Reducing the delay of phasor estimates under power system oscillations

Delay and computational complexity reduction on the phasor estimates calculated under power system oscillations is the main objective of this paper. Different approximations to the raised cosine (RC) filter lead to a reduction in delay and computational complexity of the phasor estimates. As for finite impulse response (FIR) filters, the minimum-phase implementation of the RC linear-phase FIR filter reduces by a half the delay of the phasor estimates, without introducing significant phase distortion and with shorter transients. This improvement in speed provides faster estimates for real-time monitoring and control applications. On the other hand, infinite impulse response (IIR) filters reduce the computational complexity of the estimates by a factor of ten, but without significant delay reduction, introducing a higher distortion, and with longer transient behavior. This solution could be useful for implementations in which computational complexity is a matter of serious concern. In this way, the paper provides a wide range of solutions for those who need more accurate, faster or simpler estimates.

I. Introduction

Estimation of phasors under power system oscillations is an interesting engineering challenge. Modern and future power system monitoring and control applications rely on fast and accurate phasor measurements to ensuring maximum power flow while preserving high level of power quality and security. In the design of a centralized Remote Feedback Controller (RFC), employing Synchronized Phasor Measurements (SPM), the delay between the output of the system and the controller input is a crucial factor to guarantee a robust controller [5], [6]. A robust controller is essential to mitigate the inter-area oscillations and enhance the dynamic performance of a large power system. The time delay of the output signal in the feedback loop not only depends on the data transmission system, but also on the delay introduced by the digital filters used in the Phasor Measurement Units (PMU) to estimate the phasors. The amount of delay due to the data transmission depends on the type of communication link used by the controller. The average communication delay for satellite communication links can be as long as 120ms [5], for microwave channels 40ms, but it can be as short as tens of milliseconds for fiber optic links. In fact, 4ms are defined as the Ethernet on-time message delivery capability under the Utility Communication Architecture (UCA-IEC61850). So, for fast data transmission systems, the delay between the output signal and the estimated phasors is the predominant component, and therefore a reduction in this delay becomes crucial for modern control applications.