Genetic algorithm for path planning of UAVs as a maze-solving problem

This research work addresses the problem of path planning for unmanned aerial vehicles (UAVs) in complex environments such as a maze. A genetic algorithm (GA) with variable chromosomes and gene change conditions is designed when the decision point criteria and a collision with a wall are presented; without considering large chromosomes. Our proposed GA obtains the sequence of minimum movements required to solve the maze. Then, the trajectories are generated by high-order polynomials. Finally, numerical simulations and real-time tests are carried out to validate the proposed algorithm.