DE优化T-S模糊神经网络的交通流量预测  

Traffic flow prediction based on T-S fuzzy neural network optimized by DE

作　　者：侯越[1] 

机构地区：[1]兰州交通大学电子与信息工程学院,甘肃兰州730070

出　　处：《计算机工程与设计》2013年第9期3284-3287,共4页Computer Engineering and Design

基　　金：甘肃省自然科学基金项目(1112RJZA051)

摘　　要：提出了一种差分进化算法优化T-S模糊神经网络预测交通流量的算法方法。该算法利用差分进化来弥补T-S模糊神经网络连接权值和阈值选择上的随机性缺陷,从而能发挥T-S模糊神经网络泛化的映射能力,而且能使T-S模糊神经网络具有较快的收敛性以及较强的学习能力。将该算法应用到实测交通流进行算法的有效性验证,并与传统的T-S模糊神经网络进行比较,仿真结果表明,该算法具有更好的非线性拟合能力和更高的预测准确性,在交通流量预测领域具备可行性和有效性。A prediction algorithm for traffic flow of T-S fuzzy neural network optimized differential evolution （DETSFNN） is proposed. In the algorithm, DE is used to compensate therandom defects for the thresholds and weights of T-S fuzzy neural net- work, thus it can perform mapping ability of T-S fuzzy neural network for generalization and also can make T-S fuzzy neural net- work have faster convergence and greater learning ability, The efficiency of the proposed prediction method is tested by the simu- lation of real traffic flow. The simulation results show that the proposed method has better nonlinear fitting ability and higher forecasting accuracy compared with the traditional T-S fuzzy neural network, and prove it is feasible and effective in the practical prediction of traffic flow.

关 键 词：差分进化 T-S模型 模糊神经网络 交通流量 预测 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]