Cooperative Mission Planning Using Rollout Policy Optimization

This paper proposes a novel cooperative mission planning for uncrewed aerial systems (UAS), using roll-out policy optimization and a blend of straight and double-Archimedean-spiral (DAS) paths. Typical mission tasks using UAS, possibly multiple, include visiting distinct locations/waypoints and covering specific areas of interest (AOIs). For the UAS, to guarantee the coverage of the AOI, the algorithm generates a DAS path that even accounts for the sensor’s field-of-view constraints. A scalar spiral density parameter in the algorithm effectively provides customized adjustments to the length of the DAS over an AOI while developing mission plans. A lower and upper bound value for the spiral density parameter and the dependency of the parameter on the coverage of an AOI are derived. This scalar parameter gives the users potential flexibility in the design of mission plans. Roll-out policy optimization minimizes the overall path length for the cooperative UAS mission and, as a direct consequence, minimizes time and fuel. This method leads to finding a near-optimal path in that the computation time does not grow exponentially with the number of UAS. Results from the simulation studies with several cooperative mission scenarios, including multiple UASs visiting several waypoints and AOIs starting from identical and distinct initial locations, demonstrate the efficacy of the proposed concept. The results from a data-based complex test case using the LiDAR data set further demonstrate the adaptability of the proposed mission planning approach in a real-world setting.