Measurement-based modelling of composite load using genetic algorithm

Abstract One of the major issues in simulation and control of power system dynamics is load modelling. More accurate load models in power system stability analysis increases the accuracy of simulation results. If inappropriate model is used for the load, the obtained results may contain a high degree of error. In majority of analysis, the loads are usually considered as a constant impedance element. Whereas, such a model is not only accountable for the stability analysis of power system but also may sometimes lead to opposite results. Due to the variation of the load and also the variation of the composition of the load components, it would be difficult to provide a fixed model for electrical loads similar to those of other elements of the power system. A method for modelling the power system loads via genetic algorithm is presented in this paper. This methodology is performed based on the composite load model. In order to get an accurate load model, several scenarios are considered. The particular method of this paper is that after obtaining the load model parameters corresponding to each of the scenarios, various values obtained for the parameters are averaged. Finally, the validity of the obtained parameters is testified with some other scenarios. The results reported in this paper indicate that the existing load models satisfactorily describe the actual behaviour of the physical load and can be reliably estimated using the identification techniques presented herein.