Abstract: By analyses of the advantages and limitations of the fuzzy inference-forecast model and the neural network forecast model，the ANFIS forecast model Was applyed to combinging the two models，exerting their advantages and fetching up their limitations each other．And the conclusion is demonstrated that the ANFIS forecast model can predict the trafic flow more intuitionally and more precisely．

Key words: fuzzy inference-forecast model, neural network forecast model, ANFIS forecast model, MATLAB software

摘要： 分析了模糊推理预测模型和神经网络预测模型在交通流实时预测中的优缺点，采用ANFIS预测模型将二者结 合起来，发挥两个模型各自的优势，弥补彼此的不足，通过预测结果表明该模型能够更直观、更精确地对交通流进行 实时预测.

关键词: 模糊推理预测模型, 神经网络预测模型, ANFIS预测模型, MATLAB软件