Application of Fuzzy Set Pair Analysis in Poor Shunting Early Warning for a Track Circuit

Abstract

Aiming at the problem of poor shunting failures in track circuit which seriously affect transportation safety, this paper proposes an early warning strategy based on fuzzy set pair analysis. For an example of 25 Hz electronic phase detecting track circuit, main influence factors are taken as early warning indexes through the failure mechanism analysis. In consideration of the effect of man-induced subjective factors during the index weight decision by Analytic Hierarchy Process, the grey relational analysis is applied to correct the weight to make the results more reasonable. The concept of fuzzy connection degree is applied to determine the value of discrepancy coefficient in set pair analysis theory. Measured samples and warning grades are used to compose a set pair and the connection degrees of the measured samples and the warning grades are calculated and ranked. Then through the farther same, difference and antagonism set pair analysis the poor shunting early warning for track circuit is achieved. The results and comparisons with other methods show that the fuzzy set pair analysis method is feasible and can provide a scientific basis for track circuit site maintenance.