3D Trajectory Planning of Positioning Error Correction Based on PSO-A Algorithm

• Published in: Computers, materials & continua Vol. 65; no. 3; pp. 2295 - 2308
• Main Authors: Xing, Huaixi, Zhao, Yu, Zhang, Yuhui, Chen, You
• Format: Journal Article
• Language: English
• Published: Henderson Tech Science Press 2020
• Subjects: Aircraft; Algorithms; Environment models; Error correction; Error correction & detection; Inertial navigation; Navigation systems; Optimization; Particle swarm optimization; Searching; Trajectory optimization; Trajectory planning; Yaw
• ISSN: 1546-2226; 1546-2218; 1546-2226
• DOI: 10.32604/cmc.2020.011858

Aiming at the yaw problem caused by inertial navigation system errors accumulation during the navigation of an intelligent aircraft, a three-dimensional trajectory planning method based on the particle swarm optimization-A star (PSO-A*) algorithm is designed. Firstly, an environment model for aircraft error correction is established, and the trajectory is discretized to calculate the positioning error. Next, the positioning error is corrected at many preset trajectory points. The shortest trajectory and the fewest correction times are regarded as optimization goals to improve the heuristic function of A star (A*) algorithm. Finally, the index weights are continuously optimized by the particle swarm optimization algorithm. The optimal trajectory is found by the A* algorithm under the current evaluation index, so the ideal trajectory is planned. The experimental results show that the PSO-A* algorithm can quickly search for ideal trajectories in different environment models, indicating that the algorithm has certain feasibility and adaptability, and verifies the rationality of the proposed trajectory planning model. The PSO-A* algorithm has better convergence accuracy than the A* algorithm, and the search efficiency is significantly better than the grid search A star (GS-A*) algorithm. The PSO-A* algorithm proposed in this paper has certain engineering application value. The researchers will study the realtime and systematic nature of the algorithm.