The Research And Application Of The Magnetotactic Bacteria Optimization Algorithm

As a new method of mathematical optimization,intelligent optimization algorithms require less for analyticity of the objective functions,which open a new way in optimization field for its simple principle and high efficiency in solving complex optimization problems.Recently for the characteristics of easy implementation and less parameters,power spectrum-based magnetotactic bacteria optimization algorithm is proposed based on the simulation of the interaction between its magnetic field and the geomagnetic field.The algorithm includes four operators: power spectrum calculation,bacteria rotation,bacteria swimming and bacteria replacement.Bacteria rotation and bacteria swimming operators,used for regulating the magnetic moment to adjustment the environment,can't make use of the information of the power spectrum in population.Meanwhile,bacteria replacement enhances exploration ability to some extent.So it is necessary to propose new algorithms with better convergence.This paper introduces background,development and research status of intelligent optimization algorithms and describes the optimization principle of magnetic bacteria optimization algorithm in detail.This paper intends to enhance the development ability of the algorithm and proposes two improved single objective optimization algorithms and a multi-objective optimization algorithm.The main work is as follows.Firstly,in view of the shortcomings in solving the single objective optimization problem,an improved power spectrum-based magnetotactic bacteria optimization algorithm is proposed,namely the power spectrum combined with the archives collection of magnetic bacteria optimization algorithm.Combine the bacteria rotation and bacteria swimming operators into magnetic moment regulation operator.Then replace the half of poor individuals with the random combination of the best individual and the individuals having higher cumulative contribution from the archive set by the roulette wheel selection method.Compared with the performance of the original algorithm on test functions,the improved algorithm can improve the convergence accuracy.Secondly,the improved power spectrum-based magnetotactic bacteria optimization algorithm is designed.The algorithm based on the power spectrum of the magnetic particles in the bacteria bodies improves the bacteria rotation and proposes a new replacement operator of combing chaotic mapping into bacteria replacement operator.The performance on numerical experiments verifies the effectiveness of the proposed algorithm,and the effect of the parameter is also analyzed.Finally,after studying a few calssical evolution algorithms for the multiobjective optimization problems,the power spectrum-based magnetotactic bacteria multiobjective optimization algorithm is proposed.The algorithm employs the championships mechanism of size 2 to select the non-dominated individual with bigger crowding distance as its optimal one for each individual in population,and makes use of Gauss formula to renew the magnetotactic moment based on the power spectrum and optimal moment of the population.Meanwhile a dynamic probablity variation-based strategy is adopted in the bacteria replacement operator.In the search early period adopt larger replacement parameter to accelerate the speed while in the late stage choose smaller replacement parameter to increase the diversity of the population.The results of testing the benchmark functions show that the proposed algorithm has better performance than the compared algorithms.