Design of an Adaptive Defense Scheme for the Tambora Power System Based on Artificial Neural Networks

The Tambora Power System comprises 70 kV and 150 kV transmission networks, 20 kV distribution networks, and power generation units operating at 20 kV, 70 kV, and 150 kV. Ensuring system reliability and stability is essential to prevent widespread blackouts that require significant time for recovery. Currently, the defense scheme employed is conventional, relying on under-frequency relays that are often insufficient in mitigating cascading failures. To mitigate this issue, this paper proposes the design of an Adaptive Defense Scheme (ADS) for the Tambora system based on Artificial Neural Networks (ANN). The proposed methodology involves simulating various fault scenarios using DIgSILENT PowerFactory, followed by training the ANN model with an optimization algorithm. The model considers key parameters such as generation capacity, load demand, frequency drop, and system spinning reserve. The results of the study show that the proposed method produces accurate load shedding values compared to conventional methods, while achieving a significantly faster processing time.