Adaptive evolutionary PID controller using SHADE and L-SHADE for TRIGA mark-II reactor

This paper proposes a trajectory-tracking PID (TTPID) controller optimized using Success-History Based Adaptive Differential Evolution (SHADE) and its improved variant Linear Population Size Reduction SHADE (L-SHADE) for power regulation of the ITU TRIGA Mark-II research reactor. The main novelty of the study is the integration of adaptive differential evolution mechanisms with a trajectory-tracking PID framework for nuclear reactor power control, which has not previously been reported in the literature. A constrained Integral of Time-Weighted Absolute Error (ITAE) objective function is employed to simultaneously minimize tracking error and limit excessive control-rod motion. The optimization performance of SHADE and L-SHADE is first compared, showing that L-SHADE achieves faster and more stable convergence with lower fitness fluctuations. The optimized controller is then evaluated under different initial and desired power levels ranging from 50 to 250 kW. Simulation results demonstrate that the proposed L-SHADE-TTPIDC successfully tracks the reference trajectory with negligible steady-state error and overshoot below 1.1% at a target power level of 225 kW. The controller achieves settling times below 23 s and response times as low as 17 s, outperforming previously reported FLC-PSO and MFLNN-PSO controllers in most operating scenarios. In addition, all operational constraints were satisfied while maintaining smooth control-rod movement. The results confirm that adaptive differential evolution-based PID tuning provides an effective and robust solution for nonlinear reactor power control and represents a promising alternative to existing intelligent control approaches.