WO2005015326A1 - System for monitoring the working condition of an installation - Google Patents
System for monitoring the working condition of an installation Download PDFInfo
- Publication number
- WO2005015326A1 WO2005015326A1 PCT/GB2004/003322 GB2004003322W WO2005015326A1 WO 2005015326 A1 WO2005015326 A1 WO 2005015326A1 GB 2004003322 W GB2004003322 W GB 2004003322W WO 2005015326 A1 WO2005015326 A1 WO 2005015326A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- data
- installation
- normality
- signal
- feature detectors
- Prior art date
Links
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B61—RAILWAYS
- B61L—GUIDING RAILWAY TRAFFIC; ENSURING THE SAFETY OF RAILWAY TRAFFIC
- B61L23/00—Control, warning, or like safety means along the route or between vehicles or vehicle trains
- B61L23/04—Control, warning, or like safety means along the route or between vehicles or vehicle trains for monitoring the mechanical state of the route
- B61L23/041—Obstacle detection
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B61—RAILWAYS
- B61K—AUXILIARY EQUIPMENT SPECIALLY ADAPTED FOR RAILWAYS, NOT OTHERWISE PROVIDED FOR
- B61K9/00—Railway vehicle profile gauges; Detecting or indicating overheating of components; Apparatus on locomotives or cars to indicate bad track sections; General design of track recording vehicles
- B61K9/08—Measuring installations for surveying permanent way
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B61—RAILWAYS
- B61L—GUIDING RAILWAY TRAFFIC; ENSURING THE SAFETY OF RAILWAY TRAFFIC
- B61L23/00—Control, warning, or like safety means along the route or between vehicles or vehicle trains
- B61L23/04—Control, warning, or like safety means along the route or between vehicles or vehicle trains for monitoring the mechanical state of the route
- B61L23/042—Track changes detection
- B61L23/044—Broken rails
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B61—RAILWAYS
- B61L—GUIDING RAILWAY TRAFFIC; ENSURING THE SAFETY OF RAILWAY TRAFFIC
- B61L23/00—Control, warning, or like safety means along the route or between vehicles or vehicle trains
- B61L23/04—Control, warning, or like safety means along the route or between vehicles or vehicle trains for monitoring the mechanical state of the route
- B61L23/042—Track changes detection
- B61L23/045—Rail wear
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B61—RAILWAYS
- B61L—GUIDING RAILWAY TRAFFIC; ENSURING THE SAFETY OF RAILWAY TRAFFIC
- B61L23/00—Control, warning, or like safety means along the route or between vehicles or vehicle trains
- B61L23/04—Control, warning, or like safety means along the route or between vehicles or vehicle trains for monitoring the mechanical state of the route
- B61L23/042—Track changes detection
- B61L23/047—Track or rail movements
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B23/00—Testing or monitoring of control systems or parts thereof
- G05B23/02—Electric testing or monitoring
- G05B23/0205—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults
- G05B23/0218—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterised by the fault detection method dealing with either existing or incipient faults
- G05B23/0243—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterised by the fault detection method dealing with either existing or incipient faults model based detection method, e.g. first-principles knowledge model
- G05B23/0254—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterised by the fault detection method dealing with either existing or incipient faults model based detection method, e.g. first-principles knowledge model based on a quantitative model, e.g. mathematical relationships between inputs and outputs; functions: observer, Kalman filter, residual calculation, Neural Networks
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B61—RAILWAYS
- B61L—GUIDING RAILWAY TRAFFIC; ENSURING THE SAFETY OF RAILWAY TRAFFIC
- B61L2205/00—Communication or navigation systems for railway traffic
- B61L2205/04—Satellite based navigation systems, e.g. GPS
Definitions
- the present invention relates to a system for monitoring the condition of an installation, in particular to allow fault detection in geographically extensive installations such as railway lines, roads, pipelines, boundary security devices, and also systems and installations where monitoring of faults or adverse conditions is safety critical, such as elevators and escalators.
