Research And Implementation On Key Technologies Of Flight Monitoring Based On ADS-B

With the rapid growth of aviation industry, aircraft ownership of civil aviation industry is increasing, and general aviation services are gradually liberalized. In “operations other than war”, aircrafts are taking counter-terrorism, security, disaster relief missions more frequently. Trend in the airline industry has led to a busy airport and airspace tension, greater stress air traffic control, and an increasing of factor of insecurity. Traditional air traffic control relies on radar surveillance, and radar surveillance has its own limitations. First, high radar station constructionand operating costs limit the number of radar stations; Second, the coverage of the radar signal does not work well in the western hilly areas, which is influenced by the geographical environment; Third, the accuracy, scan frequency of radar is not high, and information exchanged by secondary radar in air-ground data link is too small. We urgently need a better flight monitoring technologyto achieve the real-time tracking of aircraft in the airspace and conflict alert, which can helpthe officerto monitor the air traffic situation, and to avoid flight collisions. The main contents of this thesis include:1. Automatic Dependent Surveillance Broadcast technology. ADS-B is a new generation of flight monitoring technology, which is developed based on global satellite navigation system and air-ground data link communications technology. Compared to secondary radar, ADS-B has low construction and operating costs, and is suitable for deployment in complex terrain areas. What’s more, it has a more prominent monitoring performance.In the United States, Australia, the European Union and so on, ADS-B has been extensively studied and utilized.This thesis use the ADS-B technology to build air-ground data link, and ground stationgets broadcast packetsthrough an antenna, then ADS-B message data is parsedand filtrated to extract aircraft position and status information for data supporting in later applications.2. Trajectory prediction algorithm. Aircraft ground station will suffer packet loss, electromagnetic interference and measuring error, which results in a discontinuous, not smoothtrajectory with large errors. This thesis presents asliding window linear regression algorithm based on aircraft historical data, to make a reasonable estimate of aircraft movement patterns. Then we use the Kalman filter to predict the current aircraft position to form a continuous, stable and precise flight trajectory.3. Flight collision detection algorithms. Aircrafts in the airspace have to maintain certain safety interval to avoid flight collision, but increasing crowded airspaceleads to a higher possibility of a conflict between two aircrafts. This thesis, taking advantage of ADS-B’s high real-time, has proposed a short-term conflict detection algorithm based on cylindrical conflict detection model. The algorithm remove a large number of non-threatening aircraft pairs, and then complete the collision detection inthe potential aircraft collision pairs ina short time period.4. Flight collision algorithm parallelization. In order to improve the efficiency of the algorithm and real-time alarms, the conflict detection algorithm is transplanted to run on the GPU. With CUDA computing architecture, ADS-B message receiving, parsing and trajectory prediction algorithm are implemented on the CPU, and the flight collision detection algorithm on the GPU. Experimental results demonstrate the effectiveness of parallel algorithm, and a good speedup has been achieved.The various techniques and algorithms involved in this thesis have been implemented in "air service station alarm monitoring system".