Planning and Operational Algorithms for Autonomous Helicopter

The problem of autonomous flight of an uninhabited helicopter in an unknown environment is studied. Information regarding the instantaneous location of the obstacles is assumed to be available only during flight by sensing the environment using onboard equipment. The problem is divided into two tasks – the "path planning" task which is employed for the trajectory estimation and the "motion planning" task which is aimed towards the determination of an optimal strategy that is used in order to follow the estimated path. To ensure path feasibility, the two tasks must work in accordance, and the vehicle's dynamics must be taken into account. Hence, two methodologies for solving the problem in real-time mode are proposed: (a) an optimized method based on the potential field approach is employed for the trajectory planning and (b) improved inverse simulation technique is used for planning of the control commands for the former trajectory. Additionally, a new method that combines the above two described stages into a unified methodology is proposed. In this method, part of the method is carried out during the pre-mission phase because of the computational complexity of the problem. To test the algorithms and to show the viability of these methods, several maps of the environment were investigated.