Research On Bacterial Foraging Optimization Algorithm And Its Application

As a bionic intelligent optimization algorithm, the bacterial foraging optimization algorithm mainly simulates the foraging behavior and patterns of E. coli in the intestines of organisms. It requires more parameter settings, but some settings of the original algorithm may result in some poor performance on some specific issues. Based on the research on the original algorithm, this paper focuses on the research and improvements of two new algorithms and the main work is as follows:First, the C28-BFO algorithm is proposed based on the improvements of chemotaxis, swarming, reproduction and elimination-dispersal of the original algorithm. For chemotaxis, a decreasing composite function is used as the step size of chemokines which is larger in the early and smaller in the late. For reproduction, the introduction of Pareto’s law of the management theory improves the screening method for outstanding individuals. For elimination-dispersal, the idea of the gradient is introduced to make the stationary probability of elimination-dispersal substituted by the gradient probability of elimination-dispersal, so that bacterias on different levels are in their respective targeted elimination-dispersal activities.Second, the introduction of fine features of the particle swarm optimization based on constriction coefficient improves chemotaxis and reproduction so as to propose the CPSO-BFO algorithm. For chemotaxis, the updated velocity of the particle is used as the forward direction adjustment vector of the E. coli individual, and several special composite functions are introduced as chemotactic step size for balancing the range and precision of searching. For reproduction, the biological property of Phage preying on E. coli is introduced to select outstanding individuals, so that the population has more outstanding individuals to improve the performance of the algorithm.The effectiveness and practicability of the proposed algorithm have been proved by a large number of simulation results. Compared with the BFO algorithm, two improved algorithms have better overall convergence.The C28-BFO algorithm has a faster search speed, and the CPSO-BFO algorithm has higher search accuracy.