State identification of power-shift steering transmission based on hypersphere support vector machine

An evaluation model based on the hypersphere support vector machine for power-shift steering transmission (PSST) was developed on the basis of spectrometric oil analysis data. The pretreatment of spectrometric oil analysis data was made and the principal components were studied using principal component analysis method. The influences of parameter variation and abnormal samples on the performance of hypersphere support vector machine model were studied. It has been proved that this model is accurate in evaluating the operation state of PSST.