The traditional metering error assessment method of CVT has some limitations. In this paper

a kind of adaptive error assessment method based on WMPCA is proposed. This method adopts the moving window principle and PCA method to achieve the update of the statistical threshold of metering error Q and to compare the assessment results of traditional PCA and WMPCA. It is shown by the study that in the longterm observation

WMPCA method is

compared to the traditional principal method

more applicable to assess CVT in long term operation and the assessment accuracy is higher. The WMPCA method is applicable for the extreme temperature environment with frequent variation. The method in this paper can provide some reference for the metering error assessment of 0.2 CVT in long-term operation.