Study On CPG-based Control Method For A Rod Mechanism Of Multi-mode Locomotion

Now robot is not well adaptive to unknown environments.Building the control system for a locomotive robot is too complicated.And the control system is not modular and universal.Inspired by desert spiders,a deformable mobile robot that can roll,crawl,carwheel and run in other sports modes is designed.The spider-like robot adopts CPG bionic control method.CPG neural network is designed for the robot,whichhas excellent topology and portability.Prototype robot has achieved two gait modesof stable crawling and rolling.The main work of the dissertation are as follows:According to the movement characteristics of desert spiders,we design three gaits of rolling gait,crawling gaits and rolling gait modes for the desert spider robot.A gait transformation strategy is proposed to complete gait transitions between the tree gait modes.The CPG neural network based on Hopf oscillator is designed.The CPG network parameters of each gait are determined by trail and error in simulations.The robot locomotion is simulated Using ADAMS.The forces of the bars and the motor moments of the active joints of the spider-like robot are obtained for roll gait,crawling gait,roll gait and gait transition respectively.These results are used for prototype parts selection the control theoretical basis.Design robotic control system is designed and control system components,including motion control cards,industrial computers,drivers,motors are selected.The control softwareare programed to build a prototype experimental platform.The PC interface of the experimental platform is designed using MFC to manipulate the robot and observe the real-time motion information of the robot.The experiments in rolling gait and crawling gait are completed.In rolling gait,the robot’s speed is 0.05m/s.The robot’s walking speed is 0.017m/s.The experiments verifies the effectiveness of the CPG biomimetic control method.