Formation generation and visual simulation verification for quadrotors based on time-segmented trajectory planning

Abstract

Solving the formation generation problem in urban and field environments is crucial for collaborative flight of multiple unmanned aerial vehicles (UAVs). To ensure UAVs autonomously execute tasks in obstacle environments, a method of formation generation and visual simulation based on time-segmented trajectory planning is proposed. The adaptive, safe and reliable flight trajectory is designed, and a visual simulation approach is provided for testing and verifying the planning performance. Firstly, a particle swarm optimization (PSO) algorithm is employed for trajectory pre-planning, and the Hungarian algorithm is combined for target allocation. Furthermore, based on the curve fitting, an initial value calculation method for a Nonlinear Planning (NLP) is proposed. Then, the candidate trajectory for quadrotor UAV is represented using high-order polynomials. Based on differential flatness theory, the trajectory planning problem is transformed into a NLP problem with complex constraints. Finally, a real-time visual platform for multi-UAV collaborative formation generation in visual environment is constructed. The results show that the visual simulation platform can provide intuitive visual feedback and a simulation environment for UAV formation generation in different environments, and also verifies the effectiveness of the planning algorithm.