Two-Stage Disturbance Rejection Control Strategy for Airport Refueling Systems Based on Predictive Control

Airport refueling systems, composed of parallel pumps and pipe networks, serve as critical infrastructure within the realm of civil aviation. Although the optimal scheduling of centrifugal pumps based on the steady state or quasi-steady state of pipe networks has been extensively investigated, there is a scarcity of studies addressing the dynamic control of pipe networks considering transient characteristics. Due to random disturbances caused by aircraft refueling demand, the adjustment of pump speed needs to be further improved for safety and economics. The hydraulic transient simulation of airport refueling systems was first carried out and the accuracy of the calculation was verified. A two-stage disturbance rejection control strategy based on predictive control was proposed to adjust pump speed reasonably. At the first stage, a generalized predictive control method combined with extended state observer (ESO-GPC) was adopted for disturbance compensation and control stabilization, which was validated to be effective both under slope and general disturbance. A simplified model was employed for online prediction of the inlet pressure fluctuations of pipe networks. At the second stage, pump speed was adjusted based on the deadband judgment, eliminating unnecessary flow fluctuations that may lead to the repetitive start and stop cycles of centrifugal pumps. The proposed two-stage control strategy was implemented in the simulation of the airport refueling system and the results demonstrate that this approach can effectively ensure the safe and stable operation of the system.