Research Of Quantum Clonal Algorithm Optimization And Its Application

The intelligent optimization algorithms，which are inspired by the evolutionmechanism of nature and biology, have been widely discussed and applied to manydifferent problems which the traditional optimization can not solve. Intelligentoptimization algorithms are random search algorithm based on biological intelligenceor physical phenomenon, the new algorithm does not require the continuity of theobjective function and constraints and convexity, even if there is no analyticexpression is not required, also has a strong ability to adapt to the calculation of datauncertainty. Quantum clonal algorithm is a new optimum method that combinesquantum computing with clonal selection algorithm. It has significant value inresearch and application.In the quantum clonal algorithm, by quantum rotation gates to realize quantummutation can effectively enhance the diversity of the population, but it is morecomplex, and will significantly increase the time complexity of the algorithm. Inaddition, quantum clone algorithm in the later stage of evolution, because the geneticstructure of the same or similar individual filled the entire population, it will easilylead to the algorithm into a local optimum. Therefore, cloning of quantum clonealgorithm selection operation will only high fitness value of individual retention, theselection operation should be the purpose of living individuals of higher retaineddirectly to the next generation. In view of the above problems, The main contents ofthe dissertation are as follows:First of all, analysis of the domestic and foreign quantum clone algorithmdevelopment and the research situation at home and abroad, and detailed analysis ofthe quantum clonal algorithm design ideas and basic concepts, discussed the basicframework of quantum clone algorithm and the operators of the algorithm of common.And then, in view of the existing quantum mutation quantum clone algorithmshortcoming, proposed an improved quantum clone algorithm, this algorithm has beensuccessfully applied to function optimization problems. The simulation results showthat the Improved quantum clone algorithm designed in this paper has fasterconvergence speed and stronger global search ability, can effectively overcome thepremature convergence of genetic algorithm and avoid the phenomenon, and has goodscalability and stability, and can be widely applied to various combinatorialoptimization problems. Finally, in view of the existing clonal quantum clone algorithm selectionoperation shortcoming, proposed a new quantum clone algorithm, this algorithm hasbeen successfully applied to data mining problems in the experiment. The simulationresults show that the proposed algorithm are significantly better than the QCAalgorithm, Apriori algorithm and FP-tree algorithm in the aspact of the running time,the mining of association rules, the number of recall and accuracy.