Performance comparison of distribution system state estimation methods

Because of the characteristics associated with distribution network and scarcity of real time measurements, state estimation (SE) methods used in transmission system are not directly applicable to the distribution network. As more sensors are being employed in smart distribution networks, the amount of quasi real time data gathered by various automation systems is increasing. For customer side automations, advanced metering infrastructures (AMIs) are deployed by many utilities. AMI data can be utilized in distribution management system to establish more accurate load models. Acknowledging the uniqueness of distribution network characteristics and requirements for on-line applications, this paper presents a comparative study of three major weighted least square based distribution system state estimation methods, i.e. node voltage polar method, node voltage rectangular method, and branch current rectangular method. Performances are compared using indices that are relevant to the requirements for smart distribution grid applications, i.e. accuracy level of voltage and load estimates, numbers of iterations and execution time required under various measurement combinations and network parameters disparity.