| Particle Swarm Optimization is a new type algorithm which based on the calculation of the evolution iterated. Since the nineties of the last century, because of the advantages such as less adjustment of the parameters and easier to realize etc, it gets the attention of scholars both at home and abroad. And it is quite similar to genetic algorithm, and both of them are intelligent methods. Genetic Algorithm has inherent advantages. There has been widespread concern and research from the seventies of the last century. The genetic algorithm is an intelligent optimization algorithm, its theories have been fairly complete, and it has been applied in many fields and achieved good results after years of research and application.In this paper, we discuss the basic principles, processes and the advantages and disadvantages of Particle Swarm Optimization and Genetic Algorithm, and propose a Hybrid Genetic Particle Swarm Algorithm.The hybrid algorithm takes particle swarm optimization as the main body to optimize the parameters of the function. At the same time, in order to avoid wasting of resources which comes from the randomness of the initial stages particles, and in order to speed up the convergence rate, we introduce an abandon factor. Particles are ordered according to their fitness value and the particles which have smaller fitness value are discarded by means of the dynamic abandon factor. And it uses genetic operators to improve the fitness of the particles and uses decimal encoding to reduce the time of running.Here, this paper gives a theoretical analysis about the feasibility of the hybrid algorithm, we use it six standard test functions to test it and compares the results with standard Genetic Algorithm and Particle Swarm Algorithm Optimization to show the effectiveness of the hybrid algorithm.Then combining with the problem of image enhancement, the algorithm is applied to the gray-scale image enhancement based on the Beta function, we can get the obvious effectiveness from the images and histogram. |
