基于相空间重构理论和近邻等距及其加权预测模式,提出了多相型综合预测方法,并应用于高边坡变形监测数据的处理。该方法应用于小湾水电站2^#山梁高边坡4个GPS监测点连续430期变形时间序列的数据分析并进行预测。结果表明,该方法有效地削弱了重构参数选取算法和数据噪声等主客观因素对变形预测值的不确定影响,具有更高的预测精度和可靠性,为高边坡变形的非线性混沌预测提供了一种方法。
Based on the phase-space reconstruction theory and equal distance model combining with its weighted model of near neighborhood using the single phase-space, an integrated method is established for data processing of slope deformation monitoring. The proposed method is applied to analyze the GPS monitoring data obtained from 4 positions of the high-steep slope in Xiaowan Hydro project. The application result shows that this method can effectively weaken the effect of adopted algorithm on reconstructing parameters and the disturbance of datum noise, so that higher forecasting precision and better reliability can be achieved.