- the present invention provides a system condition monitor comprising: a data processor having: an input for receiving a signal from an installation/system; a plurality of feature detectors, each for monitoring a different aspect of said signal, each of said feature detectors comprising a model of normality for said aspect, a comparator for comparing said aspect of the received signal against the model of normality and providing an output indicative of a departure in said aspect of said received signal from said model of normality; and a display indicative of the condition of said installation/system based on said outputs from the feature detectors.
- the model of normality may be specific to each feature detector. It may comprise a threshold or multi-dimensional signature envelope against which the aspect of the received signal being monitored by that feature detector is compared. Preferably the models of normality are learnt from training data which include only normal examples of the signal.
- the different feature detectors may use different methods for detecting departures from normality. This means that a method appropriate to each aspect being monitored can be chosen.
- One or more of the feature detectors may have a model for specific fault characteristics. Such characteristics may be learnt by monitoring different faults after they have been detected as departures from normality, identified and analysed.
- the condition of the installation as detected by the feature detectors may be tracked over time.
- the feature detectors can fuse the sensed data with a time signal so that the change of the particular aspect they are monitoring can be detected. This is useful for planning when maintenance or repair becomes necessary, ie in a predictive maintenance program.
- the signal monitored is preferably "primary data” consisting of sensed condition indicative data, such as vibration or acoustic data, but the feature detectors may also be provided with “secondary data” for fusion with the primary data, such secondary data being some parameter or property of the system, for example, the location from which the signal is derived, the identity of whatever is providing the signal, the ambient conditions, or the current performance of the installation.
- the feature detectors may use data fusion and filtering techniques to synchronise the data, and the fused data is subjected to specific signal processing techniques in each feature detector including, for example, heuristic, statistics, fuzzy logic, artificial intelligence, neural network and/or expert systems.
- the signal may be derived from a mobile unit which travels along the installation, for example in the case of the installation being a railway line, the signals may be derived from a railway vehicle carrying vibration or acoustic sensors. These may be on the existing rolling stock used in normal operation of the network, or special purpose vehicles.
- the feature detectors may comprise models of normality for specific track features such as points, bridges, junctions, tunnels, passing trains etc., or models of normality may be provided for specific routes through the network. Both methods may be combined to provide complete monitoring of the condition of the network.
- secondary data can include train identification, information on the train's speed, weight, engine torque, the location and direction of the train, and the ambient conditions, such as the temperature and other weather conditions. The monitor can also monitor the condition of the train.
- the sensors may be distributed along the pipeline to provide acoustic and/or vibration signals.
- Secondary data in this case can include operating parameters of the pipeline, such as pressure, temperature and flow rate.
- the invention is also applicable to other geographically extensive installations such as cables and boundary security devices (such as fences), as well as to other safety critical installations such as elevators and escalators.
- boundary security devices such as fences
- safety critical installations such as elevators and escalators.
- FIG. 1 is a block diagram illustrating an embodiment of the invention
- FIG. 2 illustrates the application of the embodiment of FIG. 1 to a railway monitoring system
- FIG. 3 illustrates the application of the embodiment shown in FIG. 1 to a pipeline monitoring system.
- FIG. 1 illustrates in block diagram form an embodiment of the installation monitoring system.
- primary data in this case vibration data is supplied from a number of vibration sensors 3 to a data processing unit 1.
- the data processing unit also receives secondary data as mentioned above from one or more providers 5.
- the primary data is provided to an input interface 7 and is subjected to pre-processing, for example, filtering and digitisation by the signal pre-processor 9. This pre-processing may be different for each detector.
- the data may be prepared in spectral or time-series format by this pre-processor.
- feature detectors 11 within the data processing unit 1 there are a number of feature detectors 11 each with its own data fusion and signal processing capability. Each feature detector 11 is designed to monitor an individual aspect of the input data for specific event characteristics.
- each feature detector 11 the primary and secondary data are fused by data fusion device 110 and is provided to a comparator 112 which compares the data to a model of normality.
- the model of normality may comprise a threshold or may be based on novelty detection compared to a model training unit 114.
- the model training unit 114 may be a trained model, or it may be a model which is continuously trained, ie uses current data to improve its existing model.
