Study On The System And Control Algorithm For Position Force Control Of Flexible Manipulator

In recently years, there are lots of advancements in modern industry such as space technology, nucleus technology, industry automation and so on, while the development to high speed, exact precision and light weight of the robot technology, the study of flexible manipulators which representing new general robot in dynamics and control has cased much interesting. The flexible manipulators play important roles in industry and national defense fields because of its trait of higher speed, lower energy consumption, lighter weight and higher work efficiency, the study takes the values in theory and application not only in spaceflight, industry robot but also in vehicle, instrument, biologic mechanics and so on. This dissertation studies the position force control algorithm with the system of two-link flexible manipulators, and designs the hardware of experimental station's control system.Found up dynamic model is the precondition of coming true the position force control of flexible manipulators. While the flexible manipulator is a complicated nonlinear system which contains multi-input and multi-output, it has disturbing parameters and various uncertainties, it kinetic behavior is very complicated. The system must be expressed by model with infinite mode essentially, while the practical control of these disturbing parameters can only base on model with finite mode. In this thesis, based on Lagrange Equation principle and using assumed mode method, we establish the dynamic model of two-link flexible manipulators considering one rank elastic mode. From the point of view of practical control, based on the singular perturbation theory, we separate the slow and fast variables in the manipulator through introducing a small parameter. Then we obtain the slow subsystem which represent large range macrocosm movement and the fast subsystemwhich describe the small range elastic vibration. Using this method the order of the system is reduced and the design of controller is predigested.The slow subsystem can be considered as the rigid manipulator, and we design a sliding-mode adaptive controller for it. The sliding-mode control's trait is that in the switching surface, the system has a slip mode, the system isn't sensitive to the variable and disturbance in this situation, and its track in the switching surface, sliding-mode doesn't depend on parameters, the system is steady at the same time. The sliding-mode controller is easily designed; it can decouple the parameters and bring system with robust quality strongly. There are some undetermined parameters in the sliding-mode controller, it's hard to get exact values in actual instance, so we use adaptive method to get the approximate values, and optimize the control algorithm, design a sliding-mode adaptive controller in the end.For the fast subsystem, in order to compress the elastic vibration, we design an optimal controller in the fast time scale; then the combination of these control inputs yields the overall controller. By the software named MATLAB, the simulation results of this dissertation proved the feasibility and validity of the sliding-mode adaptive methods.We must simulate both numerical value and experimental station for improving mode's stability, algorithmic validity, project's application, so we need set up a dynamics and control of flexible manipulator experimental station which with trait of conciseness easily simulated. It's a important part of studying flexible manipulator's dynamics and control to optimize model and carry out advanced algorithm. This paper is based on former dynamics and control of flexible manipulator experimental station made by Robotics Laboratory of Chinese Academy of Science, and we set up the experimental station over again. The servo system of joint electromotor is the foundation of flexible manipulator's control, so we need control electromotor exactly in order to improve precision of flexible manipulator's position force control. This experimental station takes two modes of control system for various control modes, suitable in different situations and contrasting results. The experimental station consists of two control systems: one is direct control servomotor from the computer commands through numerical algorithms, which belongs to digital control system; the other is three closed-loop system, current loop, velocity loop and position loop respectively, that means analogical ones. In this paper, hardware structure and principle for this control system are introduced in details.