Hybrid Solution for UAV Multiple Target Path planning using Genetic Algorithm with Dynamic Programming

• Main Authors: Yu-MingSyu, 許育銘
• Other Authors: Chin E. Lin
• Format: Others
• Language: en_US
• Published: 2015
• Online Access: http://ndltd.ncl.edu.tw/handle/76642335772446190569

碩士 === 國立成功大學 === 航空太空工程學系 === 103 === The major objective of this thesis is to present a hybrid method by integrating, Genetic Algorithm with Dynamic Programming, for solving multiple target path planning in implementation into UAV system. In order to guarantee the safety requirement and feasibility for path planning, the flight path planning system (FPPS) necessarily consider the path security, UAV’s capability and so on. The Virtual Flight Map is built at first and the path planning would program in the map by algorithms. Between two target locations, Genetic Algorithm finds the optimal path with suitable fitness function. Then, with the distances matrix between every pair of multiple targets, Dynamic Programming (DP) expands the two targets path planning to multiple targets. The DP decide what is the visiting sequence of multiple targets which visits all targets just once and returns to the starting location. It must be the shortest distance of tour sequence. Finally, UAV system is sat up and performs flight path tracking to verify the flight path planning system.