中文摘要：

城市物流配送交通路径规划旨在选取最佳的配送路径，从而降低配送时间和成本。确立道路权重函数模型，在城市物流配送路径选择方面具有重要作用。由于配送路段车流量的不确定性，使得配送路线特征不固定。传统的建模方法，在小样本情况下难以对车流量和道路权重做出准确预测，导致建模误差较大。针对上述问题，提出了改进的GM（1，1）模型，通过将GM（1，1）模型与支持向量机相结合，对车流量进行预测，建立反映交通流量和行驶时间的道路权重函数模型，将车流量的预测值代人道路权重函数模型中，从而确定道路权重。实验结果说明，改进算法能够取得较为准确且健壮的预测结果，对道路权重进行有效建模。

英文摘要：

The main purpose of path planning in city logistics distribution is to select the best distribution path and reduce distribution costs. It is of great significance to evaluate the road weight in logistics distribution path planning. Due to the uncertainty of the traffic flow, the features of the delivery routes are ambiguous. In the conventional methods, it is difficult to predict the traffic flow accurately on the condition of the small sample data. To solve this problem, an improved GM（ 1,1 ） model was proposed to predict the traffic flow, in which the support vector machine was adopted to refine the predicted value of the original GM （ 1,1 ）. Furthermore, the model which reflects the relationships between traffic flow and time was built, thus the road weight can be determined by the predicted traffic flow. Numerous experiments were conducted to demonstrate the effectiveness and robustness of the proposed method ,which can evaluate the road weight well.