Research On Aircraft Mid-term Conflict Detection Algorithm Based On Hybrid System

The conflicts between aircraft and hazardous areas have become a major factor that impact air traffic safety with the increasing of air traffic flow and traffic complexity. As an important decision-support tool, mid-term conflict detection horizon is advanced to tens of minutes, giving air traffic controllers much more time to ring the alarm and resolution, reducing the workload of air traffic controllers, ensuring the safety of air traffic operation effectively. Accurate and efficient mid-term conflict detection algorithm is an important safeguard for the safe and efficient operation of air transport.For the comprehensive requirements of accuracy, timeliness and effectiveness, this paper proposes a stochastic reachability theory-based single-leg conflict detection algorithm considering wind disturbances which is different from the traditional probability and deterministic mid-term conflict detection algorithm. An encounter model for two aircraft flying at the different altitudes is introduced by studying the mid-term conflict detection problem. the relative position is governed by the stochastic differential equation after taking into account the correlation between the wind perturbations affecting the aircraft positions. Defining a Markov chain that converges weakly to the equation, the probability of conflict from two aircrafts can be computed quickly with variable relative velocity at all future time instants in tens of minutes. Also, the probability of conflict can be computed when the current aircraft positions are uncertain. The algorithm takes into account aircraft position errors caused by a variety of uncertainties, with speedy calculation, it meets the request of accuracy and effectiveness for mid-term conflict detection.For the mid-term conflict detection, the effect of special and temporal correlation of wind field to conflict detection is difficult to determine. Based on the above single-leg mid-term conflict detection algorithm, three different correlation models are constructed. After conflict detection in consideration of the each correlation model, comparing the level curves of estimated probability of conflict over tens of minutes, the affection to the probability of conflict from correlation is analyzed, conflict probability is different in different correlation functions. It's important to choose an appropriate correlation function for the reliability of mid-term conflict probes.For multi-route in Chinese civil aviation, a multi-route mid-term conflict detection algorithm based on hybrid system is presented based on the above studies, concerning the correlation function, an enlarged Markov chain process is introduced to solve the hybrid system, then the level curves of estimated probability of conflict over tens of minutes can be obtained, simulation results show that the probability of conflict is related to the current position of the aircraft, flight speed and the correlation function, It's important to choose an appropriate correlation function for the reliability of mid-term conflict probes.