• agv
• bms
• li-ion battery
• model based
• soc

The use of Lithium batteries for low-power consumer applications is well established for smartphones, tablets and laptops. The high density of power and energy that they are able to provide also make them very interesting for high-power applications like electric vehicles. In spite of the undoubted advantages that lithium chemistry presents, thermal stability and electric of these batteries is reduced; this makes them very sensitive to fall out of safety ranges that can sometimes cause irreversible damages to the cells. So despite the Lithium-ion batteries constitute a fairly consolidated technology, some aspects are still research object, such as the choice of materials and the development of algorithms and electronic circuits for efficient and safe use of the cells. The system that deals with the control and management of the battery is the BMS (Battery Management System), its main features are the monitoring voltage and temperature of the battery, the measurement of the load current, the estimation of parameters such as the state of charge (SOC, State of Charge) and, possibly, the balance of the cells. Providing an accurate estimate of SOC is one of the most important requirements of the BMS in order to guarantee a high reliable and efficient use of the battery; this parameter is not directly measurable, hence there is the need of a model that describes both static and dynamic behavior of the cell.
The purpose of this thesis work is the study and implementation of a SOC estimation algorithm in which the model parameters are identified online during the operational phases of the battery; the algorithm was designed for its implementation on a embedded system, i.e. a BMS, for monitoring the battery pack of an AGV (Automated Guided Vehicle).
The Li-ion batteries’s SOC estimation has been fully developed starting from the design up to the implementation on the target Electronic Control Unit (ECU).