Research And Implementation Of Six DOF Parallel Platform Control

With the development of Chinese manufacturing industry,industrial automation equipment represented by CNC machine tools has put forward higher and higher requirements for the processing accuracy of parts.The mainstream CNC machine tools on the market usually use multi-axis serial platform solution.This solution is simple to implement,but there are disadvantages such as large cumulative error between multiple axes and small structural stiffness.The parallel platform solution effectively solves these problems and theoretically more suitable for high-precision CNC machining equipment.However,due to the complexity of parallel platform control and high technical threshold,the high-precision parallel platform control system is almost monopolized by the foreign companies.In order to implement the Chinese parallel platform CNC machine tools project,our team carried out the research and development of the parallel platform control system.Based on the Stewart six-degree-of-freedom parallel platform,this thesis studies from the aspects of platform construction,servo tracking,kinematics and motion control.The main contents are as follows:Firstly,the construction of the basic parallel platform was carried out.The mechanical structure design of the parallel platform was completed,and the hardware design of the motor driver was completed based on the MCU + pre-drive chip + threephase bridge.Write the motor driver and computer programs,complete system debugging,and realize the closed-loop control effect of posture in the working space of the parallel platform.Then,the servo drive controller is optimally designed.The field-oriented control drive algorithm of brushless DC motor is studied,and a fuzzy integral + neuron PI double-loop control structure with feed-forward branch is proposed,which better completes the closed-loop task of motor speed.This algorithm applied to the parallel platform.Next,the kinematic models of parallel platforms were studied.The space geometric model of the parallel platform is established by using the space vector geometry method,and the analytical expression of the inverse kinematics of the platform is obtained.Forward kinematics algorithm based on the traditional Newton iterative is studied,the flaw of the algorithm’s poor real-time performance was found.In order to improve the real-time of the system,a new forward kinematics algorithm based on multiple polynomial regression is proposed.Simulation found that the polynomial regression algorithm had limited computational accuracy,so the algorithm was combined with Newton’s iterative algorithm.Simulation shows that the polynomial regression + Newtonian iteration algorithm greatly reduces the solution time and iteration times while ensuring accuracy.Finally,the state space model of the parallel platform is established based on modern control theory,and the parallel platform controller is designed by using the model predictive control method.A loss function is constructed to evaluate the performance of the model prediction controller,and the controller parameter optimization is completed by using the adaptive differential evolution algorithm.The algorithm was successfully deployed to the host computer and achieved good control effect.