基于改进T-S模糊神经网络的网络流量预测

摘要

为提高网络流量的预测精度,在人工蜂群算法和T-S模糊神经网络的基础上,采用一种具有差分进化搜索的蜂群算法训练T-S模糊神经网络,对网络流量进行建模预测.该算法首先利用差分进化算法的变异和交叉算子来替换人工蜂群算法中引领蜂的搜索策略,然后对人工蜂群算法中跟随蜂的搜索策略进行改进,使其在种群最优解附近产生候选食物源,该算法能较好地平衡局部搜索能力和全局搜索能力.将优化后的T-S模糊神经网络用于网络流量预测,并与T-S模糊神经网络、蜂群算法优化T-S进行比较,仿真结果表明该算法具有更高的预测准确性,从而证明该算法在预测领域的可行性和有效性.%In order to increase the precision of network traffic prediction,an improved bee colony algorithm with differential evolution is used to train the T-S fuzzy neural network.The algorithm firstly replaces the search strategy of guiding bees in artificial bee colony algorithm by variant and crossed operators with differential evolution method.Then,the search strategy of following bees in artificial bee colony algorithm is modified to generate the candidate food source around the optimum solution in population.The improved algorithm can balance the ability of local search and global search better.The improved algorithm is applied to predict the network traffic,and compared with the TSFNN and T-S fuzzy neural network optimized bee colony algorithm.Simulation results show that the proposed method has higher forecasting accuracy so that it is feasible and effective in the prediction of network traffic.