Assessment of Adaptive Coordinated Control of Overcurrent Relay Using Novel Multi-Trial Vector-Based Differential Evolution Algorithm

Abstract

In this work, we use a unique multi-trial vector-based differentiation algorithm to examine the dynamically coordinated management of over current relays (OCRs) in substations. OCRs are significant for substation protection, and their coordinated control is essential for different sections experiencing varying current levels. Numerous optimization and AI techniques have been explored by researchers to enhance OCR coordinated control performance. This study introduces adaptive control for automatic coordination time adjustment based on forecasting, particularly focusing on integrating wind energy systems with and without reactors in substations. Accurate forecasting is vital for effective adaptive OCR control; hence, a novel multi-trial vector-based evolutionary (MTDE) algorithm is proposed for wind energy forecasting. MATLAB is used for the analysis, comparing the performance of artificial neural network (ANN)-based wind energy forecasting. The proposed method employs ANN for wind energy forecasting in a 110/22KV substation, enhancing OCR coordinated control based on wind energy integration. The ANN is trained using Mean Square Error (MSE), and its weights are tuned using MTDE for improved forecasting accuracy. Simulation results show that the MTDE-tuned ANN significantly enhances wind energy forecasting accuracy compared to conventional ANN, enabling superior adaptive coordinated OCR control, and thereby improving substation protection under fault conditions.