MPC-based Adaptive Control for Primary Frequency Regulation of Nuclear Power Units with Safety Constraints under Rolling Optimization

Nuclear power units participating in primary frequency regulation face multiple safety constraints including reactor power variation rate, turbine thermal stress, and control rod actuation frequency. Existing methods mostly adopt static limits and struggle to adapt to dynamic operating conditions. This paper proposes an adaptive safety-constrained control method for primary frequency regulation of nuclear power units under MPC rolling optimization, which incorporates three types of heterogeneous safety constraints into the rolling optimization framework, designs a constraint adaptive adjustment mechanism to dynamically modify constraint boundaries, and introduces feedback correction to form closed-loop control. Simulation results show that under severe grid disturbances, the frequency deviation is kept within ±0.1 Hz, the settling time is shortened by 36% compared with the fixed-coefficient method, the peak reactor power variation rate is 4.9%/min below the limit, and both thermal stress and control rod actuation frequency meet the safety requirements, demonstrating strong robustness within the load disturbance range of 3% to 8%.