Enhancing Fault Clearing Algorithm for Renewable Energy-based Distribution Systems Using Artificial Neural Networks

Abstract

Integrating small and large-scale photovoltaic solar systems into electrical distribution systems becomes mandatory due to the increased electricity bills and the concern for limiting greenhouse gases. However, the reliable and efficient operation of photovoltaic-based distribution systems can be confronted by the intermittent and high variability of solar source and their consequent faults. In this regard, this article suggests a moderated fault-clearing strategy based on the incremental conductance-maximum power point tracking technique and artificial neural networks to enhance fault detection, localization, and restoration processes in photovoltaic-based distribution systems. The proposed strategy leverages incremental conductance-maximum power point tracking to ensure optimal power generation from the photovoltaic solar system, even in the presence of faults. By tracking the maximum power point, the algorithm maintains the system’s performance and mitigates the impact of faults on the output power. Furthermore, an artificial neural network is employed to improve fault detection and localization accuracy. The developed artificial neural network-based moderated fault-clearing strategy is trained using historical data and fault scenarios, enabling it to recognize fault patterns and make informed decisions through extensive simulations and comparisons with traditional fault-clearing methods. To accomplish this study benchmarks in photovoltaic-based distribution systems are constructed and employed using the MATLAB®/Simulink® software package. Moreover, to validate the efficacy of the developed artificial neural network-based moderated fault-clearing strategy a real case study of 1 MW photovoltaic-based distribution systems in an industrial field located in Giza governorate, Egypt, is tested and investigated. The obtained results demonstrate the effectiveness of incremental conductance- maximum power point tracking and artificial neural network-based moderated fault-clearing strategy in achieving faster fault detection, precise fault localization, and efficient restoration in photovoltaic solar-based distribution systems while preserving maximum power extraction under small and large system disturbances. Furthermore, incremental conductance- maximum power point tracking based on an artificial neural network achieves an average power of 98.556 kW and 299.632 kWh energy availability, whereas the incremental conductance-maximum power point tracking based on proportional-integral controller achieves 95.7996 kW and 283.4036 kWh, and classical perturb and observe maximum power point tracking algorithm achieves 92.2657 kW and 276.8014 kWh.