Nonlinear adaptive control of a transcritical Organic Rankine Cycle

This paper presents an adaptive control design for a transcritical Organic Rankine Cycle (TORC) working in both supercritical and subcritical regions. Even though supercritical operation increases the overall efficiency, previous research has focused on developing controls for ORCs working in the subcritical region only. This work first develops a model for TORC suitable for control system design. A decentralized control strategy using three single-input, single-output control loops provides the basis for a stable control, justified using a relative gain array. In simulations, a nonlinear approximate adaptive control utilizing the Cerebellar Model Articulation Controller outperforms a Proportional-Integral control, both in tracking set points changes and in rejecting disturbances due to a fluctuating heat source temperature.