A method for speed control of permanent magnet synchronous motor in electric vehicles based on a fixed-time observer

By adopting a dual closed-loop control method based on a fixed-time observer, the problem of rotational inertia fluctuation caused by load changes in the permanent magnet synchronous motor of electric vehicles is solved. Real-time estimation and compensation of mechanical parameter fluctuations and nonlinear disturbances are achieved, thereby improving the dynamic performance and robustness of the system.

CN123122368APending Publication Date: 2026-11-17YANCHENG INST OF TECH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611462490.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-09-20
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Existing technologies are insufficient to effectively address changes in equivalent rotational inertia caused by load variations in permanent magnet synchronous motors for electric vehicles, which affects speed control performance. Furthermore, traditional methods are not robust enough when faced with complex coupled disturbances.

Method used

A dual-closed-loop control method based on a fixed-time observer is adopted. The nominal control gain and lumped disturbance term are separated by feedback linearization. The fixed-time observer is used to estimate and compensate for mechanical parameter fluctuations and nonlinear disturbances in real time, and a cascade control structure of speed outer loop and current inner loop is constructed.

Benefits of technology

It enables real-time estimation and compensation of mechanical parameter fluctuations and nonlinear disturbances under complex road conditions, improving the system's dynamic response speed and anti-disturbance capability, and possessing good dynamic performance and strong robustness.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123122368A_ABST
    Figure CN123122368A_ABST
Patent Text Reader

Abstract

This invention discloses a speed control method for a permanent magnet synchronous motor (PMSM) in an electric vehicle based on a fixed-time observer. The method includes: constructing a fixed-time observer based on an arbitrary-order fixed-time differentiator to estimate the lumped disturbance term, ensuring that the estimation error converges to a neighborhood of near zero within a fixed time; constructing a cascade control structure comprising an outer speed loop and an inner current loop, wherein the inner current loop uses PI control and the outer speed loop uses a controller based on the fixed-time observer; constructing a speed outer loop control law based on the nominal control gain, the desired speed and its derivative, the actual speed, and the state estimate output by the fixed-time observer; using the quadrature-axis current reference value generated by the speed outer loop control law as the quadrature-axis current reference value; performing closed-loop control on the direct-axis and quadrature-axis currents respectively; and generating corresponding d-q axis voltage commands to drive the PMSM.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the field of electric vehicle technology, specifically relating to a speed control method for a permanent magnet synchronous motor in an electric vehicle based on a fixed-time observer. Background Technology

[0002] In the transportation sector, electric vehicles have attracted widespread attention due to their advantages of zero direct emissions and rapid response [1,2]. Currently, the types of motors that can be used to drive electric vehicles mainly include permanent magnet synchronous motors, asynchronous motors, and synchronous reluctance motors. Among them, permanent magnet synchronous motors have become a hot topic in current research and engineering applications due to their advantages such as wide speed range, high power density, high reliability, and excellent operating efficiency [3-5].

[0003] However, in actual operation, road conditions are complex and changeable, and the permanent magnet synchronous motor itself has strong nonlinear characteristics, which puts higher demands on the control performance of the speed controller. Traditional proportional-integral controllers (TPICs) rely on mechanical parameters such as friction coefficient, load torque, and equivalent moment of inertia when designing the speed loop; while the current loop design requires electrical parameters such as permanent magnet flux linkage, stator resistance, and inductance [6-8]. Various mechanical parameters (especially equivalent moment of inertia and load torque) will change with the operating conditions, and the fluctuation of equivalent moment of inertia is particularly obvious under load variation conditions [9-12]. In contrast, the fluctuation amplitude of electrical parameters is relatively small, so traditional proportional-integral current loop controllers usually have strong robustness [10,11]. To achieve high-performance speed control, the controller needs to have two capabilities: first, satisfactory dynamic performance, including fast response and no steady-state error; second, the ability to effectively compensate for the effects of parameter uncertainty and load torque fluctuations [9,13]. It is evident that accurately acquiring mechanical parameters has become one of the key issues in improving speed control performance [14,15].

[0004] Regarding the identification of mechanical parameters, various methods have been proposed in existing research. Early studies mostly identified only a single parameter, such as the fast identification of equivalent moment of inertia based on sinusoidal disturbance signals

[10] , acceleration and deceleration identification method

[16] , recursive least squares method

[17] , and identification method based on sliding mode observer

[18] . However, in practical applications, the equivalent moment of inertia and load torque often change simultaneously, and it is difficult to guarantee the overall control performance of the control system by identifying only a single parameter [7]. In response to this problem, subsequent studies have successively proposed methods that can realize the synchronous estimation of multiple parameters, including various sliding mode observers and their improved algorithms [2,7,8,14,19-22]. Although the above methods have made some progress in parameter identification, there are still many shortcomings. For example, there is a lack of comprehensive testing and analysis of the dynamic performance of the control system[2]; insufficient analysis of the dynamic coupling relationship between parameters[7,8]; offline identification methods are difficult to adapt to real-time parameter changes[14,19]; the global convergence of the multi-observer system still needs further verification[20,21], and such methods also require prior knowledge of perturbation of the upper bound

[22] .

