Path Planning of UAV in Uncertain Static Environment Using Matrix Based Genetic Algorithm (MGA)

Genetic algorithm (GA) is widely used for path planning problems of mobile robots and UAVs. To enhance the performance of conventional GA in terms of convergence speed and computation time for path planning of UAV in 3D environment, a new variant of GA based on matrix operation is proposed in the current work. The unique characteristic of proposed Matrix based genetic algorithm (MGA) is that it implies the matrix operations in chromosomes selection, crossover and mutation stage which helps in reducing the computation time. MGA is implemented for UAV path planning in indoor environment which allows us to not consider the inertial effect of the wind. The sensors attached on the periphery of the UAV collect the environment data and send to controller for processing. Performance of proposed technique is validated by performing the simulation and experimental study. For simulation the virtual environment is developed in MATAB with static obstacles. The outcomes demonstrate that employing the MGA controller enables the UAV to navigate the shortest path efficiently and safely without encountering collisions in the shortest possible time. The performance of proposed MGA is also compared with other developed AI techniques like Probabilistic Road map (PRM) The result shows that MGA outperforms PRM in terms of short and smooth path.