- the feature detectors provide outputs to a condition and alarm display 13 and to a more detailed diagnosis unit 15 which can relate specific detected features to particular faults (such as loose fishplate bolts on a railway line for example) and to a data storage device 17.
- the feature detectors can monitor the state of some particular aspect over time. By fusing the primary data with, for example, location and time data, the state of a particular track feature on a route, such as a set of points, can be logged every journey and any change over time can be recorded. The same is true for other features such as junctions and bridges etc. Also where a feature detector is provided to detect a particular fault condition, such as a cracked rail or missing fastener, for example, any change, such as a deterioration, in the fault can be monitored.
- a particular fault condition such as a cracked rail or missing fastener
- the feature detectors 11 may use a heuristic, statistics, fuzzy logic, artificial intelligence, neural network and/or expert system to analyse the data and provide the condition, alarm and diagnostic information.
- the model of the normality is learnt by providing data which is known to be normal. Clearly the broader the training data the better the model of normality will be. Anomalies in received data can then be determined through a combination of pre-determined thresholds including rates of change, N-dimensional signature envelopes (e.g. established through pre-trained neural networks) or on-line learned envelopes (manual or automatic).
- Novelty detection techniques may be employed within each feature detector.
- the principle of novelty detection offers an approach to the problem of identifying abnormalities which only requires the normal class to be defined.
- a model of normality is learnt by including normal examples only in the training data.
- Novelty detection techniques may include support vector machines, Gaussian (and other) mixture models, neuro-scales, generative topographic maps and clustering/nearest neighbour techniques, as well as non-linear dynamic modelling.
- the alarm and diagnostic information may be provided to both local and remote sources, and appropriate visualisation techniques may be used to display the data in a meaningful and easily understood way.
- FIG. 2 illustrates schematically an application of the system of FIG. 1 to railway monitoring.
- the system is included in trains travelling over the railway network. These need not be special railway monitoring vehicles, but instead can be ordinary passenger and freight trains.
- Primary data is provided by acoustic or vibration sensors 3 which may be in one or more of the units of the train. Suitable sensors include a piezoelectrical or piezoresistive accelerometer such as a Monitran (TM) Series 1120 sensor.
- Secondary data, such as position may be supplied from GPS device 205 and also operating parameters, such as engine torque, train speed and weight, train identity etc can be supplied to the data processing unit.
- Each of the feature detectors analyses a different aspect of the incoming data using a model of normality appropriate for its aspect.
- Feature detectors are provided for different track features, such as points, bridges, junctions, tunnels, passing trains etc.), and for specific routes through a network, the combination of these then forming the basis for monitoring the entire railway network. Extraordinary or anomalous events are detected and diagnosed allowing the railway line to be examined for the cause of the anomaly.
- FIG. 3 An alternative application of the system is illustrated schematically in FIG. 3 to the monitoring of pipelines.
- primary data by way of vibration or acoustic data is fed from vibration sensors of either the piezoelectric or piezoresistive type mechanically secured to the pipeline 300 at intervals along it.
- Secondary data such as the temperature, pressure, flow rate or the open/closed state of any valves is also supplied from the secondary data providers 5.
- feature detectors are again provided each with their own data fusion and signal processing techniques as described with reference to FIG. 1. Each detector is designed to monitor an individual aspect of the data for specific events. Examples of different events to be detected include leaks, third-party intervention (for example distortion or breakage of the pipeline, for example unauthorised tapping), or malfunction of pipeline equipment.
- the pipeline can be fitted with electromechanical (or other) vibration stimulation devices. This may be appropriate for pipelines which transport, for example, low flow liquids.
- a further alternative is to use only the secondary input data (ie. no vibration data).
- the data is still analysed by the feature detectors to detect anomalous situations, such as leaks for example.
- the location of such leaks can be then determined by further analysis, monitoring the transmission of pressure waves for example.
- the invention is also applicable to other safety critical installations. Escalators, elevators and moving walkways often form an integral part of the communicational route of enlarged buildings. It is important to maximise their availability by ensuring that routine maintenance is conducted appropriately, and also to monitor for potentially unsafe conditions.
