基于IOWA算子的短时交通流预测方法研究

本文首次将诱导有序加权平均（IOWA）算子应用到短时交通流预测中，建立了以整体预测误差平方和最小为目标的组合预测模型。在分析短时交通流预测模型的基础上，本文选取了指数平滑法、季节自回归求和移动平均模型（SARIMA）、BP神经网络模型对短时交通流进行预测，再用IOWA算子将这三种模型进行组合预测。最后进行实例验证，通过MAE、MSE和MAPE三项指标比较分析四种模型的预测效果。结果证明，IOWA算子组合预测模型明显优于其他的预测模型，有效地提高了短时交通流的预测精度。

Research on short-term traffic flow forecasting Method Based on IOWA Operator

The induced ordered weighted averaging(IOWA) operator is applied to the short-term traffic flow forecasting at the first time, and a combination forecasting model is established as a goal of minimizing the whole error square. By the analysis of short-term traffic flow forecasting model, this paper selects the exponential smoothing method , seasonal autoregressive integrated moving average model (SARIMA) and BP neural network model to forecast the short-term traffic flow, then uses IOWA operator to combine the three models to forecast the short-term traffic flow. Finally, taking the practical traffic data as the example, this paper compares and analyses the effects of four kinds of prediction models by the three indices :MAE,MSE and MAPE. It proves that IOWA operator combination prediction model is superior to other models, effectively improving the prediction accuracy of short-term traffic flow.