论文部分内容阅读
基于老采空区上方地表实际监测沉降数据,提出了利用混沌理论相空间重构及预测技术研究老采空区地表沉陷特征及时变规律的新思路.研究了沉陷系统的相空间重构、关联维数的确定、Lyapunov指数的计算和预测模型的构建及应用等问题.采用最大Lyapunov指数法及混沌最小二乘支持向量机法对老采空区地表沉陷量进行了建模和预测.结果表明:该沉陷系统是一个混沌系统,沉陷影响因素有3~6个,变形理论最大预测期数为18期.构建的混沌预测模型预测结果相对误差均控制在10%以内,预测精度良好,可为老采空区废弃土地再利用决策提供支持.
Based on the actual monitoring data of ground settlement above the old goaf, a new idea to study the characteristics of surface subsidence and the time-varying law of the old goaf by using the phase space reconstruction and prediction techniques of chaos theory is proposed.The phase space reconstruction and correlation of the subsidence system Dimension determination, calculation of Lyapunov exponent, construction and application of prediction model, etc. The maximal Lyapunov exponent method and chaos least square support vector machine (SVM) method were used to model and forecast the ground subsidence in the old goaf.The results show : The subsidence system is a chaotic system with 3 to 6 factors affecting the subsidence and the maximum number of forecasting periods for deformation theory is 18. The relative errors of the prediction results of the constructed chaotic prediction models are both within 10% and the prediction accuracy is good Old abandoned goaf abandoned land reuse decision-making support.