Application of Fuzzy-RBF-CNN Ensemble Model for Short-Term Load Forecasting

Table 1

Relevant work summary report in terms of techniques, aim, drawbacks, and remarks.

Techniques	Aim	Drawback	Remarks
Load forecasting based on weather information for bulk power system	Improvement in forecasting accuracy	Suitable for bulk power system only	Incorporating exogenous variables, the performance of bulk and distributed power systems improves
Residential forecasting using DRNN	Enhance the user’s comfort level by reliable electricity availability	Model complexity increases	Residential energy forecasting is possible by sharing the load data of consumers to energy regulation commission
LSTM-RNN for residential forecasting	Improved accuracy	Increased in accuracy only for meter level forecast	Improvement in accuracy not in convergence rate
IoT-based load forecasting	Improved operation of power system with accuracy	Large complex framework	Impact on convergence rate
Forecasting based on big data approach	Accuracy improved for scalable models	Complex structure with less convergence	High complexity with improved accuracy
Week ahead forecasting using deep model with denoising auto encoders	Improved accuracy	Model performance is affected with reduced data size	Convergence rate is affected, but accuracy improved with large data size
Artificial intelligence-based load forecasting	Reduction in MSE with improved accuracy	Accuracy with high convergence rate	Sigmoid function reduced convergence rate
Intelligent hybrid model for load forecasting	Day-ahead load forecasting	Effective management of grid operation	Reliability is improved with high model complexity