Walking pattern generation based on energy function and inverse pendulum model for biped robot

摘要

The main propose of this paper is to make comparison between the ZMP-based Gravity-Compensated Inverted Pendulum Model(GCIPM) joint trajectory generator and the unequal height of hip joint trajectory generator proposed in our previous work based on energy efficiency function. Because the battery equipped on biped robot is very height and cannot maintain the power supply status for a very long period, the walking pattern of biped robot needs to be not only human-like but energy-efficient. Our hip trajectory pattern is proved to be able to work with good stability and energy-efficient in our simulation environment. And we also investigate the robot is more energy efficient with the proposed trajectory generator than GCIPM trajectory generator with large step length walking pattern.