Research On Bean Optimization Algorithm Based On Abundance Distribution Patterns

Different species in nature have provided beautiful and impressive ways to solve large-scale,high-dimensional,and nonlinear complex problems through long-term survival and reproduction,adaptation,and evolution of behaviors.Nature is an important source of intelligent optimization calculations.Swarm Intelligence(SI)is a kind of highly efficient intelligent optimization calculation method developed by bionics on the emergence and cooperation of social biological groups in nature.It has become a hot research frontier in the field of intelligent computing.A large number of researchers are still working on developing more efficient and practical SI optimization algorithms.The Bean Optimization Algorithm(BOA)is a new SI algorithm based on the evolution law of plant population distribution in nature.It is a combination of the natural evolution strategy of plant population and limited random search.It is mainly optimized by two core strategies of father bean selection and population distribution evolution,which is different from the existing swarm intelligent optimization algorithms.At present,through the study of the spatial distribution patterns of plant populations,BOA has developed a variety of typical algorithm models,including piece-wise function model,normal distribution model,and negative binomial distribution model.This study will discuss the BOA from the perspective of plant population abundance patterns.Based on the typical log-normal distribution in the population abundance patterns,a Log-Normal Bean Optimization Algorithm(Log-BOA)is proposed,a population distribution model and algorithm structure for the proposed algorithm are designed,and then the Log-BOA is implemented.Through the optimization experiments of typical benchmark functions(including uni-modal and multi-modal functions),the performance comparison with typical SI algorithms such as particle swarm optimization(PSO)and BOA is performed to verify the effectiveness and excellent performance of the proposed algorithm.To verify the stability of the algorithm,the Markov chain model of Log-BOA is constructed,and the convergence of the Log-BOA algorithm is preliminarily analyzed according to the convergence criteria of the random search algorithm proposed by Solis and Wets.Finally,the validity of the algorithm is verified in the optimization of ranking project recommendation model of Dangshan Pear Germplasm Resource Nature Reserve.a fuzzy preference relationship model is built by categorizing the survey data according to user roles,Log-BOA is applied to quickly obtain the ranking and recommendation of user's preferences,and the personalized information recommendations are provided for all kinds of people in Dangshan Pear Nature Reserve.Finally,all the work of the thesis is summarized,and future research on BOA is prospected.