Research On Router Planning For Unmanned Air Vehicles

The UAV(Unmanned Aerial Vehicle)route planning is to search the optimal fight route to satisfy the UAV machine performance and battlefield environment under special restrict condition. And the UAV route planning is the key technology guarantee that UAV can flight by itself. In this thesis, three different tasks of the route planning are discussed: static route planning, dynamic route planning and multi-UAVs cooperative route planning. According to automatic search methods of UAV route, the thesis addressed the following problems: (1) Modeling of mission planning; (2) Static route planning of UAV; (3) Dynamic route planning of UAV; (4) Cooperative route planning of multi-UAVs.In the modeling of route planning, three works are conducted: Firstly, according to the characteristics of route planning, a representation of the planning space and flight route is proposed; Secondly, a representation that is convenience of calculation and realistic of the cost function is established, which maximizes the survival probability of the vehicle, minimizes route length and time of flight; lastly, a new representation of route point which is based on Polar coordinates is given, and the method has better features than that based on Cartesian coordinates.In the static route planning of UAV, a new planner that is based on Heuristic Genetic Algorithm Route Planning (HGARP) is proposed. In the algorithm, the representation of the route is based on polar coordinates, and in the process of the route initialization and iteration, the constraints of the UAV are considered. In this way, the search space can be reduced effectively and without smoothing process, the route that is generated is feasible route. In addition, to improve the search efficiency, an idea of variable step length according to the search space's environment is used. The experiences show that the algorithm can generate an optimal route in a short time.In dynamic route planning of UAV, an idea of planning during flight is adopted. The route planner is a two-stage plan: first a rough route is planned as a"beacon", and then a fine route is planned by using of this beacon. In the stage of fine route planning, the rough route points are used as local target points, which can reduce the search space and improve search efficiency. Accordingly, a heuristic algorithm based route planning in real time (RT-HGARP) is designed.At last, the coordination route planning of multi-UAVs is discussed. In connecting with the synchronization and collision-free of the multi-UAVS route, the concept of multispecies co-evolution is introduced. In the algorithm, every UAV is regarded as a different species, and evolves only within its own sub-population. The selection of the UAVs group's routes is to select outstanding individuals of every sub-population firstly, then to evaluate the co-cost of the combination of the routes, and lastly, to choose the best route group.According to this idea, a heuristic genetic algorithm based multi-UAVs route planning (HGAMRP) is designed.