Subspace based widely linear frequency estimation in sustainable energy systems

Frequency estimation has become a great challenge in power systems for sustainable provision of energy. Renewable energy systems like wind turbines and solar energy offer variable power output and are highly unpredictable. Smart grid systems, which incorporate several renewable energy resources, have varying topology and frequent switching between the main grid and micro-grids take place due to supply-demand mismatch. This results in frequent variation in system frequency. Thus, effective frequency estimation is necessary to control the future state of the power system. In this paper, we have presented subspace based techniques for real-time frequency estimation in balanced and unbalanced power systems by exploiting widely linear modeling. Comparison has been performed with recently developed CLMS and ACLMS techniques. It was observed that subspace based methods demonstrate significantly better performance compared to CLMS and ACLMS techniques in terms of accuracy and time resolution.