Research On Self-adaptive Bacteria Foraging Optimization Algorithm

Inspired by the physiological phenomenon of nature, swarm intelligence algorithm was designed to simulate the characteristics of the natural biological groups. bacterial foraging optimization algorithm simulates the foraging behavior of the Escherichia coli, it is a typical representative of swarm intelligence algorithm. The search and optimization are the process of biological evolution and foraging, and the objective function is the biological adaptation to the environment value. The advantages of bacterial foraging optimization algorithm are parallel search and novel mechanism, etc. The algorithm has been successfully applied to a variety of combinatorial optimization problems, and achieved good results. But the bacterial foraging optimization algorithm has relatively slow convergence and speed low accuracy. According to these disadvantages, this paper will improve the algorithm based on self-adaptive methods, to improve its performance in solving problems.Firstly, this paper introduces the basic principle and mathematical model of bacterial foraging algorithm, and analyzes the three key operation of bacterial foraging algorithm. Secondly, here are some problems in bacteria foraging optimization algorithm, it is easy to fall in local optimum and it has relatively low accuracy and slow convergence speed. A new algorithm based on self-adaptive method was proposed to solve these problems. This paper mainly focused on improving two key steps of BFO, chemotaxis and reproduction. The swimming step-size was adjusted adaptively to make the algorithm rapidly converge to the global optimum; and the roulette wheel selection was introduced into the reproduction step. And select four classical test functions to test the performance of modified algorithm. Experimental results show that the modified algorithm has high convergence speed and accuracy. Finally, the RFID reader network optimization problem was used to verify the modified bacterial foraging algorithm. According to the characteristics of RFID reader network optimization problem, we selected the appropriate evaluation index to establish the simulation experiment. Experimental results show the effectiveness of the modified algorithm.