Mechanical transmission is one of the earliest transmission modes in human society. With the continuous progress of science and technology, effective simulation and calculation research on mechanical transmission has gradually become an important link in the study of mechanical transmission. In the actual engineering practice, reliable and accurate data are difficult to obtain due to the complexity and low accuracy of the traditional mechanical transmission process. Machine learning (ML), a model trained by data, was used to analyze the response of the system through different parameters and drew scientific and reasonable conclusions. ML is more intuitive, easier to operate, and faster in calculation than the traditional methods. In many mechanical structures, due to the large number of processing parts and data, numerical simulation of this important equipment requires a considerable time to adjust and optimize accordingly.