Table nbsp Comparison with benchmarking

The multi-objective framework is implemented on the data corresponding to wind farms located in the eastern region of United States. Several case studies are performed for different forecast horizons, nominal confidence levels on different data sets. The ability of the methodology to generate a range of pareto-optimal solutions is clearly observed through these Ginsenoside Re studies. The best compromised solution determined by a fuzzy approach is shown to generate high quality PIs indicated by low coverage errors and higher sharpness of the generated PIs. The quality of the PIs is also compared with those generated by some benchmarking techniques and superiority of the proposed method is observed in all test cases. The PIs are more useful from practical aspects and can be incorporated by the wind farm owners into the decision making process. The proposed framework can be further improved by incorporating a dedicated feature selection technique for determining more relevant input features. The technique can also be extended to longer look-ahead forecasts and also for other renewable energy sources such as solar and tidal.