Research On Nonlinear Control Of Harmonic Drive Systems

The harmonic drive system is widely used in aerospace and industrial applications due to its compact structure,stable transmission,and high precision.However,the harmonic drive system is a flexible system,and there exist nonlinear dynamic problems such as kinematic error,hysteresis,and friction,which seriously affect the transmission performance of the harmonic drive system in the application process.The dynamic characteristics of the harmonic drive system are studied and an effective control strategy is proposed to achieve high-precision control of these nonlinear dynamic problems.It has certain theoretical significance and engineering application value.Firstly,the kinematic error of the harmonic drive is compensated based on back-stepping method.This method guarantees the asymptotic stability of the closed-loop system and has high control precision.In order to further reduce the steady-state error,the method of state expansion is used to derive the design idea of the integral back-stepping method.An adaptive integral back-stepping controller is designed to deal with unknown model parameters.Secondly,ignoring flexibility can only simplify the model but does not meet the actual situation.This paper presents an adaptive integral back-stepping sliding mode control for hysteretic nonlinear high-order systems.Hysteresis of the harmonic drive is regarded as a type of bounded disturbance and compensated by sliding mode control.The back-stepping method is based on the Lyapunov function and guarantees the asymptotic stability of this type of flexible coupled complex system.In order to weaken the chattering phenomenon introduced by sliding mode control and the "calculation expansion" problem caused by back-stepping,an adaptive integral back-stepping quasi-sliding mode dynamic surface controller is designed.Thirdly,an adaptive friction compensation method is proposed for harmonic drive systems considering the uncertainty of model parameters and friction.The friction in the system is described by the LuGre model.The nonlinear state observer is used to estimate and compensate friction.The adaptive control is used to solve the parameter uncertainty.Adaptive law are given by the stability analysis.A first-order filter is used to solve the "calculation expansion" problem caused by back-stepping.An adaptive integral back-stepping dynamic surface controller based on a state observer is designed.The simulation results verify the effectiveness of the friction compensation method based on the LuGre model.The adaptive parameter estimations solve the problem of parameter uncertainty.The proposed method has good tracking performance.Finally,the high-precision control strategy of harmonic-driven manipulators with multiple nonlinear transmission problems such as external disturbance,input saturation,nonlinear friction and modeling error is studied.We design an adaptive neural network integral back-stepping sliding mode controller with a modified LuGre friction compensator.Sliding mode control is used to compensate the complex external disturbances,and a nonlinear state observer based on the modified LuGre model is used to estimate and compensate friction.The hyperbolic tangent saturation function limits the control input to the specified range.The radial basis neural network is used to estimate the part beyond the saturation range,and a robust term is introduced to eliminate the approximation error of the neural network.Thereby,we can solve the problem of input saturation.The simulation results show that the proposed control algorithm not only can effectively compensate the system friction,but also has good robustness and anti-saturation capability.