Design Of Robot Bionic Eye System Based On Parallel Structure

Robot technology,as a representative of high technology,has been developing rapidly in recent years and applied in more and more fields.As one of its important functions,the development of vision function of the robot is also very necessary.In this paper,a design scheme of bionic eye system is proposed referring to the movement mechanism of human eyes.The bionic eyes imitated the movement mode of human eyes,which are moved by the traction of extraocular muscles.based on the research status of bionic eye mechanism and the human eye movement mechanism,the 3-RPU parallel platform is proposed as the movement platform of the bionic eye mechanism,and the following functions are realized:(1)control the bionic eye mechanism to reach any posture in the reachable domain of the upper platform;(2)control the upper platform to move according to the given motion trajectory;(3)control the upper platform to track the non-high-speed moving targets,and make the target stay in the central area of the camera image of the upper platform during the tracking process.To achieve these functions,the bionic eye system can be divided into three parts:eye movement mechanism,control system and visual system.In this thesis,a real object platform for the bionic eye movement mechanism is designed and maked,called the 3-RPU parallel platform.And the control algorithm for this platform is then studied.In the vision system part,the solution of the platform posture on the motion mechanism through images is realized,and the target recognition problem in the target tracking is solved.Firstly,the design scheme of the bionic eye movement mechanism is proposed and the selection of each part of the mechanism is carried out.The electric push rod is used as the branch chain,and the upper and lower ends are connected with the upper and lower platform by rotating pair and hooker hinge respectively,so as to complete the construction of the mechanism.Then the kinematics analysis and dynamic modeling of the bionic eye movement platform are carried out.Then,the inverse solution of the parallel mechanism is completed in the kinematics analysis,and the relationship between the position and chain length of the upper platform is obtained,so that the upper platform can reach the specified position and pose by controlling the length of the branch chain.Then,the dynamic model of the mechanism is establishedby using Lagrange method,and the platform is simulated on MATLAB to lay a foundation for subsequent control.Then the control algorithm of 3-RPU mechanism is studied.The commonly used control algorithms of parallel mechanism include PID control of hinge points and torque control of computation,and as it is difficult to measure the posture of the upper platform directly,the loop is usually opened at the posture of the upper platform by taking the rod length as feedback.In this paper,the position and posture of the platform on the mechanism are calculated through the pictures taken by the camera,and the overall closed loop of the system is realized.In this paper,the hinge point PID control is selected,and the mechanism is controlled by the branch chain length feedback PID control,the upper platform posture feedback PID control and the fuzzy PID control algorithm,and the simulation is completed by MATLAB.Finally,the target tracking algorithm is studied.According to the movement characteristics of eye mechanism and the need of target tracking,a simplified tracking algorithm is proposed to improve the real-time performance of the system.Finally,the whole control system is built and tested.With the PC as the host computer,STM32F103 as the controller,which controls the parallel mechanism through the motor drive module.PC sends control signal to STM32 with the serial port,and generates PWM wave at GPIO port to drive the parallel platform by motor driver,and STM32 reads the branch chain displacement by ADC module and sends it to the host computer to achieve the desired function.