This study presents a new optimal supplementary neuro-controller design for unified power flow controllers (UPFCs) using wide-area signals. The new design is oriented by the dual heuristic programming (DHP) method, which is a powerful adaptive critic procedure. The proposed controller injects additional signals to the UPFC series inverter to enhance power system stability. A model, action and critic neural networks are also designed for optimizing the neuro-controller. To have an accurate dynamic response of multi-machine power systems, the concepts of two-machine equivalent model (TMEM) and center of inertia (COI) are used to train the adaptive critic design (ACD) neural networks. The ANFIS structure is applied to the suggested DHP technique based on the selected input signals to improve the dynamic performance of the applied DHP controller. The designed optimal controller is applied on a real interconnected power system between Tehran and Khuzestan regions in Iran power grid. The proposed ANFIS-DHP controller performance is compared with various controllers, such as PI-Genetic, PI-Lyapunov and other intelligent approaches. The dynamic responses of the proposed controller are found to be the most effective in increasing system stability and damping out system inter-area oscillations.

本研究为使用广域信号的统一潮流控制器 (UPFC) 提出了一种新的最佳补充神经控制器设计。新设计以双重启发式编程（DHP）方法为导向，这是一种强大的自适应批评程序。建议的控制器向 UPFC 系列逆变器注入额外的信号，以提高电力系统的稳定性。还设计了模型、动作和评论神经网络来优化神经控制器。为了获得多机电力系统的准确动态响应，使用两机等效模型 (TMEM) 和惯性中心 (COI) 的概念来训练自适应批评设计 (ACD) 神经网络。基于选定的输入信号，ANFIS 结构应用于建议的 DHP 技术，以提高所应用的 DHP 控制器的动态性能。设计的最优控制器应用于伊朗电网德黑兰和胡齐斯坦地区之间的真实互联电力系统。将提出的 ANFIS-DHP 控制器性能与各种控制器进行比较，例如 PI-Genetic、PI-Lyapunov 和其他智能方法。发现所提出的控制器的动态响应在提高系统稳定性和抑制系统区域间振荡方面是最有效的。