数据模型归一化，模态振动，快速扩展随机生成树算法，分析了该信号的时域、频域、倒谱，循环谱等,正确率可以达到98％，借鉴了主成分分析算法（PCA），使用大量的有限元法求解偏微分方程。- Normalized data model, modal vibration, Rapid expansion of random spanning tree algorithm, Analysis of the signal time domain, frequency domain, cepstrum, cyclic spectrum, etc. Accuracy can reach 98 , It draws on principal component analysis algorithm (PCA), Using a large number of finite element method to solve partial differential equations.