Model Predictive Control for an Organic Rankine Cycle System applied to a Heavy-Duty Diesel Engine

Abstract

Innovative internal combustion engine (ICE) concepts are in the focus of current research to further increase the engine's efficiency and decrease the greenhouse gas emissions. Only one third of the fuel's energy can be converted to mechanical power. The remaining two thirds leave the engine via exhaust gases and the coolant system as losses. Due to the high exergy level of the exhaust gas, a recovery of its energy with the help of a waste heat recovery system is possible. One promising technology for the use in commercial on-road vehicles is the organic Rankine cycle (ORC). The working principle is as follows: A working fluid is fed by a pump to a heat exchanger in which the fluid is vaporized. The vapor is led through an expansion machine converting the fluid's energy into mechanical energy. This paper presents a model predictive control (MPC) concept for a waste heat recovery system based on an ORC system applied to a heavy-duty diesel engine. The reduced-order modeling approach described in this study is based on physical equations. The resulting model is real-time capable and suitable for the use within the MPC scheme. For validation, the control algorithm is implemented on a rapid control prototyping hardware and tested on a heavy-duty diesel engine test bench equipped with the ORC system.