Research On Indexing And Forcasting Of Trajectory Data Of Carrier Aircraft

With Internet technology and mobile computing technology, moving object data management and analysis of the growing demand, promote the development of the mobile object database-related research. In addition, aircraft carrier and other military simulation system also faces the problem of moving objects trajectories handle large data .How to management these data effectively has become a more and more important problem. For the trajectory data generated in aircraft carrier simulation system, the R-tree and several index method based on R-tree index were widely used. But due to the characteristics of aircraft carrier simulation trajectory data itself, it will generate coverage between MBRs when applying the MBR-based data indexing methods to the aircraft track, which will reduce indexing efficiency either.Therefore, this thesis based on the actual characteristics of the ship simulation data, two trajectory data indexing methods were proposed. The first one is layered - indexing method combines a fixed grid. By separation spatial dimensions in which the track data changes even,this method use different sub-index methods to enhance the ability in distinguishing temporal and spatial. The second method index the security level sequence of track data on reference points, proposed a indexing method for security level sequence based on the Q-tree. The method provides query functions for trajectory global features. In addition in this thesis, a pretreatment method of carrier aircraft trajectory data was given out, and the dimensionality reduction methods for glide path data have been studied. Proposed a reference point sequence selection algorithm based on the main element. And a main element selection algorithm of kinetic parameters of the glide path. The algorithm has been implemented and verification experimental has been finished. After the algorithm processing, forming a sequence of reference points characteristic of the original data. On top of this, this article introduced researching methods in the field of time series mining into carrier aircraft trajectory data study.Using BP neural network model, the research of forecasting results based aircraft landing and glide path data classification was conduct. Established the corresponding network model and finished the experimentally verified using carrier-based aircraft simulation data.