Search

The authors have tested, for about 6 years, four different types of lead-acid batteries under cycling, floating and self-discharge conditions. A correlation has been established between the operating data (current, voltage) and the capacity measured during these tests by a fuzzy cluster approach. The data sets were categorised in three fuzzy clusters. These three fuzzy clusters: (1) starting phase (new phase); (2) working phase; and (3) capacity drop phase should able to describe the ageing behaviour of the battery sufficiently. For all fuzzy analysed data, they found a continuous correlation between the fuzzy classifications and the operation time of the battery, which themselves correlate to the development of the battery state of health. This fuzzy technique allows the prediction of the future behaviour of the battery by measuring the operating parameters (current, voltage) without explicitly checking the capacity, if the fuzzy set is determined once as a kind of calibration curve.

Reliability and costs of an energy storage system are two very important parameters for uninterruptible power supplies (UPS) and other battery applications. The increasing of battery life and the prediction of battery failure are therefore two important features of a battery management system (BMS). A couple of other helpful features can be implemented in a BMS. In the case of advanced batteries, special features like thermal management (i.e. high temperature batteries) or safety management (i.e. lithium batteries) is necessary. Therefore the BMS must be optimized to the battery type and the application. In smaller systems the BMS is connected very closely to the battery or inside the battery and an operation without the BMS is not possible.