Development and Validation of Artificial Neural Network-Based Tools for Forecasting of Power System Inertia With Wind Farms Penetration

Increased penetration of power electronics interfaced renewable energy sources-based generation, for instance, wind farms, that displaces much of the conventional synchronous generation, has a profound effect on the inertia of modern/future power system networks. This article presents the development and validation of artificial neural network (ANN)-based tools, utilizing the power system variables measured by phasor measurement units through wide-area measurements systems, for estimation/forecasting of power system inertia with high penetration of wind farms. The development stage involves the correlation analysis to identify the best power system variables that can be nominated as inputs, and the training of the proposed inertia forecasting tools with the best-nominated inputs that are highly correlated with the power system inertia. Whereas, in the validation stage, the functionality of the trained ANN-based inertia forecasting tools have been validated using the hardware-in-the-loop testing facility developed at the University of Manchester. The development and validation procedures of the proposed inertia forecasting tools have been demonstrated on the IEEE 9-bus modified test system. The validation results revealed the effectiveness of the proposed inertia forecasting tools in estimating the inertia of modern/future power system networks with high penetration of wind farms.