Search Results

Since the vehicle AMT(Automatic Manual Transmission) has been developed to meet the requirements of high quality vehicles as well as hybrid vehicles, there are more and more tasks running on the ECU (Electronically Controlled Unit). Concerned with both function correctness and timing correctness, the scheduling of those tasks becomes important. A perfect scheduler, which distributes the proper task working during the proper time, can not only make the best use of the ECU, but also predict its capacity rationally. Compared with ordinary algorithm such as clock driven scheduler, static and dynamic priority-driven scheduler as well as round robin scheduler, this paper proposes an offline-designed Blended Timer-triggered Scheduler (BTTS) for AMT control. BTTS can coordinate periodical and event-triggered tasks in one envelope. It designates the available time slice for each task, which can make each critical task run at the proper time.

Real-time theory is introduced to analyze reliability and robustness for ECU (Electronically Controlled Unit) of vehicle AMT (Automatic Mechanical Transmission). Real time analysis architecture is proposed. Analysis and simulation for the four working stages of AMT are illustrated. The results show, the real time system model with 9 tasks working on 4 processors is reasonable, and the designed Blended Timer-Triggered Scheduler is feasible as well as robust, which provide the system more extendibility and reliability.