Global Optimization Problem Of The Continuous Function Based On Estimation Of Distribution Algorithms

Evolutionary computation concepts derived from Darwinian theory of evolution, specifically referring to optimization theory, methods and techniques according to the Darwinian theory of evolution invention. At present, in the field of evolutionary computation, the effective algorithm of a new solution of engineering problems, estimation of distribution algorithms. Estimation of distribution algorithm concept was formally proposed in1996,followed later in the new millennium has been the rapid development of very effective for solving engineering problems, it has become the cutting edge of evolutionary computation research.With belonging to the evolutionary computation algorithm, however, compared to the traditional genetic evolutionary algorithm, estimation of distribution algorithm uses a new mode of evolution in the estimation of distribution algorithms, it is replaced by the dominant groups in the selection, no longer using cross over and mutation traditional computing operations, but a combination of a scientific theory of mathematical statistics, statistical learning the means of repeated evolutionary computation to obtain the optimal solution.Estimation of distribution algorithms creates a probability model from a macro point of view, using the probability model to describe the distribution of the solution space, by evolutionary computation to obtain the advantage of individuals. At present, the discrete distribution estimation algorithms are already quite mature, but continuous estimation of distribution algorithm progresses slowly. This article uses the idea of uniform distribution to narrow the sampling field for continuous optimization problems, designing a new estimation of distribution algorithm. Experimental data show that this kind of estimation of distribution algorithms is valid.