Research And Design Of Control System For Bionic Quadruped Robot

Bionic quadruped robot has amazing potential applications as an intelligent mobileplatform in the field of scientific exploration, mountain transportation, disaster relief andmilitary operations. To continuously improve the mobile ability of the robot in wildenvironment has become one focus of research in the field of robotics. High adaptability tocomplex environment and highly dynamic reaction for external interference as well as motionbalance control under high load are main bottlenecks for quadruped robots to practicalapplications. Demands on real-time control system and robustness of the motion controlalgorithm are highly increased.In this thesis, the quadruped robot’s structure model and kinematic equation is createdbased on a four-legged mammal bionics model. Two motion control algorithms are designedaccording to the characteristics of trot gait. One of them is based on position control, and theother is based on impedance control. Considering the bionic control model and missionrequirement of the control system, a hierarchical architecture is selected as the generalframework of quadruped robot control system.In the overall framework of the control system, the hardware of the control system isdesigned. Design requirement of the supervisory controller and slave controller is proposed.The selection of hardware for the supervisory controller, design of the slave controller and theconditioning circuit of sensors are accomplished.On the basis of hardware design, respectively, the overall frameworks of control softwarefor the supervisory controller and slave controller have been designed. QNX is selected as thereal-time operating system for the supervisory controller. A detailed description and design ofthe multithreaded programs of supervisory controller are achieved. Also the CAN networkcommunication program between the computer and the supervisory computer are analyzedand designed.Finally, the experimental prototype platform is introduced. Experimental test andverification on the hardware and software of the control system as well as the motion controlalgorithms are realized. Experimental results show that the control system agrees with thecontrol requirements.