Adaptive PMU-based Distribution System State Estimation exploiting the Cloud-based IoT paradigm

TL;DR

This paper introduces an adaptive, cloud-based distribution system state estimation method that dynamically adjusts measurement and estimation rates based on system dynamics, improving efficiency and responsiveness.

Contribution

It proposes a novel adaptive DSSE framework leveraging virtualized PMUs and local decision algorithms for dynamic rate adjustment in a cloud-IoT environment.

Findings

Effective detection of system dynamics using voltage and ROCOF metrics.

Improved estimation accuracy during unexpected system changes.

Bandwidth-efficient data transmission in cloud-based architecture.

Abstract

This paper presents an adaptive Distribution System State Estimation (DSSE) which relies on a Cloud-based IoT paradigm. The methodology is adaptive in terms of the rate of execution of the estimation process which varies depending on the indications of the distributed measurement system. The system is composed, in particular, of Phasor Measurement Units (PMUs). PMUs are virtualized with respect to the physical devices and the corresponding virtualizing modules run in the communication network edge (i.e. closer to the physical objects). PMUs are set at a higher measurement rate, while the estimation process works at a given slower rate, for example once per second, in normal operative conditions. A local decision algorithm implemented in the virtualized module, monitors the measured quantities in order to detect and address possible unexpected dynamics. In particular, different metrics can be applied: the variations and the trend of variation of the rms voltage values, but also the Rate Of Change Of Frequency (ROCOF) of the monitored signals can be used to trigger rate variation in the DSSE.In case dynamics are detected, the measurement data is sent to the DSSE at higher rates and the estimation process runs consequently on a finer time scale. In the considered system only application level entities are located in the Cloud, thus allowing to obtain a bandwidth-efficient and smart data transmission. The results obtained on a 13-bus systems prove the goodness of the proposed methodologies.