[0005] Considering the complex coupling relationships among various mechanical parameters, another way to improve system robustness is to treat multi-parameter disturbances as lumped disturbances and then estimate and compensate them in real time. Related research methods mainly rely on various observers [12,15,23-28]. Such methods have achieved certain results in improving system dynamic performance and enhancing robustness, but further improvements are still needed. For example, assumptions that are difficult to fully satisfy in practical applications

[12] ; insufficient testing of equivalent moment of inertia (EMI) changes [23,24]; insufficient dynamic performance testing and analysis

[25] ; and the observation convergence speed also needs to be further improved [26,27]. References [15,28] point out that to achieve estimation and compensation of lumped disturbances, it is first necessary to obtain EMI. Summary of the Invention

[0006] Purpose of the invention: To address the issue that sudden load changes in electric vehicles (EVs) can cause variations in the equivalent moment of inertia (EMI), severely impacting speed control performance, this invention proposes a speed control method for permanent magnet synchronous motors in electric vehicles based on a fixed-time observer.

[0007] Technical Solution: This invention proposes a speed control method for permanent magnet synchronous motors in electric vehicles based on a fixed-time observer, comprising the following steps:

[0008] A state-space model of a permanent magnet synchronous motor in the synchronous rotating coordinate system of the dq axis is established; the quadrature axis current is selected as the system input and the motor speed is selected as the system output. Based on the state-space model, the mechanical transmission system of the motor is linearized by feedback, and the nominal control gain and lumped disturbance term are separated.

[0009] Based on an arbitrary-order fixed-time differentiator, a fixed-time observer is constructed to estimate the lumped perturbation term, ensuring that the estimation error converges to the neighborhood of zero within a fixed time.

[0010] A cascade control structure comprising an outer speed loop and an inner current loop is constructed. The inner current loop employs PI control, with the direct-axis current reference value set to 0. The outer speed loop uses a controller based on a fixed-time observer. Based on the nominal control gain, the desired speed and its derivative, the actual speed, and the state estimate output by the fixed-time observer, a speed outer-loop control law is constructed. The quadrature-axis current reference value generated by the speed outer-loop control law is used as the quadrature-axis current reference value, and closed-loop control is performed on the direct-axis and quadrature-axis currents respectively to generate corresponding dq-axis voltage commands to drive the permanent magnet synchronous motor.

[0011] Furthermore, based on the state-space model, the mechanical transmission system of the motor is linearized to separate the nominal control gain and lumped disturbance term, including:

[0012] For a permanent magnet synchronous motor system, the feedback linearization expression of its mechanical transmission system is as follows:

[0013]

[0014]

[0015]

[0016] In the formula, , Indicates intermediate parameters, J represents the moment of inertia, and p represents the number of pole pairs. T represents permanent magnet flux linkage. L k represents the unknown load torque. f Where ω is the coefficient of viscous friction and ω is the rotor speed. T represents the unknown disturbance torque. td Indicates the total disturbance torque;

[0017] The corresponding linearized system expression is:

[0018]

[0019]

[0020] Where v is the control input, y r Let y be the expected reference value of the output quantity y, and k1 be the gain of the linear controller;

[0021] Given that all system nonlinear terms are unknown, the expression for the disturbance term is constructed as follows:

[0022]

[0023] Where Ψ is the disturbance term, B0 = B(x)| x=x(0) Represents the nominal control gain:

[0024]

[0025] Where, ψ f0 J0 and J0 are the permanent magnet flux linkages ψ f The nominal value of the moment of inertia J.

[0026] Furthermore, the construction of a fixed-time observer based on an arbitrary-order fixed-time differentiator to estimate the lumped disturbance term includes:

[0027] The perturbation term Ψ is decomposed into non-differentiable and differentiable components: ,in For non-differentiable parts, It is a differentiable part;

[0028] Define the state vector z = [z1, z2] T Where z1=y, z2= By performing state augmentation on the mechanical transmission system, we can obtain:

[0029]

[0030] Design a fixed-time observer to estimate the state vector z based on an arbitrary-order fixed-time differentiator:

[0031]

[0032] Where λ1 and λ2 are both positive constants, It is a nonlinear composite function. , These are the estimated values ​​of the first derivatives of z1 and z2.

[0033] Furthermore, the nonlinear composite function The expression is: , where φi is represented as:

[0034]

[0035] in ,parameter and It can take any positive value, r 0,i With r ∞,i The solution is then obtained using the following recursive method:

[0036]

[0037]

[0038] Wherein, parameters d0 and d ∞ Satisfying −1≤d0≤d ∞ <0.5.

[0039] Furthermore, based on the nominal control gain, the desired speed and its derivative, the actual speed, and the state estimate output by the fixed-time observer, the outer-loop control law for the speed is constructed, expressed as:

[0040]

[0041] In the formula, This is the reference value for the quadrature-axis current. The derivative of the desired rotational speed, For the desired rotational speed, This is the actual motor speed. The disturbance state estimate output by the fixed-time observer.

