Prediction of environmental severity in Tibet based on the polycarbonate(PC) aging behaviors

Through the analysis of aging behaviors of PC in 10 environments test stations in Tibet for 1 year, the spatio and temporal distributions of severity in Tibet were evaluated and predicted based on the Chromatic aberration. The sensitive environmental factors that highly related to PC aging were screened according to the grey correlation analysis and regression analysis. The "environment-material" mapping model with excellent training precision and generalization ability was established by the Back Propagation Artificial Neural Network (BP-ANN). By inputting the environment data of 28 cities in Tibet to the well-built models, the severity were predicted and visualized to form spatial distribution maps through the Griddate interpolation method. Results showed the areas with low altitude in the east of Tibet presented low severity. By inputting the monthly mean values of the sensitive environments data, the severity maps of the whole 12 month were obtained respectively. Results showed that the severity in summer was much higher than that in winter, and the severity of the northwest area presented high value even in winter. The exact quantitative evaluations of the severity play significant role in the safety service for the equipment or facilities in Tibet.