Research on Robustness of Lightweight SLAM Framework in Unknown Environment

In this paper, a lightweight event-triggered back-end optimization algorithm of distributed attitude relaxation architecture is proposed to solve the failure problem of cooperative SLAM in multi-robot systems due to lack of computing resources and extreme communication conditions. This paper simulates extreme network scenarios in space environment, uses a decentralized algorithm with low computing requirements in high packet loss rate, optimizes and verifies that map drawing can be corrected by using a single loop data packet in the environment, which reduces the error caused by sensor drift, and determines the robustness and accuracy of the algorithm in extreme environments.