Flocking in Dynamical Networks Using Fuzzy Logic

In this study, flocking motion control algorithm of multi-agent network with dynamical topology is studied using fuzzy logic. Cooperative control algorithm is proposed for a group of autonomous agents to achieve flocking formations following a virtual agent. The control law consists of three terms. The first term is a gradient-based one designed using fuzzy logic; The second and third ones are for velocity consensus and tracking purposes, respectively. The virtual agent is introduced to represent a group objective for tracking or motion planning purposes. Smooth graph Laplacian and Smooth attractive/repulsive potential designed using fuzzy logic are used to overcome the difficulties in stability analysis. The theoretical result is presented to indicate the achievement of flocking motion (cohesiveness, collision avoidance and velocity matching). Finally, simulation example is given to validate the theoretical analysis.