Research On Active Control Of Flexible Manipulator

With the development of science and technology,the demand for industrial manipulators is also increasing.But rigid manipulators,which are normally used for producting,are too heavy to move itself quick.Rigid manipulators also could not satisfy the industrial requires.The flexible manipulators,by contrast,which could move itself more quick than rigid manipulators,attracts more attention.Compared with the rigid manipulators,the flexible manipulators has the characteristics of low energy consumption and heavy load.However,due to the existence of flexibility,the dynamic characteristics of the flexible manipulators are more complicated,and it is more difficult to control it to tracking expectation position.Therefore,it has always been a hot issue in robotics.Two different structural flexibility of flexible manipulators should be noticed,joint-flexibility and link-flexibility.This paper research the dynamic modeling of flexible joint manipulators and flexible link manipulators respectively.Then two different control strategys are presented to control two different manipulators in this paper.Finally,considering the influence of joint-flexibility and link-flexibility on the dynamic characteristics of manipulator,a control strategy is prensrnted.The effectiveness of the proposed control strategies are verified by computer simulation and physical simulation.First at all,considering the structural characteristic of flexible joint manipulators,voltage control strategy is used to replace the commonly used torque control strategy,selecting joint motor voltage as system control input and hub position as system control output,for tracking the trajectory of the flexible joint manipulator.The controller is designed based on the neural network theory,the controller can be built without parameters of the flexible joint manipulators.Secondly,considering the structural characteristics of flexible link manipulators,the assumed modes method is used to describe the elastic vibration of flexible link manipulators,and the minimum phase characteristic of the system is solved by redefining the system output.The controller is designed based on nonlinear state observer and feedback linearization theory.The controller can be used to track the trajectory offlexible link manipulators.The control parameters are optimized by genetic algorithm to get better vibration suppression effect.The control strategy is relatively simple,and the flexible link manipulators system responds quickly and smoothly.Third,considering the influence of joint-flexibility and link-flexibility on the dynamic characteristics of manipulator,singular perturbation method was exploited to obtain the two-time-scale simpler subsystem.The slow subsystem controller is designed based on neural network,the fast subsystem controller is designed based on linear-quadratic theory.Computer simulation results show that the control strategy can track the ideal trajectory quickly.Finally,the testbed of the planar two link manipulators is built.The control system is built in PC.PC can indirectly control the two link manipulators by motion control card.