Abstract: In view of the fact that different variable weight functions have different application scopes, when the decision-making evaluation object contains multiple sets of state values with large difference in configuration imbalance degree, the traditional variable weight function has low adaptability. A hybrid state variable weight function based on configuration imbalance degree and state value was established.In practical application, due to the division of excitation and punishment balance interval, the state value dispersion cannot accurately reflect the configuration imbalance, therefore, a measurement scheme of configuration imbalance degree was proposed based on the distance between the state value and the boundary of excitation and punishment balance interval. The TOPSIS method was combined with the improved variable weight function to form a variable weight evaluation model, which was applied to the risk assessment of anchorage slope instability based on multi-stage monitoring data.The research results show that the proposed variable weight evaluation model is suitable for the risk assessment of anchorage slope instability. Meanwhile, its variable weight adjustment degree and the configuration imbalance degree change direction can be completely consistent,which is more sensitive to the advantages and disadvantages.With higher applicability and better effect, the proposed model provides a reference method for decision evaluation with multiple sets of state values.

Key words: geotechnical engineering;TOPISIS, state variable weight function, anchorage slope;monitoring data, risk assessment

摘要： 针对不同的变权函数有不同的适用范围，当决策评估对象中包含组态失衡度相差较大的多组状态值时，传统的变权函数适应性较低的问题，构造了一种基于组态失衡度和状态值的混合型状态变权函数。在实际应用中，因激惩平衡区间划分，使得状态值离散度无法准确体现组态失衡度，为此提出了一种基于状态值到激惩平衡区间边界的距离的组态失衡度度量方案，将TOPISIS法与改进的变权函数相结合构成变权评估模型，并应用于基于多期监测数据的锚固边坡失稳风险评估中。研究结果表明：所提出的变权评估模型适用于锚固边坡失稳风险评估，同时其变权调节度与组态失衡度大小变化方向能保持完全一致，对优势和劣势状态敏感度更高，适用性更高、效果更好，为包含多组状态值的决策评估提供一种方法参考。

关键词: 岩土工程, TOPISIS, 状态变权函数, 锚固边坡, 监测数据, 风险评估