Two Improved Cultural Algorithms And Their Applications

The optimization problem and its solution have become an indispensable subject for scientific research and production in various fields.Intelligent optimization algorithm,which has less dependence on information and conditions of specific problems,has replaced traditional optimization methods as the mainstream tool for solving optimization problems.Since the second half of last century,various kinds of intelligent optimization algorithms have been proposed one after another.Most of their design inspiration comes from biological evolution,physical phenomena and animal or plant behavior,such as genetic algorithm,simulated annealing algorithm,chaotic optimization algorithm and bird swarm algorithm.Most intelligent algorithms use population as the optimizing search space.Among them,cultural algorithm is one of the few optimization algorithms with double evolution space.This algorithm simulates the accelerating role of culture in human evolution,and gives a highly inclusive parallel evolution framework,which can be combined with different algorithms and derive different cultural improvement algorithms.In this paper,two improved cultural algorithms are proposed to solve the optimization problems with discrete and continuous variables respectively.The evolutionary mechanism of individuals,which belongs to the population space of culture algorithm,is a hybrid strategy that composed of genetic algorithm and simulated annealing algorithm.This paper verify the effectiveness and superiority of the culture hybrid strategy by solving four traveling salesman problems.Compared the mixed method with particle swarm optimization,ant colony algorithm and genetic hybrid algorithm,the optimization results of the culture hybrid algorithm can be reduced by 0.6% to 13.01%,and the optimization process curve proved that the mixed strategy of cultural optimization has stronger global search ability and faster convergence speed.Cultural bird swarm algorithm.Put the algorithm that simulate the behavior of bird swarm into the cultural algorithm framework.The belief space of cultural algorithm converts the partial individuals' foraging skill into the survival experience of the whole swarm.Combining the actual data of river quality models with the optimal results of the cultural birdswarm algorithm,the minimum error of two sets data shows the feasibility of the cultural bird swarm algorithm for solving continuous optimization problems.The performance of the algorithm is validated by perturbing the population size,the maximum permission iterative number,the acceptance ratio of the better individuals and the range of each variable.All the optimization results show that the cultural bird swarm algorithm can rapidly converge to the global optima.Generally,cultural bird swarm optimization is a new improved algorithm with better performance,which promote the advantages and improve the disadvantages of the bird swarm algorithm,such as high convergence accuracy,strong search ability,poor stability,premature convergence and easy to fall into local optimum.