- the invention may be applied to the monitoring of such systems by using vibration sensors to provide the primary data in the same way as the embodiments above, and secondary data by way of equipment identification and performance data (for example direction, speed, motor torque, ambient temperature/conditions etc.).
- equipment identification and performance data for example direction, speed, motor torque, ambient temperature/conditions etc.
- the feature detectors learn appropriate models of normality so that they can then diagnose the signal characteristics which indicate deterioration of the system. This allows maintenance to be appropriately timed, and for failing components to be diagnosed before failure occurs.
- the digital processing unit can control the installation to terminate its operation should an extreme safety critical situation arise.
Abstract
A system for monitoring the condition of installation such as railways, pipelines, escalators and elevators etc. comprises a data processor which includes a plurality of separate feature detectors, each for monitoring a specific aspect of data obtained from the installation. The feature detectors include a model of normality, which may be learned from training data sets, and then compare the input signals to the model of normality to detect departures from normality.
Description
SYSTEM FOR MONITORING THE WORKING CONDITION OF AN INSTALLATION
The present invention relates to a system for monitoring the condition of an installation, in particular to allow fault detection in geographically extensive installations such as railway lines, roads, pipelines, boundary security devices, and also systems and installations where monitoring of faults or adverse conditions is safety critical, such as elevators and escalators.
Systems monitoring, particularly for detection of faults or adverse conditions, is a common requirement and many different monitoring systems and arrangements exist for different applications. Typically some parameter of the system is monitored and compared to a pre-set threshold, with an alarm indicating when the threshold is exceeded. However, such pre-set thresholds are not always an appropriate way in which to monitor condition. The relationship between the sensed data and system condition may be complex. It can be difficult to decide an appropriate threshold, for example the same threshold may not be appropriate for all operating conditions of the system - monitoring vibration levels in a train without knowing the train speed does not yield meaningful data. Also, natural variability in systems which are installed for long term usage can result in either false alarms if the thresholds are set too low, or insufficient alarm, and consequently safety concerns, if the thresholds are set too high. Proper monitoring of the condition of an installation is also difficult where the installation extends over a large geographical area. For example, it is difficult to monitor the condition of an entire rail network, pipeline or boundary security device.
Some systems have been proposed for monitoring the condition of railway networks. In a railway or tram network, the railway and tram lines (and associated fixing plates and fixings) suffer from deterioration through fatigue, stress and damage which can ultimately result in failure and catastrophic incidents. For example, DE 19,858,937 and US 5,529,267 disclose systems in which sensors are mounted on railway lines to detect passing trains and to indicate possible abnormality in the conditions when the
train passes. However, such systems are inevitably only positioned at certain parts of the rail network and so cannot provide complete coverage. WO 01/86227 and GB 2,077,822 disclose railway vehicles which include measuring apparatus for measuring the position of the rails in the track. These provide, however, only limited information about the track and will only provide an alarm in preconceived circumstances. Another example of a geographically extensive installation which requires monitoring is a pipeline. Particularly in the case of the pipelines for transporting hazardous materials, there is a growing requirement for pipeline integrity monitoring (PIM) to provide adequate safety for the public and environment, as well as to safeguard against product and revenue loss. For example, corrosion is a threat to the integrity of older pipelines and unauthorised tapping of the pipeline to remove product can be a problem. However, there are many other reasons for the development of faults in the pipeline, such as fatigue, stress, hydrogen induction, material failure, and external influences. A large number of pipeline faults, for example, are caused by third party intervention (TPI), for example where construction work adjacent to a buried pipeline accidentally damages the pipeline.
There is therefore a need for an economical system for monitoring the condition of installations such as those mentioned above, but which provides increased flexibility in the type of conditions which it can detect. This will improve considerably the safety and security of such installations.
