Gear engagement control for transmissions in hybrid electric vehicles

Doctoral thesis, 2024

Hybrid and electric vehicles have become very important for vehicle manufacturers due to stricter legislation, government incentives and customers’ growing awareness of the environmental issues. For conventional vehicles, Dual clutch transmissions (DCT) have better shift quality and efficiency than conventional automatic transmissions. The hybridization of a DCT leads to several challenges to the shift quality. The most important aspects of shift quality are smoothness, shift time as well as noise and wear during a shift.

This thesis presents control methods for improving shift quality in a hybridized DCT during torque interrupt shifts. During these gear shifts, the traction source is disconnected from the wheels, thus long shift times adversely affect drivability. Firstly, a method to minimize the shift time during a torque interrupt shift is presented. A simulation method is also presented which demonstrates the relationship between dog teeth contacts during a shift and noise and wear. To reduce noise and wear, model-based open-loop and model-based feedback control methods are designed that minimize the intensity of the dog teeth contact. A time-optimal control method, which minimizes both the shift time and the intensity of the dog teeth contact, is developed. Performance of the control methods is verified by simulations.

The second part of the thesis focus on a dedicated hybrid transmission (DHT). The DHT in this research is shifted from series mode to parallel mode, by the engagement of a dog clutch. To minimize noise during the mode switch, an LQI controller has been designed for the speed synchronization of the dog clutch. An algebraic method of determining the feedback gains of the LQI controller, based on the physical parameters and functional requirements of the system, has been developed, greatly reducing the need for manual tuning.

The control methods defined in this thesis focus on implementation in the existing control hardware and software, so the complex calculations are done offline and control algorithm that must be embedded in real-time systems has been kept simple. Using this approach, optimum performance from the mechanical synchronizer or dog clutch system can be achieved without extending vehicles’ existing control systems.