Enhanced Symbiotic Organism Search Algoritm And Application Research

Symbiotic organism search(SOS)algorithm has been widely concerned by scholars at home and abroad because of its simple structure,easy to implement,less adjustable parameters and strong stability.At present,symbiotic organism search algorithm has been successfully applied in many fields,such as power system,traffic scheduling,and intelligent manufacturing and so on.Although some success has been achieved in the theory and application of the algorithm,researchers found that the convergence speed and solution accuracy of the algorithm still need to be further improved to obtain better performance when solving complex optimization problems.In this thesis,aiming at the shortcomings of symbiotic organism search algorithm,such as solution accuracy,late convergence speed,and local mining ability and so on,on the basis of basic symbiotic organism search algorithm,the performance of optimization algorithm is enhanced by multi strategy fusion.The purpose is to improve the theoretical basis of the algorithm and expand its application range.The main work of this paper is as follows:(1)Aiming at the multimodal error plane in IIR filter design,the traditional optimization algorithm is easy to be misled by the local minimum error plane.Using the global search ability of symbiotic organism search algorithm to find the minimum error plane can avoid falling into the local optimum prematurely and improve the search accuracy at the same time.The experimental results show that using symbiotic organism search algorithm to minimize the error plane can provide more robust filter parameters for IIR filter design,making the system identification performance optimal.(2)Inspired by the biological immune characteristic in nature,an anti parasitic immune algorithm is embedded into parasitism stage,commensalism stage,parasitism and commensalism stage of symbiotic organism search to explore and enhance the influence of algorithm performance.An anti parasitic immune enhanced symbiotic organism search algorithm is proposed and applied to some engineering optimization problems.The experimental results show that the immune parasite enhanced symbiotic organism search algorithm has good performance.(3)The search mechanism of symbiotic organism search algorithm is mapped to binary discrete space by an adaptive transformation function,and a binary symbiotic organism search algorithm is proposed to solve the problem of feature selection and classification.The experimental results show that the binary symbiotic organism search algorithm can search for the minimum number of feature combinations.And the classification model constructed by the searched features can ensure high classification accuracy.