Analysis of the use of pattern recognition networks in the application of intelligent electrical networks

• Autor(a) principal: García Talledo, Raúl Andrés
• Data de Publicação: 2022
• Outros Autores: Cuenca Álava, Lenin
• Tipo de documento: Artigo
• Idioma: spa
• Título da fonte: Sapienza (Curitiba)
• Texto Completo: https://journals.sapienzaeditorial.com/index.php/SIJIS/article/view/408
• topic: Variáveis de entrada, Consumo de energia, Reconhecimento de padrões da rede neural; Input variables, energy consumption, The Neural Net Pattern Recognition; Variables de entrada, Consumo energético, The Neural Net Pattern Recognition

Pattern recognition is based on recognizing the unique properties that identify an individual from others of the same species. The methodology was based on knowing the load and energy consumption generated by the Faculty of Mathematics, Physics and Chemistry of the Technical University of Manabí, in the same way the structure of "The Neural Net Pattern Recognition" was used. pattern recognition), which were trained to classify the inputs according to their classes, carried out using the Matlab R2017b software, with which the training of the network was done with different numbers of neurons in the hidden layer, where values ​​of 10,15,20,25 and 30 were used to obtain the lowest error, the following objectives were set: Know the load and the energy consumption that it generates in the FCMFQ, select the variable that will be established as inputs to the network, training of the smart network using the Matlab software and analyzing the results obtained with the training, The comparison between the different trainings of the network with the mentioned values ​​of the neurons was made, choosing to choose 30 neurons, obtaining the lowest error (0). In conclusion, the use of an intelligent electrical network with the implementation of the ANN technique is beneficial, since, if an intelligent electrical network is implemented throughout the UTM campus, it will be possible to obtain more profitable energy efficiency.