Control of Organic Rankine Cycle for waste heat recovery based on an optimized predictive model

The Organic Rankine Cycle (ORC) is a promising technique to recover low grade waste heat, and thus helps to improve the overall thermal efficiency of a process and reduce the environmental impact of large consumption of fossil fuels. A proper control strategy is a key for the safe and efficient operation of the ORC systems. In this paper, the key components in the ORC system are introduced first. Then a constrained generalized predictive control (CGPC) is designed to control the process. A controlled auto-regressive integrated moving average (CARIMA) model is used as the online self-tuning predictive model for the GPC controller, while the structure of the CARIMA model is optimized by the fast recursive algorithm (FRA). Simulation results confirm that the developed ORC control strategy is capable of achieving desirable set-point tracking performance, while also has a satisfactory disturbance rejection capability.