Prioritization of Medium-Voltage Wiring Actions With Artificial Intelligence Algorithm

Danilo  Pereira; Carlos Frederico  Meschini Almeida; Fillipe Matos de Vasconcelos; Nelson Kagan; James Júnior; Fabricio Expedito  Viana; José Dorlando de Souza Junior; Alexandre  Dominice

doi:10.48011/sbse.v1i1.2175

Danilo Pereira Universidade de São Paulo
Carlos Frederico Meschini Almeida Universidade de São Paulo
Fillipe Matos de Vasconcelos Universidade de São Paulo
Nelson Kagan Universidade de São Paulo
James Júnior EDP São Paulo Distribuição de Energia S.A.
Fabricio Expedito Viana EDP São Paulo Distribuição de Energia S.A.
José Dorlando de Souza Junior EDP Espírito Santo Distribuição de Energia S.A.
Alexandre Dominice EDP São Paulo Distribuição de Energia S.A.

DOI: https://doi.org/10.48011/sbse.v1i1.2175

Keywords: Optimization, Genetic algorithms, Maintenance planning, Asset management, Quality of service, Electric power distribution

Abstract

In power distribution networks, keeping quality of service indexes at high levels (SAIDI and SAIFI) can be accomplished through preventive maintenance interventions, which are scheduled by utility’s maintenance planning professionals. They are responsible for planning maintenance actions consisting of replacing damaged and obsolete devices and installing new devices, while respecting a certain budget availability. Among all possible actions, spacer cable (SC), phase separators (FS) and tree pruning (TP) are aimed to correct issues on medium-voltage wiring (MVW). Through these actions, impending wiring-related failures are mitigated or even avoided. Currently, utilities’ planning professionals may inaccurately determine the annual set of maintenance actions, due to the use of recorded measurements, data from multiple convoluted spreadsheets and personal experience. This paper presents the development of an automated computational tool aimed to prioritize MVW-related maintenance actions, assisting planning professionals in optimizing the use of annual available budget.