[0042] Beneficial Effects: This invention employs a dual closed-loop structure: the speed loop uses a controller based on a fixed-time observer, while the current loop uses traditional proportional-integral control. This method integrates load fluctuations, equivalent moment of inertia changes, and system nonlinear disturbances into a lumped disturbance term, achieving real-time disturbance estimation and dynamic compensation through a fixed-time observer. Compared with existing technologies, this invention has the following advantages:

[0043] (1) The present invention models the mechanical parameter fluctuation and system nonlinearity as a lumped disturbance term and uses a fixed-time observer to estimate and compensate for it in real time, which effectively avoids the analysis complexity caused by the coupling effect between mechanical parameters;

[0044] (2) The fixed-time observer designed in this invention has fixed-time convergence characteristics, which can ensure that the disturbance estimation error achieves fixed-time convergence, effectively improving the system response speed and anti-disturbance capability. This invention can effectively suppress various disturbances, adapt to system parameter perturbations, and has excellent dynamic speed tracking performance.

[0045] (3) The controller proposed in this invention not only has good dynamic performance, but also has strong robustness to mechanical parameter disturbances, and also has the advantage of fast convergence speed. Attached Figure Description

[0046] Figure 1 This is a schematic flowchart of a method for controlling the speed of a permanent magnet synchronous motor in an electric vehicle based on a fixed-time observer, as proposed in this invention.

[0047] Figure 2 Block diagram of a permanent magnet synchronous motor (PMSM) speed control system;

[0048] Figure 3 This is a schematic diagram of the experimental dynamic response under a step change in rotational speed without disturbance. Figure 3 In this context, (a) represents method 1. Figure 3 In the text, (b) represents method 2. Figure 3 In this context, (c) represents method 3, which is the method proposed in this invention;

[0049] Figure 4 The experimental dynamic response is given when the total disturbance torque T_td changes abruptly under constant speed conditions, where... Figure 4 In this context, (a) represents method 1. Figure 4 In the text, (b) represents method 2. Figure 4 (c) in the text represents method 3;

[0050] Figure 5 To consider the experimental dynamic response under constant velocity conditions when mechanical parameters (moment of inertia J) change, where, Figure 5 In this context, (a) represents method 1. Figure 5 In the text, (b) represents method 2. Figure 5 (c) in the text represents method 3;

[0051] Figure 6 To consider the experimental dynamic response under constant velocity conditions and the changing electrical parameters (R_s, L_(sd,q), ψ_f), where, Figure 6 In this context, (a) represents method 1. Figure 6 In the text, (b) represents method 2. Figure 6 (c) in the text represents method 3. Detailed Implementation

[0052] The technical solution of the present invention will now be further described in conjunction with the accompanying drawings and examples.

[0053] Fixed-time observers (FxTOs) have been successfully applied in several control fields in recent years, such as torque estimation of robotic arms

[29] , disturbance estimation of quadcopters

[30] , and lumped disturbance estimation of unmanned helicopters

[31] . The core advantage of fixed-time observers is that they can limit the system convergence time to a fixed range, ensuring that the estimation error converges to zero within a fixed time, and that this convergence time is not affected by the initial state of the system. This characteristic provides a new solution to the disturbance estimation problem in the speed control system of permanent magnet synchronous motors. Therefore, this invention proposes a speed control method for permanent magnet synchronous motors of electric vehicles based on fixed-time observers, such as... Figure 1 As shown, the main steps include:

[0054] Step 1: Establish the state-space model of the permanent magnet synchronous motor in the synchronous rotating coordinate system of the dq axis; select the quadrature axis current as the system input and the motor speed as the system output. Based on the state-space model, perform feedback linearization on the mechanical transmission system of the motor to separate the nominal control gain and lumped disturbance terms.

[0055] As one implementation method, a state-space model of the dq-axis of a permanent magnet synchronous motor is established. This dq-axis state-space model includes stator voltage, current, electromagnetic torque, and mechanical motion equations. It is represented as:

[0056]

[0057] in:

[0058]

[0059]

[0060]

[0061]

[0062]

[0063] Where, x∈R 3 u∈R 2 Let f(x) be the state vector and g(x) be the input vector, respectively, and let f(x) and g(x) be smooth vector fields. sd,q For stator voltage, i sd,q Where J is the stator current, J is the motor moment of inertia, and L is the stator current. s =L sd,q For stator inductance, p is the number of pole pairs, ψ f For permanent magnet flux linkage, R s For stator resistance, k f ω is the coefficient of viscous friction, ω is the rotor speed, ωe (=pω) is the electromagnetic rotational speed, Te is the electromagnetic torque, and T L The load torque is unknown.

[0064] The load torque, viscous friction, equivalent moment of inertia change, and nonlinear terms are unified and merged into a total aggregate disturbance.

