Impact of Measurement Accuracy on Fault Detection Obtained with Compressive Sensing

The paper presents a study focused on the modelization of measurement uncertainty in fault location methods and on the evaluation of its impact on the performance. In particular, algorithms based on a compressive sensing methodology are considered. Fault detection applications are very important in power systems control and protection, nevertheless, most of the methods proposed in the literature do not appropriately consider the overall accuracy of the measurement chain involved in the process both in the definition and testing. In this context, this paper faces the problem of the accuracy actually available in the measurement devices working under the dynamic conditions, which occur during faults, and its impact on fault location applications. The paper will discuss, by means of the results obtained on a small distribution network, how the operating conditions during fault and their effects on the accuracy performance of both instrument transformers and measurement devices affect the results.