基于LMD-Elman神经网络模型的建筑物沉降预测  

Prediction of building settlement based on LMD-Elman neural network model

作　　者：周鸿芸 ZHOU Hongyun(Guangzhou Huangpu District Real Estate Registration Center, Guangzhou Guangdong 510700, China)

机构地区：[1]广州市黄埔区不动产登记中心,广东广州510700

出　　处：《北京测绘》2022年第6期811-815,共5页Beijing Surveying and Mapping

摘　　要：为提高建筑物沉降变形预测精度,准确掌握建筑物变形趋势,发挥局部均值分解(LMD)算法与Elman神经网络模型在数据处理、数据预测中的优势,提出一种新的LMD-Elman神经网络模型。该组合预测模型有效实现建筑物沉降预测的流程为:①通过LMD方法将沉降序列分解为若干的不同尺度具有物理意义的乘积函数;②发挥Elman神经网络模型在数据预测中的优势,针对不同分量建立预测模型得到各分量预测值;③将各分量预测值重构得到最终预测结果。将组合预测模型应用于实测建筑物沉降数据预测中,结果表明,相较于GM(1.1)模型与单一的Elman神经网络模型,本文提出组合预测模型预测结果与实际监测值具有较高的一致性,预测精度更高。该组合预测模型能够充分发掘建筑物沉降数据本身所蕴含的物理机制与物理规律,提高了建筑物沉降变形的预测精度。In order to improve the prediction accuracy of building settlement deformation and accurately grasp the building deformation trend,this paper made used of both advantages of local mean decomposition(LMD)algorithm and Elman neural network model in data processing and data prediction,and puts forward a new LMD Elman neural network model.The process of building settlement prediction effectively realized by the combined prediction model was as follows:firstly,the settlement sequence was decomposed into several product functions(PF)with physical significance in different scales by LMD method;Secondly,taking advantage of Elman neural network model in data prediction,a prediction model was established for different components to obtain the predicted values of each component;Finally,the predicted values of each component were reconstructed to obtain the final prediction results.The combined prediction model was applied to the prediction of measured building settlement data.The results showed that compared with GM(1.1)model and single Elman neural network model,the prediction results of the combined prediction model proposed in this paper had higher consistency with the actual monitoring values and higher prediction accuracy.The combined prediction model could fully explore the physical mechanism and physical law contained in the building settlement data,and improve the prediction accuracy of building settlement deformation.

关 键 词：局部均值分解 Elman神经网络模型 组合模型 沉降预测 GM(1.1)模型 

分 类 号：P258[天文地球—测绘科学与技术]