交通分布预测模糊重力模型

Fuzzy gravity model of traffic distribution forecast

为了解决交通分布预测中的模糊不确定性,以重力模型与模糊数学理论和方法为基础,基于三角模糊数建立了模糊重力模型.采用模糊线性最小二乘法标定了模型中参数.利用模糊幅度对观测值的比、观测值与估计值的隶属函数的不重合部分的面积与其面积之和的比这2种方法进行模型精度检验.研究了实现交通分布预测结果的发生、吸引平衡处理的模糊Frator增长率迭代平衡方法.通过实例详细说明了发生量、吸引量和交通阻抗均具有模糊不确定性时的模糊重力模型的应用过程,提出了根据模糊不确定性大小程度来利用模糊预测值的建议.模糊重力模型可以实现交通分布预测中的模糊不确定性.

In order to solve the fuzzy uncertainty of the traffic distribution, using the gravity model and fuzzy mathematics theory, a fuzzy gravity model is established based on the triangular fuzzy number. The parameters are calibrated by the fuzzy linear least-squares regression method. Two methods are used to test the model accuracy through the ratio of fuzzy magnitude to observation value and the ratio of the un-overlapped area of the observations and estimated subjection function to the sum area of the two. Research on fuzzy Frator iterative approach is made to balance the production-attraction results of traffic distribution forecast. Finally, an example is shown to validate the model practicability when production-attraction and traffic impedance values are uncertain. Discussion about how to use the fuzzy forecast results according to the fuzzy uncertainty degree is also presented. The fuzzy gravity model can realize fuzzy uncertainty of the traffic distribution prediction.