A Bearing Fault Diagnosis Using a Support Vector Machine Optimised by the Self-Regulating Particle Swarm

Table 9

Accuracy of bearing fault recognition.

Case number	Optimization type	Support vector machine parameters			Average value of classification accuracy
		1	PSO optimized multiclass SVM + HE [5]					97.75%
2	SVM with parameter optimized by ICD [10]				97.91–100%
3	Improved PSO + LS-SVM [14]				89.50%
4	The proposed theory	0.5975	8.3853	50.2970	99.72%

The expression of accuracy is the ratio of the number of correctly classified sets to all sets (including the number of correctly classified sets and the number of incorrectly classified sets).