Over the last century, the energy storage industry has continued to evolve and adapt to changing energy requirements. To run an efficient energy storage system two points must be considered. Firstly, precise load forecasting to determine energy consumption pattern. Secondly, is the correct estimation of state of charge (SOC).
In this project there is a model introduced to predict the load consumption based on ANN implemented by MATLAB. The Designed intelligent system introduced for load prediction according to the hypothetical training data related to two years daily based load consumption of a residential area.
For another obstacle which is accurate estimation of SOC, two separate models are provided based on ANN and ANFIS for Lithium-ion batteries as an energy storage system. There are several researches in this regard but in this project the author makes an effort to introduce the most efficient based on the MSE of each performance and as a result the method by ANN is found more accurate.
Remaining energy estimation for lithium-ion batteries via Gaussian mixture and Markov models for future load prediction