[0065] For the state-space model of the dq axis of the permanent magnet synchronous motor, considering the unknown disturbance torque ΔT(x) and mechanical parameter perturbations, u=i sq Let y = ω be the system input and y = ω be the system output. Therefore, the mechanical transmission system of the permanent magnet synchronous motor can be represented as:

[0066]

[0067] in:

[0068]

[0069]

[0070] Among them, T td (T) td =T L +k f ω+ΔT(x)) is the total disturbance torque. T represents the unknown disturbance torque. L This represents the unknown load torque. In a practical permanent magnet synchronous motor system, the number of pole pairs p and the permanent magnet flux linkage ψ... f Since both the moment of inertia J and the rotational inertia J are non-zero, the system satisfies B(x)≠0 at all rated operating points.

[0071] For a permanent magnet synchronous motor system, the feedback linearization expression of its mechanical transmission system is as follows:

[0072]

[0073]

[0074] The corresponding linearized system expression is:

[0075]

[0076]

[0077] Where v is the control input, y r Let y be the expected reference value of the output quantity y, and k1 be the gain of the linear controller.

[0078] Given that all system nonlinear terms are unknown, the expression for the disturbance term is constructed as follows:

[0079]

[0080] Where Ψ is the disturbance term, B0 = B(x)| x=x(0) Represents the nominal control gain.

[0081]

[0082] Where, ψ f0 J0 and J0 are the permanent magnet flux linkages ψ f The mechanical transmission system of a permanent magnet synchronous motor, relative to the nominal value of its moment of inertia J, can be rewritten as follows:

[0083]

[0084] Step 2: Based on an arbitrary-order fixed-time differentiator, construct a fixed-time observer to estimate the lumped disturbance term, ensuring that the estimation error converges to the neighborhood of zero within a fixed time.

[0085] As one implementation approach, the parameters within the disturbance term Ψ may undergo abrupt changes during system operation, meaning these parameters may not possess time differentiability. Because the disturbance term Ψ may experience abrupt parameter changes during system operation, it may not possess time differentiability, thus making direct observation using an extended state observer impossible. Ψ can be decomposed into non-differentiable and differentiable components: .in For non-differentiable parts, Let be a differentiable part. Now we make the following assumptions:

[0086] Assumption 1: Both h are bounded, that is, they satisfy... L1 and All are positive real numbers.

[0087] Define the state vector z = [z1, z2] T Where z1=y, z2= By performing state augmentation on the mechanical transmission system, we can obtain:

[0088]

[0089] To estimate the state z in the above equation, based on an arbitrary-order fixed-time differentiator, the following FxTO is designed:

[0090]

[0091] Both λ1 and λ2 are positive constants.

[0092] For the composite function of φi:

[0093]

[0094] in ,parameter and It can take any positive value, r 0,i With r ∞,i The solution is then obtained using the following recursive method:

[0095]

[0096]

[0097] It is determined by the parameter −1≤d0≤d ∞ <0.5 is confirmed.

[0098] Define the estimation error vector ,in , Combining the above formula, we can deduce that:

[0099]

[0100] Theorem 1: Under the condition that Assumption 1 holds, the estimation error ξ can converge to the neighborhood of zero in a fixed time.

[0101] Step 3: Construct a cascade control structure comprising an outer speed loop and an inner current loop. The inner current loop uses PI control, with the direct-axis current reference value set to 0. The outer speed loop employs a controller based on a fixed-time observer. Based on the nominal control gain, desired speed and its derivative, actual speed, and the state estimate output by the fixed-time observer, a speed outer loop control law is constructed. The quadrature-axis current reference value generated by the speed outer loop control law is used as the quadrature-axis current reference value. Closed-loop control is then applied to both the direct-axis and quadrature-axis currents to generate corresponding dq-axis voltage commands to drive the permanent magnet synchronous motor. This ensures that the permanent magnet synchronous motor in the electric vehicle maintains high-speed control performance under time-varying and complex road conditions.

[0102] The outer loop control law for rotational speed is expressed as:

[0103]

[0104] In the formula, This is the reference value for the quadrature-axis current. The derivative of the desired rotational speed, For the desired rotational speed, This is the actual motor speed. This is the disturbance state estimate output by the fixed-time observer. It is worth noting that the designed controller only needs to use parameters J0 and ψ. f0 And the measured value of ω, where ω rThis is a reference value for ω.

[0105] To verify the excellent dynamic performance and disturbance rejection capability of the proposed controller based on FxTO (FxTOC) strategy applied to the speed control system of permanent magnet synchronous motors in electric vehicles, this paper uses Matlab / Simulink to build a model and conducts simulation experiments under three different operating conditions. Operating condition 1 (no disturbance, step change in speed): verifies the system's tracking performance for rapidly changing speeds under no disturbance conditions; Operating condition 2 (total disturbance torque T at constant speed) td Step change): Verify the system's ability to resist unknown disturbance torque; Condition 3 (considering system parameter perturbation in Condition 2): Verify the system's strong robustness to multiple disturbances of unknown disturbance torque and parameter perturbation. In this embodiment, TPIC (Method 1), non-cascaded feedback linearization controller (Method 2)

[35] and the proposed FxTOC control strategy (Method 3) are compared and simulated. The parameters of FxTOC controller and permanent magnet synchronous motor system are shown in Table 1 and Table 2, respectively. The performance index results of the three control methods are detailed in Table 3 and Table 4.

