chemotaxis-swarm-algorithm

bionic chemotaxis swarm algorithm for cooperative exploring during undersea remote sensing

Due to the harsh conditions of the seabed environment, such as high water depth pressure, low temperature, and weak light, it is difficult for traditional geological exploration techniques to obtain seabed mineral information directly. With the development of robotic technology, the image information acquired by AUVs, ROVs, and other equipment plays a vital role in the exploration tasks of polymetallic nodules. This paper proposes a bioinspired bionic AUHs swarm algorithm to find PMNs exposed outside the sediment by learning the stochastic exploration behavior of bacteria. The algorithm relies on pheromone concentration and gradient to achieve faster and more robust multi-source exploration by combining the bacterial and swarm behavior. A set of simulation experiments are conduct to compare the performance under different chemotaxis, probability guidance strength, scale robot team, and starting position in the propsed simulation environment. In addition, the proposed algorithm are compared with the ergodic searching algorithm and general chemotaxis algorithms in the exploring paths and exploration time . The results consistently show that the algorithm can efficiently and robustly complete multi-source exploration tasks in large-scale areas with no prior information, and can also obtain information that is of reference significance for PMN abundance assessment.