摘要

A decentralized control approach is presented for a network of robotic sensors to achieve the global coordination: group aggregation or consensus flocking. Each robotic sensor obeys three local interaction rules: group cohesion, collision avoidance and velocity alignment. Accordingly, the resultant control law for each robotic sensor is composed of three associated control components. The cohesion control component is based on geometry and additionally enables maintaining initial interconnection of network. The alignment control component ensures that all members in an interconnected network eventually reach to a common velocity. The control component for collision avoidance among members, as a passive role, is embedded in cohesion control. Simulation results demonstrate the capability of the proposed control approach in achieving the coordination behaviour with an initially interconnected network. However, we observed that an initially non-interconnected network may also achieve the coordination behaviour depending on the initial distribution and velocities. Initial interconnection of a network is only a sufficient but not essential condition for the network to achieve the coordination behaviour for our control approach.