地面沉降预测的灰色-马尔柯夫模型

天津滨海新区地处京津和环渤海两大城市带的交汇点,与日本、韩国隔海相望,是我国对外开放的重要通道。滨海新区广泛分布有软弱土层,属于欠固结地层,不仅会自然压密,而且在荷载的长期作用下容易产生次固结变形;长期以来,由于地表水资源短缺,大幅超采地下水的形势严峻。自然因素与人为活动的综合作用,使得滨海新区地面沉降现象普遍发生,形势不容忽视。应用一定的理论与方法,预测地面沉降可能的发展趋势,以寻求有效的防治措施,是天津滨海新区地面沉降研究的最紧迫任务之一。地面沉降是一种渐变性地质灾害,可以预测其发展趋势。文章结合灰色系统理论预测模型及马尔柯夫预测模型特点,提出了地面沉降预测的灰色-马尔柯夫模型,并应用实例演示灰色-马尔柯夫模型预测过程。为便于大量监测点的预测,采用C＋＋编程对天津滨海新区170个监测点进行预测,结果证明灰色-马尔柯夫模型对天津滨海新区地面沉降趋势预测具有实际意义,表明此模型对随机性强,波动性大的时间数列预测具有较好的精度。在实际应用中,灰色-马尔柯夫模型一般需要较多的原始数据,原始数据越多,预测精度越高。

Grey-Markov Model for prediction of land subsidence

Tianjin Binhai New Area is situated in the meeting place of Beijing and Tianjin, facing Japan and Korea across the straights, it is an important passage for China＇ s reform and opening to the outside world. Binhai New Area widely distributed soft soil is under-consolidated ground, not only will natural pressure consolidation, but also easily produce deformation under long-term load. For a long time, the situation of groundwater overexploitation is severe for the shortage of surface water. The land subsidence of Tianjin Binhai New Area is the result of natural factors and human activity comprehensively, the situation is indispensable. Applied certain theory and method to forecast the possible development trend of land subsidence and seek effective control measures is an urgent task of the land subsidence study of Tianjin Binhai New Area. Land subsidence is a kind of gradually changeable geological disaster, can be forecast its development trend. This paper has adopted grey system predictive model in combined with Markov model, provided grey-Markov forecasting model for prediction of land subsidence, and its calculation process is shown in an example. In order to easily calculate mass data, We wrote C ＋ ＋ program to forecast land subsidence of 170 monitoring points The results showed that grey- Markov model has practical significance for land subsidence prediction in Tianjin Binhai New Area, and also showed that it has higher precision of prediction in the forecasting of time series with strong randomicity and high fluctuation. In the application, grey-Markov model need more original data, the more data, the higher precision.