Implementing Fuzzy Reasoning Petri-Nets for Fault Section Estimation

Fuzzy reasoning Petri-nets (PNs) is a promising technique to tackle the complexities of power system fault section estimation. This paper addresses several key issues in implementing fuzzy reasoning PNs for fault section estimation, which include optimal design of structure of diagnosis models to avoid large matrix size, utilization of fuzzy logic parameters to effectively handle uncertainties, realization of matrix execution algorithm to achieve parallel reasoning and adaptability, and integration of more reliable input data to enhance estimation accuracy. Case studies are presented to demonstrate the estimation capability under complex scenarios. An implementation solution residing in a control center is proposed.

I. Introduction

A power system is composed of lots of sections such as generators, transformers, bus bars and transmission lines. These sections are protected by protection systems comprised of protective relays, circuit breakers, and communication equipment. When a fault occurs on a certain section, the protection devices will reach certain statuses accordingly. To identify the faulted section of a power system based on a set of observed statuses of protection devices is called fault section estimation. This is a vital task for system operators because it provides the most fundamental information for restorative actions. The task is stressful, time consuming, and the accuracy is restricted when multiple faults, failures of protection devices, and false data are involved. When all mix up, a large number of scenarios can be hypothesized and the possibility of each scenario needs to be examined. The complexity of fault section estimation increases significantly.