Optimization of R-rod control strategy for PWR based on rod position input rating

With the increasing demand for flexible resources in the power system, when nuclear power participates in peak - shaving and load variation, the frequent changes in the opening of the main steam valve disturb parameters such as the average temperature of the primary coolant in the primary loop. Additionally, the power mismatch channel in the reactor R - rod control system has difficulty meeting the requirements of rapid control.  [Purpose]: This study focuses on exploring a suitable optimization strategy for the R-rod control system, aiming to significantly improve the control performance of nuclear power units under variable load conditions. [Methods]: Firstly, based on the dynamic model of large pressurized water reactor nuclear power units, the rod position input rating strategy was adopted, that is, a feedforward link with the load dispatching instruction as the input signal was introduced in front of the rod speed control unit of the traditional R-rod control system. In terms of specific implementation, on the one hand, the theoretical derivation method was used. On the basis of reasonably simplifying the complexity of the R-rod control system, the transfer function of the feedforward link was precisely calculated with the help of Mason's formula. On the other hand, the particle swarm optimization algorithm was adopted. Relying on the efficient data interaction between MATLAB and Simulink, the undetermined coefficients in the transfer function of the feedforward link were directly optimized. Finally, the transfer functions obtained by the two methods were respectively incorporated into the model, and ±5%, ±10% and ±20% load steps were introduced under 80% working conditions to obtain the changes of various parameters. [Results]: Under six step load change conditions, the feedforward link obtained through theoretical derivation reduces the adjustment time of the primary coolant temperature and the main steam pressure by 53.7% to 89.6%, and the volatility is reduced by 22.8% to 100%. On the other hand, the feedforward link calculated by the particle swarm optimization algorithm optimizes the adjustment time of the primary coolant temperature and the main steam pressure by 64.4% to 95.3%, and the volatility is reduced by 23.4% to 100% [Conclusions]:The results demonstrate that under the same operating conditions, the optimization effects of the theoretical derivation method and the particle swarm optimization algorithm are comparable. Both methods significantly enhance the regulation speed of the average temperature of the primary coolant and the main steam pressure, while effectively reducing system fluctuations, thereby achieving the desired control objectives. Among them, the theoretical derivation method, owing to its clear physical significance, low computational cost, and straightforward operational procedures, exhibits greater practical value in engineering applications.