Fast Flight of the Flying Robot With Fuzzy Decision and Multimodal Control Tackling Uncertainties

Abstract

To address the challenges of flying at high speed and maintaining safety in uncertain environments, a novel autonomous flight control approach is proposed based on fuzzy decision and multimodal control. This approach first develops an uncertainty assessment model that functionalizes the uncertainties included in target capture, environmental modeling, and flight control. On this basis, a fuzzy decision on flight control is formulated comprising two parts: the control mode decision and the speed decision. In a certain environment, the control mode decision strategy selects an optimal control mode that can minimize the uncertainties in the subsequent flight according to the aforementioned uncertainty assessment model. With the determined control mode, the speed decision continuously evaluates the maximum velocity that ensures safety regarding the lumped uncertainties. Subsequently, multimodal control was developed to facilitate fast flight using the determined control mode and flight speed. Finally, simulations and experiments demonstrate that the proposed method can effectively adapt to uncertain environments and the flying robot can fly at the appropriate decided speed while maintaining safety. In a typical environment, the proposed method can approximately reduce the time cost by 20% and increase the maximum speed by 121% while ensuring safety, achieving a fast and safe flight in a complex environment with a maximum speed of 3.2 m/s.