ESO-based Adaptive Event-Triggered Load Following Control Design for a Pressurized Water Reactor with Samarium–promethium Dynamics
Jiuwu Hui,Yi-Kuen Lee,Jingqi Yuan
DOI: https://doi.org/10.1016/j.energy.2023.127058
IF: 9
2023-01-01
Energy
Abstract:Conventional control systems used in nuclear reactors rely on real-time information transmission between the controllers and the actuators, which is unnecessary and imposes a heavy communication burden. To this end, this paper proposes an extended state observer-based adaptive event-triggered control (ESO-AETC) scheme for the load following problem of a pressurized water reactor (PWR) in the presence of uncertainties and limited communication resources. Different from most of the previous controllers presented in the literature, the samarium–promethium dynamics are considered in the constructed PWR model in this paper. Based on the PWR model with the load-dependent parameters and the samarium–promethium dynamics, an ESO, which can estimate lumped uncertainties and unmeasured system states including the relative density of the delayed neutron precursor, the average fuel temperature, the xenon concentration, the iodine concentration, the samarium concentration, the promethium concentration, and the total reactivity, is designed. With the help of the ESO and the event-triggered techniques, an AETC scheme is developed to avoid the real-time communication in the controller-to-actuator channel of the PWR system. It is rigorously proved that the control error can converge to any prescribed residual set in a finite time according to Lyapunov stability theory, excluding the Zeno behavior. Finally, simulation studies involving comparisons with existing works, estimation performance verification, and sensitivity analysis are conducted to assess the proposed ESO-AETC scheme. Simulation results reveal that (1) relative to the pre-existing model-free controller (MFC) and the disturbance observer-based sliding mode controller (DO-SMC), the proposed ESO-AETC scheme is found to provide better load following performance with at least 56.12% lower maximum absolute error (MAE), at least 15.05% lower integral absolute error (IAE), and 87.58% reduced communication resources in the controller-to-actuator channel, (2) the ESO possesses strong estimation capability for system states and lumped uncertainties, and (3) the proposed ESO-AETC scheme yields a low sensitivity with respect to the change of the controller parameters.