Performances in network-based synchronization depend on several related factors, including the instability of local clocks, the rate at which timing information is exchanged, and the accuracy of the resulting correction estimates. This paper analyzes these effects and their relationships, showing how these may affect the design of an IEEE 1588 Precision Time Protocol synchronization scheme. This paper introduces a state-variable clock model for which realistic parameters can be obtained for different kinds of clocks from experimental measurements of Allan variance plots. A Kalman-filter-based clock servo employing this model is developed, and a simulation analysis of the behavior of clock regulation and the effect of parameter variations on its performances is presented

This paper analyzes the effects on synchronization performances of the accuracy of skew and offset estimates, the short and long-term stability of the unregulated clock, the rate at which timing information is exchanged within the network. The work makes use of a simulated clock based on the state-variable model, for which realistic parameters are obtained from experimental measurements of Allan variance plots for different kinds of clocks

No abstract

Synchrophasor measurements, performed by phasor measurement units (PMUs), are becoming increasingly important for power system network monitoring. Synchrophasor standards define test signals for verification of PMU compliance, and set acceptance limits in each test condition for two performance classes (P and M). Several PMU algorithms have been proposed to deal with steady-state and dynamic operating conditions identified by the standard. Research and discussion arising from design, implementation, testing and characterization of PMUs evidenced that some disturbances, such as interharmonic interfering signals, can seriously degrade synchrophasor measurement accuracy. In this paper, a new compressive sensing (CS) approach is introduced and applied to synchrophasor measurements using a CS Taylor-Fourier (TF) multifrequency (CSTFM) model. The aim is to exploit, in a joint method, the properties of CS and the TF transform to identify the most relevant components of the signal, even under dynamic conditions, and to model them in the estimation procedure, thus limiting the impact of harmonic and interhamonic interferences. The CSTFM approach is verified using composite tests derived from the test conditions of the synchrophasor standard and simulation results are presented to show its potentialities

No abstract