Power Flow Parameter Estimation in Power System Using Machine Learning Techniques Under Varying Load Conditions

Abstract

In power transmission network state estimation is more complex and the measurements are critical in nature. Estimation of power flow parameters such as voltage magnitude and phasor angle in a power system is challenging when the loads are varying. The objective of the work is to estimate the voltage magnitude and phase angle using machine learning techniques. Some of the Machine learning techniques are decision trees (DT), support vector machines (SVM), ensemble boost (E-Boost), ensemble bags (E-bag), and artificial neural networks (ANN) are proposed in this work. Among these methods, the best machine learning techniques are selected for this study based on performance metrics. Performance metrics are Mean Square Error (MSE), Mean Absolute Error (MAE), and Root Mean Square Error (RMSE). Neural network produces minimum error when compared to other ML Techniques. Among three Performance Metrics MSE provides minimum error and is used to predict the exact model in this work. Therefore, it is concluded that the neural network can predict voltage and angle accurately under various load conditions in the power system effectively. Neural Network (NN) is applied to different load condition, and performance metrics are computed. To validate the proposed work, IEEE 14 and IEEE 30 bus systems are considered. The predicted value is compared to the actual value for all the load variation and residues are measured. The regression learner software in MATLAB is used to implement ML approaches in this work. The Outcome of this proposed work is used in phasor measurement units. The predicted value of voltage and angle using a neural network can be used to minimize the voltage magnitude and phase angle error in phasor Measurement Units (PMU).