Real-Time Battery Bank Charge-Discharge Using Neural Sliding Mode Control

Batteries are one of the most important technologies for energy storage, and an important part of the renewable energy generation systems. Nowadays, there exist diverse types of batteries with different characteristics and it is very important to be able to charge and discharge the battery at a safe and effective rate to maximize the battery life and fulfill the requirements of each application. Different power electronics circuits can be used to control battery charge and discharge; one of the most used is the DC-DC Buck-Boost converter. Typically, this type of converter only allows for energy transfer in one direction; otherwise a different topology and two controllers are needed, which must commute between charging and discharging mode. The objective of this work is to develop a single neural sliding modes control with on-line identification for both battery charge and discharge processes. The proposed controller does not depend on the converter parameters; for this reason, it can be used under different voltage and current requirements without the need of considerable changes in its implementation.