Under the background of digitalization of big data and information, the national “One Belt and One Road” initiative has produced a new huge effect on the ancient Silk Road between Xinjiang and Eurasia, how to revitalize the traditional music culture of ethnic minorities and how to make the traditional music culture resources of ethnic minorities be effectively protected and developed in the process of protection and inheritance of intangible cultural heritage in the world? Based on the analysis of ethnic elements in Xinjiang music works based on the PPM model under the perspective of “One Belt, One Road”, this paper first introduces the concept of PPM, the Bayesian structure algorithm, constructs the PPM model, and analyzes the data using by the structural equation modeling method, and explores the evaluation of PPM model under different sample sizes and different indirect The performance of four methods, Bayesian structural equation modeling method, structural equation modeling method, classical regression method and Bayesian regression method, was evaluated for the analysis of ethnic elements in Xinjiang music works under different sample sizes and different indirect effect truth values. The results show that the standard errors under the four models become larger as the regression coefficient b between Y and X increases, but in the case of (a=0, b=0.59), the standard error of the structural equation modeling method for the analysis of ethnic elements in Xinjiang music works is 0.048, which is the smallest compared with the standard errors of other algorithms. The structural equation algorithm of the PPM model is very applicable when dealing with practical problems. It can not only be used in theoretical research but also has a great influence on realistic guidance, which has an important theoretical significance and practical value for the survival and development of the traditional music culture of ethnic minorities in China.