[0106] Operating Condition 1: Undisturbed step change in rotational speed

[0107] To verify the rapid response and accurate tracking capability of the proposed FxTOC under conditions of abrupt speed changes, speed step commands were applied at 0.1s and 0.6s, respectively. Figure 3 As shown in Table 3, Method 2 and Method 3 exhibit the shortest settling time (Ts) of 46.7 ms and the smallest overshoot of 8 rpm, respectively, while Method 1 has the longest settling time (240.9 ms) and the largest overshoot (68 rpm). When the system operating state changes, Method 1, based on a single operating point design, requires a certain amount of time to reach the optimal operating state. Compared to Method 2 without a cascade structure, Method 3, which uses a cascade structure, requires a longer response and settling time. Table 4 shows the performance index of the integral of the time multiplied by the absolute error (ITAE). The index value of Method 1 is significantly higher than the other two control methods, while the index values ​​of Method 2 and Method 3 are relatively close. Simulation results indicate that both Method 2 and Method 3 possess excellent characteristics of rapid response and accurate tracking when facing rapid changes in rotational speed.

[0108] Operating Condition 2: Total Disturbance Torque T at Constant Speed td Step change

[0109] To verify the proposed FxTOC's ability to handle unknown total disturbance torque T tdStrong disturbance rejection performance during step changes; the motor speed is set to be stably maintained at 1000 rpm, and the total disturbance torque T is reduced in 0.2 seconds. td The flow rate increases from 20 Nm to 80 Nm and remains constant. From Figure 4 As shown in Table 3, all three control methods are affected by step changes in the total disturbance torque. Method 1 has the longest recovery time (Tr) at 277.3 ms, with a maximum speed fluctuation of 157 rpm. Method 2 has the shortest recovery time at only 80 ms, with a maximum speed fluctuation of 35 rpm, higher than Method 3's 11 rpm. Table 4 shows the ITAE performance indicators, with Method 1 having the highest indicator value and Method 3 the lowest. When the total disturbance torque undergoes a step change, Method 1 requires a certain amount of time to recover to a stable operating state. The control system of Method 2 is designed based on precisely known system parameters and state values. When faced with a sudden change in unknown total disturbance torque, the lack of precise information about the relevant state variables leads to tracking errors. However, this method has the shortest recovery time of only 80 ms. Method 3 has the smallest speed fluctuation due to its disturbance compensation capability. However, because the observer needs time to estimate the lumped disturbance and its inherent cascade structure, its recovery time is slightly longer than that of Method 2. Simulation results show that Method 3 has strong robustness under the condition of step change in unknown total disturbance torque, and its overall control performance is better than that of Method 1 and Method 2.

[0110] Operating Condition 3: Considering the system parameter perturbations in Operating Condition 2

[0111] To further demonstrate the effectiveness of the proposed FxTOC strategy in suppressing various unknown disturbances, parameter perturbations were introduced in operating condition 2, and tests were conducted on mechanical and electrical parameter perturbations. It is worth noting that this invention primarily proposes a speed control system with strong robustness to unknown disturbance torques and mechanical parameter perturbations. For electrical parameter perturbations, the current controller was not specially designed and still employs the TPIC control strategy.

[0112] When the load changes drastically, EMI may fluctuate significantly. Therefore, in the simulation test, the EMI was increased to five times the original value in 0.2 seconds. (Comparison) Figure 4 , Figure 5As shown in Table 3, due to the impact of EMI changes on the system's operating state, the recovery time of Method 1 increased from 277.3 ms to 457.4 ms. The recovery time of Method 2 increased from 80 ms to 170.9 ms. Method 2 not only requires precise system state variables but also relies on accurate system parameters, leading to a further decline in control performance when facing EMI changes. Compared with Methods 1 and 2, Method 3 has the shortest recovery time of 115.2 ms, which remains almost unchanged, while the maximum speed fluctuation remains at 11 rpm. According to the indicators in Table 4, Method 3 has the smallest ITAE value and the smallest change compared to the value in Condition 2, with a relative error of only 6.6%, far lower than the 24.5% of Method 1 and the 34.6% of Method 2. The simulation test results show that Method 3 has strong robustness under conditions of unknown disturbance torque step change and EMI change. This is mainly because it can compensate for lumped disturbances in real time, ensuring that its speed control performance is significantly better than that of Methods 1 and 2.

