Solution of cannonball dispersion evenness based on estimation of distribution algorithm

Cannonball dispersion evenness is the main parameter for the design and optimization of future air window. The key of solution of cannonball dispersion evenness is the computations of the biggest value and smallest value of a 2D continuous function, which has many maximum values and minimum values. Moreover, as it need to compute the biggest value and smallest value for many times in the design and parameter computation of FAW, the solution algorithm is required global convergence and quick convergence speed. Many randomly generated individuals are introduced in each generation to avoid local convergence, and in order to speed up the convergence speed, probability distribution models are founded to guide the searching process. By computer simulation verifying, the algorithm has the advantages of quick convergence speed and high precision of solution in certain generations compared to conventional genetic algorithm. The algorithm is also suitable for other optimizing solutions of 2D continuous function.