Distributed nonlinear state estimation using adaptive penalty parameters with load characteristics in the Electricity Reliability Council of Texas

Abstract As the industries transit towards industrial integration and informatization, the many advantages from interdisciplinary collaborations come with added technical challenge, especially in large scale and complex systems. Different from typical objects, the interconnected power system is the largest system ever built in industrialized world. Since the development of Power System State Estimation (PSSE), it has predominantly been a centralized process that relies on consistent measurement data availability. In a centralized architecture, a single point of failure can impact the entire system. While in distributed topology, the damage could be decreased with exchanging information between neighboring sub-systems. In other fields, distributed architectures have been widely used to avoid this issue, however shallow number of works are reported in PSSE literature. This paper presents a distributed nonlinear PSSE innovation based model that uses an adaptive penalty parameter to improve the convergence and accuracy of the PSSE output such as bus voltage and bus phase. The alternating direction method of multipliers is modified and used to optimize the distributed PSSE while an innovation-based nonlinear model is used to represent the sub-areas composed measurement error. The distributed PSSE algorithm is tested on the IEEE-14 and 118-bus systems using load characteristics from the Electricity Reliability Council of Texas (ERCOT). Numerical results show that the penalty parameter successfully adapts to optimal condition and the objective function has better performance compared to state-of-the-art models after convergence. Easy-to-implement model towards industrialization, built on the weighted least squares (WLS) solution, without hard-to-design parameters, highlight potential aspects for real-life implementation.