Software-enabled adaptive mode transition control for autonomous unmanned vehicles

Typically, complex large-scale systems such as unmanned aerial vehicles are required to operate in a finite number of operational modes that necessitate robust, stable and smooth transitions between them. In this work, an on-line adaptation scheme is proposed for adapting the parameters of mode transition controllers designed off-line via the method of blending local mode controllers. The adaptive mode transition control algorithm is enabled via a software architecture that accommodates functionalities such as dynamic reconfigurability, plug and play extensibility, interoperability and openness. Through its reconfiguration management, virtual resource network and real-time distributed computing, it allows the UAV to execute agile and extreme performance maneuvers.