Combining Learning-based Locomotion Policy with Model-based Manipulation for Legged Mobile Manipulators
This work addresses the challenge of mobile manipulation for legged robots, offering a hybrid approach that is incremental in integrating model-based methods with existing learning techniques.
The paper tackled the problem of combining robust locomotion policies for legged robots with precise manipulation control by incorporating external dynamics plans into learning-based policies, resulting in zero-shot adaptation for unseen manipulators and stable hardware locomotion with external wrench prediction.
Deep reinforcement learning produces robust locomotion policies for legged robots over challenging terrains. To date, few studies have leveraged model-based methods to combine these locomotion skills with the precise control of manipulators. Here, we incorporate external dynamics plans into learning-based locomotion policies for mobile manipulation. We train the base policy by applying a random wrench sequence on the robot base in simulation and adding the noisified wrench sequence prediction to the policy observations. The policy then learns to counteract the partially-known future disturbance. The random wrench sequences are replaced with the wrench prediction generated with the dynamics plans from model predictive control to enable deployment. We show zero-shot adaptation for manipulators unseen during training. On the hardware, we demonstrate stable locomotion of legged robots with the prediction of the external wrench.