基于检测器数据的快速路路阻模型标定

针对交通规划中的路阻模型变量的要求，探讨了检测器数据的处理方法以及模型的标定方法。同时结合交通规划软件中路阻函数模型的要求和形式，参数标定采用了适用于非线性、非二次型逆的修正BFGS算法的非线性回归方法。其中，步长一维搜索采用Wolfe准则。由于非线性问题对样本数据质量非常敏感，同时，提出了对调查（检测器）数据质量检查的方法。利用上海快速路检测数据进行回归分析，从结果看，该模型对快速路具有很好的适应性，并具有较高的精度.

Calibration of Volume-delay Model of Expressway Using Detector Data

Considering variables of volume-delay model in traffic planning, methods of data analysis and the model calibration for the model are discussed, According to the type and the requirements of volume-delay model, the non-linear regression method, which is suitable for the non-linear, non-quadric revertible and modified BFGS algorithm is presented. Considering the model sensitivity to the quality of data in non-linear regression, the check method of samples are put forth.The results show that the model is fairly adaptable and accurate for expressway.