基于MSHHO—BP神经网络的装配式建筑成本预测模型研究  

RESEARCH ON COST PREDICTION OF PREFABRICATED BUILDINGS BASED ON THE MSHHO-BP NEURAL NETWORK

作　　者：班代锐 BAN Dai-rui(School of Management Science and Engineering,Anhui University of Technology,Ma′anshan 243032,China)

机构地区：[1]安徽工业大学管理科学与工程学院,安徽马鞍山243032

出　　处：《南阳理工学院学报》2025年第2期41-47,共7页Journal of Nanyang Institute of Technology

基　　金：国家自然科学基金青年项目(72204001)。

摘　　要：为了对装配式建筑项目成本进行准确预测,保障装配式建筑项目顺利实施,预防成本失控。针对当前BP神经网络在成本预测方面的相关缺陷,提出了一种基于多策略改进的哈里斯鹰算法(Multi-Strategy Harris Hawks Optimizer,MSHHO)优化BP神经网络预测模型。文中将影响装配式建筑成本的主要因素作为预测模型的输入层,单方造价成本为输出层,构建MSHHO—BP神经网络的装配式建筑成本预测模型。以采集的45组装配式建筑项目数据为实例,分别利用BP预测模型、PSO-BP预测模型和MSHHO-BP模型进行验证。结果表明,相较于BP模型和PSO-BP模型,经MSHHO算法优化后的BP神经网络具有更高的预测精度,为装配式建筑建造成本的准确估算提供了一种有效途径。To accurately predict the cost of assembly building projects,to ensure the smooth implementation of assembly building projects,and to prevent the cost from getting out of control.Aiming at the current BP neural network′s relevant defects in cost prediction,this paper proposes an optimized BP neural network prediction model based on the Multi-Strategy Harris Hawks Optimizer(MSHHO).In the paper,the main factors affecting the cost of assembly building are taken as the input layer of the prediction model,the cost of one-sided cost is the output layer,and the MSHHO-BP neural network prediction model of assembly building cost is constructed.Taking 45 sets of data collected from assembly building projects as examples,the BP prediction model,PSO-BP prediction model,and MSHHO-BP model were utilized for validation respectively.The results show that compared with the BP model and PSO-BP model,the BP neural network optimized by the MSHHO algorithm has better prediction accuracy,which provides an effective way to accurately estimate the construction cost of assembly buildings.

关 键 词：装配式建筑 MSHHO-BP神经网络 预测模型 

分 类 号：TU723.3[建筑科学—建筑技术科学]