基于SSA-HKELM算法的装配式建筑项目风险评估方法研究  

Research on Risk Assessment Method of Prefabricated Building Project Based on SSA-HKELM Algorithm

作　　者：刘全滢 汪新宇 LIU Quanying;WANG Xinyu(School of Economics and Management,Shenyang University of Chemical Technology,Shenyang 110142,China)

机构地区：[1]沈阳化工大学经济与管理学院,辽宁沈阳110142

出　　处：《建筑经济》2024年第S01期361-365,共5页Construction Economy

摘　　要：本文研究装配式建筑项目的风险评估方法。首先,该文提出一套装配式建筑项目风险评估指标体系。其次,该文采用麻雀搜索算法(SSA)以最小化模型输出的均方差为适应度函数,对混合核极限学习机(HKELM)模型的超参数进行自主优化,构建一种SSA-HKELM混合模型,用于装配式建筑项目的风险评估。实验结果表明,该模型能够实现对装配式建筑项目风险的准确评估,普适性更强、评估过程简单、成本更低。This paper studies the risk assessment methods of prefabricated building projects.Firstly,this paper puts forward a set of risk assessment index system of prefabricated building projects.Secondly,this paper uses Sparrow Search algorithm(SSA)to minimize the mean variance of model output as fitness function,and independently optimizes the hyperparameters of hybrid core Extreme Learning Machine(HKELM)model,and constructs an SSA-HKELM hybrid model for risk assessment of prefabricated building projects.The experimental results show that this model can achieve accurate risk assessment of prefabricated construction projects,this method is more universal,simple assessment process and lower cost.

关 键 词：装配式建筑项目 风险评估 风险评价指标体系 麻雀算法(SSA) 混合核极限学习机(HKELM) 

分 类 号：TU714[建筑科学—建筑技术科学]