Search

This paper presents the wideband system identification (WSI) technique, i.e., an online method to identify power impedances over a wide frequency range from which the corresponding parametric impedance can be calculated online as well. The WSI technique exploits an existing custom 25-kW power electronic converter on the top of its power conversion function, which serves as power amplifier of an existing power hardware-in-the-loop (PHiL) simulation setup. The PHiL simulation technique allows connecting a real device under test (DUT) with the real-time simulated rest of system (ROS) at power level. An interface algorithm (IA) on simulation side and a power amplifier (the 25-kW power electronic converter) connect ROS and DUT. This paper shows the impact of the uncertainties in the WSI chain on the accuracy of the impedance identification and highlights how the WSI technique can be combined with the damping impedance method IA to improve both accuracy and stability of the PHiL test bench. The application of the method is illustrated for the scenario of a PHiL test of a dc microgrid with a passive load. [ABSTRACT FROM PUBLISHER]

This paper proposes a control method to coordinate multiple energy storage systems (ESS) spread across an AC microgrid. Each storage unit controls its own local node voltage (i.e., where the storage is connected in the network) as well as a global node voltage (i.e., a remote location in the network) that is controlled by all ESSs together. The storage power output is controlled in a way that the node voltage is restored and maintained in case it drops or rises outside the predefined limits. This can occur due to disturbances, such as load changes, loss of distributed generators (DGs), or power limitations in the DGs. In order to determine the power reference for each storage unit, a state of charge (SoC) based adaptive deadband droop control is used. In this decentralized control, the storage power output is inversely proportional to its SoC. By using this approach, the ESS with higher SoC will inject more power whereas storage units with lower SoC will inject less power, leading to more balanced SoC levels among the ESSs. The controller is tested in a microgrid in both island- and grid-connected modes in different scenarios. The microgrid includes several DGs that operate in different control modes depending on the microgrid operation. The storage unit configurations differ in their response time and capacity. This has significant implication in dynamic response of the storage units which is scrutinized in this paper with time-domain simulation. [ABSTRACT FROM PUBLISHER]