CPG Gait And Reflex Motion Control Of Pneumatic Muscle Quadruped Robot

Bionic quadruped robots have the characteristics of high rugged terrain adaptability,strong dynamic motion control and other characteristics,which make them have a good application prospect in unstructured terrain environment,such as field exploration,material transportation,high-risk environmental operation,disaster rescue and other fields.In this thesis,Bionic muscles are used as the actuators of quadruped joints,the rhythmic motion gait and bioreflective motion of the quadruped Central Pattern Generator(CPG)is studied,and achieving the motion control of the bionic quadruped robot.Firstly,from the perspective of biology,quadruped robots adopt the advantages of CPG rhythm gait for motion control.The outputs map of the existing Kimura CPG model is further improved.The knee and hip signals can be simultaneously generated by the oscillator,and the phase relationships can be freely combined without the need for complex knee-to-hip mapping operations.Based on this,the rhythmic gait of a pneumatic quadruped robot is planned.Using MATLAB simulation software to analyze the impact of CPG parameters on the model's period,amplitude and convergence characteristics.Then,compared with the hydraulic drive and the motor drive,the pneumatic muscle quadrupedal robot has quick response and flexibility,and it is suitable for reflective motion control.Using the bio-reflection motion mechanism,the reflection motion of the bionic robot is modeled and the robot's adaptability to the complex environment is realized.The stretch reflex is regulated by the central nervous system.Through the setting of the stretch threshold and gain,the position control of the foot robot is achieved,making the movement of the foot robot more stable.Besed on the biological neural reflection mechanism and the Virtual Model Control(VMC),using the limb swing phase and foot contact information as the state switching conditions,and the cycloid function is used to plan the foot trajectory,so as to establish an adaptive gait generation method for the gully terrain;Using the relationship between robot body attitude angle and joint balance position,a vestibular reflex model for achieving stable slope motion of a quadruped robot is established,and the integration of the vestibular reflex model with CPG is realized,improving the ability of quadruped robot.Finally,a quadruped robot simulation platform is built.Using the modified CPG to construct the trot rhythm gait,the quadruped robot is simulated by crossing the gully and slope.With the "PC + Motion Controller" solution,the control tasks of the real-time module are mainly implemented by the Beckhoff embedded controller.Based on the Twin CAT programming environment,the upper computer control interface is developed to realize the rhythm gait generation and reflection operation planning,and the rhythm gait tracking status is displayed in real time.The quadruped robot experimental platform is perfected and realizing quadruped robot cross the gully experiment.Through simulation and experiments,the adaptability of the reflex movements based on the CPG model planning to unstructured environments such as gully and slope is verified.