A method on path optimization modeling of UAV based on probabilistic model checking

Abstract

To optimize the design of flight path on unmanned aerial vehicle (UAV for short) and improve its efficiency and safety, a method on path optimization modeling of UAV based on probabilistic model checking is provided. In the proposed method, the problem for the uncertainty of the flight path of UAV is studied, it is caused by arbitrary angle selection and repeated round-trip flight of UAV. When UAV interacts with the operator on duty, the influence of operator’s characteristics are considered. In this paper, the scene of road network monitoring by UAV is taken as a case study, accepting the image form UAV, the operator on duty judges whether the image is qualified according to monitoring points and non-monitoring points. Besides, the dangerous area of UAV is avoided and repetitive flight of the initial node in the flight strategy of UAV is limited. Finally, Markov decision process model is extended to specify the formal model for UAV by the target monitoring scene, and multi-objective attributes for UAV is verified by the model checking tool for PRISM. Formal verification results indicate the workload of the operator is reduced, and the efficiency for the operator and the safety of UAV is improved in the proposed method, which provides an excellent way for establishing a safety flight model for UAV.