Advanced Air Path Control in Diesel Engines Accounting for Variable Operational Conditions

Abstract

In this paper, we develop an extended linear parameter-varying (LPV) model to design an LPV controller for air path system control in diesel engines. The objective is to use a widely used and accepted nonlinear diesel engine air path model, minimize simplifying assumptions on the model, and then design a model-based gain scheduling controller to work in a wide range of engine operating points. To that end, we transform the nonlinear model to linear parameter-varying form by defining state-dependent inputs and scheduling parameters. Some of the defined state-dependent scheduling parameters are synthetic in the sense that they are not obvious from the model and are created through algebraic operations. The control law we design is parameter-dependent and allows a large range of operating points to be considered. The robust performance of the controller (with respect to parametric uncertainties in the control design model) under variable operating points (depending on engine speed, fuel flow rate, and intake-exhaust manifold temperatures) is tested on simulations by tracking reference exhaust manifold pressure and compressor air mass flow signals. Finally, the performance of the designed extended LPV controller is compared to an H∞ controller and to an LPV controller from the existing literature to see its superior performance under variable operating points.