Collision Prevention Strategy Problem Research In Air Traffic Management

Collision prevention strategy in air traffic management (ATM) is not only a short-term coordination between safety and efficiency, but also a long-term planning for national policy of airspace and air transportation system. The optimization of scheme should be based on good command of equipment, staff, procedure and operation restriction to meet the real-time requirement and integrity. Essentially, collision prevention strategy in ATM is to seek synthetically and effective automatized avoidance methods in order to reduce possibility of dangerous approach or collision, of which conflict detection, conflict resolution and resolution trajectory optimization are the key technology. The thesis gives a deeply research on the standard of intelligent prevention of collision, and theory and methodology of optimization from systems engineering perspective. The thesis detailedly describes the decision-making procedure of air traffic control (ATC), establishes the optimized target for ATC's decision-making, and puts forward an optimization of conflict detection and conflict resolution between several aircrafts in four-dimensional space. Thus it lays a solid theoretical foundation for new generation of intelligent ATC system. It is of practical significance for improving the service quality of in ATC system.The main research work of the dissertation includes the following several respect:Firstly, the thesis focuses on the dynamic identification of the aircraft position, analyzes and compares the differences and similarities of operational principles and mechanism among three kinds of collision prevention methods. The collision prevention strategy can be divided into medium-and-long-term and short-term, considering timing sequence, and determinates the decision-making sequence as the altitude, the speed and then the heading. Thus the parameters related with ATC decision-making analysis is further quantized.Secondly, the research simplifies the technique in conflict detection in four-dimensional space. The initial confirmation of the minimum safety spacing implies the evaluation of the probability of collision. Therefore, the deterministic conflict detection arithmetic based on minimum safety spacing is tested and verified. The dynamic identification of the aircrafts' position laid a solid foundation for implementing collision detection and resolution. Studying on Kalman Filtering can be used to track the position and trajectory of aircraft. In the condition of perturbation, the speed and displacement of aircraft related to time can be represented by nonlinear function in artificial neural network. Through derivation calculus to S (t), we can obtain the minimum distance between two aircrafts. By compare the result with the minimum safety spacing as published, we can detect the collision between the two aircrafts. To solve the problems of "curse of dimensionality" in process of conflict detecting among aircrafts by using Delaunay Triangle characteristics, the thesis establishes the Delaunay network corresponding to the position of n aircrafts. By using this method, it reduces the comparing times and scale of computing. On the premise of normal practice and decision-making procedural in ATC, the thesis studies on collision resolution caused between several aircrafts in four-dimensions and gave an optimization algorithm. "Divide and Conquer" not only simplifies the original problems, but also keeps the correctness and integrality. Based on the speed control by neural networks, the discussion about occupation of the flight level is converted to the generalized assignment problem. The model is quantized properly, such as profits, performance of aircrafts and human factors. The parameters can be adjusted along with the variation of tense and orientation value. Hungarian Algorithm increases the efficiency of the optimization model in a vertical profile. The example proves that this arithmetic is effective for the ATC's decision-making.Fourthly, the horizontal separation control among aircrafts can be achieved by adjusting speed or heading. First of all, it makes nonlinear Programming models with the adjustment of speed as parameter. Then it can solve the problem using MATLAB.Fifthly, according to the characteristics of collisions happened in airport, the dynamic programming theory can be used to gain an optimized sequence of approach, which lays a solid foundation for the simulation of collision prevention strategy. By making simulation testing using VB, MATLAB and geometrical graphics, it proves that the methods and models put forward in the paper can meet the ATC requirement when dealing with the collision.The application of "Divide and Conquer" is the most important innovation. The medium and long term collision prevention strategy through adjusting speed or altitude and short-term collision prevention strategy through adjusting heading makes the intelligent ATC system a perfect one. The nonlinear programming,0-1 programming, multi-objective programming and dynamic programming have significant value when putting to practice. This study has a better application prospect.