Model predictive optimization for imitation learning from demonstrations

Yingbai Hu, Mingyang Cui, Jianghua Duan, Wenjun Liu, Dianye Huang, Alois C. Knoll, Guang Chen

Abstract: Highlights•We propose the MPDMP+ for imitation learning from multiple demonstrations•Our method incorporates the obstacle avoidance in DMP for real-time control.•We apply proximal augmented Lagrangian solver to improve the computation speed.•We verify the effectiveness of the proposed method by simulation and experiments.