Fault Diagnosis for Compensating Capacitors of Jointless Track Circuit Based on Dynamic Time Warping

Table 5

Description of the diagnosis (classification) experiment based on SVM.

Item	Description
Vector dimension (the number of sampling points in a curve)	156
The number of classes	17
Training data set	1 vector per class (without noise)
Testing data set	50 vectors per class (with ±0.005 V uniformly distributed measurement noise and smoothing filtering with 3 points)
SVM tool	LIBSVM [29]
Scaling scheme [30]	Vectors in training data set are linearly consistently scaled to the range [0,1], and testing data set has the same scaling proportion as the training data set
SVM type	C-SVM [29]
Kernel type for SVM	Linear kernel
Penalty factor of the error term in C-SVM	𝐶=1
Testing result	Classification accuracy = 100%