Data-Based Control Structure Selection for RCCI Engines with Electrically Assisted Turbocharger

Internal Combustion Engines (ICEs) are getting more complex in order to meet the limits set by tightening emission regulations. As a result, the control problem of ICEs is also getting more complex, due to an increased number of subsystems and actuators. In this paper, an input-output scaling and pairing procedure based on measured input-output data will be presented. The input-output scaling method is based on the Sinkhorn-Knopp algorithm, while the input-output pairing is determined using a Gramian-based interaction measure. The applicability of the presented procedure is demonstrated on a simulated RCCI engine with an electrically assisted turbo.