Research And Implementation Of Route Planning For Unmanned Aerial Vehicle Based On Genetic Algorithm

Computer route planning is a key technology to achieve autonomous navigation of unmanned aerial vehicles. Under a given space planning and constraints, Route planning searches for the optimal or feasible trajectory from the starting point to the target point for the unmanned aerial vehicles. These constraints include terrain, aircraft performance, the threat model and the task requirements. Genetic algorithm is suitable for solving complex nonlinear problems with multiple constraints. This paper focuses on the route planning issues, designs a GA-based route planning program, and verifies the correctness and availability of the design. This paper analyzes and researches on the encoding of GA, evaluation function, cost weight and genetic parameters. The main works are as follows:Aiming to solve the problem of traditional encoding method which has complex operations and low efficiency, we design the dynamic doubly linked list to improve the chromosome encoding. This approach reveals the essence of the route planning problem more closely, and can improve the efficiency of encoding and genetic evolution.Based on the fitness of genetic algorithm, we design the route evaluation function, which handles different threat problems with different ways, optimizes the threat cost model, builds a height cost model, and improves the computing efficiency of route cost.We use AHP, which transforms the multi-objective route cost weight problem into multi-level single-objective problem, to meet the requirements of route planning in different environments. The experiment shows that the AHP can solve the route cost weight problem effectively.We design an online route re-planning algorithm based on adaptive genetic search. This algorithm, which adaptively modifies the crossover and mutation probability of genetic parameters, can effectively avoid the limitation of traditional genetic algorithm that is likely to early mature and fall into local optimal solutions, achieves the avoidance of unpredictable threats, and meets the requirements of online route re-planning.In this paper, we research on the NP route planning problem with GA, and verify the validity of GA by experiments.