PSO Rapid Ascending Trajectory Planning Method Based on Neural Network Trajectory Surrogate Model
To reduce the time cost of Runge-Kutta multi-step prediction calculation process of PSO direct shooting method in solving trajectory planning problem, a fast trajectory planning method based on neural network surrogate model converting multi-step prediction process into single-step prediction is proposed. In this method, the solution space samples consisting of feasible trajectories are designed to approximate the real feasible trajectory solution space, and the training set of neural network is constructed by feasible trajectory database. The samples generation time is reduced by trajectory reusing, and the incremental method of distinguishing reference state vectors at different times is used to reduce the complexity of the model so as to facilitate the training of the model of the neural network. The simulation results show that the method is fast, feasible and adaptable.
KeywordsSurrogate model Neural network Particle swarm optimization Ascending trajectory planning Shooting method
This work was supported in part by the National Nature Science Foundation of China (Grant nos. 61873319, 61803162 and 61573161).
- 2.Changwan, M., Jianping, Y.: Introduction of military aircraft route planning. Flight Dyn. 4 (1998)Google Scholar
- 5.Verma, A., et al.: Neural dynamic trajectory design for reentry vehicles. In: AIAA Guidance, Navigation and Control Conference and Exhibition (2007)Google Scholar
- 6.Julian, K.D., Kochenderfer, M.J.: Neural network guidance for UAVs. In: AIAA Guidance, Navigation, and Control Conference (2017)Google Scholar
- 10.Rao, A.V.: A survey of numerical methods for optimal control. Adv. Astronaut. Sci. 135(1), 497–528 (2009)Google Scholar
- 14.Geiger, B., Horn, J.: Neural network based trajectory optimization for unmanned aerial vehicles. In: 47th AIAA Aerospace Sciences Meeting Including the New Horizons Forum and Aerospace Exposition (2009)Google Scholar
- 15.Xue, M.: UAV trajectory modeling using neural networks. In: 17th AIAA Aviation Technology, Integration, and Operations Conference (2017)Google Scholar
- 16.Murillo, O., Lu, P.: Fast ascent trajectory optimization for hypersonic air-breathing vehicles. In: AIAA Guidance, Navigation, and Control Conference (2010)Google Scholar
- 17.Sagliano, M., Mooij, E., Theil, S.: Onboard trajectory generation for entry vehicles via adaptive multivariate pseudospectral interpolation. In: AIAA Guidance, Navigation, and Control Conference (2016)Google Scholar