基于改进型拉开档次法的滑坡治理方案优选

(西安公路研究院,陕西 西安 710065)

滑坡方案优选; 特征值法; 拉开档次法; 风险态度因子

Optimization of landslide control scheme based on improved scatter-degree approach
LEI Xiaofeng,YANG Shaowen,WANG Lei

(Xi'an Highway Institute, Xi'an 710065,China)

Landslide scheme optimization; eigenvalue method; scatter degree method; risk attitude factor

DOI: 10.15986/j.1006-7930.2020.02.009

备注

滑坡是工程建设中常见的地质灾害,安全又经济的治理方案可以降低滑坡发生的概率,减轻对工程建设的影响,如何从系统工程的角度对方案进行排序是近年来的研究热点.以延安某滑坡的治理工程为例,在初步设计阶段提出的四种滑坡治理方案的基础上,结合同类工程经验,构建了滑坡方案优选的层次模型; 并采用特征值法求出各评价指标的主观权重系数,对评价指标进行“权化处理”; 再采用拉开档次法对权化处理后的数据进行客观赋权,突出各评价对象间的整体差异; 最后采用加权集结模型对各滑坡治理方案进行综合评价,确定出最优方案.同时引入风险态度因子的概念,分析了不同风险态度因子对方案排序的影响,研究了滑坡治理方案排序结果对风险态度因子的敏感性.结果表明:当风险态度因子ε∈[-0.5,-0.2]时,决策者倾向于接受风险,方案四的综合属性效用值明显高于其他三个方案,方案四为优选方案; 当风险态度因子ε∈[-0.1, 0.5]时,决策者倾向于规避风险,方案一的综合属性效用值最大,方案一为最佳方案; 一般工程治理时决策者所持的风险态度均为规避风险,所以推荐方案1作为该滑坡治理的首选方案.该方法弥补了单独使用主观或者客观赋权法的不足,使评价结果更为准确,为该类方案优选设计提供了有效的思路及参考.

Landslide is a common geological hazard in engineering construction. Safe and economical treatment schemes can lower the occurrence probability of landslides and reduce the impact on engineering construction. How to rank the schemes from the perspective of system engineering is a research hotspot in recent years. Taking the landslide control project in Yan'an as an example, on the basis of four landslide control schemes proposed in the preliminary design stage and in combination with the experience of similar projects, an optimized hierarchical model for landslide schemes is constructed. The subjective weight coefficients of each evaluation index are obtained with eigenvalue method, and the evaluation indexes are “weighted”. The weighted data are empowered objectively with scatter-degree approach to highlight the overall differences among the evaluation objects. Finally, the weighted aggregate model is adopted to evaluate the landslide control schemes comprehensively and determine the optimal scheme. At the same time, the concept of risk attitude factor is introduced to analyze the impact of different risk attitude factors on the scheme sequencing and to research the sensitivity of the sequencing results to risk attitude factors in landslide control schemes. Results show that when the risk attitude factor is ε∈[-0.5,-0.2], decision makers tend to accept the risks, and the comprehensive attribute utility value of scheme 4 is significantly higher than that of the other three schemes, and scheme 4 is an optimal scheme. When the risk attitude factor is ε∈[-0.1, 0.5], decision makers tend to avoid risks. Scheme 1 with the biggest comprehensive attribute utility value should be the optimal scheme. Generally, in engineering management, decision makers are to avoid the risk, so scheme 1 is recommended as the preferred scheme for the landslide control. This method compensates for the shortcomings in independent use of subjective or objective weighting method, makes the evaluation results more accurate, and provides an effective thought and reference for optimal design of such schemes.