Multi-objective particle swarm optimization with multi-mode collaboration based on reinforcement learning for path planning of unmanned air vehicles

Highlights：

• Multi-mode collaboration strategy is developed with three modes to update the particle positions, including the exploration and exploitation modes, as well as the hybrid update mode.

• The proposed MCMOPSO-RL is used to find the optimal flight path for the single and multiple UAVs in the complex situations, by regarding the constraint conditions of UAV path planning as the multiple objective functions.

• Numerical experiments are carried out on various test scenarios and the results show the proposed MCMOPSO-RL algorithm can effectively solve the multi-UAVs path planning problem in the complex three-dimensional environment, and also is more efficient and robust than several existing state-of-the-art optimization algorithms.

摘要

•This paper proposes a multi-objective particle swarm optimization algorithm with multi-mode collaboration based on reinforcement learning (MCMOPSO-RL), where the reinforcement learning is introduced to enable the proposed algorithm to choose the suitable position updated mode to achieve the high performance.•Multi-mode collaboration strategy is developed with three modes to update the particle positions, including the exploration and exploitation modes, as well as the hybrid update mode.•The proposed MCMOPSO-RL is used to find the optimal flight path for the single and multiple UAVs in the complex situations, by regarding the constraint conditions of UAV path planning as the multiple objective functions.•Numerical experiments are carried out on various test scenarios and the results show the proposed MCMOPSO-RL algorithm can effectively solve the multi-UAVs path planning problem in the complex three-dimensional environment, and also is more efficient and robust than several existing state-of-the-art optimization algorithms.