Intelligent and Soft Computing Systems for Green Energy

3Electrical Load Forecasting Using Bayesian Regularization Algorithm in Matlab and Finding Optimal Solution via Renewable Source

Chinmay Singh, Yashwant Sawle, Navneet Kumar, Utkarsh Jha* and Arunkumar L.

Department of Electrical Engineering, Vellore Institute of Technology, Vellore, Tamil Nadu, India

Abstract

As the smart grid development advances in India, load forecasting of power is a must. Electric power demand forecasting is critical for energy providers and energy consumers in the electrical industry sector, i.e., power generation, transmission, and its distribution to electric markets as it helps organizations maintain a demand-supply power equilibrium. Precise models in electrical load forecasting have been basic to the activity and planning of any company-based organization. Machine Learning algorithms like Artificial Neural Networks (ANNs) are incredibly powerful and are only limited by their implementation and execution. This paper focuses on minimizing the error between predicted and actual load demand for the industry. That means we will use the ANN method to minimize the Mean Absolute Percentage (MAP) Error between predicted load demand and the actual load values to overcome MAP_Error through graphs and show the advantages of using a Bayesian network over other forecasting techniques like using standard analytical functions. We are using a Bayesian algorithm neural network through MATLAB Simulink. Also, we are focusing on implementing this concept into a renewable ...