2010 · 13 citations · 0 references
ComputingComputer ScienceElectrical EngineeringArtificial IntelligenceForecastingMachine LearningForecasting MethodsKalman FilteringShort-term ForecastingState EstimationTime Series ForecastingTime Series PredictionDynamic Forecasting ModelsNeural ForecastingNeural Network-based EstimationNeural Time Series ForecastingState PredictionState VariablesState Vector EstimationState Vector Representation
This paper is devoted to the technique of forecasting all state variables for a short term. Kalman filter-based algorithms of dynamic state estimation and learned artificial neural networks are used to forecast the state vector components. The trend should be taken into consideration to forecast the state vector components for more than 5 min. The trend is forecasted based on the special table of trends that is filled beforehand for the studied state variable by using two artificial neural networks.