Predictive Control Method Based on Linear Varying Parameter Model for Space Nuclear Reactors

Space nuclear reactors exhibit strong nonlinearity and significant time delay characteristics. To address the challenge of traditional PID control methods failing to meet the control performance requirements of space nuclear reactors amidst complex load demand fluctuations, a predictive control approach based on a linear varying parameter model for space reactors is proposed. To tackle the pronounced nonlinearity, a linear varying parameter model is employed as the predictive model to anticipate the operational characteristics of space reactors across the entire power spectrum. To mitigate the notable time delay characteristics, the cascade control theory is combined with the model prediction method to optimize actuator actions, thereby enhancing the response speed of the controlled variable. Predictive control systems for electrical power and coolant temperature are individually designed, and control performance simulations are conducted under typical operational scenarios. The results indicate that compared to the cascade PID control system, the proposed predictive control system reduces power adjustment time by 68. 7% and coolant temperature adjustment time by 81. 7%. The nuclear power overshoots of predictive control systems for electrical power and coolant temperature are reduced by 22. 3% and 13. 0%9 respectively? ensuring reactor safety. The control system demonstrates robust applicability and commendable control performance across various power levels. Additionally, a significant decrease in actuator actions is achieved? with a peak voltage drop of 84. 1% for the drive motor? thereby extending its lifespan and enhancing system efficiency. Therefore? the proposed predictive control method based on the cascade control theory and linear varying parameter model can be used to optimize the control system of space reactors. © 2024 Xi'an Jiaotong University. All rights reserved.