Mechanism Design And Active-passive Control Research For Humanoid Robot Foot

Biped robot research is a hot topic in the field of robotics, over the past decade, several anthropomorphic robots have been constructed. But compared with the people, those robots have a big different in the walking stability, gait, walking speed and so on. In the traditional design, bipedal robot often ignored the role of the forefoot in the robot walking. In recent years, though the study of same scholars, found the forefoot of robot plays an important role of improving robot walking speed, increase the length and range of step. In this paper, we design a robot foot mechanism with forefoot, and make a study of its passive control problem.Through the analysis of the geometric configurations of robot walking, proved the forefoot has many advantages in making robot get natural gait, larger step, higher lift height, faster speed, etc. And deduced the forefoot joint’s torque of the robot walking by compare the robot inverted pendulum model with ZMP stability criterion.Through analysis several different suitable compliant robot foot, the foot mechanism design scheme is determined. Deduced the output torque of foot mechanism by using the important structure and mechanic parameters of foot mechanism. The optimization model is established under the different robots’ forefoot rotation Angle and angular velocity, and make the RMS and maximum error of the output torque and the forefoot theory demand smallest as a goal, at last the genetic algorithm was used to solve the model. After get the optimized mechanism and structure parameters, completed the structure design of foot mechanism. At last using the software FLUENT analysis variable damper which is the most important part on the foot mechanic.The status of active and passive of foot mechanism control were divided, and the controller of foot mechanism was designed by using the method man-imitating intelligent PID. In order to get the reference joint trajectory, the ZMP track and ankle track was planned, and joint movement of the robot was got by using the method of control theory and solving inverse kinematics. Established the virtual prototype system of the robot with forefoot and active and passive control systems separately by using ADAMS and MATLAB/Simulink. In the simulation, the angle of forefoot joint can successfully follow the designed joint trajectory, proved the effectiveness of the control strategy.