3-D path planning with multi-constrains based on genetic algorithm

Path planning for low attitude penetration of UAV is a complicated optimization problem with multiple constrains. A new approach of three-dimensional path planning was presented and some simulations were completed. It is very efficient by applying terrain smoothing technology to construct a safe terrain-follow surface for dealing with constrains of ascending/descending angle and normal acceleration. Then a new planning strategy based on azimuth variation of flight path in horizontal plane was adopted. It takes into account of minimal path step and minimal turning radius. At last a satisfying flight path was obtained by using genetic algorithm and geometry computation. The cost function of the algorithm is about distance, threat and height. Simulation result shows that this approach is very efficient by taking into account of all kinds of constrains and the path generated is flyable.