文物裂隙趋势预测模型研究

露天不可移动文物常年遭受风化侵蚀和人为破坏,保存环境十分恶劣,掌握文物健康状况并进行趋势预测,是"预防性"文物保护工作的重要内容。文物健康监测数据具有不确定性和贫信息性,传统概率统计方法和模糊数学方法不能满足预测要求。针对文物裂隙数据的特点,以陕西唐顺陵天禄石刻为研究对象,提出了采用灰色系统理论来预测文物裂隙发展趋势,建立了裂隙GM((1,1)模型和Verhulst模型,分别确定了灰色模型的灰色作用量和发展系数,并且对唐顺陵石刻文物的裂隙进行了预测。实验用前12个月的裂隙月均值,对随后的7个月的裂隙进行了定量预测,发现天禄石刻的裂隙有缓慢增大的趋势。与同期真实裂隙监测值相比,GM((1,1)模型和Verhulst模型预测值的平均相对误差满足裂隙预测精度要求。研究结果为文物健康趋势研究提供了定量分析的理论依据。

Study on crack prediction models for cultural relics

Perennially suffering from weathering erosion and manmade damage, openair immovable cultural relics face severe problems.Present condition based trend prediction is part of the "preventive" protection concept. However, sometimes, because of uncertainty in the prediction method, little useful information is obtained.Traditional statistical probability methods and fuzzy comprehensive evaluations are not suitable for making prediction models. We used the Tang Dynasty Shunling Tianlu stone carvings in Shaanxi as examples for predicting fracture damage based on grey system theory. Two models, the GM(1,1) and the Verhulst models of settlement are presented. The grey coefficient and the development coefficient are calculated by ordinary least squares. Experimental results show that the average predicted relative errors for the two models are 6.23% and 4.40%, respectively,and meet the expectations of crack prediction accuracy. This research provides a quantitative basis for assessing the health of relics and for guiding future research.