Dynamic State Estimation Based on Time Series and Exponential Smoothing

Dynamic State Estimation (DSE) is one of the most important parts of Wide Area Monitoring and Control Systems because other functions are based on the results of the state estimator. DSE is based on a statistical predictive method. For this reason, different statistical methods such as time series, Kalman filter, exponential smoothing, regression, and etc. are used and each of which has benefits and drawbacks. This paper proposes a method for DSE on the basis of linear combination of time series and exponential smoothing. The predicted state estimates for different methods are compared using mean square errors. The important advantage of the proposed combinational method is the reduction of error comparing to each method.

Keywords:

Dynamic State Estimation , Weighted Least Squares (WLS) , Time Series , Exponential Smoothing , Combinational Prediction