Numerical and experimental study of feedforward and feedback control for microchannel cooling system

A lumped model for a cooling system incorporated with a microchannel heat exchanger is presented, based on which control strategies with feedforward and feedback controllers are developed. By comparing different control strategies numerically and experimentally, the robustness and accuracy of the proposed physical model are examined, a reliable and efficient control method is designed for achieving high operation efficiency and quick response. Validation of the lumped model shows that the maximum model prediction inaccuracies for wall temperature and pressure drop are within 1%. For the feedforward control system, the response is relatively fast compared with the feedback control system, which has a long settling time due to the heating process of cartridge heaters and heat transfer of the microchannel block. The system with feedforward and feedback control has a fast response and desired steady-state value. For a system undergoing large heat load variations, a scheduled setpoint control method could ensure that the pump work at its middle range of frequency. The numerical and experimental comparisons for different control methods show that the inaccuracies are within 2% for steady states. The application of the feedforward and feedback control strategy could address the deviation of system output from setpoints caused by model inaccuracies, disturbance variations, or perturbations in the operating conditions.