Goud B Srikanth, Kalyan Ch Naga Sai, Rao Gundala Srinivasa, Mohapatra Bhabasis, Kuppireddy Narsimha Reddy, Pulluri Harish, Reddy Ch Rami, Shorfuzzaman Mohammad, Zneid Basem Abu, Pushkarna Mukesh
Department of Electrical and Electronics Engineering, School of Engineering, Anurag University, Hyderabad, 500088, India.
Department of Electrical and Electronics Engineering, Vasireddy Venkatadri International Technological University (VVITU), Guntur, 522 508, India.
Sci Rep. 2025 Jun 4;15(1):19677. doi: 10.1038/s41598-025-04833-4.
Power quality has prominently gained its importance in power systems with the advancement of technology. Voltage sags/swells, harmonics, and other disturbances are the major issues causing most of the technical and financial damages which are reducing the quality of energy supplied. To overcome these challenges design of a unified power quality conditioner (UPQC) plays a vital role in mitigating the PQ issues. In this paper, an advanced neural network base approach is developed to manage UPQC to maintain a constant power supply for the end users. DC link of UPQC is taken from PV, fuel, and battery at a specific range. The compensator DC link is linked in a smart grid with nonlinear load. On the other hand, the switching pulse was performed with the use of the Gated Recurrent Unit (GRU) controller technique. Various fault conditions are created to make a dataset that is utilized to design the GRU that analyses the load voltage and current at each second to generate a pulse for UPQC. The performance is evaluated utilizing an advanced controller under various conditions, including swell, sag, harmonics, and combined three-phase faults. The low harmonic content of voltage is 0.04%, 0.25%, and 0.98%. The suggested controller is accessible with 99.5% specificity, 99% sensitivity, and 98% accuracy. The proposed controller provides low harmonic content while operating in a highly secure, dependable, and efficient manner.
随着技术的进步，电能质量在电力系统中已显著变得重要起来。电压骤降/骤升、谐波及其他干扰是导致大部分技术和经济损失的主要问题，这些问题正在降低所供应电能的质量。为克服这些挑战，统一电能质量调节器（UPQC）的设计在减轻电能质量问题方面起着至关重要的作用。在本文中，开发了一种先进的基于神经网络的方法来管理UPQC，以确保为终端用户维持恒定的电力供应。UPQC的直流母线取自特定范围内的光伏、燃料和电池。补偿器直流母线与带有非线性负载的智能电网相连。另一方面，开关脉冲是利用门控循环单元（GRU）控制器技术执行的。创建了各种故障条件以生成一个数据集，该数据集用于设计GRU，GRU每秒分析负载电压和电流，以生成UPQC的脉冲。在包括骤升、骤降、谐波和组合三相故障等各种条件下，利用先进的控制器对性能进行评估。电压的低谐波含量分别为0.04%、0.25%和0.98%。所建议的控制器具有99.5%的特异性、99%的灵敏度和98%的准确率。所提出的控制器在以高度安全、可靠和高效的方式运行时，能提供低谐波含量。
