For gain parameters, and The selected current-limited control barrier function, u nom For nominal speed prediction controllers that do not consider current safety u op .

Images