| In the research field of robotics, micromanipulator has now become one of thehot topics. The micromanipulator can achieve fine operation, positioning precision atsubmicron even to the nanometer level, and has been widely used in biology, medical,aerospace and other fields. Because having the advantages of high rigidity, strongbearing capacity etc, the parallel mechanism has a good application in the structuraldesign of micromanipulator.On the basis of the analysis of 3-5R parallel micromanipulator, this paper notonly researches its position and stiffness, but also designs the control experimentsystem of the mechanism.The analysis of position and stiffness are more important two parts of parallelmicromanipulator. According to the characteristics of micromanipulator, this paperanalyses position and stiffness respectively by using differential method and flexibilitymatrix method, establishes the characteristics matrix between input and output andstiffness model, then uses the finite element method for validation, and establishesfoundation for controlling the micromanipulator.Since the control algorithms of PID and CMAC neural network have their ownadvantages and disadvantages, this paper uses neural network PID control methodafter researching the two control algorithms. Through the simulation analysis, PID andCMAC compound control has a good control effect, and can meet the requirements ofthe application of micromanipulator.Micromanipulator is a special mechanism, and its movement is implemented bythe flexible deformation of the material, so in the design of experiment system, we cannot adopt ordinary motor drive. This paper designs an experimental system of PZTdrive power, PZT actuators and micrometer etc, they constitute the hardware, and theprogramming of software is accomplished by LabVIEW.At last, the paper analyses the trajectory planning and the result of controller’sexperiment, which proves the validity of this control system and establishes the foundation for further study. |
PDF Full Text Request