Fault Line Detection of Leakage Protection System of Mine Based on Rough Sets and Support Vector Machine

Abstract

For the complexity and multiformity of leakage faults of mine, the fault information has the uncertainty. With the capability of the solving the uncertain problem, Rough Sets theory can determine the fault timely and accurately. The paper adopted Rough Sets to extract the attribute characteristics of leakage fault signal and then to build the decision table. As the training sample of Support Vector Machine, the criterion by decision rules between the leakage fault signals and line detection method to detect fault line accurately can be gotten. The test results show that adopting the Rough Sets and Support Vector Machine to detect fault line is simple, efficient and good robustness.