The multi-field coupling mechanism of a continuously variable transmission (CVT) system has a greater impact on its overall friction loss. The friction loss of steel rings represents the internal loss of the metal belt occurring between the steel ring group and friction plate, and it is the main reason for the loss of CVT energy. This study is based on the dissipative wear model of the CVT steel ring friction pair, and it innovatively establishes a steel ring wear depth model that relates the wear depth to the macro- and micro-factors. Here, a mixed lubrication model between the friction pairs of steel rings is constructed, and the change law of oil film thickness, pressure, and temperature between the contacting bodies is obtained. On the basis of the wear depth model of the steel ring, the results of the wear depth of the steel ring are obtained via simulation using torque, speed, fretting frequency, and other operating parameters. The surface wear of the steel ring at different fretting frequencies is observed via SEM, and the simulation and test results are compared. Subsequently, the obtained results of the wear depth and oil film thickness tests are used to determine the safety margin of the application efficiency of the steel ring friction pair in the CVT working range via the interpolation processing method. According to the simulation results, the overall performance area of the safety margin of the application efficiency of the steel ring friction pair is between −1 and 1. When the torque is greater than 130 N•m, the safety margin value decreases below 0, the safety failure probability of the steel ring increases significantly, and the safety margin decreases gradually. This research can provide new insights into solving the reliability and service life of CVTs.

This study proposes and experimentally validates an optimal integrated system to control the automotive continuously variable transmission (CVT) to achieve its expected transmission efficiency range. The control system framework consists of top and bottom layers. In the top layer, a driving intention recognition system is designed on the basis of fuzzy control strategy to determine the relationship between the driver intention and CVT target ratio at the corresponding time. In the bottom layer, a new slip state dynamic equation is obtained considering slip characteristics and its related constraints, and a clamping force bench is established. Innovatively, a joint controller based on model predictive control (MPC) is designed taking internal combustion engine torque and slip between the metal belt and pulley as optimization dual targets . A cycle is attained by solving the optimization target to achieve optimum engine torque and the input slip in real-time . Moreover, the new controller provides good robustness. Finally, performance is tested by actual CVT vehicles. Results show that compared with traditional control, the proposed control improves vehicle transmission efficiency by approximately 9.12%–9.35% with high accuracy.