[0113] Electrical parameter perturbations can also affect the operating performance of a control system to some extent. This invention provides simulation test results for electrical parameter perturbations, such as... Figure 6 As shown in Tables 3 and 4, compared to the nominal values, the magnetic flux decreased by 10%, while the resistance and inductance both increased by 20%. The impact of electrical parameter perturbations on the overall control strategy was evaluated using simulation results. The comparison of the three methods shows that Method 3 has the shortest recovery time (123 ms) and the smallest speed fluctuation (only 15 rpm). Compared to the simulation results for condition 2, Method 1 is less affected by electrical parameter perturbations. However, Method 2 exhibits a steady-state error (STE), and its control performance deteriorates sharply under electrical parameter perturbations.

[0114] Table 1: Parameters of Fixed-Time Observer Control (FxTOC)

[0115]

[0116] Table 2: Parameters of Permanent Magnet Synchronous Motor (PMSM) Used for Simulation Studies

[0117]

[0118] Table 3: Simulation Results: Comparison of Control Performance Index I under Different Scenarios

[0119]

[0120] Table 4: Simulation Results: Comparison of Control Performance Index II under Different Scenarios

[0121]

[0122] In summary, the embodiments of this invention not only improve the system's robustness to load changes and parameter perturbations but also maintain excellent dynamic performance under time-varying and complex road conditions. In the proposed controller, the current loop uses a TPIC (Transient Component Interchange) and the speed loop uses an FxTO-based controller. The speed loop controller first linearizes the mechanical transmission system, then uses a designed FxTO to estimate and compensate for lumped disturbances, including mechanical parameter perturbations and system nonlinearities, in real time. Finally, precise speed control is achieved through a designed linear controller. This speed loop controller only needs to know the motor's moment of inertia, the nominal value of the permanent magnet flux linkage, and the measured speed. Simulation under condition 1 confirms its excellent dynamic response capability, and experiments under condition 2 verify its robustness to unknown disturbance torque. Simultaneously, experiments under condition 3 verify the controller's strong robustness to mechanical parameter perturbations and demonstrate the entire control system's ability to suppress disturbances when electrical parameters and disturbance torque change. Future research will mainly focus on further improving the system's response speed and dynamic performance by designing a fixed-time controller without cascading structures. Sensorless speed control strategies will also be included in future research directions.

[0123] [1]ElhajA,AlzayedM,ChaouiH.DirectvoltageMTPAcontrolofinteriorpermanentmagnetsynchronousmotordrivenelectricvehicles.ControlEng.Pract.2025;156:106197.

[0124] [2]ChangH,WangZ,YangJ,LiS,LuS.SynchronousestimationschemeofPMSMmechanicalparameters:adecoupledextendedsliding-modeobserver.IEEETrans.PowerElectron.2025;40(8):10877-10888.

[0125] [3]NicolaM,NicolaC,Selis,teanuD,S,endrescuD.SensorlessPMSMcontrolsystembasedonmulti-agentreinforcementlearningandfractionalorderLADRCforrapidcontrolprototyping.AlexandriaEng.J.2025;130:292-310.

[0126] [4]XuX,WangL,ZhangZ,WuZ.Anovelrobustcontrolforpermanentmagnetsynchronousmotorintegratingdeadbeatpredictivecurrentcontrolandslidingmodeobserver.AlexandriaEng.J.2025;130:57-67.

[0127] [5]MengD,WuQ,ZhangJ,LiY,DiaoL.Restartingcontroloffree-runninginteriorpermanentmagnetsynchronousmotorunderpositionsensorlesscontrol.AlexandriaEng.J.2025;115:83-93.

[0128] [6]YouZ,BianY,ZhangY,ChenL.Anintelligentoptimizationalgorithmwithnovelfitnessfunctionforhigh-performancePMSMFOC.AlexandriaEng.J.2025;115:286-296.

[0129] [7]TangS,ShiT,CaoY,LinZ,WangZ,YanY.SimultaneousidentificationofloadtorqueandmomentofinertiaofPMSMbasedonvariablestructureextendedslidingmodeobserver.IEEETrans.PowerElectron.2024;39(7):8585-8596.

[0130] [8]TangS,CaoY,ShiT,YanY,XiaC.Onlineestimationofloadtorqueandmomentofinertiaincorporatingextendeddisturbanceobserverwithtrigger.IEEETrans.PowerElectron.2025;40(4):5731-5742.

[0131] [9]TarczewskiT,SzczepanskiR,ErwinskiK,HuX,GrzesiakLM.AnovelsensitivityanalysistomomentofinertiaandloadvariationsforPMSMdrives.IEEETrans.PowerElectron.2022;37(11):13299-13309.

[0132]

[10] LiuK,ZhuZ.Fastdeterminationofmomentofinertiaofpermanentmagnetsynchronousmachinedrivesfordesignofspeedloopregulator.IEEETrans.ControlSyst.Technol.2017;25(5):1816-1824.

[0133]

[11] FanY,ChenJ,ZhangQ,ChengM.AnimprovedinertiadisturbancesuppressionmethodforPMSMbasedondisturbanceobserverandtwo-degree-of-freedomPIcontroller.IEEETrans.PowerElectron.2023;38(3):3590-3599.

