基于非线性时间延迟扰动估计的永磁同步直线电机无模型鲁棒位置跟踪控制

Model-Free Robust Position Tracking Control of Permanent Magnet Synchronous Linear Motor Based on Nonlinear Time Delay Disturbance Estimation

为了提高永磁同步直线电机在载荷变化工况下的伺服性能,该文提出了一种基于非线性时间延迟扰动估计的无模型鲁棒位置控制策略。首先,建立一种新型的动力学阶次模型,消除了传统动力学模型中的电机参数和非线性项;其次,引入终端吸引因子设计了期望非线性误差动力学,确保位置跟踪误差的高精度有限时间收敛;然后,采用时间延迟扰动估计技术在线估计载荷变化引起的不确定性并前馈补偿到控制回路,同时结合非线性阻尼项克服时间延迟扰动估计的误差,利用李雅普诺夫理论分析了闭环控制器的稳定性和有限时间收敛性;最后,在不同工况下与传统的比例微分控制和滑模控制进行对比,仿真和实验结果验证了所提方法的有效性和优越性。

In practical permanent magnet synchronous linear motor(PMSLM)servo drive applications,uncertainties arising from time-varying payloads,such as mismatched inertia and nonlinear friction,can significantly degrade the tracking accuracy of the servo system.Traditional robust control techniques can enhance the servo performance of linear motors but often rely on the motor’s precise models and parameters.To diminish reliance on the motor system model,a model-free control approach grounded in the ultra-local model concept has been introduced in recent years.Nevertheless,the uncertainty estimation within the ultra-local model typically involves complex algebraic identification and observer methods,which are intricate and fail to account for the effects of estimation errors.This paper presents a model-free robust position control strategy predicated on nonlinear time delay estimation,streamlining the implementation process and enhancing its universality.Firstly,a dynamic order model for PMSLM is constructed on the foundation of the ultra-local concept.This model eliminates motor parameters and nonlinear terms in dynamic models,furnishing a framework for model-free control.Secondly,the terminal sliding mode attraction factor is introduced to design the expected nonlinear error dynamics,ensuring high-precision finite-time convergence of position tracking errors.Then,the uncertainty caused by payload changes is estimated online using the time-delay disturbance estimation technology,and feedforward is compensated to the control loop.Meanwhile,nonlinear damping terms are integrated to counteract the time-delay disturbance estimation error.Finally,the stability of the closed-loop controller is examined via Lyapunov theory,and guidelines for controller parameter selection are articulated.The simulation and experimental results of the position step response are 0.114 s and 0.202 s with the proposed method,0.163 s and 0.392 s with the traditional proportional-derivative(PD)control,and 0.132 s and 0.273 s with the sliding mode control(SMC).In addition,the control input current of PD is the smoothest,while the of SMC exhibits a severe jitter phenomenon.The of the proposed method contains noises due to the presence of the acceleration term.Regarding the payload variation,the position response curves of both PD and SMC begin to diverge from the designated positioning trajectory,with great loads leading to more pronounced deviations.In contrast,the tracking error bounds(TEBs)measured by the proposed method are 0.78 mm,0.86 mm,and 0.94 mm,consistently ensuring accurate tracking of the positioning trajectory.Random noise with an amplitude of 0.1mm is added to the feedback position signal to further verify the impact of position sampling disturbance on the proposed method.The results indicate that the impact of these noises is not significant.The conclusions are as follows.(1)Compared with PD and SMC controls,the proposed method guarantees finite-time convergence of the system and effectively mitigates uncertainties from payload fluctuations.(2)The time-delay disturbance estimation technique is independent of the motor's precise model and intrinsically serves as an equivalent low-pass filter.Thus,an additional filter design is optimal.(3)Incorporating a nonlinear damping term founded on the sliding mode surface curtails estimation errors associated with delayed disturbance compensation,thereby enhancing the control precision of the system.