基于深度学习的纸张老化速度预测方法

Deep Learning-Based Paper Aging Rate Prediction Method

为了解决传统纸张老化速度预测误差大的问题,提出基于深度学习的纸张老化速度预测方法。分析纸张研究样本的初始成分和结构组成情况,在此基础上分别从结构聚合度、耐破、撕裂、拉伸以及颜色等方面,测定纸张的初始性能。分析纸张的老化机理,确定纸张老化的影响因素。评估当前纸张的老化状态,并利用深度学习算法确定纸张老化规律。以纸张老化状态的评价结果为起始状态,在考虑环境影响因素的情况下,得出纸张老化速度的预测结果。通过对比实验得出结论:与传统纸张老化速度预测方法相比,设计预测方法得出的老化速度更加接近实际的老化速度,且将该方法应用到实际的纸张保存管理工作中,能有效的延长纸张的使用寿命,在应用性能方面具有明显优势。

In order to solve the problem of large error in traditional paper aging rate prediction, a paper aging rate prediction method based on deep learning was proposed. The initial composition and structural composition of the paper samples was analyzed, and on this basis, the initial properties of the paper from the aspects of structural polymerization, breaking resistance, tearing, stretching and color, etc. were measured. The aging mechanism of paper and determine the influencing factors of paper aging were analyzed. The aging status of paper was evaluated, and the deep learning algorithm were used to determine the aging law of paper. Taking the evaluation results of paper aging status as the starting state, and considering the environmental factors, the paper aging rate prediction results are obtained. The results showed that: Compared with the traditional paper aging speed prediction method, the design and prediction method of aging speed is more close to the actual aging, and this method was applied to the actual paper conservation management work, the service life of the paper is effectively extended, the application has obvious advantages in terms of performance.