Development and evaluation of a prototype for remote voltage monitoring based on artificial neural networks

• Published in: Engineering applications of artificial intelligence Vol. 57; pp. 50 - 60
• Main Authors: Silva, Alex S., dos Santos, Ricardo C., Bottura, Fernando B., Oleskovicz, Mário
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.01.2017
• Subjects: Artificial neural networks; Hardware implementation; Power quality; Remote voltage monitoring; Smart grid; Hardware implementation; Power quality; Smart grid; Remote voltage monitoring; Artificial neural networks
• ISSN: 0952-1976; 1873-6769
• DOI: 10.1016/j.engappai.2016.10.012

This work presents a new solution for remote voltage monitoring in distribution systems based on Artificial Neural Network (ANN). The main advantage of this method is the possibility of monitoring the voltage amplitude in different locations of the distribution system, by using only signals from the utility substation. Initially, the proposed scheme is designed to monitor the voltage amplitude in a specific location of the distribution system. Thus, a methodology to define and evaluate ANNs, as well as to validate the proposed scheme is presented. As a next step, five additional locations are chosen to be monitored and the extended proposed scheme is defined and validated. Alternative Transients Program (ATP) is used to simulate the distribution system and Matlab is used to specify ANNs. Finally, a hardware prototype is developed in order to validate the proposed scheme for practical applications. The development environment LabVIEW and its data acquisition platforms (real-time and non-real-time) are used for the hardware implementation, while real signals are used in the evaluation process. Results from simulations and experiments are compared, proving the effectiveness of the scheme presented in this paper. •A new solution based on ANNs for remote voltage monitoring is presented.•A FPGA hardware prototype to validate the ANN based algorithm is developed.•The proposed scheme is able to monitor voltage sags by using only local signals.•Neural networks are able to support or replace conventional power-quality meters.•Different locations can be monitored by using the proposed ANN based scheme.