[0134]

[12] HouQ,DingS.Finite-timeextendedstateobserver-basedsuper-twistingslidingmodecontrollerforPMSMdriveswithinertiaidentification.IEEETrans.Transp.Electrif.2022;8(2):1918-1929.

[0135]

[13] ZhuZ,XuY,LiL,LiC.ParametersestimationforPMSMofelectro-mechanical-brakingsystemunderlowadhesionbraking.IEEETrans.Transp.Electrif.2026;1(12):1430-1440.

[0136]

[14] LianC,XiaoF,GaoS,LiuJ.Loadtorqueandmomentofinertiaidentificationforpermanentmagnetsynchronousmotordrivesbasedonslidingmodeobserver.IEEETrans.PowerElectron.2019;34(6):5675-5683.

[0137]

[15] XieM,SheJ,LiuZ,LiD.Onlineestimationofmomentofinertiaincorporatingdisturbanceestimation.IEEETrans.Ind.Electron.2024;71(7):7610-7620.

[0138]

[16] JiangF,YangF,SunS,YangK.ImprovedlinearactivedisturbancerejectioncontrolforIPMSMdrivesconsideringloadinertiamismatch.Energies2022;15(3):1169.

[0139]

[17] YangC,SongB,XieY,LuS,TangX.Speed-controller-independentmechanicalparameteridentificationinSPMSMdriveachievedviasignalinjection.IEEETrans.Ind.Electron.2023;70(2):1282-1297.

[0140]

[18] LuW,ZhangZ,WangD,LuK,WuD,JiK.Anewloadtorqueidentificationslidingmodeobserverforpermanentmagnetsynchronousmachinedrivesyetem.IEEETrans.PowerElectron.2019;34(8):7852-7862.

[0141]

[19] ZhangX,LiZ.Sliding-modeobserver-basedmechanicalparameterestimationforpermanentmagnetsynchronousmotor.IEEETrans.PowerElectron.2016;31(8):5732-5745.

[0142]

[20] YangC,SongB,XieY,ZhengS,TangX.AdaptiveidentificationofnonlinearfrictionandloadtorqueforPMSMdrivesviaaparallel-observer-basednetworkwithmodelcompensation.IEEETrans.PowerElectron.2022;38(5):5875-5897.

[0143]

[21] XiangD,YangJ,HaoY,XuG.Parallel-cascadedparameteridentificationschemeforPMSM-drivenservosystemsduringselfcommission.IEEETrans.Ind.Electron.2025;72(2):1914-1924.

[0144]

[22] YangC,SongB,XieY,TangX.OnlineparallelestimationofmechanicalparametersforPMSMDrivesviaanetworkofinterconnectedextendedsliding-modeobservers.IEEETrans.PowerElectron.2021;36(10):11818-11834.

[0145]

[23] GaoZ,YuanH,LvG,DaiR,WangH,LiuY.Acompositefinite-timecontrolschemewithdisturbanceobserverforrobustspeedtrackingofpermanentmagnetsynchronousmotor.Electr.PowerSyst.Res.2026;254:112620.

[0146]

[24] TangS,ShiT,CaoY,YanY,XiaC.Cascadedextendedstateobserver-basedspeedcontrollerdesignofPMSMconsideringvariablecontrolgainanduncertaindisturbance.IEEETrans.PowerElectron.2026;41(1):303-314.

[0147]

[25] XiongJ,FuX.Extendedtwo-stateobserver-basedspeedcontrolforPMSMwithuncertaintiesofcontrolinputgainandlumpeddisturbance.IEEETrans.Ind.Electron.2024;71(6):6172-6182.

[0148]

[26] XiaH,LiS,XiangZ,ChenD,LiZ.AdaptiveparameterestimationbasednonlinearactivedisturbancerejectioncontrollerforaPMSMsystem.IEEETrans.Ind.Inf.2025;21(9):7311-7321.

[0149]

[27] FuX,XiongJ.MultielementdisturbanceobserverforseparatelyestimatingperiodicandnonperiodicdisturbancesinPMSMcontrolsystems.IEEETrans.PowerElectron.2026;41(3):3388-3398.

[0150]

[28] WangT,LuoG,ChenZ,TuW,LiuC.AnimprovedrobustmodelpredictivespeedcontrolwithinertiaidentificationforPMSMdrivesintheelectrohydrostaticactuator.IEEETrans.PowerElectron.2023;38(11):13825-13841.

[0151]

[29] MaJ,DengW,ChenW,HuJ,YaoJ.Fixed-timeadaptiveslidingmodemomentumobserverforexternaltorqueestimationofroboticmanipulators.Mechatronics2026;113:103429.

[0152]

[30] XiaY,GongM,MaD,ShenG.Saturatedoutputfeedbackcontrolforquadrotortrajectorytrackingviafixed-timeobservers.ControlEng.Pract.2026;168:106713.

[0153]

[31] WangH,ShiS,ChenZ.Disturbanceobserverbasedfixed-timetrackingcontrolforunmannedaerialhelicopters.J.FranklinInst.2024;361(16):107184.

