In the thesis relations between time series forecasting schemes and dynamic state estimation methods are investigated. It is shown in the thesis that dynamic state estimation problems can be considered as time series problems. Different time series forecasting model using nonlinear autoregressive network with exogenous inputs (NARX) and long short-term memory (LSTM) schemes for dynamic state estimation purposes are developed and these models are validated in the IEEE-39 bus systems.
|Qualification||Doctor of Philosophy|
|Award date||29 Sept 2020|
|Publication status||Unpublished - 2020|