The quadruped robot excels in locomotion on rough terrains locomotion because of its higher flexibility and better environment adaptability. The gait planning in complex environments becomes an active research in the field of robotics. Based on the central pattern generator, this research focuses on the gait planning of quadruped robots in confined environments. The main work is summarized as follows. The Hopf oscillator based central pattern generator, utilized in this research, can generate rhythmic control signals with different frequencies for the swing phase and the stance phase. Considering the phase anomaly caused by gait transition, a phase delay model is proposed by modifying the Hopf oscillator. The simulation results show that the quadruped locomotion is stable during gait transition. Aiming at passing through tight spaces with free-collision, an adjustment approach of the Center of Gravity is designed to achieve stable quadruped walking. The Center of Gravity is adjusted through the rhythmic medium value transition, which automatically generates smooth trajectories according to preset parameters. A rhythmic bias compensation method is proposed for maintaining the rhythmic medium positions of legs vertical to the ground to alleviate the bump caused by the body pitch. Meanwhile, an adjustment technique for the hip motion amplitude is developed to maintain the uniform motion when the robot posture changes. A gait planning method is developed for crossing planar obstacles. The method includes locomotion generation and gait planning. Locomotion generation is composed by the central pattern generator model based on the Holf oscillator to output the standard oscillation signals, and the motion amplitude adjustment for controllable oscillation amplitudes of the negative part and positive part. The gait planning algorithm outputs a sequence set of footholds that guarantees the stability and validity of the locomotion. To improve the locomotion speed, the strategy of gait pattern transition is adopted. Gait pattern is selected according to robot postures and available area. The performances of proposed methods have been verified on a simulated quadruped robot model and a real quadruped robot.
修改评论