摘要
高边坡系统的演化过程表现出复杂的非线性动力学特性,由于边坡经常受到外界因素的扰动,而使整个监测资料时间序列具有以突变点为分界的跳跃性,因此,有效辨识测值突变位置,是提高高边坡位移监控模型拟合和预测精度的关键问题。组合应用相空间重构、云模型和最大Lyapunov指数等数值分析手段,研究了高边坡位移突变辨识的实现方法,探讨了考虑动力学结构突变影响的位移预测模型构建原理与算法。由于该模型依据最近一次位移突变后的监测资料,着重考虑突变后相对稳定的高边坡动力系统特性,因而可以有效提高监控模型的拟合和预测精度。以某水电工程为例,论证了利用该方法进行高边坡监测的有效性与准确性。
The evolution of a high slope system has complex nonlinear dynamic characteristics and is often disturbed by external conditions,as a result,the time sequences of displacement monitoring of high slope have the jumping property that is demarcated by mutation points.Therefore,effectively identifying the mutation location of measured values is the key to improving the fitting and prediction precision of the high slope displacement monitoring model.We study the realization method for mutation identification of high slope displacement values through comprehensively applying the phase space reconstruction,the cloud model,the largest lyapunov exponent and other numerical analysis methods.Furthermore,we discuss the establishing principle and algorithm of the displacement prediction model with consideration of dynamic structure mutation effects.The proposed model depends on up-to-date monitoring data after the last displacement mutation,and considers the relatively stable dynamic system characteristics of high slope formed after mutation,so the fitting and prediction precision of the high slope displacement monitoring model can be effectively improved.The effectiveness and precision of the presented model for high slope monitoring are verified by an engineering case of a hydropower project.
出处
《人民长江》
北大核心
2012年第24期86-90,共5页
Yangtze River
基金
水利部公益性行业专项项目"农村小河流综合治理关键技术研究与示范"(201201016)
江西省科技支撑项目(2010BSA16800)
