首页|期刊导航|徐州工程学院学报(自然科学版)|基于模糊神经网络的装配式建筑吊装工艺优化及风险分析

基于模糊神经网络的装配式建筑吊装工艺优化及风险分析OA

Optimization and Risk Analysis of the Prefabricated Building Hoisting Process Based on Fuzzy Neural Networks

中文摘要英文摘要

中国对建筑业高质量发展的要求,使得装配式建筑因其施工效率高、绿色环保、节约人力成本等优势,受到政府的高度重视,并被大力推广.但吊装施工是安全风险较高的环节,其风险分析面临巨大的挑战.针对上述问题,研究通过系统理论事故模型分析风险因素,并提出动态模糊神经网络结合改进遗传算法的装配式建筑吊装工艺优化方法.研究结果显示,研究方法仅需迭代 97 次就可达到稳定状态,并且对应预测误差也最低,为0.043.在实际应用中,研究确定的最佳方案总体上可降低12.5%的总成本,并将设备利用率提升 11%.上述结果说明研究方法能有效解决装配式建筑吊装工艺中的不确定性和模糊性问题,为装配式建筑吊装工艺的优化提供科学依据.

The pursuit of high-quality development in China's construction industry has prompted the government to vigorously promote prefabricated buildings,owing to their notable advantages in terms of construction efficiency,environmental sustainability,and labor cost reduction.However,hoisting operations are a high-risk phase of construction,making their risks,making risk analysis particularly challenging.To address this issue,this study employs the STA(System-Theoretic Accident)framework to systematically identify and analyze risk factors,and further proposes an optimization method for the prefabricated building hoisting process,which integrates a dynamic fuzzy neural network with an improved genetic algorithm.The experimental results show that the proposed method converges to a stable state within only 97 iterations,achieving a minimum prediction error of 0.043.In practical application,the optimal scheme identified by this method can reduce total costs by 12.5%and increase equipment utilization by 11%.These findings indicate that the proposed method can effectively accommodate the uncertainty and fuzziness inherent in prefabricated building hoisting operations,thereby offering a scientific basis for process optimization decision-making.

王其燊

福州理工学院 智能建造学院,福建 福州 350000

信息技术与安全科学

模糊神经网络装配式建筑吊装工艺优化风险分析改进GA算法

fuzzy neural networkprefabricated buildingshoisting process optimizationrisk analysisimproved genetic algorithm

《徐州工程学院学报(自然科学版)》 2026 (3)

44-49,6

福建省中青年教师教育科研项目(科技类)(JAT231169)

评论