Memetic Algorithm And Its Application

All along, human beings have been inspired by nature. Through the observation and thinking of natural phenomena found many laws of nature, people use the natural law with their own thinking get a lot of inspiration to solve all kinds of problems. In 1975, American professor John Holland from the biological survival of the fittest natural law presents a randomized algorithm, the algorithm is later known as genetic algorithm. In 1989, Pablo Mocsato proposed memetic algorithm firstly. It is a new intelligent algorithm combined with genetic algorithm and local search strategy. Since then, the memetic algorithm has attracted more and more researchers’ attention. The algorithm also has been greatly developed for the continuous efforts of many researchers.Memetic algorithm is an effective evolutionary algorithm. The basic idea is to obtain the optimal solution by simulating the evolution of biological evolution. At first memetic algorithm is considered for the improvement of genetic algorithm, so it is also called "hybrid genetic algorithm". With the deepening of research, memetic algorithm has evolved into a global search strategy and local search strategy of optimization algorithm framework. The algorithm has high flexibility. By adopting different search strategies can constitute different memetic algorithm, and can be applied to the major field, compared with the traditional optimization algorithm, the optimization algorithm has strong searching ability, less number of iterations, fast operation, etc. This paper mainly studies the theory of memetic algorithm and its application in the university timetable problem and intelligent test paper generation problem.First of all, the paper introduces the occurrence and the development history of memetic algorithm, the algorithm of the basic principle and some professional terms in detail, and puts forward some improvement ways and methods of the algorithm. Secondly, the memetic algorithm apply to the problem of university exam schedule and the problem of intelligent test paper, at the same time, we have carried out simulation experiments on these two problems. The results show that, through the optimization of the global search strategy and the local search strategy, the operation efficiency of the algorithm is greatly improved, the number of iterations is reduced, and satisfactory results are achieved. Finally, the full text is summarized and prospects, and the future research direction has carried on the detailed planning of the algorithm.