Accordingly the present invention provides a system condition monitor comprising: a data processor having: an input for receiving a signal from an installation/system; a plurality of feature detectors, each for monitoring a different aspect of said signal, each of said feature detectors comprising a model of normality for said aspect, a comparator for comparing said aspect of the received signal against the model of normality and providing an output indicative of a departure in said aspect of said received signal from said model of normality; and
a display indicative of the condition of said installation/system based on said outputs from the feature detectors.
The model of normality may be specific to each feature detector. It may comprise a threshold or multi-dimensional signature envelope against which the aspect of the received signal being monitored by that feature detector is compared. Preferably the models of normality are learnt from training data which include only normal examples of the signal. The different feature detectors may use different methods for detecting departures from normality. This means that a method appropriate to each aspect being monitored can be chosen.
One or more of the feature detectors may have a model for specific fault characteristics. Such characteristics may be learnt by monitoring different faults after they have been detected as departures from normality, identified and analysed.
The use of novelty detection to detect departure of the received signal from normality means that it is only necessary to define for each feature detector the normal situation. Thus it is not necessary to pre-conceive different fault conditions which are to be detected.
The condition of the installation as detected by the feature detectors may be tracked over time. The feature detectors can fuse the sensed data with a time signal so that the change of the particular aspect they are monitoring can be detected. This is useful for planning when maintenance or repair becomes necessary, ie in a predictive maintenance program.
The signal monitored is preferably "primary data" consisting of sensed condition indicative data, such as vibration or acoustic data, but the feature detectors may also be provided with "secondary data" for fusion with the primary data, such secondary data being some parameter or property of the system, for example, the location from
which the signal is derived, the identity of whatever is providing the signal, the ambient conditions, or the current performance of the installation.
The feature detectors may use data fusion and filtering techniques to synchronise the data, and the fused data is subjected to specific signal processing techniques in each feature detector including, for example, heuristic, statistics, fuzzy logic, artificial intelligence, neural network and/or expert systems.
The signal (the primary data) may be derived from a mobile unit which travels along the installation, for example in the case of the installation being a railway line, the signals may be derived from a railway vehicle carrying vibration or acoustic sensors. These may be on the existing rolling stock used in normal operation of the network, or special purpose vehicles. In this application the feature detectors may comprise models of normality for specific track features such as points, bridges, junctions, tunnels, passing trains etc., or models of normality may be provided for specific routes through the network. Both methods may be combined to provide complete monitoring of the condition of the network. In this application secondary data can include train identification, information on the train's speed, weight, engine torque, the location and direction of the train, and the ambient conditions, such as the temperature and other weather conditions. The monitor can also monitor the condition of the train.
In an alternative application, to a pipeline installation, the sensors may be distributed along the pipeline to provide acoustic and/or vibration signals. Secondary data in this case can include operating parameters of the pipeline, such as pressure, temperature and flow rate.
The invention is also applicable to other geographically extensive installations such as cables and boundary security devices (such as fences), as well as to other safety critical installations such as elevators and escalators.
The invention will be further described by way of example by reference to the accompanying drawings in which:-
FIG. 1 is a block diagram illustrating an embodiment of the invention; FIG. 2 illustrates the application of the embodiment of FIG. 1 to a railway monitoring system; and FIG. 3 illustrates the application of the embodiment shown in FIG. 1 to a pipeline monitoring system.
FIG. 1 illustrates in block diagram form an embodiment of the installation monitoring system. As shown in FIG. 1 primary data, in this case vibration data is supplied from a number of vibration sensors 3 to a data processing unit 1. The data processing unit also receives secondary data as mentioned above from one or more providers 5. The primary data is provided to an input interface 7 and is subjected to pre-processing, for example, filtering and digitisation by the signal pre-processor 9. This pre-processing may be different for each detector. The data may be prepared in spectral or time-series format by this pre-processor. Within the data processing unit 1 there are a number of feature detectors 11 each with its own data fusion and signal processing capability. Each feature detector 11 is designed to monitor an individual aspect of the input data for specific event characteristics. Within each feature detector 11 the primary and secondary data are fused by data fusion device 110 and is provided to a comparator 112 which compares the data to a model of normality. The model of normality may comprise a threshold or may be based on novelty detection compared to a model training unit 114. The model training unit 114 may be a trained model, or it may be a model which is continuously trained, ie uses current data to improve its existing model. Depending on the comparison of the received data with the model, the feature detectors provide outputs to a condition and alarm display 13 and to a more detailed diagnosis unit 15 which can relate specific detected features to particular faults (such as loose fishplate bolts on a railway line for example) and to a data storage device 17.
