Design of distributed energy system based on artificial neural network approach

Designing distributed energy system (DES) is a complex task due to large varieties and combinations of energy generation, conversion, and storage technologies as well as time-varying energy supplies and demands. In this article, an artificial neural network (ANN) is trained by known DES design samples. Results have shown that after training, ANN can approximate the complex DES mathematical model and yield similar new DES designs to the mathematical model, given new conditions of energy supplies and demands. The advantages of using ANN to design DES lie in the simple structure of ANN and the learning ability from practical as well as updated samples.