基于非支配排序遗传算法的系统分析程序优化技术OA

System analysis codes are widely used in the safety analysis of nuclear reactor systems.However,the model-ing process of reactor systems largely depends on the trial-and-error experience of technical personnel,which is time-consuming and highly empirical.To enhance the efficiency and accuracy of modeling in system analysis codes,the non-dominated sorting genetic algorithm Ⅱ is applied to optimize the system analysis process.Using the ORNL-THTF ex-perimental facility as the modeling object,sensitivity analysis was conducted with a sampling-based analysis tool to iden-tify key parameters and their ranges.An automated optimization program was established,with the objective function de-fined as the relative error between simulation results and experimental data.The ORNL-THTF model was optimized for both steady-state and transient conditions.The optimized simulation results were then compared with experimental data for validation analysis.The results indicate that,under steady-state conditions,the relative error after optimization was reduced to within 1%,while under transient conditions,the errors decreased by 9.76%and 4.56%,and the simulated trends became more consistent with the experimental data.The proposed method demonstrates advantages in achieving effective optimization and a fully automated optimization process.