Research On Some Control Problems Of High Performance Servo Drive System

The servo drive system with high-performance that makes the output of servo motor obey the input can accurately track or reproduce the motion process by various control strategies based on the principle of negative feedback.Thereby it has been common widely applied in high-end equipment fields such as robots,CNC machine tools,aerospace,etc.However,the servo drive system is a complex nonlinear system with multivariable and strong coupling,which has poor stability.There are often chaotic oscillations,time-delays,unknown control gains,violating input and output constraints,parameter variations and uncertain disturbances.Designing a high-quality controller is a pivotal technique to ensure the high-performance(fast response,high precision,small overshoot,strong robustness and good tracking performance)of servo drive system operating.Intelligent adaptive control is a popular tool in controlling nonlinear systems with complex nonlinear terms and external disturbances.Hence considering the above problems,it is of great theoretical significance and engineering value to research the adaptive tracking control method to ensure the high-performance operating of servo drive system.Foremost,clued by some existing overseas approaches,the state-of-the-art of servo drive system control was concluded and analyzed,the shortcomings of some methods are summarized,and then the research contents were decided.Then based on backstepping and intelligent adaptive control theory,some adaptive tracking controllers for ensuring the high-performance of servo drive system operating are designed,which comprehensively focus on the output constraints,preset performance constraints,time-delays,unknown control gains,uncertain disturbances,chaotic oscillations and parameter perturbations.Firstly,based on the dynamic model of permanent magnet synchronous motor(PMSM),the chaotic dynamics of the system are revealed.Then,the adaptive dynamic surface controller with uncertainty bounded disturbance and output constraints,adaptive tracking controller with asymmetric input-output constraints and unknown control gains,tracking error performance constraints and time-delays are studied.The main contributions of the paper are organized as:(1)Aiming at the chaotic PMSM system with output constraints,parameter variations,unknown gain value,and bounded uncertainty disturbances,an adaptive dynamic surface tracking control method based on Radial Basis Function Neural Networks(RBFNNs)is studied.Firstly,barrier Lyapunov function(BLF)and nonlinear damping technique are used to deal with the output constraints and uncertain bounded disturbances.Secondly,the unknown nonlinear function and control gain value are estimated by RBFNNs and the adaptive parameter estimation method,respectively.Thirdly,a first-order filter is adopted to deal with the "explosion of complexity," and a RBFNNs-based dynamic surface controller is designed.Meanwhile,it is proved that all signals in the entire closed-loop are bound,and the system output constraints are not violated.Finally,the results of simulation experiments and scheme comparisons show the effectiveness of the proposed control scheme.(2)The adaptive tracking controller for servo drive system with asymmetric inputoutput constraints,unknown control gains,chaotic oscillations,bounded disturbances and parameter perturbations is studied.Firstly,a novel unified BLF(BLF without piecewise formulations)is proposed to realize the asymmetric constraint on the system output.Secondly,two auxiliary power dynamic signals are designed and combined into the tracking error variables of the last two steps of the controller design to realize the asymmetric constraint on the input of the servo drive system.Thirdly,Nussbaum type gain function and tracking differentiator are used to deal with the control gain unknown and "explosion of complexity," respectively.Subsequently,Chebyshev Neural Networks(CNNs)are used to approximate the uncertainty function in the controller design.Then,an adaptive neural tracking controller is designed,which successfully deals with asymmetric input and output constraints.Finally,the results of simulation experiments and scheme comparisons verify the superiority of the proposed control scheme.(3)A neural adaptive control algorithm for servo drive system with output tracking error performance constraints,unknown control gains,chaos and time-varying delays is studied.Firstly,a new prescribed performance TBLF(Prescribed performance-tangent BLF,PP-TBLF)is designed,which ensures the output constraint and the preset performance constraint of tracking error,thus enhancing the transient performance of the system.Secondly,a Lyapunov-Krasovskii functional is constructed to compensate the time-varying delay effect in the system state.Thirdly,by integrating Nussbaum type gain function,tracking differentiator and CNNs,a neural adaptive control algorithm is designed,and the stability of the system is proved,and the constraints are not violated.Finally,the effectiveness of the proposed method is confirmed by the simulation experiments and scheme comparisons.(4)A neural adaptive controller is designed for the servo drive system with asymmetric tracking error constraints,unknown control gains,chaotic oscillations and time-varying delays.Firstly,a novel unified prescribed performance TBLF is proposed to ensure that the asymmetric performance constraint of the system output tracking error and the system output are not violated at the same time.Secondly,combining with the time delay terms of dynamic equations,a more universal Lyapunov-Krasovskii functional is designed to compensate the time-delays.Then,a backstepping neural adaptive controller is designed by combining Nussbaum type gain function,tracking differentiator and CNNs.The system stability is proved by Lyapunov-Krasovskii stability theory,and the output constraints and asymmetric performance constraints are not violated.The effectiveness of the designed controller is verified through the comparison and analysis of simulation examples and common algorithms.