Digital Transformation in Mini Power Plants: Optimization of Hydroelectric Turbines with Neural Networks

Murillo Manrique M.; Loayza Jaqui R.; Vidal Barrena V. · Proceedings of the LACCEI international Multi-conference for Engineering, Education and Technology (2024) · DOI: 10.18687/laccei2024.1.1.1403

Resumen

This research aimed to optimize the parameters of hydraulic turbines using artificial neural networks (ANNs) in a mini power plant operating with various water volumes. Several phases were conducted, including data collection of flow rate, water head, and rotation speed (rpm), which were tabulated and utilized to train the ANNs. An intelligent system based on ANNs was implemented, consisting of three input layers and five output layers using the backpropagation algorithm. Additionally, an instruction sequence was designed, and the ANNs were trained using NeuroShell 2 software. The selection of turbines was performed by the intelligent system based on expert knowledge and previous calculations. The results obtained showed significant accuracy in predicting power (in HP and KW), impeller type, and output speed (rpm), ensuring stable and reliable values of electrical energy for the mini hydroelectric plant. In conclusion, this study highlights the effectiveness of neural networks in optimizing hydroelectric turbines, emphasizing their importance in improving efficiency and reliability in electricity generation.

Instituciones: Universidad Ricardo Palma