Abbas Raftari

Decoupling of engine subsystems for modular control design

In this paper, the problem of optimal decoupling of engine idle speed and emission subsystems is studied for facilitation of modularization of control tasks, enhancement of robustness, reduction of control complexity and calibration time, and improvement of engine system performance. A new design method is introduced which employs feedforward compensation to achieve optimal decoupling of engine subsystems over large frequency ranges. The design method is applied to the decoupling of engine idle speed and emission subsystems which demonstrate significant reduction of their interconnections.

Optimal H-infinity decoupling of engine idle speed and emission subsystems

In this paper, the problem of optimal decoupling of engine idle speed and emission subsystems is studied for facilitation of modularization of control tasks, enhancement of robustness, reduction of control complexity and calibration time, and improvement of engine system performance. A new design method is introduced which employs feedforward compensation to achieve optimal decoupling of engine subsystems over large frequency ranges. The design method is applied to the decoupling of engine idle speed and emission subsystems which demonstrate significant reduction of their interconnections.

Robust control and coordination of engine systems

In this paper, a novel two-step design procedure is introduced for robust control and coordination of engine idle speed and emission systems. The procedure starts with the optimal broad-band decoupling of engine subsystems via feedforward compensation. The problem can be reduced to a standard H/sup /spl infin// optimization problem and solved by using Nevanlinna-Pick interpolation. A robust decentralized control in H/sup /spl infin// is then employed to achieve robust stabilization and disturbance attenuation of the compensated systems. A computer search algorithm for optimal decentralized PI controllers is developed. Benefits of the two-step design include modularization of control tasks, enhancement of robustness and performance, reduction of calibration time and complexity.