Integrated Decision-Making Approach for Automated Diversion Runway Selection and Flight Path Generation

The automated in-flight airport diversion trajectory generation in case of emergency is, to this day, an open problem and a real need for general and civil aviation. This problem is often addressed by solving three subproblems independently in cascade: i) selection of diversion airport, runway, and approach mode; ii) obstacle-free path planning; and iii) flyable trajectory generation under given aircraft performance. This paper proposes an integrated decision-making approach for subproblems (i) and (ii) by performing multitarget minimum-risk path planning. The [Formula: see text]-best runway candidates are first identified based on runway-risk criteria. Then the Monte Carlo tree search algorithm is applied to simultaneously search for obstacle-free paths to these [Formula: see text] candidates while minimizing a global risk by aggregating the en route and runway risks. Unlike the cascaded approaches, the proposed integrated approach can evaluate the en route risk more precisely and efficiently and use it in deciding the most suitable diversion runway. Moreover, the planning difficulty is also reflected in the runway selection. The proposed approach is tested in simulation for 1000 randomly generated missions. The results are compared with those of cascaded approaches, demonstrating a higher success rate of feasible path generation in difficult mission configurations.