基于改进T-S模糊神经网络的网络流量预测

发布时间：2018-01-11 22:10

本文关键词：基于改进T-S模糊神经网络的网络流量预测　出处：《济南大学学报(自然科学版)》2015年04期 　论文类型：期刊论文

【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。
[Abstract]:In order to improve the accuracy of network traffic prediction, based on artificial bee colony algorithm and T-S fuzzy neural network, a colony algorithm with differential evolution search is used to train T-S fuzzy neural network. The algorithm uses the mutation and crossover operator of differential evolution algorithm to replace the search strategy of artificial bee colony algorithm. Then the search strategy of following bee in artificial bee colony algorithm is improved to make it produce candidate food source near the optimal solution of population. This algorithm can balance the local search ability and the global search ability. The optimized T-S fuzzy neural network is used to predict the network traffic, and it is compared with the T-S fuzzy neural network. Compared with T-S optimization, the simulation results show that the proposed algorithm has higher prediction accuracy, which proves the feasibility and effectiveness of the algorithm in the field of prediction.
【作者单位】：兰州交通大学电子与信息工程学院;
【基金】：甘肃省自然科学基金(1310RJZA027) 甘肃省科技支撑项目(1204FKCA025) 甘肃省高校基本科研业务费基金(213056)
【分类号】：TP183;TP393.06
【正文快照】： 随着Internet和网络技术的飞速发展,网络规模日益庞大,网络拓扑结构越发复杂,网络性能和网络服务质量方面的问题日益突出。在网络资源有限的情况下,建立网络流量预测模型,实时对网络流量进行预测,将会极大地提高网络性能和网络服务的质量。网络流量具有非线性、长相关、时变性

【参考文献】