Source seeking of multi-UAV based on extremum seeking algorithm

In this paper, a simple cooperative control strategy for multi-unmanned aerial vehicle (UAV) seeking the source of a scalar signal is developed. This strategy based on extremum seeking algorithm(ESA) and just takes the average of the environmental measurements of each agent as a demodulation signal for the ESA control loop to every UAV in the group. The scalar signal field may represent the strength of a spatially distributed signal. We assume that the signal field is the strongest at the source and decays away from it but the functional form of the field is not available to our UAVs. We point out that only rotocraft UAVs are developed as the flight platforms in this paper. It is also necessary to note that the ESA used in this paper is slightly different from the classic ESA, we call it non-steady state oscillation, which is obtained by adjusting the structure of the classical ESA to make the sinusoidal excitation signal amplitude proportional to the square of the extremum estimation error. The resulting method is efficient and precise. A simulation study involving 3 UAVs is used to verify the effectiveness of this method.