Study Of Prediction Of Three-Dimensional Information And Path Planning For Aircraft

As important research subjects in vehicle path planning study, three-dimensional information prediction based on sensor data and vehicle path planning not only provide a new planning model to the existing route planning, which is adapted to the characteristics of new aircraft as high-speed and long endurance, but also make vehicle real-time planning possible. This research project mainly focuses on two major themes, namely, three-dimensional prediction of the spatial information based on sensor data and route planning for air vehicle.In three-dimensional information prediction, three-dimensional space data by the detector are laminated, and those from the most representative layer (characteristic layer) are selected for data processing, then through correlation operation of the matrix in characteristic layer and the pattern matrix and after hypothesis testing, predictors are selected to build prediction matrix, and field information of unknown region is predicted finally through a weighted processing. Thus it provides the available preliminary data for vehicle route planning which ensures the optimal safe flight.The vehicle path planning based on three-dimensional data includes point-to-point path planning which is based on moving targets, cooperative planning of multi-UAV and multi-objectives and track processing of non-steady region. For point-to-point path planning based on moving targets, through an analysis of A* algorithm and genetic algorithm, the paper puts forward the concept of using improved A* search algorithm for ground planning and genetic algorithm based on terrain matching to make real-time correction during the flight time. While in multi-UAV and multi-objectives path planning, the paper brings forth the combined pretreatment of regional data resampling and clustering of target points, and then proposes a TSP problem solving model based on the interference operator after a detailed analysis of this algorithm. Finally, the author constructs a algorithm about qualitative-change-model for the non-steady region, and offers the method to use this model to smooth the route.The study not only gives a new algorithm of three-dimensional prediction of the spatial information and qualitative-change-model, but also improves some relative intelligent algorithm which was used in vehicle path planning. Experiments show that those intelligent path planning model based on the theory extemly decrease the cost of time in the process of path planning and strengthen the ability of avoiding threats, which establish a new profess-ional platform of vehicle path planning.