Model Predictive Control of Vapor Compression Cycle for Large Transient Heat Flux Cooling

Critical heat flux (CHF) avoidance is crucial to the safe operation of two-phase cooling systems, such as a vapor compression cycle (VCC). Due to the relatively slow dynamics, feedback control alone may be insufficient to reject large transient heat loads. When such heat loads are known beforehand, e.g., due to planned operations, model predictive control (MPC) is an effective tool to preemptively steer the VCC to a more advantageous operating point. In order to compensate for the valve nonlinearity and save computational cost, we adopt a two-loop control structure. The outer-loop MPC operates on a slower update frequency and optimizes the input trajectories for efficiency and disturbance rejection. The inner-loop feedback control tracks the refrigerant mass-flow-rate setpoint using electronic expansion valves (EEV). Experimental validations indicate successful CHF avoidance control for large transient heat-loads.