文物保护与考古科学2026，Vol.38Issue(2)：27-38,12.DOI:10.16334/j.cnki.cn31-1652/k.20250803711
基于多因素耦合的馆藏铁质文物风险因子分析及腐蚀预测模型构建
Risk factor analysis and corrosion prediction modeling for museum iron cultural relics based on multi-factor coupling
摘要
Abstract
The corrosion and deterioration of iron cultural relics in museum collections,under the coupled effects of multiple factors,pose a severe challenge for cultural heritage conservation.This study systematically quantified the synergistic corrosion effects of environmental factors,including temperature,humidity,SO2,NO2,O3,HCOOH,and Cl-on iron cultural relics through simulated exposure experiments.A hybrid prediction model was developed based on a Random Forest(RF)-Long Short-Term Memory(LSTM)network.This study introduced a"Corrosion Index(IP)"to correlate environmental stress with the condition of cultural relics.By combining feature engineering of a temperature-humidity interaction term(T_RH)and a pollutant synergy term(SO2_RH),the approach enabled accurate time-segmented prediction of the corrosion rate(60-day prediction R2=0.940).Feature importance analysis identified Cl-as the core risk factor(with a long-term contribution of 0.32),while the temperature-humidity synergy(T_RH)consistently dominated the corrosion process(contribution of 0.28±0.03).Based on corrosion kinetic thresholds,a four-tier risk classification model was established,and a graphical user interface(GUI)supporting real-time assessment was developed,providing technical support for the preventive conservation of museum iron art culture relics.关键词
铁质文物/多因素耦合/风险分析/风险因子Key words
Iron cultural relic/Multi-factor coupling/Risk analysis/Risk factor分类
社会科学引用本文复制引用
李丹丹,吴来明,黄河,蔡兰坤,周浩,徐方圆..基于多因素耦合的馆藏铁质文物风险因子分析及腐蚀预测模型构建[J].文物保护与考古科学,2026,38(2):27-38,12.基金项目
中国博士后科学基金第76批面上(2024M761909)、上海市"超级博士后"(2024214)资助 （2024M761909）
