Digital Transformation in Mini Power Plants: Optimization of Hydroelectric Turbines with Neural Networks (#1403)

Authors

Murillo Manrique, Margarita F.

Loayza Jaqui, Raul

Vidal Barrena, Victor

Abstract

The objective of the research was to optimize the parameters of hydraulic turbines using the artificial neural network (ANN) for a mini-power plant that operates with different volumes of water. For the procedure, different phases were carried out, which included the collection of data on the flow rate, height of the water fall and rotation speed (rpm), which were tabulated and used to train the ANN, the implementation of an intelligent system based on ANN composed of three input layers and five output layers with the bakpropagation algorithm, the design of a sequence of instructions and the training of the ANN using the NeuroShell 2 software. The selection of the different turbines was carried out by the intelligent system based on expert knowledge and prior calculations. The results obtained showed a significant accuracy in the prediction of power (in HP and KW), the type of impeller and the output speed (ns), which guaranteed stable and reliable values of electrical energy for the mini hydroelectric plant. The conclusions of this study highlight the effectiveness of the application of neural networks in the optimization of hydroelectric turbines, underlining their importance to improve efficiency and reliability in the generation of electrical energy.