Method and system for health assessment of a track circuit and/or of a track section

1. Method for health assessment of a track circuit and/or of a track section, the track circuit being configured for detecting the presence of a vehicle on the track section between a transmitting end and a receiving end of the track circuit, the transmitting end and the receiving end being associated to respective ends of the track section, the method being implemented by an electronic device, and comprising:

obtaining, from a sensor device placed near the receiving end, samples of an electrical parameter of electric signal pulses transmitted between the transmitting end and the receiving end of the track circuit, forming a temporal series of received samples,

processing the temporal series of received samples to obtain estimated pulse features of the received samples,

applying an automatic clustering algorithm on the estimated pulse features of the received samples to separate the received samples in a predetermined number of clusters,

selecting one of said clusters and determining, for the selected cluster, a first peak value of the received samples classified within the selected cluster, and

computing a track circuit health indicator depending on the first peak value determined for the selected cluster.

2. The method of claim 1 , further comprising:

obtaining, from a sensor device placed near the transmitting end, samples of the electrical parameter of the electric signal pulses transmitted between the transmitting end and the receiving end of the track circuit, forming a temporal series of transmitted samples,

processing the temporal series of transmitted samples to obtain estimated pulse features of the transmitted samples,

applying the automatic clustering algorithm on the estimated pulse features of the transmitted samples to separate the transmitted samples in a predetermined number of clusters,

selecting one of said clusters and determining, for the selected cluster, a second peak value of the transmitted samples classified within the selected cluster,

wherein the track circuit health indicator further depends on the second peak value.

3. The method of claim 2 , wherein the electrical parameter is a current, the received samples are received current samples, the transmitted samples are transmitted current samples, the first peak value is a first peak current value and the second peak value is a second peak current value.

4. The method of claim 3 , wherein the track circuit health indicator is either:

a ballast resistance, expressed in Ohms per thousand feet and computed by the following formula:

R

ballast

=

TL

×

(

V

tx

+

V

rx

)

2

×

1000

×

(

I

tx

-

I

rx

)

where TL is the track section length, V rx is a first voltage value of the electric signal at the receiving end, V tx a second voltage value of the electric signal at the transmitting end, I rx the first peak current value and I tx , is the second peak current value; or

a rail resistance, expressed in Ohms per feet and computed by the following formula:

R

rail

=

2

×

(

V

tx

-

V

rx

)

TL

×

(

I

tx

+

I

rx

)

where TL is the track section length, V rx is a first voltage value of the electric signal pulse at the receiving end, V tx is a second voltage value of the electric signal pulse at the transmitting end, I rx is the first peak current value and I tx , is the second peak current value.

5. The method of claim 1 , wherein the automatic clustering algorithm applied is a k-means algorithm.

6. The method of claim 5 , further comprising, before applying the k-means algorithm, automatically computing a number K of clusters to use for classifying the temporal series of samples.

7. The method of claim 5 , wherein the automatic clustering algorithm comprises:

computing a P-dimensional feature vector associated to each sample or sub-group of samples of the temporal series of samples.

8. The method of claim 7 , wherein the automatic clustering algorithm further comprises:

a) obtaining K centroids, each centroid being associated to a cluster,

b) computing a distance between each feature vector and each centroid, and assigning the feature vector to a cluster associated to the closest centroid according to the distance computed;

c) for each cluster, obtaining an updated centroid computed as a mean value of all feature vectors assigned to the cluster associated to the centroid,

d) repeating steps b) and c) until a stop criterion is met.

9. The method of claim 1 , further comprising obtaining a first voltage value of the electric signal pulse at the receiving end and a second voltage value of the electric signal pulse at the transmitting end.

10. The method of claim 9 , wherein the track circuit health indicator is a ballast resistance and/or a rail resistance.

11. System for health assessment of a track circuit and/or of a track section, the track circuit being configured for detecting the presence of a vehicle on the track section between a transmitting end and a receiving end of the track circuit, the transmitting end and the receiving end being associated to respective ends of the track section, the system comprising an electronic device comprising at least one processor configured to:

obtain, from a sensor device placed near the receiving end, samples of an electrical parameter of electric signal pulses transmitted between the transmitting end and the receiving end of the track circuit, forming a temporal series of received samples,

process the temporal series of received samples to obtain estimated pulse features of the received samples,

apply an automatic clustering algorithm on the estimated pulse features of the received samples to separate the received samples in a predetermined number of clusters,

select one of said clusters and determine, for the selected cluster, a first peak value of the received samples classified within the selected cluster, and

compute a track circuit health indicator depending on the first peak value determined for the selected cluster.

12. The system of claim 11 , wherein the processor is further configured to:

obtain, from a sensor device placed near the transmitting end, samples of the electrical parameter of the electric signal pulses transmitted between the transmitting end and the receiving end of the track circuit, forming a temporal series of transmitted samples,

process the temporal series of transmitted samples to obtain estimated pulse features of the transmitted samples,

apply the automatic clustering algorithm on the estimated pulse features of the transmitted samples to separate the transmitted samples in a predetermined number of clusters,

select one of said clusters and determine, for the selected cluster, a second peak value of the transmitted samples classified within the selected cluster, wherein the track circuit health indicator further depends on the second peak value.

13. The system of claim 12 , wherein the electrical parameter is a current, the received samples are received current samples, the transmitted samples are transmitted current samples, the first peak value is a first peak current value and the second peak value is a second peak current value.