Observer-Based Control for a Torque Converter Lock-Up Clutch Utilizing a Steady-State Torque Map

Abstract

Torque converter (TC) slip control is one of the most important ways to improve vehicle fuel economy and ride comfort without changing hardware. Several model-based feed-forward controllers have been proposed to improve control performance. However, model-based controllers depend on a torque map (lookup table) measured in a steady state; thus, they have errors in transient situations. Therefore, this article proposes an observer-based feed-forward controller that utilizes a steady-state torque map. To overcome the limitation of the torque map and unknown road torque, a novel observer is proposed that makes it possible to use both the advantages of a torque map and a vehicle driveline model. The proposed observer can adjust dependency on the map in real time with the observer index. A fuzzy logic-based observer index can obtain near-optimal gain with minimal tuning and low computation. In addition, a real-time TC map-update algorithm using a recursive least square (RLS) method is proposed. As a result, the feed-forward controller is designed considering the error of torque maps. The effectiveness of the proposed observer and feed-forward controller was verified using a simulation based on experimental data from an actual vehicle.