基于优化的灰色GM模型的滑坡预测  

Forecast for landslide based on Optimum Grey model

作　　者：肖云[1] 李先福[1] 

机构地区：[1]武汉工程大学环境与城市建设学院,湖北武汉430074

出　　处：《武汉工程大学学报》2012年第1期31-35,共5页Journal of Wuhan Institute of Technology

摘　　要：滑坡发生的时间预报是目前滑坡预报的关键方向之一,而滑坡预测模型的建立是滑坡时间预测的核心.由于滑坡的模糊性和灰色不确定性特点,采用灰色预测模型适用有效.本文在灰色系统理论的基础上,研究了灰色GM(1,1)的建模和精度检验过程,根据残差对模型进行了高阶优化.同时结合工程实例,选择了有效合理的监测数据,进行了滑坡临滑预报模型的研究,并将传统GM模型与优化GM模型的精度进行了对比,结果显示优化的GM模型预测精度大大提高.在灰色优化的GM(1,1)模型研究基础上,对临滑时间进行了预报,结果显示预测时间较早,可以起到提前预报作用.根据预测模型分析,提出了一些有益结论,供今后滑坡预报工作的参考.The sliding time is the key to forecast landslide, and landslide prediction model is the core of landslide prediction. Grey prediction model was applicable and effective because of the ambiguity and grey uncertainty of landslide. Based on grey system theory, the modeling process, precision examination and higherorder optimization on residual error were studied. Accorcling to effective monitoring data selected, the sliding model was studied in light of an engineering example. By comparison with traditional GM model and optimized GM model, the latter had high analog precision. The sliding time was predicted, which showed that the time was so early to play a role in landslide warning. Through an analysis of the prediction model, some significant and helpful conclusions were obtained, which may be a reference to landslide prediction for future.

关 键 词：灰色预测 优化GM(1 1)模型 黄茨滑坡 

分 类 号：X43[环境科学与工程—灾害防治]