An Efficient Regression Prediction Model for Daily Solar Power Generation of Microgrid based on the Collaborative Optimization of Dual-Module GA-BP

Weimin Wu, Zhaoqin Liu, Yun Wu, Wanjun Yan, Xiufan Yang, Dianxi Zhang

Abstract


Combined the backpropagation (BP) neural network with the genetic algorithm (GA), an efficient dual-module regression prediction model is presented for power generation prediction in distributed solar intelligent microgrids. Two years of operation data were collected from a local microgrid system with eight environmental factors as input data. The input data is selected as input of the module, and the cumulative daily power generated by monocrystalline silicon module is used for prediction. The structural framework of the module adopts the dynamic mechanism of switching models. In normal mode, the shallow module (a single-hidden-layer BP network with the GA optimized) is adopted for rapid inference within 0.5 s, and the root mean square error (RMSE) is 62.02W. In more intricate or variable situations, the subsystem activates a deep module (a dual-hidden-layer network trained on 150 generations of GA), and the results are highly accurate and decrease the RMSE to 20.93W and increase the coefficient of determination (R2) to 0.943. The onset time of light intensity precedes that of ambient temperature by one to two hours, resulting in an average relative prediction error within 3%. The efficiency of power scheduling can reach more than 30% by adjusting power allocation and optimize energy curtailment. The proposed framework can obtain effective balance between computation and forecasting precision, providing a realistic and reliable solution for on-line energy management in microgrid applications.

Keywords


Daily power generation prediction; Distributed solar micro-grid; Dual-module dynamic switching; GA-BP collaborative optimization; Nonlinear regression modeling

Full Text:

PDF

References


Hassan Q., Hsu C.Y., Mounich K., Algburi S., Jaszczur M., Telba A.A., Viktor P., Awwad E.M., Ahsan M., Ali B.M., Al-Jiboory A.K., Henedy S.N., Sameen A.Z., and Barakat M., 2024. Enhancing smart grid integrated renewable distributed generation capacities: Implications for sustainable energy transformation. Sustainable Energy Technologies and Assessments 66: 103793. https://doi.org/10.1016/j.seta.2024.103793.

Dawn S., Ramakrishna A., Ramesh M., Das S.S., Rao K.D., Islam M.M., and Selim Ustun T., 2024. Integration of renewable energy in microgrids and smart grids in deregulated power systems: a comparative exploration. Advanced Energy and Sustainability Research 5(10): 2400088. https://doi.org/10.1002/aesr.202400088.

Suresh M.P., Yuvaraj T., Thanikanti S.B., and Nwulu N., 2024. Optimizing smart microgrid performance: Integrating solar generation and static VAR compensator for EV charging impact, emphasizing SCOPE index. Energy Reports 11: 3224-3244. https://doi.org/10.1016/j.egyr.2024.02.050.

Kumar M.J., Sampradeepraj T., Sivajothi E., and Singh G., 2024. An efficient hybrid technique for energy management system with renewable energy system and energy storage system in smart grid. Energy 306: 132454. https://doi.org/10.1016/j.energy.2024.132454.

Arkhangelski J., Abdou-Tankari M., and Lefebvre, G., 2021. Day-ahead optimal power flow for efficient energy management of urban microgrid. IEEE Transactions on Industry Applications, 57(2): 1285-1293. https://doi.org/10.1109/TIA.2020.3049117.

Zhang B., Yan L., He Y., Huang H., Feng J., Lu L., Zhang E., Che Y., Jiang M., Liu L., and Li Z., 2021. Research on Multi-Energy Complementary Linkage Technology and Economic Evaluation. 2021 4th International Conference on Information Systems and Computer Aided Education. 2021: 1871-1878. https://doi.org/10.1145/3482632.3484058.

Elgammal A. and C. Boodoo. 2021. Optimal energy management strategy for a DC linked hydro-PV-wind renewable energy system for hydroelectric power generation optimization. European Journal of Energy Research 1(3): 9-18. https://doi.org/10.24018/ejenergy.2021.1.3.16.

Kumar G.B. and K. Palanisamy. 2022. Energy management of renewable energy-based microgrid system with HESS for various operation modes. Frontiers in Energy Research 10: 995034. https://doi.org/10.3389/fenrg.2022.995034.

