Adaptive PD Controller Modeled via Support Vector Regression for a Biped Robot

Abstract

The real-time balance control of an eight link biped robot using a zero moment point (ZMP) dynamic model is difficult due to the processing time of the corresponding equations. To overcome this limitation, an intelligent computing control technique is used. This technique is based on support vector regression (SVR). The method uses the ZMP error and its variation as inputs, and the output is the correction of the robot's torso necessary for its sagittal balance. The SVR is trained based on simulation data and their performance is verified with a real biped robot. The ZMP is calculated by reading four force sensors placed under each robot's foot. The gait implemented in this biped is similar to a human gait that is acquired and adapted to the robot's size. Some experiments are presented, and the results show that the implemented gait combined with the SVR controller can be used to control this biped robot. The SVR controller performs the control in 0.2 ms.