Analysis and Processing of Electrical Fault Signals With Modified Adaptive Prony Method
Rongxiang Yuan · Proceedings of the CSEE · 2010
Prony method (PM) is unsuitable for processing the fault signals containing components with uncontinuous or abrupt changes. The adaptive PM (APM), by adaptively segmenting a signal, improves PM’s performance, but it takes fixed step to search the subsegments’ dividing points, so efficiency is low and execution time is long. This paper proposes a modified APM (MAPM) that introduces variable-step measure and greatly promotes the searching efficiency. Simulation results show that MAPM’s signal fitting and parameters estimation are of high accuracy, and the signal interpolation and prolongation based on it are adaptive while the traditional methods are not. A real-life fault signal is analyzed and processed by MAPM, again showing its superior performance and establi- shing its basis for extensive use in electrical engineering.