自适应骨干细菌觅食优化算法的研究

In the replication operation of bacterial foraging optimization algorithm, half of the high-quality bacteria are replicated, which reduces the population diversity, thus affecting the optimization performance of the algorithm. Inspired by the backbone particle swarm, this paper proposes four skeleton construction methods in the replication operation of bacterial foraging optimization algorithm. By using the backbone idea, the Gaussian variation of the barycenter and variance information of half of the high-quality bacteria is adjusted in real time, which increases the diversity of the population on the basis of searching the neighborhood of the semi-population elite and improves the overall performance of the bacterial foraging optimization algorithm. The fixed step size in the chemotaxis operation leads to slow convergence speed, and the swimming step size decreases with the increase of swimming times in the chemotaxis operation of the bacterial foraging optimization algorithm, hoping to keep the balance search between the large step size in the early stage focusing on global exploration and the small step size focusing on local development in the later stage. In order to compare and analyze the comprehensive performance of four adaptive optimization algorithms for backbone bacteria foraging, CEC2014 benchmark test function was used for simulation experiments.