【作者】 张宇；

【导师】 王丽梅； 苏凯；

【作者基本信息】 沈阳工业大学 ， 电气工程， 2025， 硕士

【摘要】 随着高端制造业的不断发展,在数控机床、交通运输、航空航天等领域对控制系统的精度要求也逐渐变高。永磁直线同步电机(PMLSM)作为制造业的重要部件,有着响应速度快、推力大等优点,在直线运动的设备上得到了广泛应用。但是,PMLSM省去了中间环节,没有了机械传动的缓冲,这样就会导致非线性摩擦力、负载扰动以及端部效应等不确定因素直接作用在直线伺服系统上,从而严重影响鲁棒性能和跟踪性能。因此,本文将PMLSM作为研究对象,以提高控制系统的位置跟踪精度和抗扰性能为目标,设计了基于超前校正的智能分数阶自抗扰控制器。首先,对PMLSM的工作原理和基本结构进行了阐述,在坐标变换和矢量控制的基础上,建立了PMLSM的数学模型和矢量控制系统,分析了系统不确定因素对系统产生的影响及其产生的原因。其次,阐述了自抗扰控制(ADRC)和分数阶微积分理论的基本原理,对PMLSM伺服系统的状态方程进行了推到,然后利用自抗扰控制器其对系统具有快速的协调性和不错的超调性能以及不需要精确的系统模型等优点,将自抗扰控制器应用到了直线电机的位置环控制器上,并引入分数阶微积分理论,设计分数阶自抗扰控制器(FO-ADRC),来提高控制系统的鲁棒性和响应速度。由于分数阶扩张状态观测器(FO-ESO)在观测扰动时存在相位滞后的问题,通过将超前校正环节引入FO-ESO中,设计一种分数阶相位超前扩张状态观测器(FO-PLESO),使观测扰动时不但具有简便的调节方式并且还能使相位滞后减小,并在Matlab/Simulink中进行仿真验证,证明此方法的可行性。最后,本文设计的分数阶自抗扰控制器是通过带宽法进行设计的,从而解决传统自抗扰控制器参数整定复杂的问题,但是,带宽法在确定控制器参数时,通常只能通过经验法来确定带宽,为此,在带宽法的基础上引入了鲸鱼优化算法。针对传统的鲸鱼优化算法易陷入局部最优解、易早熟收敛等问题,通过折射反向学习机制对鲸鱼优化算法进行改进。采用改进后的鲸鱼优化算法,并结合带宽法,在一定范围内对控制器中的关键参数进行寻优,使控制器获得更好的控制效果。通过Matlab/Simulink验证基于超前校正的智能分数阶自抗扰控制器的跟踪精度和抗扰性能均有提高。更多还原

【Abstract】 With the continuous advancement of high-end manufacturing industries,the precision requirements for control systems in fields such as CNC machine tools,transportation,and aerospace have been progressively increasing.The permanent magnet linear synchronous motor(PMLSM),as a vital component in the manufacturing sector,boasts advantages such as fast response speed and high thrust,finding widespread application in rectilinear motion equipment.However,the elimination of intermediate components in PMLSM,which removes the buffering effect of mechanical transmission,leads to the direct impact of uncertain factors such as nonlinear friction,load disturbances,and end effects on the linear servo system,severely affecting its robustness and tracking performance.Therefore,this thesis takes PMLSM as the research object and is aimed at improving the position tracking accuracy and disturbance rejection performance of the control system by designing an intelligent fractional-order active disturbance rejection controller based on lead compensation.Firstly,the working principle and basic structure of PMLSM are explained.Based on coordinate transformation and vector control,a mathematical model and vector control system for PMLSM are established,and the impact of system uncertainties on the system and their causes are discussed.Secondly,the fundamental principles of energetic disturbance rejection control(ADRC)and fractional-order calculus theory are elaborated.The state equation of the PMLSM servo system is derived,and leveraging the advantages of ADRC,such as its rapid coordination,good overshoot performance,and lack of reliance on an accurate system model,the ADRC is applied to the position loop controller of the linear motor.Fractional-order calculus theory is introduced to design a fractional-order functional disturbance rejection controller(FO-ADRC)to enhance the robustness and response speed of the control system.Due to the phase lag issue in the fractional-order extended state observer(FO-ESO)when observing disturbances,a fractional-order phase-lead extended state observer(FO-PLESO)is designed by incorporating a lead compensation link into FO-ESO.This design not only provides a convenient adjustment method but also reduces phase lag.Simulation verification in Matlab/Simulink proves the feasibility of this approach.Finally,the fractional-order operational disturbance rejection controller designed in this thesis is developed using the bandwidth method to address the complexity of parameter tuning in traditional ADRC.However,the bandwidth method typically relies on empirical methods to determine the bandwidth when specifying controller parameters.To address this,a whale optimization algorithm is introduced based on the bandwidth method.To overcome the issues of traditional whale optimization algorithms,such as their tendency to gain trapped in local optima and premature convergence,an improved whale optimization algorithm is proposed by incorporating a refraction opposition-based learning mechanism.The improved whale optimization algorithm,combined with the bandwidth method,is utilized to optimize key parameters in the controller within a certain range,achieving better control performance.Simulation verification in Matlab/Simulink shows that the tracking accuracy and disturbance rejection performance of the cerebral fractional-order active disturbance rejection controller based on lead compensation are improved.更多还原