Organic Rankine Cycle (ORC) system stands out in low-grade waste heat recovery for its outstanding performance. Considering the fluctuating nature of the heat source, model predictive control (MPC) is used to control the ORC systems and has achieved good performance. However, with the increase of the number of variables and the expansion of the control/prediction horizon, MPC will take longer calculation time. This may bring about the incapacity to obtain optimization problem solutions in one sampling instant and finally, lead to the failure of the setpoint tracking. To address this problem, in this article, a dual-mode fast dynamic matrix control (FDMC) algorithm grounded on singular value decomposition (SVD) and dynamic matrix control (DMC) is proposed. The presented algorithm features in its calculation speed improvement on the premise of ensuring tracking performance. Two simulations on disturbance rejection and setpoint tracking are performed to verify the effectiveness and rapidity of the proposed algorithm in comparison with the traditional DMC approach.
Read full abstract