An Expert Committee Evaluation for Load Forecasting in a Smart Grid Environment

Electricity markets as well as integration of renewables and electric vehicles impose new requirements for efficient energy management, turning load forecasts into an essential tool for safe system operation at the lowest cost. However, the diverse load dynamics identified in a distribution system necessitate the development of novel forecasting models capable of accurate load forecast provision in a smart grid environment. The load forecasting system under evaluation is an expert committee which comprises four prediction models (PMs) and one combination model (CM) and provides forecasts in a 24-h ahead forecasting horizon. The load forecasting system performance is validated using data from four actual load time series characterized by large diversity. The expert committee exhibits satisfactory performance on all load profile tested, effectively capturing the load dynamics.