Low frequency oscillations (LFOs) occur in a system of interconnected generators connected by weak interconnection. A power system stabilizer (PSS) is commonly used to improve the capacity of the power system dampening. Under a variety of operating conditions, traditional PSSs fail to deliver superior damping. To address this issue, a Farmland Fertility Algorithm (FFA-PSSs controller) was used to solve an optimization problem for optimal design of PSSs system parameters, and its performance efficiency was compared to GA and PSO-based PSSs controllers. In addition to PSS, flexible current transmission (FACTS) devices are widely used. PSSs controllers and FACTS devices are frequently constructed in tandem to improve the dampening efficiency of the system. In this study, an Interline Power Flow Controller (IPFC) FACTS device will be added to the PSSs controller to improve the power system’s oscillatory stability. PSSs optimal design and supplemental controller of power fluctuations for IPFC were conducted out on WSCC multi-machine test systems using a system linear model. Using time-domain simulations and quantitative analysis, the proposed IPFC model was compared to the FFA-PSSs controller in terms of performance and efficiency. The main disadvantage of this technique is the difficulty in designing a dynamic IPFC model in test systems, as well as the burden of IPFC coordinated PSSs optimization. In both PSSs design using FFA method and FFA-optimized PSS with IPFC cases, rise in the computational and simulation costs was found unavoidable. To compensate for these flaws and obtain the research contribution, this paper proposes a Neuro-Fuzzy Controller (NFC) developed as a damping controller that can take the place of the two controllers (research objectives three). The application of the NFC substitute the computational and simulation cost involved in designing multi-machine PSS and IPFC-FACTS systems simultaneously. With the availability of NFC in SIMULINK, a dynamic model of the WSCC three-machine system was developed under a variety of operating situations. Quantitative analysis results from the WSCC test system simulation show that when comparing the proposed NFC model to the IPFC model for the WSCC test system, the proposed NFC model was found to be 149 percent and 0 percent efficient in terms of the time to settle of rotor angle respond for G2 and G3, respectively, but 394 percent efficient when compared to the uncontrolled model. The decreased settling time values ensured the proposed NFC model’s efficacy in damping down the LFO and achieving superior stability over the two controllers. The proposed NFC model has shown significant performance improvement in both the transient and steady-state areas than when the system was design with the two damping controllers.

In this study, a novel photovoltaic power forecasting system that utilizes a deep Convolutional Neural Network (CNN) structure and an input signal decomposition algorithm is proposed. The proposed CNN architecture extracts deep features to forecast short-term power using transfer learning-based AlexNet. The historical power, solar radiation, wind speed, and temperature data are selected as the input. The signal decomposition algorithm called Empirical Mode Decomposition (EMD) is utilized to decompose the historical power signal into sub-components. In order to extract deep features, all input parameters are converted to 2D feature maps and feed to the input of the CNN. The experiments are realized on a grid-tied Photovoltaic Power Plant (PVPP) that has 1000 kW installed capacity located in Turkey. The experiments are performed under four weather conditions as partial cloudy, cloudy-rainy, heavy-rainy, and sunny days to show the effectiveness of the proposed method. The obtained results are compared with the benchmark regression algorithms. When the results are analyzed, the proposed method gives the highest Correlation Coefficient (R) and the lowest Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and SMAPE values under all horizons and weather conditions. For 1-h to 5-h ahead, the average R values of the proposed method are obtained as 97.28%, 95.77%, 94.49%, 93.61%, and 92.62%, respectively. The average RMSE values are observed as 4.90%, 6.30%, 7.50%, 8.00%, and 9.17% for 1-h to 5-h ahead. The experimental results confirm that the proposed method outperforms the conventional regression algorithms and reveals effective results with its competitive performance.