基于混合智能优化算法的建筑物沉降预测应用研究  

Application Research of Building Settlement Prediction Based on Hybrid Intelligent Optimization Algorithm

作　　者：马子俊 姚彦夫 MA Zijun;YAO Yanfu(Hainan Nonferrous Engineering Survey and Design Institute,Haikou Hainan 570206,China)

机构地区：[1]海南有色工程勘察设计院,海南海口570206

出　　处：《北京测绘》2018年第9期1102-1107,共6页Beijing Surveying and Mapping

摘　　要：为解决现代工程施工中对沉降预测精度提出的更高要求,本研究提出了一种基于粒子群算法(PSO)和支持回归机(SVR)的混合预测模型。依据山西某工程建筑施工沉降监测数据,将其分为建模和测试两部分,分别建立了SVR、PSO-SVR、BP神经网络和多元回归预测模型。最后对模型做试算分析,结果表明:PSO-SVR模型预测精度绝对占优,对实际值具有更强的逼近能力,多元回归模型预测精度相较于PSO-SVR模型略低,但明显优于其它两类智能模型。本文提出的混合模型对解决实际工程中遇到的沉降预测问题非常实用,值得进一步推广应用。A hybrid prediction model based on particle swarm optimization（PSO）and support regression machine（SVR）is proposed in this paper to meet the higher requirement of settlement prediction accuracy in modern engineering construction.The SVR,PSO-SVR,BP neural network and multiple regression prediction models are established respectively according to the monitoring data of the construction settlement of a project in Shanxi,which is divided into modeling phase and testing part.Finally,the model is tested and analyzed.The results show that the prediction accuracy of the PSO-SVR model is absolutely dominant and has stronger approximation ability to the actual value.The prediction accuracy of the multiple regression model is slightly lower than that of the PSO-SVR model,but it is obviously superior to the other two kinds of intelligent models.The hybrid model proposed in this paper is very practical to solve the settlement prediction problem in practical engineering,and is worthy of further popularization and application.

关 键 词：沉降预测 粒子群算法 支持回归机 回归分析 

分 类 号：P258[天文地球—测绘科学与技术]