Avoidance Navigation Based on Offline Pre-Training Reinforcement Learning

TL;DR

This paper introduces an offline pre-training deep reinforcement learning approach for mobile robot avoidance navigation that significantly reduces training time and demonstrates strong generalization in various environments.

Contribution

The study proposes a novel offline pre-training strategy with expert experience collection, improving training efficiency and generalization for robot avoidance navigation without maps.

Findings

Reduced training time by 80%

Doubled the reward compared to baseline DRL

Achieved collision-free navigation in multiple environments

Abstract

This paper presents a Pre-Training Deep Reinforcement Learning(DRL) for avoidance navigation without map for mobile robots which map raw sensor data to control variable and navigate in an unknown environment. The efficient offline training strategy is proposed to speed up the inefficient random explorations in early stage and we also collect a universal dataset including expert experience for offline training, which is of some significance for other navigation training work. The pre-training and prioritized expert experience are proposed to reduce 80\% training time and has been verified to improve the 2 times reward of DRL. The advanced simulation gazebo with real physical modelling and dynamic equations reduce the gap between sim-to-real. We train our model a corridor environment, and evaluate the model in different environment getting the same effect. Compared to traditional method navigation, we can confirm the trained model can be directly applied into different scenarios and have the ability to no collision navigate. It was demonstrated that our DRL model have universal general capacity in different environment.