考虑影响公路边坡稳定性信息的不完整性、随机性和模糊性，以最大熵原理和工程模糊集理论为基础，并采用综合赋权法确定指标权重，提出了一种改进的公路边坡稳定性模糊评价方法。该模型充分利用影响因素的信息，以广义加权距离表示待评价样本与标准样本的差异，从优化条件中获得相对隶属度，并依据加权平均原则对边坡样本等级进行识别。该模型应用于常吉和吉怀高速公路边坡稳定性评价中，将评价结果与模糊综合模型和属性识别模型的结果以及边坡实际情况进行对比验证，体现了其较好的一致性，并且结果更合理。同时，该模型具有较小的信息熵值，表明其评价的不确定性小，可靠性较高，是一种边坡稳定性评价分析新方法，可在相关工程领域中应用推广。

In consideration of the imperfection, randomness and ambiguity of the information for stability analysis of highway slopes, an improved fuzzy evaluation method is put forward to assess the stability of highway slopes based on the maximum entropy principle and the engineering fuzzy set theory. In this method, the comprehensive empowering method is adopted to determine the weight of index. The procedure makes full use of the information of the factors influencing stability of highway slopes, and uses the generalized weighted distance to represent the difference between slope sample and standard sample. The relative membership degree is obtained by conditions’ optimization, the grades of the slope samples are identified according to the weighted average principle. The procedure is applied to Changde-Jishou and Jishou-Huaihua highway in Hunan province to analyze the stability of rock slopes. It is found that the assessment results agree well with the results obtained with the fuzzy comprehensive evaluation method，the attribute recognition method and the practical condition of slopes, showing that the new procedure is more reasonable. Furthermore, the Shannon entropy value of the model is smaller, which indicates less uncertainty and higher reliability of the evaluated results.