Abstract The control of a nuclear power plant is important due to its nonlinear dynamics. An optimized proportional–integral–derivative (PID) controller is a popular tool to this end. PID is tuned by various methods, including a genetic algorithm (GA). The quality and speed of GA optimization depend on the objective function as the performance index. In this paper, integral of absolute error (IAE), integral of the square error (ISE), integral of time-absolute error (ITAE), and integral of time-square error (ITSE) objective functions are evaluated. The results comparison is done based on overshoot/undershoot and the number of function evaluations (NFE) criteria. The PID gains are tuned to control of a typical pressurized water reactor (PWR), based on the point-kinetics model. The results show the high control performance compared to an empirically tuned PID. Also, ISE and ITSE are better performance indexes than IAE and ITAE in steady-states as critical situations. Also, these performance indexes have signified a good robustness against dynamics parametric uncertainties compared to an empirically tuned PID in every power mode. This comparative approach provides a good idea for experts to choose an appropriate performance index.

中文翻译：

---

压水堆型核反应堆点动力学模型GA调谐PID控制系统误差积分性能指标比较

摘要 由于核电厂的非线性动力学特性，核电厂的控制非常重要。优化的比例积分微分 (PID) 控制器是实现此目的的流行工具。PID 通过各种方法进行调整，包括遗传算法 (GA)。GA 优化的质量和速度取决于作为性能指标的目标函数。本文对绝对误差积分（IAE）、平方误差积分（ISE）、时间绝对误差积分（ITAE）和时间平方误差积分（ITSE）目标函数进行了评估。结果比较基于过冲/下冲和函数评估 (NFE) 标准的数量。基于点动力学模型，PID 增益被调整以控制典型的压水反应堆 (PWR)。结果表明，与经验调整的 PID 相比，控制性能更高。此外，ISE 和 ITSE 在作为临界情况的稳态下是比 IAE 和 ITAE 更好的性能指标。此外，与在每种功率模式下凭经验调整的 PID 相比，这些性能指标表明对动态参数不确定性具有良好的鲁棒性。这种比较方法为专家选择合适的性能指标提供了一个好主意。