On-line Bushing Condition Monitoring

摘要

An on-line bushing condition monitoring framework is presented, the framework is able to adapt as new data are introduced. Furthermore, it can accommodate new classes that are introduced by incoming data. The framework is implemented using an incremental learning algorithm that uses MLP as a weak Learner. The performance of the on-line bushing condition monitoring is compared to that of an MLP trained off-line. The proposed framework is able to adapt as new data are introduced and is able to accommodate new classes. The testing results improved from 67.5% to 95.8% as new data are introduced and the testing results improved from 60% to 95.3% as new classes are introduced. On average the confidence value of the framework on its decisions is 0.92.