基于优化TSVR的混凝土重力坝变形预测模型

Deformation Prediction Model of Concrete Gravity Dam Based on Optimized TSVR

以某混凝土重力坝为研究对象,基于孪生支持向量机算法(TSVR)优化算法,建立大坝变形预测模型,分析不同因素对混凝土重力坝水平位移的影响,并将预测结果与实际监测结果进行对比,以验证该算法的准确性。结果表明,混凝土重力坝的上部位移情况较不稳定,随着外部条件及监测时间的变化,位移情况波动较大;混凝土重力坝的下部位移情况较为稳定,随着监测时间的增大,变形量较为稳定。PL5坝段受库水位升降的影响较大,PL4坝段受库水位升降的影响较小。对比库水位升降对混凝土重力坝的影响可知,温度变化对混凝土重力坝水平位移的影响较大,不同监测点的水平位移受温度变化的影响程度具有一定的差异性。采用TSVR算法得出的模型残差较小,随着监测时间的变化,模型残差变化趋势较为平稳,表明采用TSVR算法得出的混凝土重力坝水平位移的准确性较高。

In this study,a concrete gravity dam is taken as the research object.Based on the optimization algorithm of the Twin Support Vector Machine(TSVR)algorithm,a dam deformation prediction model is established to analyze the impact of different factors on the horizontal displacement of the concrete gravity dam.The prediction results are compared with the actual monitoring results to verify the accuracy of the algorithm.The results show that the upper displacement of the concrete gravity dam is unstable,with the change of external conditions and monitoring time,its displacement fluctuates greatly;the lower displacement of concrete gravity dam is relatively stable,with the increase of monitoring time,its deformation is relatively stable.PL5 dam section is greatly affected by the rise and fall of reservoir water level,while PL4 dam section is less affected by the rise and fall of reservoir water level.By comparing the influence of reservoir water level rise and fall on concrete gravity dam,it can be concluded that temperature change has a great influence on the horizontal displacement of concrete gravity dam,and the influence of temperature change on the horizontal displacement of different monitoring points is different.The residual error of the model obtained by TSVR algorithm is small.With the change of monitoring time,the variation trend of the residual error of the model is relatively stable,indicating that the accuracy of the horizontal displacement of concrete gravity dam obtained by TSVR algorithm is high.