Localization of multiple odor sources via selective olfaction and adapted ant colony optimization algorithm

This paper presents an asynchronous method for localizing multiple odor sources one by one. We use adapted ant colony optimization algorithm and flux divergence based idea for plume tracing and source declaration, respectively. By selective olfaction, we mean that the concentration sensors are halted when the robots are searching in the declared areas. Thus, the robots can successfully jump out of the local concentration maxima in declared areas, and converge to other concentration maxima that may contain real sources. It is unnecessary to employ more robots to localize more simultaneously releasing odor sources in our method. Simulation results show the proposed method can localize multiple odor sources in a large ventilated outdoor environment with considerably high accuracy.