Research On Model Identification And Force Control Algorithm Of 6PUS-UPU Redundant Actuated Parallel Manipulator

With the development of parallel robot technology and its application in the field of industrial production is becoming more and more widely. Therefore, the related research on parallel robot has attracted more and more attention. Compared with the serial robot, parallel robot has the characteristics of compact structure, high rigidity, strong bearing capacity, but also has smaller working space, easy to produce coupling force, appear singularity and other shortcomings, which makes the application of parallel robot is limited. In the proposed solution of many researchers, redundant drive has become a kind of effective solution.The 6PUS-UPU redundant driven parallel robot is our research object in this paper. We identified its dynamic model, and studied the lag, controller design and improvement of the redundant driving branch, so as to achieve the optimization of the internal force of the structure, improve the stability of the system. The specific research contents are as follows:Firstly, we make fuzzy identification to the dynamic model of the parallel robot and design the controller for the redundant branch. For most of complex systems, it is difficult to explain the mathematical model of the controlled object. The dynamic model of parallel robot is a multi input and multi output nonlinear mechanism, and its internal structure is very difficult to be described accurately by the mathematical model, so the model of system identification is needed. Here we construct a fuzzy T-S model to identify, and based on the model we design a controller for the redundant branches, to achieve the purpose of the optimization of the driving force.Secondly, we study the lag problem of redundant branches, and design the Smith predictive compensation control structure. Hysteresis is an important feature of the process control system, which can lead to instability of the system. In this paper if the redundant branch has a lag, the impact on the overall performance of the system is huge, so we use the prediction compensation method proposed by Smith to eliminate the influence of lag in the closed-loop system through the compensation link. After solving the lag problem, we improved the control algorithm of the redundant branch, and designed a fuzzy PI controller. We use the joint simulation of ADAMS and MATLAB to verify the correctness of the controller, and realized the further optimization control of the driving force.Finally, considering the fuzzy controller cannot self learning, we add the neural network structure with learning ability, and design a fuzzy neural network controller to improve the control effect of the redundant branches, so as to improve the performance of the whole system. Finally, we validate the design in the joint simulation of MATLAB and ADAMS.