Integrated Li-ion battery management system

The generation and demand shifts in the power industry due to the introduction of renewable alternatives has dramatically shaped the development of technology in this industry over the years. This shift has brought embedded electronics deeper into the implementation of solutions to power industry problems. This dissertation is part of a hybrid system, one that is connected to both renewable resources and also the main grid. A common issue faced by renewable energy harvesters is the issue of intermittency. Typically batteries are connected to resolve this issue. However batteries also have their own respective laws of operation, typically when to tum on; when to tum off; when to charge, depending on the type of material used. As the battery connected to the hybrid grid by NTU is Lithium Iron Phosphate (LiFePO4) base type, that will be the focused battery material type for this dissertation. This dissertation is broken into two parts: A study of battery management technology and building of a battery management system (BMS). Under the study of battery management technology five cell monitoring methods (three high impedance and two low impedance techniques) and three cell balancing techniques (active, passive and charging) are reviewed. Their theory of operation, implementation and pros and cons are discussed briefly. The second part of the dissertation covers the process of building the BMS. As this project is not a new project (there were four attempts prior to this author's attempt), previous attempts were investigated, commented and failures discussed. Resolving the previous attempts issues, leads to the first of three design phase: Prototyping stage. The second phase was accomplished by upgrading the sensor IC from LTC6802-1 to LTC6803-1. The third phase was the actual PCB fabrication and firmware development stage. A cost and stability analysis study was done in the fabrication part of the third phase that resulted in purchasing 3 LTC6803-1 evaluation kits. The firmware developed was built upon the predecessor's code, optimised with a state deterministic approach. This ensured that the system would not slip into an unknown state. The end result was a 30 channel BMS capable of monitoring, balancing and protecting a 30 cell battery pack connected either in series, parallel or a mixed configuration. The algorithm uses passive balancing based on the maximum and minimum battery reading obtained real time via the monitoring portion of the system. Future recommendations suggested for to this project was to develop the communications aspect of the system or to test other types of cell balancing techniques.Master of Science (Power Engineering