Adeoye, O. S,, Akinwole, O. O, Olofinjana, A.S, Ogunlowo, M.O, Ogunsakin,O, Bankole,T.S, Omojoyegbe, M. O, and Oluyemi, F, Oyedele, O.A, Aliu A. A, a,, Erinle, T.J , b, Ologunde, C.A, Akinwande, J.T
Abstract
Accurate prediction of hybrid inverter performance is essential for improving the reliability, efficiency, and energy management of residential renewable energy systems. This study presents a data driven performance prediction and assessment of a 5kVA hybrid inverter installed in an academic office building in Ado-Ekiti, Ekiti State, Nigeria, using Specialized Neural Network and a Least Squares Regression model. Operational parameters, including input voltage, output voltage, frequency, and power factor, were collected and used for model development and validation. The experimental set up comprised of 5.2 kVA hybrid inverter, photovoltaic modules, and a Fluke 1736, three phase power loggers for electrical parameter acquisition. The power logger was configured to measure input and output voltage, current, frequency and power factor. A laptop computer running MATLAB 2025a was used for data pre-processing, model development, and performance evaluation. The predictive performances of the models were evaluated using Mean Average Error , Mean Squared Error , Root Mean Squared Error , and the coefficient of determination (R2). The SNN achieved an MSE of 0.012, 0.0002, 0.000001 and 0,00162 for voltage, frequency, power factor and current predictions respectively. The MAE for predicted values of voltage, frequency, power factor and current were 0.090, 0.014, 0.0009 and 0.040 respectively. For the predicted values of voltage, frequency, power factor and current, the RMSE values were 0.108, 0.016, 0.0010 and 0.04025 respectively. The predicted values of the voltage, frequency, power factor and current, R2 values were 0.9999, 0.96, 0.999 and 0.9996 respectively. The lower prediction errors and higher coefficient of determination obtained by SNN demonstrate its superior capability in capturing the non-linear operating characteristics of hybrid inverter. The findings indicate that specialized neural networks provide more accurate and reliable IJEMT performance prediction than conventional regression methods, making them suitable for intelligent monitoring, predictive maintenance, and optimal energy management in residential hybrid renewable energy systems.
References
Adaramola, M. and Adelabu, M. (2017): Performance Analysis of Grid-Tied Sine Wave Inverters in a Hybrid Power System, Vol.7. Issue 6, 47-58. Adeoye,O.S. and Olulope, P.K. (2018). Investigation and Analysis of Power Quality in Single Phase, Low Voltage Consumers in Ado-Ekiti Metrolopolis. International Journal of Research Granthaalayah, Vol.6, Issue 5, 177-189. Adeoye, O.S., Yusuf, B.M. and Ogunsakin, O.O. (2023). Scientific Assessment of Locally and Factory Built 2kVA Modified Sine Wave Solar Powered Inverters. Indian Journal of Microprocessors and Microcontroller, 2(1), 1-12. Afandi, I.M. ( 2020). Management of Voltages in Low Voltage Networks Final Reports. Australian Power Quality and Reliability Centre, University of Wollongong, Australia, 2020 ( 2), 1-22. Alquithami, T., Sreerama, K., R., and Shaikh, (2020). Mitigation of Voltage Rise due to High Solar Penetration in Saudi Distribution Network. Electrical Engineering, .2I (2), 881-890 Ganesh Babu, B., Harshini, B. and Kapoor, R. (2024): Enhancing Selective Harmonics Eliminating in Cascaded H-Bridge Multilever Inverters through Dragonfly Algorithm Optimization. 2nd International Conference on Intelligent Cyber Physical System and Internet of Things, ICOICT-2024 Proceedings, Issue ICOICT, 1278-1284. Gupta, R. and Pandey, M. (2021). Load Adaptability of Hybrid Inverters in Residential Applications. Energy Reports, 8(1), 2124-2135. Gupta, R., Sharma, N. and Verma, (2021). Grid Stability and Power Quality Considerations for Hybrid Inverters. Energy Reports, 8(1), 453-467. Karimi, M., Mokhlis, K., Uddin, S., and Bakar, A.H.A. (2016). Photovoltaic Penetration Issues and Impacts in Distribution Network. Renewable and Sustainable Energy Reviews, 53 (2016), 594-605. Kumar, R. and Singh, P., (2021). Smart Inverter Technologies for Renewable Energy Integration. Renewable Energy Journal, 45(3), 112-124. Li. Y. and Zhang, X. (2019). Solar Array Sizing and Efficiency Optimization for Hybrid Systems. Solar Energy Journal, 11(2), 144-151. Liu, Y. and Wang, Z. (2020). Performance Analysis of Hybrid Inverters for Grid Connected Solar Systems. IEEE Transactions on Power Electronics, 35(3), 22452257-22452260. Mishra, N., Singh, B., and Cheng, K. (2023). Power Quality Assessment of CPUC Based Solar Photovoltaic System with NILM for Grid Tied Applications. IEEE Transactions on Power Delivery, .(38(01), 655-669. Po, L., Ruigi. L. and Haifeng, F. (2019). Total Harmonic Distortion Oriented Finite Control Set Model Predictive Control for Single Phase Inverters. Energies, 2 (3467), 1-15. Sharma, V., Aziz, S., and Haque, M. (2020) Effects of High Solar Photovoltaic Penetration on Distribution Feeders and the Economic Impacts. Renewable and Sustainable Energy Review, Vol. 31(07), 110021-110032. Singh, S., Patel, K. and Mishra, P. (2020). Load Capacity and Efficiency Evaluation of Hybrid Inverters. Journal of Electrical Energy Studies, 7(2), 75-83. Sinnela, R., Abo-Al-Ez, K., and Kahn, M. (2019): Total Harmonic Distortion with PV System Integration in Smart IJEMT Grids. Proceedings of 27th International Conference on Domestic Use of Energy Conference, 2019(1), 102-108. Wang, L., Ke, W.J.Y., and Wu, R. X. (2023). Stability Evaluation of a Grid-Tied Hybrid Wind /PV Farm joined with a Hybrid Energy Storage System. Sustainable Environment Research, 33(21), 1-5. Zhao, Y., Yang, S., and Zhang, X. (2020). Development of Advanced Lead Carbon Batteries for Renewable Energy Storage. Journal of Energy Storage, 28 (1), 101194-101198. Zhou, P., Yang, T., and Sun, L. (2020). Power Management Strategies in Hybrid Inverters Systems. Renewable Energy and Power Electronics Journal, 20(4), 134-149.
