Early Fault Diagnosis Model Design of Reciprocating Compressor Valve Based on Multiclass Support Vector Machine and Decision Tree

Table 2

Feature vector of signal sample.

Fault type	Sample number	Crest coefficient	Kurtosis index	Skewness index	Effective value	Standard deviation	Fault number
Normal valve	1	0.0303	0.0029	0.0458	0.7609	0.79257	①
	2	0.1023	0.2164	0.3279	0.5665	0.9026	①
	3	0.0402	0.0367	0.1469	0.6367	0.9881	①
	4	0.6129	0.0294	0.3205	0.6137	0.8351	①
	5	0.1425	0.0561	0.3361	0.4821	0.7093	①
Valve fracture	1	0.0478	0.0463	0.0917	0.84221	0.00135	②
	2	0.6190	0.0637	0.3275	0.6347	0.9027	②
	3	0.4543	0.3219	0.3575	0.5342	0.9434	②
	4	0.2011	0.0676	0.3821	0.5038	0.9221	②
	5	0.3579	0.1041	0.3528	0.5674	0.8657	②
Valve plate wear	1	0.9581	0.6287	0.4292	0.5821	0.5238	③
	2	0.9893	0.7953	0.3453	0.6281	0.8671	③
	3	0.7859	0.6513	0.4217	0.6915	0.9877	③
	4	0.9139	0.7385	0.5446	0.6609	0.8592	③
	5	0.8768	0.6054	0.4573	0.6581	0.9107	③
Valve spring failure	1	0.3082	0.2787	0.2168	0.7635	0.4891	④
	2	0.1027	0.1669	0.2461	0.7623	0.9135	④
	3	0.4103	0.3373	0.1931	0.8938	0.9437	④
	4	0.3841	0.2673	0.2037	0.7589	0.9276	④
	5	0.3724	0.1976	0.2524	0.7095	0.9471	④