An approach based on multi-objective evolutionary algorithm and Monte Carlo Method for optimized monitoring of voltage sags in electricity distribution systems

摘要

Among common disturbances in Electric Power Quality, the most relevant and with greatest rate of occurrence are voltage sags, which cause substantial economical loss to concessionary companies and to customers. Constant monitoring is an essential part of identifying existing disturbances. However, the costs involved make monitoring the complete system infeasible, thus only a reduced number of monitors are available, which need to be installed in strategic positions in order to cover the greatest number of possible events. This work presents an approach for solving the problem of allocating monitors of electric power quality, considering various aspects of the problem, such as topological coverage, voltage sags that have happened but have not been monitored and total cost of equipment installed. The Monte Carlo Method was used for modeling the time series of faults in the distribution system and the Non-Dominated Sorting Genetic Algorithm II (NSGA-II) was used in the construction of this model. The approach was submitted to IEEE 13, 34 and 37-bus distribution systems, which were simulated using DigSILENT Power Factory 15.1 software. The results allow the user to make decisions regarding the amount and the position of the monitors to be installed, in order to seek adequacy to the financial reality of the power supplies' and as to avoid unnecessary costs that would not result in improvements in monitoring performance.