Imitation Learning from MPC for Quadrupedal Multi-Gait Control

TL;DR

This paper introduces a novel imitation learning algorithm based on MPC-Net that trains a single policy to control multiple gaits in a quadrupedal robot, demonstrating effectiveness on hardware and various terrains.

Contribution

It extends MPC-Net with new loss functions and a mixture-of-experts network to enable multi-gait control in a single policy, improving over prior methods.

Findings

Single policy successfully controls multiple gaits.

Policy outperforms Behavioral Cloning and original MPC in terrain tests.

Hardware validation confirms practical applicability.

Abstract

We present a learning algorithm for training a single policy that imitates multiple gaits of a walking robot. To achieve this, we use and extend MPC-Net, which is an Imitation Learning approach guided by Model Predictive Control (MPC). The strategy of MPC-Net differs from many other approaches since its objective is to minimize the control Hamiltonian, which derives from the principle of optimality. To represent the policies, we employ a mixture-of-experts network (MEN) and observe that the performance of a policy improves if each expert of a MEN specializes in controlling exactly one mode of a hybrid system, such as a walking robot. We introduce new loss functions for single- and multi-gait policies to achieve this kind of expert selection behavior. Moreover, we benchmark our algorithm against Behavioral Cloning and the original MPC implementation on various rough terrain scenarios. We validate our approach on hardware and show that a single learned policy can replace its teacher to control multiple gaits.