Research on method of main reducer assembly quality evaluation based on K-L transform and support vector machine

This paper proposes a quality evaluation method for main reducer assembly based on K-L transform and support vector machine hybrid algorithm. The high-dimensional sample data collected from main reducer are compressed into low-dimensional independent eigenvector by K-L transform, and the classifier designed by support vector machine completes quality evaluation. Experimental results show that the method of combination eigenvector decomposed by K-L transform with SVM can evaluate quality of the main reducer assembly effectively and accurately. This method provides a new approach to the intelligent development of time domain analysis of vibration signal diagnosis.