Barua S. and N. Mohammad. 2023. Investigating heuristic and optimization energy management algorithms to minimize residential electricity costs. 2023 International Conference on Electrical, Computer and Communication Engineering (ECCE). IEEE, 2023: 1-6. https://doi.org/10.1109/ECCE57851.2023.10101630.

Mohan H.M. and S.K. Dash. 2024. Optimized power flow management based on Harris Hawks optimization for an islanded DC microgrid. Energy Harvesting and Systems 11(1): 20220153. https://doi.org/10.1515/ehs-2022-0153.

Siddaraj S., Yaragatti U.R., and Harischandrappa N., 2023. Coordinated PSO-ANFIS-based 2 MPPT control of microgrid with solar photovoltaic and battery energy storage system. Journal of Sensor and Actuator Networks 12(3): 45. https://doi.org/10.3390/jsan12030045.

Elalfy D.A., Gouda E., Kotb M.F., and Sedhom B. E., 2023. Frequency Regulation Based on Antlion Optimization Technique for Microgrids with Electric Vehicles. 2023 24th International Middle East Power System Conference (MEPCON). IEEE, 2023: 1-7. https://doi.org/10.1109/MEPCON58725.2023.10462385.

Lv Y., Li K. Zhao, H., and Lei H., 2024. A multi-stage constraint-handling multi-objective optimization method for resilient microgrid energy management. Applied Sciences 14(8): 3253. https://doi.org/10.3390/app14083253.

Agupugo C.P., Kehinde H.M., and Manuel H.N.N., 2024. Optimization of microgrid operations using renewable energy sources. Engineering Science and Technology Journal 5(7): 2379-2401. https://doi.org/10.51594/ESTJ.V5I7.1360.

Ersöz B., Taşdelen M.C., Eren S., Sagiroglu S., and Öter A., 2024. Solar energy forecasting using ensemble learning method. 2024 13th International Conference on Renewable Energy Research and Applications (ICRERA), IEEE, 283-287. https://doi.org/10.1109/ICRERA62673.2024.10815583.

Alagappan C., Ramasamy K., and Velusamy D., 2025. Bagging ensemble of artificial neural networks with weighted averaging and grey wolf optimization for solar photovoltaic power generation prediction. International Journal of Green Energy 22(8): 1587-1601. https://doi.org/10.1080/15435075.2024.2436535.

Al-Dahidi S., Ayadi O., Adeeb J., Alrbai M., and Qawasmeh B.R., 2018. Extreme learning machines for solar photovoltaic power predictions. Energies 11(10): 2725. https://doi.org/10.3390/en11102725.

Al-Dahidi S., Ayadi O., Alrbai M., and Adeeb J., 2019. Ensemble approach of optimized artificial neural networks for solar photovoltaic power prediction. IEEE Access 7: 81741-81758. https://doi.org/10.1109/ACCESS.2019.2923905.

Alrbai M., Abubaker A.M., Ahmad A.D., Al-Dahidi S., Ayadi O., Hjouj D., and Al-Ghussain L., 2022. Optimization of energy production from biogas fuel in a closed landfill using artificial neural networks: A case study of Al Ghabawi Landfill, Jordan. Waste Management 150: 218-226. https://doi.org/10.1016/j.wasman.2022.07.011

Al-Dahidi S., Muhsen H., Sari M.E.S., Alrbai M., Louzazni M., and Omran N., 2022. An adaptive approach-based ensemble for 1 day-ahead production prediction of solar PV systems. Advances in Mechanical Engineering 14(3): 16878132221089436. https://doi.org/10.1177/16878132221089436

Al-Ghussain L., Darwish Ahmad A., Abubaker A.M., Alrbai M., Ayadi O., Al-Dahidi S., and Akafuah N.K., 2023. Techno-economic assessment of photovoltaic-based charging stations for electric vehicles in developing countries. Energy Sources, Part A: Recovery, Utilization, and Environmental Effects, 45(1): 523-541. https://doi.org/10.1080/15567036.2023.2171517.

Al-Dahidi S., Adeeb J., Ayadi O., Alrbai M., and Al-Ghussain L., 2022. A feature transformation and extraction approach-based artificial neural network for an improved production prediction of grid-connected solar photovoltaic systems. Energy Sources, Part A: Recovery, Utilization, and Environmental Effects 44(4): 9232-9254. https://doi.org/10.1080/15567036.2022.2128475.




DOI: https://doi.org/10.64289/iej.26.0103.7975642