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OACHSSCD

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.