DOA: A Degeneracy Optimization Agent with Adaptive Pose Compensation Capability based on Deep Reinforcement Learning

TL;DR

This paper introduces DOA, an adaptive reinforcement learning agent designed to improve 2D-SLAM in degenerate indoor environments by dynamically optimizing sensor contributions and enhancing generalization across different settings.

Contribution

The paper presents a novel PPO-based degeneracy optimization agent with a specialized reward function and transfer learning, addressing data, quality, and annotation challenges in SLAM.

Findings

DOA outperforms SOTA methods in degeneracy detection and optimization.

The agent effectively adapts sensor contributions based on degeneracy factors.

Transfer learning enhances the generalization of the agent across environments.

Abstract

Particle filter-based 2D-SLAM is widely used in indoor localization tasks due to its efficiency. However, indoor environments such as long straight corridors can cause severe degeneracy problems in SLAM. In this paper, we use Proximal Policy Optimization (PPO) to train an adaptive degeneracy optimization agent (DOA) to address degeneracy problem. We propose a systematic methodology to address three critical challenges in traditional supervised learning frameworks: (1) data acquisition bottlenecks in degenerate dataset, (2) inherent quality deterioration of training samples, and (3) ambiguity in annotation protocol design. We design a specialized reward function to guide the agent in developing perception capabilities for degenerate environments. Using the output degeneracy factor as a reference weight, the agent can dynamically adjust the contribution of different sensors to pose optimization. Specifically, the observation distribution is shifted towards the motion model distribution, with the step size determined by a linear interpolation formula related to the degeneracy factor. In addition, we employ a transfer learning module to endow the agent with generalization capabilities across different environments and address the inefficiency of training in degenerate environments. Finally, we conduct ablation studies to demonstrate the rationality of our model design and the role of transfer learning. We also compare the proposed DOA with SOTA methods to prove its superior degeneracy detection and optimization capabilities across various environments.