将都市圈客运量样本数据集分为训练集、测试集和检验集,采用最小最终误差预测准则确定预测值的损失函数参数与惩罚因子,选取ε-不敏感损失函数与高斯核函数减小预测复杂性,构建了基于支持向量机的都市圈客运量预测模型,并通过逐渐改变损失函数、惩罚因子与高斯核函数参数的取值,对京津冀都市圈客运量进行了预测。预测结果表明:客运量预测的平均相对误差为0.15%,预测值与实测数据拟合良好,整体变化趋势一致,反映了预测模型的可靠性。 The metropolitan area passenger volume sample data set is divided into training set, test set and test set, the minimum final error prediction criterion is used to determine the prediction function of the loss function parameters and penalty factors, ε-insensitive loss function and Gaussian kernel function reduction prediction Complexity and the forecasting model of metropolitan area passenger volume based on support vector machine is established. The passenger volume of Beijing-Tianjin-Hebei metropolitan area is predicted by gradually changing the values ​​of loss function, penalty factor and Gaussian kernel function parameters. The prediction results show that the average relative error of passenger traffic forecast is 0.15%, the predicted value is well fitted with the measured data, and the overall trend of change is consistent, which reflects the reliability of the forecasting model.