Mechanical Transmission Model and Numerical Simulation Based on Machine Learning

Introduction

With the continuous improvement of technology and industrial production level, mechanical transmission technology has been widely used. Specifically, it is used to connect motion transmission belts of various large mechanical equipment and engineering machinery in industrial production, as well as the transmission between various industries and machinery. It is among the indispensable fields with great development potential in current industrial applications given its advantages, such as high transmission accuracy, good reliability, low noise, and few moving parts. The method based on machine learning (ML) provides a theoretical basis and technical means for optimizing transmission systems, exhibiting great importance for various complex transmission systems.

Mechanical transmission is a widespread concern for the industrial manufacturing community, and many scholars have carried out research on this subject. Zhang et al. (2019) proposed a power return hydromechanical transmission system and modeled the speed ratio, torque ratio, efficiency, capacity, and other characteristics of the return hydromechanical transmission system to ensure the dynamic torque ratio and improve the efficiency of the automatic transmission system. Compared with the hydraulic mechanical split transmission, the return flow hydraulic mechanical transmission system can achieve higher efficiency while ensuring the starting torque ratio. Blagonravov et al. (2017) considered the stepless mechanical transmission with the vibration motion of internal components and outlined the basic principle of the automatic control of transmission with internal force function according to the vibration amplitude of the internal components. Asgari and Yazdizadeh (2018) proposed a robust fault diagnosis scheme for wind turbine generator sets and established a comprehensive mathematical model of the mechanical transmission system and gearbox dynamics of wind turbine generator sets operating in a wind farm. Farrage and Uchiyama (2018) believed that the friction occurring in mechanical systems is an important issue to achieve high-precision performance. They also believed that friction would not only adversely affect the motion accuracy of the drive shaft but also consume excessive energy and that the sliding mode control verified the effectiveness of the proposed friction model in the dual axis feed drive system. Kim et al. (2019) analyzed the transmission strength of a tractor’s transmission gear by using the equivalent torque during plowing. The load measurement system consists of engine speed sensor, torque gauge, four-shaft speed sensor, and a pressure sensor of two hydraulic pumps. The analysis method using equivalent torque showed lower stress and higher safety factor than that using maximum torque. Therefore, the equivalent torque method would support a more reliable product development when designing tractors with an actual working torque. Yang et al. (2019) mentioned that the hydraulic mechanical transmission (HMT) is suitable for high-power vehicles and established the speed model of HMT in power shift, considering HMT as the research object. Baek et al. (2020) proposed a quantitative method of reliability distribution to solve the reliability distribution problem of the mechanical transmission system. Their case analysis results proved that the method can provide certain reference for reliability distribution of the mechanical transmission system. The research theory of the mechanical transmission is relatively rich, but its application is still limited.