Online Robust Gait Generator of Biped Robots Inspired by Human Anti-disturbance Strategies

In order to enhance the anti-disturbance ability of biped robots, a novel linear model predictive control framework is proposed in this paper. We integrate the step duration, footstep location, and angular momentum into the objective function while the center of pressure(CoP) is located in the supporting polygon. The contributions of this paper are as follows. First, four anti-disturbance strategies are applied online simultaneously. The anti-disturbance ability is improved compared to the methods that only consider three or less strategies. Second, we use known initial values to avoid the nonlinear constraints caused by adjusting the step duration. The optimization problem is converted to standard quadratic programming(QP), thus reducing the computational complexity. Simulation results show that this approach has stronger anti-disturbance ability than the previous linear anti-disturbance methods while ensuring low time cost.