Research on Light Pollution Model Based on Clustering Algorithm

Abstract

Over the past decade, artificial light sources have brightened the night sky so much that the number of stars visible to the naked eye has plummeted, the US journal Science reports. In addition, light pollution can also destroy bio-diversity and even affect the ecological balance. Therefore, it is very important to establish an evaluation index of light pollution and put forward effective treatment strategies.In this paper, we set up a light pollution evaluation model (Model I) through the integration of a large number of data and principal component analysis, which is used to evaluate the degree of light pollution in a region, and it is applied to four types of regions, and the clustering algorithm is used for analysis.Through statistical analysis of a large number of samples, light pollution was first divided into four risk levels: no pollution, light pollution, moderate pollution and heavy pollution, and the four risk levels were assigned a score range. Then, we selected seven parameters affecting light pollution, and made statistical analysis of parameters in four regions to obtain eight groups of data. Through principal component analysis, we reduced their dimensions, and finally got the relationship between the grade fraction of light pollution and the three principal components. So we used the score as a measure of universality to determine the level of light pollution risk in a place.