Real-time risk assessment method for multi-aircraft interaction based on potential field theory

Abstract

To assess the operational risk of multi-aircraft interactions in complex air traffic scenarios, this paper proposes the 'multi-aircraft interaction risk potential field,' inspired by the similarities between aircraft risk and potential fields. The method further establishes risk potential fields for interactions both among multiple aircraft and between aircraft, waypoints, and routes. This paper develops real-time risk assessment models for both the aircraft level (microscopic state) and the airspace sector level (mesoscopic state), based on potential risk metrics. Comparisons with the air traffic controllers' subjective risk metric (Risk_SE) and the traditional conflict-time-based metric (Risk_ATSR) in a simulated visual environment of a designated airspace sector affirm the efficacy of the model. The results show that the proposed model substantially aligns with Risk_SE, displaying heightened sensitivity in specific intervals, evidenced by the mean absolute error rates, with Error_PE at 0.073, significantly lower than Error_ATSR at 0.121. Concurrently, the sector risk metric (Risk_S) realistically shows a delayed growth in assessment compared to the microscopic-state metric. Therefore, our method offers enhanced precision in representing operational risks for aircraft and airspace sectors. It also serves as a vital reference for decision-making in intricate air traffic scenarios and supports the sophisticated refinement of trajectory-based operations (TBO).