Zhang Yanfen, Mo Haijun
School of Architecture and Engineering, Guangdong Polytechnic of Science and Technology, Zhuhai, Zip code:519000, China.
Sengkei Engineering Company Limited, Macau, China.
Heliyon. 2024 Sep 7;10(18):e37525. doi: 10.1016/j.heliyon.2024.e37525. eCollection 2024 Sep 30.
This study aims to address the challenges of capturing design changes, supply chain fluctuations, and labor cost variations to improve the accuracy and real-time nature of intelligent building construction cost predictions. It seeks to accurately forecast and optimize project costs. The study innovatively constructs an intelligent building construction cost prediction model based on Building Information Modeling (BIM) and Elman neural networks (ENNs), denoted as the BIM-ENN model. The BIM-ENN model first introduces BIM technology to digitize and visualize information related to building structures, electromechanical systems, and pipelines. The digitized data obtained through BIM technology is then used as input data for the ENN, which optimizes the neural network parameters to predict and optimize intelligent building construction costs. Finally, the BIM-ENN model is experimentally evaluated. The results demonstrate that the prediction value of the construction cost of the intelligent building by this model closely matches the original information price, with a prediction accuracy of 95.83 %. Compared with the single ENN, the root mean squared error of the BIM-ENN model is less than 75, and the determination coefficient is above 0.95. This indicates that this model can explain more than 95 % of the construction cost prediction results, making it a feasible solution for actual prediction problems and achieving satisfactory results. The intelligent building construction cost prediction model reported here exhibits high accuracy and reliability. It can successfully forecast construction costs, providing robust decision support for the digitalization and intelligent development of construction enterprises. The practical significance of this study lies in providing the construction industry with an accurate cost management tool that helps enterprises optimize budget control and resource allocation, enhancing risk assessment and management capabilities. Moreover, the potential impact of the BIM-ENN model is its ability to elevate prediction standards within the construction industry, promote technological integration and innovation, enhance enterprise competitiveness, and drive the industry's transition towards digitalization and intelligent sustainable development.
本研究旨在应对捕捉设计变更、供应链波动和劳动力成本变化等挑战，以提高智能建筑工程造价预测的准确性和实时性。它力求准确预测和优化项目成本。该研究创新性地构建了一种基于建筑信息模型（BIM）和埃尔曼神经网络（ENN）的智能建筑工程造价预测模型，即BIM-ENN模型。BIM-ENN模型首先引入BIM技术，将与建筑结构、机电系统和管道相关的信息进行数字化和可视化。然后，通过BIM技术获得的数字化数据被用作ENN的输入数据，ENN对神经网络参数进行优化，以预测和优化智能建筑工程造价。最后，对BIM-ENN模型进行了实验评估。结果表明，该模型对智能建筑工程造价的预测值与原始信息价格紧密匹配，预测准确率为95.83%。与单一的ENN相比，BIM-ENN模型的均方根误差小于75，决定系数高于0.95。这表明该模型可以解释超过95%的工程造价预测结果，使其成为实际预测问题的可行解决方案，并取得了令人满意的结果。本文报道的智能建筑工程造价预测模型具有很高的准确性和可靠性。它能够成功预测工程造价，为建筑企业的数字化和智能化发展提供有力的决策支持。本研究的实际意义在于为建筑业提供一种准确的成本管理工具，帮助企业优化预算控制和资源配置，增强风险评估和管理能力。此外，BIM-ENN模型的潜在影响在于其能够提升建筑业的预测标准，促进技术集成与创新，增强企业竞争力，并推动行业向数字化和智能可持续发展转型。
