| Estimation of distribution algorithm (EDAs) is a new field of a kind of stochastic optimization method of computational intelligence. No crossover and mutation operation in EDAs, but by sampling to generate new individuals of dominant groups to establish the probability model. This is a genetic algorithm based on the probability model. Distribution estimation algorithm based on statistical information to establish the probability model of the solution space, and then randomly sampled to generate new individuals, so repeatedly, to achieve the evolution of groups. The main contents of this paper are three parts.First, a hybrid particle swarm distribution algorithm (THP-EDA) based on tabu search is proposed, which includes the EDAs algorithm, particle swarm optimization (PSO) and tabu search algorithm (TSA). The heuristic algorithm PSO is used to maintain the diversity of the EDAs population; the tabu search updates the global optimal value at each generation.Second, a hybrid DE and EDA algorithm (DEEDA) is proposed to solve the optimization problem, and the evolutionary mechanism of EDA is incorporated into the DE algorithm. DEEDA uses the global information obtained by EDA and the differential information in DE to obtain the optimal solution, which is better than the DE recombination (crossover and mutation), which can explore the search space more effectively.Third, the application of the improved distribution estimation algorithms in the 7 test functions, and the application of power network planning problem. |
