Application of mean-variance mapping optimization for parameter identification in real-time digital simulation

Abstract

This paper deals with the process of identifying the parameters of the dynamic equivalent (DE) load model of an active distribution system (ADN) simulated in RTDS using mean-variance mapping optimization (MVMO) algorithm. MVMO is an emerging variant of population-based, evolutionary optimization algorithm whose features include evolution of its solutions through a unique search mechanism within a normalized range of the sample space. Due to the prominent large-scale integration of DG in low and medium voltage networks, it is important to develop equivalent models that are suitable for representing the resulting active distribution network in dynamic studies of large power systems. This would significantly reduce the computational demands and simulation time. Moreover, only a defined portion of a system is usually studied, which means that the external system can be substituted with DE thereby allowing the detailed modelling of the focus area. The IEEE 34-Bus distribution system was modified and used as the reference network where measurement data were gathered for identification of the parameters of its developed DE. An optimization-enabled simulation involving MATLAB, which host the MVMO algorithm and RTDS, which simulates the models was established. The reactions of the detailed network and the DE were compared upon subjecting them to different disturbances in the retained system. The effectiveness of the MVMO algorithm in identifying DE parameters based on its unique mapping function is reflected through the results of the response comparison.