The aim of the present project is the design of optimal flight trajectories for an automomous aerial vehicle which is expected to reach the desired locations in the operational environment expressed in terms of planned waypoints. The navigation must be performed with the vehicle's best effort, i.e. with the lowest cost. Hence, we want to minimize the input energy, a function of the inputs for the mathematical model which describes the dynamics of the vehicle. The trajectory must satisfy all the constraints and pass through all the planned waypoints. Assuming the vehicle as a point mass model, the best solution has been investigated through a genetic algorithm search procedure. The optimisation problem has been solved by modifying a micro-genetic algorithm software which was initially developed by D.L. Carroll. Between all the possible trajectories we select the more "realistic" connections among the waypoints. First of all, we have left out the trajectories with discontinuity in the derivatives as these are not feasible by the real aircraft. The polynomial spline is a suitable candidate to solve our problem. The algorithm splits the trajectory in sub-trajectories which join a sequence of three waypoints. Starting from the first three waypoints, the following sub-trajectories are superimposed keeping the first waypoint coincident with the last of the previous sub-trajectory. The sequence of polynomials is initialized assuming that jumps in the direction of flight are avoided pointing the heading angle in the presumed direction of flight. The optimal trajectory is a trade-off amongst three factors: the required energy cost, the minimum distance from the required waypoint and the feasibility of the trajectory. Results obtained with this optimization procedure are presented
[1]
Alberto Bemporad,et al.
Reference governor for constrained nonlinear systems
,
1998,
IEEE Trans. Autom. Control..
[2]
Giorgio Guglieri,et al.
Flight control system design for a micro aerial vehicle
,
2006
.
[3]
S. LaValle,et al.
Randomized Kinodynamic Planning
,
2001
.
[4]
Giorgio Guglieri,et al.
Flight control system design and optimisation with a genetic algorithm
,
2005
.
[5]
B. Faverjon,et al.
Probabilistic Roadmaps for Path Planning in High-Dimensional Con(cid:12)guration Spaces
,
1996
.
[6]
D. Carroll.
GENETIC ALGORITHMS AND OPTIMIZING CHEMICAL OXYGEN-IODINE LASERS
,
1996
.
[7]
E. Mosca,et al.
Nonlinear control of constrained linear systems via predictive reference management
,
1997,
IEEE Trans. Autom. Control..