Research on the Composition Analysis and Identification of Ancient Glass Products

Abstract

Chinese glass has a history of thousands of years, and ancient glass is valuable material evidence of early cultural exchanges and trade between China and the West. With the continuous development of economy and technology, Chinese scholars have made deeper research on glass composition, but at present, most of the works in China focus on the chemical point of view of ancient glass composition, detection technology and cultural exchanges, and less on the statistical law of chemical composition, identification and classification. In this paper, the relationship between weathering, ornamentation, colour and type of glass surface is analyzed by correspondence analysis. Then the Mann-Whitney test is used to test the content of each chemical substance of different types of glass before and after weathering, and the statistical law of the chemical composition content of the surface of cultural relics samples with or without weathering is expounded. Then, the classification criteria of high-potassium glass and lead-barium glass were obtained by using a decision tree and random forest model, and the sub-classification method of four types of glass was given by using K-means clustering algorithm based on standardized data, and each type of glass was divided into two categories. Finally, the neural network is used to analyze and identify the unknown glass. It provides a reference for the further study and identification of the chemical composition of ancient glass.