[0154]

[32] WuX,LuQ,SheJ,SunM,YuL,SuC.-Y.Onconvergenceofextendedstateobserversfornonlinearsystemswithnondifferentiableuncertainties.ISATrans.2023;136:590-604.

[0155]

[33] LiuD,ChenJ,CangJ,ZhangP,BaiX,LiuH,LongF.Amaximumpowerpointtrackingcontrolstrategyoftidalenergyconversionsystembasedonfixed-timeobservertechnology.OceanEng.2024;309:118386.

[34] MorenoJA.Arbitrary-orderfixed-timedifferentiators.IEEETrans.Autom.Control2021;67(3):1543-1549.

[0156]

[35] ChenJ,YaoW,ZhangC,RenY,JiangL.DesignofrobustMPPTcontrollerforgrid-connectedPMSG-Basedwindturbineviaperturbationobservationbasednonlinearadaptivecontrol.Renew.Energy2019;134:478-495.

Claims

1. A method for controlling the speed of a permanent magnet synchronous motor in an electric vehicle based on a fixed-time observer, characterized in that: Includes the following steps: Establish a state-space model of the permanent magnet synchronous motor in the synchronous rotating coordinate system of the dq axis; The quadrature-axis current is selected as the system input, and the motor speed is selected as the system output. Based on the state-space model, the mechanical transmission system of the motor is linearized by feedback to separate the nominal control gain and the lumped disturbance term. Based on an arbitrary-order fixed-time differentiator, a fixed-time observer is constructed to estimate the lumped perturbation term, ensuring that the estimation error converges to the neighborhood of zero within a fixed time. A cascade control structure comprising an outer speed loop and an inner current loop is constructed. The inner current loop employs PI control, with the direct-axis current reference value set to 0. The outer speed loop uses a controller based on a fixed-time observer. Based on the nominal control gain, the desired speed and its derivative, the actual speed, and the state estimate output by the fixed-time observer, a speed outer-loop control law is constructed. The quadrature-axis current reference value generated by the speed outer-loop control law is used as the quadrature-axis current reference value, and closed-loop control is performed on the direct-axis and quadrature-axis currents respectively to generate corresponding dq-axis voltage commands to drive the permanent magnet synchronous motor.

2. The method for controlling the speed of a permanent magnet synchronous motor in an electric vehicle based on a fixed-time observer according to claim 1, characterized in that: Based on the state-space model, the mechanical transmission system of the motor is linearized using feedback, separating the nominal control gain and lumped disturbance term, including: For a permanent magnet synchronous motor system, the feedback linearization expression of its mechanical transmission system is as follows: In the formula, , Here, J represents the moment of inertia, and p represents the number of pole pairs. T represents permanent magnet flux linkage. L k represents the unknown load torque. f Where ω is the coefficient of viscous friction and ω is the rotor speed. T represents the unknown disturbance torque. td Indicates the total disturbance torque; The corresponding linearized system expression is: Where v is the control input, y r Let y be the expected reference value of the output quantity y, and k1 be the gain of the linear controller; Given that all system nonlinear terms are unknown, the expression for the disturbance term is constructed as follows: Where Ψ is the disturbance term, B0 = B(x)| x=x(0) Represents the nominal control gain: Where, ψ f0 J0 and J0 are the permanent magnet flux linkages ψ f The nominal value of the moment of inertia J.

3. The method for controlling the speed of a permanent magnet synchronous motor in an electric vehicle based on a fixed-time observer according to claim 2, characterized in that: The construction of a fixed-time observer based on an arbitrary-order fixed-time differentiator to estimate the lumped disturbance term includes: The perturbation term Ψ is decomposed into non-differentiable and differentiable components: ,in For non-differentiable parts, It is a differentiable part; Define the state vector z = [z1, z2] T Where z1=y, z2= By performing state augmentation on the mechanical transmission system, we can obtain: Design a fixed-time observer to estimate the state vector z based on an arbitrary-order fixed-time differentiator: Where λ1 and λ2 are both positive constants, It is a nonlinear composite function. , These are the estimated values ​​of the first derivatives of z1 and z2.

4. The method for controlling the speed of a permanent magnet synchronous motor in an electric vehicle based on a fixed-time observer according to claim 3, characterized in that: The nonlinear composite function The expression is: , where φi is represented as: in ,parameter and It can take any positive value, r 0,i With r ∞,i The solution is then obtained using the following recursive method: Where, parameters d0 and d ∞ Satisfying −1≤d0≤d ∞ <0.

5.

5. The method for controlling the speed of a permanent magnet synchronous motor in an electric vehicle based on a fixed-time observer according to claim 4, characterized in that: The outer-loop control law for the speed is constructed based on the nominal control gain, the desired speed and its derivative, the actual speed, and the state estimate output by the fixed-time observer, and is expressed as follows: In the formula, This is the reference value for the quadrature-axis current. The derivative of the desired rotational speed, For the desired rotational speed, This refers to the actual motor speed. This is the disturbance state estimate output by the fixed-time observer.