The feature detectors can monitor the state of some particular aspect over time. By fusing the primary data with, for example, location and time data, the state of a particular track feature on a route, such as a set of points, can be logged every journey and any change over time can be recorded. The same is true for other features such as junctions and bridges etc. Also where a feature detector is provided to detect a particular fault condition, such as a cracked rail or missing fastener, for example, any change, such as a deterioration, in the fault can be monitored.
The feature detectors 11 may use a heuristic, statistics, fuzzy logic, artificial intelligence, neural network and/or expert system to analyse the data and provide the condition, alarm and diagnostic information. The model of the normality is learnt by providing data which is known to be normal. Clearly the broader the training data the better the model of normality will be. Anomalies in received data can then be determined through a combination of pre-determined thresholds including rates of change, N-dimensional signature envelopes (e.g. established through pre-trained neural networks) or on-line learned envelopes (manual or automatic).
Novelty detection techniques may be employed within each feature detector. The principle of novelty detection offers an approach to the problem of identifying abnormalities which only requires the normal class to be defined. A model of normality is learnt by including normal examples only in the training data.
Abnormalities are then identified by testing for novelty against this description.
Novelty detection techniques may include support vector machines, Gaussian (and other) mixture models, neuro-scales, generative topographic maps and clustering/nearest neighbour techniques, as well as non-linear dynamic modelling.
This means that it is not necessary to preconceive or pre-analyse fault conditions, making the system much more flexible.
The alarm and diagnostic information may be provided to both local and remote sources, and appropriate visualisation techniques may be used to display the data in a
meaningful and easily understood way.
FIG. 2 illustrates schematically an application of the system of FIG. 1 to railway monitoring. As illustrated in FIG. 2A the system is included in trains travelling over the railway network. These need not be special railway monitoring vehicles, but instead can be ordinary passenger and freight trains. Primary data is provided by acoustic or vibration sensors 3 which may be in one or more of the units of the train. Suitable sensors include a piezoelectrical or piezoresistive accelerometer such as a Monitran (TM) Series 1120 sensor. Secondary data, such as position, may be supplied from GPS device 205 and also operating parameters, such as engine torque, train speed and weight, train identity etc can be supplied to the data processing unit. Each of the feature detectors analyses a different aspect of the incoming data using a model of normality appropriate for its aspect. Feature detectors are provided for different track features, such as points, bridges, junctions, tunnels, passing trains etc.), and for specific routes through a network, the combination of these then forming the basis for monitoring the entire railway network. Extraordinary or anomalous events are detected and diagnosed allowing the railway line to be examined for the cause of the anomaly.
Of course it may be that faults developing on the train itself cause anomalies, and thus the system is also applicable to monitoring of the condition of the train. For example, deterioration in the condition of the drive unit, bogeys or brakes etc. may cause a departure from the normal acoustic and/or vibration signals.
An alternative application of the system is illustrated schematically in FIG. 3 to the monitoring of pipelines. Again, primary data by way of vibration or acoustic data is fed from vibration sensors of either the piezoelectric or piezoresistive type mechanically secured to the pipeline 300 at intervals along it. Secondary data such as the temperature, pressure, flow rate or the open/closed state of any valves is also supplied from the secondary data providers 5. Within the data processing unit 1
several feature detectors are again provided each with their own data fusion and signal processing techniques as described with reference to FIG. 1. Each detector is designed to monitor an individual aspect of the data for specific events. Examples of different events to be detected include leaks, third-party intervention (for example distortion or breakage of the pipeline, for example unauthorised tapping), or malfunction of pipeline equipment.
