Benefits of model predictive control for gasoline airpath control

Abstract

One effective possibility to reduce pollutant emissions and fuel consumption of an internal combustion engine is the use of improved process control. It is made viable by the implementation of additional actuators and sensors which allow to operate the process more flexible. For full exploitation of the setup an appropriate control algorithm is necessary. Classical engine control structures rely on the use of many calibration parameters which result in high demands on the calibration time. Model predictive control (MPC) is an advanced control algorithm which is able to overcome this drawback. It allows to use a mathematical plant model for control synthesis which reduces calibration time and makes reusability possible. The present paper introduces the MPC algorithm and discusses the benefits of MPC for engine control. A special emphasis is put on the application for gasoline airpath control. For clarification, the benefits are demonstrated by numerical simulation studies. For the example of gasoline two stage turbocharging, the advantage concerning control performance are shown as well as the possibility to easily adapt to changed specifications for the closed-loop control dynamics.

Dual-mode predictive control of a rotor suspension system

“…Theorem 1. If Assumptions 1-3 hold, the asymptotical stability of the closed-loop rotor-AMB system composed of (6), (9), (16) and 17can be guaranteed.…”

“…Proof. We assume that optimization problem (16) and 17can be solved; then, from Assumption 2, ξ(k) always exists such that x(k+1) ∈ S * in (21) if x(k) ∈ S * . At time k+1, a feasible choice for ξ(k+1) is given by ξ(k+1|k) = L 2 ξ(k) according to (10)…”

“…(13); Step 3. Set the value interval of the system state, calculate the optimal values ξ * (k) based on (16) and (17), and iteratively establish a table Ξ of the optimal solutions for all the states;…”

“…To address this challenge, in this paper, a novel active controller based on a model predictive control (MPC) technique is developed to stabilize a constrained AMB system. The MPC and dual-mode predictive control (DMPC) techniques are common control strategies in manufacturing industries [16][17][18][19] because of their capability to overcome nonlinearities and system uncertainties and to deal with constraints [20][21][22]. Basically, in the framework of MPC, with the system dynamics identified by an internal model, an optimal control input trajectory can be obtained by solving the receding horizon optimization problem online, which minimizes a given performance index [23][24][25].…”

A robust integrated model predictive iterative learning control strategy for batch processes

“…Therefore, the tracking performance is dependent only on the learning rate or weighted parameters. In order to obtain better control performance, traditional ILC is combined with other control strategies like model predictive control (MPC), which has found widespread application in the field of control, such as shared control framework [3] and gasoline airpath control [4]. In [5], the combination of MPC strategy and ILC algorithm was studied for the constrained multivariable control of batch processes.…”