This technical report details the development and experimental validation of a flexible laboratory setup designed to test battery systems and evaluate battery models for microgrid and economic analyses. The setup incorporates a bidirectional DC/DC converter, battery system, measurement equipment, a Texas Instruments digital signal processor (DSP), and a PC-based human-machine interface (HMI). The DSP enables real-time current and voltage control, supervisory charger and discharger operation, and comprehensive battery-state monitoring, including state of charge (SOC), state of energy (SOE), capacity, efficiency, state of health (SOH), and open-circuit-voltage (OCV)-based SOC correction. The HMI facilitates system control, automated experimental protocols, real-time visualization, and experimental data acquisition.
Experimental verification compares battery-state quantities calculated in real time by the DSP with independent offline calculations from recorded measurements. The results confirm the consistency of the implemented SOC and SOE calculations and demonstrate the importance of accounting for battery losses. In particular, neglecting energy efficiency produces cumulative SOE deviation during repeated charge–discharge operation, whereas introducing an experimentally determined efficiency correction substantially improves the stability and repeatability of the estimated energy state. The SOC experiments also show how OCV-based correction compensates for accumulated estimation errors during battery rest and how estimated battery capacity and SOH affect long-term SOC accuracy.
The experimental results verify that the developed laboratory setup provides a suitable platform for battery characterisation, control-algorithm verification, parameter identification, and evaluation of simplified battery models. Accurate representation of SOC, SOE, usable capacity, efficiency, and SOH is particularly important when the resulting battery model is used for microgrid simulation and economic optimisation, since errors in the estimated available battery energy can directly affect calculated grid-energy exchange, battery utilisation, operating costs, and optimal storage capacity.