Decoupling control for parallel robot based on equivalent-input-disturbance approach

This paper discusses the problem of posture control for a parallel robot. The parallel robot features multiple variables and large nonlinearities. The control accuracy is strongly affected by the parameters which are coupled with each other, and by exogenous disturbances. A decoupling control strategy is presented based on the equivalent-input-disturbance (EID) approach to improve control performance. First, decoupling control is used to eliminate the main coupling effect between multiple variables. Then, the EID method is used to further improve decoupling effect and suppress the external disturbances. Simulation results show that the control strategy is effective for the decoupling control of parallel robot.

1 Introduction

Parallel robot with many advantages, such as high rigidity, high force-to-weight ratio and high positioning accuracy, has been widely used in motion simulator, parallel machine tool, vibration isolator and other industrial field. At present, the research of parallel robot mainly focuses on kinematics, dynamics and control strategy. The kinematics of the parallel mechanism mainly contains forward and inverse kinematic problem, workspace and singularity analysis. Solving the platform pose according with the length of each branch is forward kinematic, and vice versa is inverse kinematic. In this paper, the purpose of control is realizing platform positioning certain posture via control very hydraulic rod, which belongs to inverse kinematic. Dynamic model of the parallel robot plays an important role in the analysis, design, synthesis of control law, and simulation. The common modeling methods contain Lagrange equation, Newton-Euler method, virtual work principle and so on. Newton-Euler method needs to calculate the force of all the joints. Kane's law and the principle of virtual work are relatively simple, but the calculation process is more abstract, it is difficult to understand its physical meaning. The Lagrange equation systematically expresses system's dynamic characteristics by the generalized coordinates, and has clear physical meaning. In this paper, the system dynamic model is derived by Lagrange equation. The dynamic models derived by different methods are equivalent to each other.