Hot rail wheel bearing detection system and method
Hot rail car bearings or wheels are identified by sensing an infrared radiation from the hot surface and determining whether features of the sensed signals are indicative of hot rail car surfaces. The features may include the signals themselves, with distances or correlations being established between the signals and signals of known hot bearings or wheels. The features may be analyzed in a feature or decision space, with boundaries being established that identify hot bearings or wheels, or that establish false positive features or noise. The identification may also be implemented as a matched filter.
1. A method for detecting a moving hot bearing or wheel of a rail car comprising:
(a) establishing features of sensor signals in a multidimensional decision space;
(b) establishing a multidimensional threshold for discriminating between abnormally hot bearings or wheels and bearings or wheels that are not abnormally hot, wherein the multidimensional threshold is based on an average power of the sensor signals and a normalized fourth moment of the sensor signals;
(c) receiving signals representative of temperature of the moving bearing or wheel; and
(d) determining whether the bearing or wheel is likely hotter than desired based upon the multidimensional threshold and the signals.
2. The method of claim 1 , wherein establishing the multidimensional threshold for discriminating between abnormally hot bearings or wheels and bearings or wheels that are not abnormally hot includes establishing the multidimensional threshold based on analysis of clustering of the features in the multidimensional decision space.
3. The method of claim 1 , wherein establishing the multidimensional threshold for discriminating between abnormally hot bearings or wheels and bearings or wheels includes identifying a region in the multidimensional decision space in which the features are indicative that a bearing or wheel is hotter than desired.
4. The method of claim 3 , wherein establishing the multidimensional threshold for discriminating between abnormally hot bearings or wheels and bearings or wheels includes identifying a region in the multidimensional decision space in which the features are not indicative that a bearing or wheel is hotter than desired.
5. The method of claim 3 , further comprising adjusting a boundary of the region.
6. The method of claim 5 , wherein the decision boundary is adjusted based upon a FIFO analysis of decisions.
7. The method of claim 1 , further including determining a set of features indicative that a bearing or wheel is hotter than desired, and determining a distance between sampled sensor signals and the set of features.
8. The method of claim 1 , further including determining a set of features indicative that a bearing or wheel is hotter than desired, and determining a correspondence between sampled sensor signals and the set of features.
9. The method of claim 1 , further including establishing a matched filter having an impulse response that provides an output indicative that a bearing or wheel is hotter than desired.
10. The method of claim 1 , wherein the features include at least one of signal amplitude, signal persistence at an elevated level, a waveform shape, and average power.
11. The method of claim 1 , wherein establishing the multidimensional threshold includes determining the multidimensional threshold based on at least one of signal amplitude, duration of the signal at an elevated level, the presence of peaks in the signal, average power, and known false positive patterns.
12. The method of claim 1 , wherein the features include sampled sensor signals.
13. A method for detecting a moving hot bearing or wheel of a rail car comprising:
(a) establishing features of sensor signals in a multidimensional decision space;
(b) identifying a multidimensional region in the multidimensional decision space in which the features are indicative that a bearing or wheel is hotter than desired, including identifying a multidimensional decision threshold for discriminating between abnormally hot bearings or wheels and bearings or wheels that are not abnormally hot, wherein the multidimensional decision threshold is based on an average power of the sensor signals and a normalized fourth moment of the sensor signals;
(c) receiving signals representative of temperature of the moving bearing or wheel; and
(d) determining whether the bearing or wheel is likely hotter than desired based upon the multidimensional decision threshold and the signals.
14. The method of claim 13 , wherein step (b) includes identifying the multidimensional decision threshold by analysis of clustering of the features in the multidimensional decision space.
15. The method of claim 13 , further including identifying a region in the multidimensional decision space in which the features are not indicative that a bearing or wheel is hotter than desired.
16. The method of claim 13 , further comprising adjusting a boundary of the region.
17. The method of claim 16 , wherein the decision boundary is adjusted based upon a FIFO analysis of decisions.
18. A system for detecting a moving hot bearing or wheel of a rail car comprising:
a sensor configured to detect radiation from a moving hot bearing or wheel and to generate a signal representative of the radiation; and
processing circuitry configured to receive signals from the sensors and to determine whether the bearing or wheel is likely hotter than desired based upon a relationship between the features in a multidimensional decision space and a multidimensional threshold, wherein the multidimensional threshold is based on an average power of the sensor signals and a normalized fourth moment of the sensor signals, and the features permit discriminating between abnormally hot bearings or wheels and bearings or wheels that are not abnormally hot.
19. The system of claim 18 , wherein the processing circuitry is configured to establish the relationship between the features in the multidimensional decision space.
20. The system of claim 18 , wherein the features include at least one of signal amplitude, signal persistence at an elevated level, a waveform shape, and average power.