Urban Air Mobility (UAM) is new air transportation concept where highly autonomous aircraft could safely and efficiently transport passengers and cargo within the city by rising above the traffic congestion on the ground. Conflict avoidance is a core research problem to enable UAM for high density operations. As the outer layer of protection, pre-departure planning can optimize trajectory under physical and operation constraints, coordinate flights with existing flight plans, alleviate the burden of inner layer sense and avoid, and increase the predictability of an operation. Dynamic geofence is a moving region of reserved airspace for an aircraft, which provides a safety buffer for various uncertainties including flight operation, localization and disturbances from weather. This paper is an endeavor to do pre-departure conflict-free trajectory planning for flight with dynamic geofence in structured airspace. We propose an efficient, scalable, two-phase, linear programming based algorithm. In the first phase of the algorithm, we discretize the flight into time periods and formulate an integer programming model, which is proved to be Totally Unimodular (TU), to avoid conflict with already filed flight plans and to determine the location of dynamic geofence in each time period. In the second phase, we propose a velocity profile smoothing model to make the trajectory easier to fly. This main advantage of this algorithm is its verifiable safety guarantee, which can be valuable to the UAM community.
展开▼