Steady-state inertia estimation using a neural network approach with modal information

The inertia of a power grid plays a significant role in maintaining the stability of a system. If the inertia is large enough, stable operating conditions can be maintained during small scale events. As the percentage of power supplied by renewable energy sources increases, the value of inertia in a system will decrease. Therefore, it has become necessary to accurately estimate the inertia in the system. Traditional methods of estimating the inertia make use of fault conditions to allow for the dynamics in the system to be accurately observable. However, this is not optimal as fault conditions are infrequent and undesirable. The method detailed makes use of modal information which can be obtained via synchrophasor measurements to estimate the inertia during steady-state conditions. The results show that while the estimation is not accurate for individual buses, the values calculated for regional and system inertias are more accurate.