Asynchronous and Decentralized Multiagent Trajectory Planning in Dense Environments

This paper proposes an online decentralized and asynchronous multiagent trajectory planning algorithm in dense environments. In our algorithm, the optimization problem is transformed into a quadratic programming (QP) problem to reduce the computational complexity by constructing the optimal linear flight corridors (OLFC). A cooperation-based deconfliction framework is designed to ensure the safety and feasibility under the decentralized and asynchronous architecture. We conduct a large number of simulations to verify the reliability and efficiency of our algorithm in dense environments with higher success rate, less computational time and total navigation time, which is more aggressive and cooperative.