Research On Scene Grid Modeling And Application Method For Urban Space Management And Control

With the rapid development of general aviation and unmanned aerial vehicle industries and the opening of low-altitude fields,the number of aircraft continues to increase,and more and more missions need to be completed at low altitudes or even in cities,which puts increasing pressure on space management;On the one hand,intelligent warfare will subvert the traditional war mode,and UAV group operations will become an important part of future warfare.The existing space management and control application system is an object-oriented planning method,which is difficult to meet the efficient management needs of a large number of space objects.The grid management method is based on space.This method does not consider the number of objects,but considers the occupancy of the grid,turning the calculation problem into a query problem,and reducing the complexity of the object management method in principle.It provides a new solution to the problem of space management and control.This paper is based on the idea of grid space management.The research includes 3D urban scene grid modeling and data organization,3D non-rigid grid coding and dynamic object identification,and 3D path planning of UAV based on grid points.The innovations are as follows:2.Aiming at the problems of the existing modeling methods,such as low computational efficiency and large data redundancy,a unified grid framework was established and a multi-scale grid modeling method was designed to realize unified grid expression of objects in different data forms.On this basis,aiming at the low efficiency of 3D data organization,a grid-based data organization method is proposed,which can realize the efficient index and query of urban buildings and other data.Experimental results show that the method presented in this paper can realize efficient multi-scale grid modeling of different types of data,and the data organization method is better than Oracle Spatial in efficiency and stability in Spatial query of different query areas and different quantity data.The method can better meet the demand of data index in the case of big data.3.Aiming at the problems existing in the existing rigid grid structure,a non-rigid grid coding method is designed,which encodes the grid elements,edges,points and surfaces simultaneously with the same set of coding,so that any eight adjacent grids can be non-rigid aggregation.The grid point identification method is designed,and the grid point,grid edge and grid surface identification ability are added,which is used to solve the identification accuracy problem of dynamic objects in grid framework.The method of grid points representing straight line is studied to improve the fitting degree of grid and spatial track data.The experimental results show that compared with the rigid grid identification method,the proposed method is more accurate to the target identification and more suitable to the scale of the object.Compared with the Bresenham algorithm,the method presented in this paper has a higher fitting degree for spatial straight lines and a lower track deviation degree.In flight conflict detection,the query detection method is more efficient than the computational detection method.4.Aiming at the problem of inaccurate positioning of UAVs and unflyable trajectory in the traditional A* grid algorithm,an A* algorithm based on grid point optimization is proposed to realize the three-dimensional trajectory planning of UAVs in a grid framework.This method is based on grid points and completes trajectory planning through grid point search.Flight constraints are considered in the planning,and a better trajectory cost function is designed to improve the safety and flyability of the trajectory under real conditions.Experiments show that: compared with the A* grid algorithm,the method in this paper can realize obstacle avoidance and planning in three-dimensional space,and the planned route is shorter and better.