As an alternative to relying upon the flow medium within the pipe to stimulate vibrations which are detected by sensors 3, the pipeline can be fitted with electromechanical (or other) vibration stimulation devices. This may be appropriate for pipelines which transport, for example, low flow liquids.
A further alternative is to use only the secondary input data (ie. no vibration data). The data is still analysed by the feature detectors to detect anomalous situations, such as leaks for example. The location of such leaks can be then determined by further analysis, monitoring the transmission of pressure waves for example.
Although in this application all of the sensors are shown connected to a single central data processing unit, it may be appropriate to have a number of such data processing units for different parts of the pipeline where the pipeline is very large.
The invention is also applicable to other safety critical installations. Escalators, elevators and moving walkways often form an integral part of the communicational route of enlarged buildings. It is important to maximise their availability by ensuring that routine maintenance is conducted appropriately, and also to monitor for potentially unsafe conditions. The invention may be applied to the monitoring of such systems by using vibration sensors to provide the primary data in the same way as the embodiments above, and secondary data by way of equipment identification and performance data (for example direction, speed, motor torque, ambient temperature/conditions etc.). In these applications the feature detectors learn
appropriate models of normality so that they can then diagnose the signal characteristics which indicate deterioration of the system. This allows maintenance to be appropriately timed, and for failing components to be diagnosed before failure occurs.
In all the embodiments above in addition to providing the condition and alarm information, the digital processing unit can control the installation to terminate its operation should an extreme safety critical situation arise.
Claims
1. A system for monitoring the condition of an installation, the system comprising: a data processor having: an input for receiving a signal from said installation; a plurality of feature detectors, each for monitoring a different aspect of said signal, each of said feature detectors comprising a model of normality for said aspect, a comparator for comparing said aspect of the received signal against the model of normality and providing an output indicative of a departure in said aspect of said received signal from said model of normality; and the system further comprising a display indicative of the condition of said installation based on said outputs from the feature detectors.
2. A system according to claim 1 wherein said model of normality is specific to each feature detector.
3. A system according to claim 1 or 2 wherein at least one of said models of normality comprises one of a threshold or a multi-dimensional signature envelope against which said aspect of the received signal is compared.
4. A system according to claim 1 , 2 or 3 wherein at least one of said models of normality is learnt from training data comprising normal examples of said signal.
5. A system according to claim 1, 2, 3 or 4 wherein said comparators in at least two of said feature detectors use different methods for detecting departure from said models of normality.
6. A system according to any one of the preceding claims wherein at least one of said comparators uses novelty detection to detect departure in said aspect of said received signal from said model of normality.
7. A system according to any one of the preceding claims wherein said signal comprises primary data consisting of sensed condition indicative data.
8. A system according to claim 7 wherein a plurality of different types of sensed condition indicative data are provided.
9. A system according to claim 7 or 8 wherein said sensed condition indicative data comprises vibration data received from a vibration sensor.
10. A system according to claim 7 or 8 wherein said sensed condition indicative data comprises acoustic data received from an acoustic sensor.
11. A system according to any one of the preceding claims wherein said data processor further comprises an input for receiving secondary data for fusion in said feature detectors with said signal.
12. A system according to claim 11 wherein said secondary data comprises data representing at least one of location, identity, ambient conditions or installation performance.
13. A system according to any one of the preceding claims wherein changes over time in the output of the feature detectors are tracked.
14. -A system according to claim 13 wherein said tracked changes over time form inputs to a predictive maintenance plan.
15. A system according to any one of the preceding claims further comprising a mobile unit for travelling along said installation, said mobile unit comprising at least one sensor for obtaining said signal from said installation.
16. A system according to any one of the preceding claims wherein said installation is a railway line.
17. A system according to claim 16 wherein a sensor for providing said signal is mounted on a railway vehicle.
18. A system according to claim 16 or 17 wherein said feature detectors comprise models of normality for specific track features.
19. A system according to claim 16, 17 or 18 wherein at least one of said feature detectors comprises a model of normality for a route through a railway network.
20. A system according to claim 16, 17, 17 or 18 wherein data comprising at least one of train identification, train speed, train weight, engine torque, ambient temperature, location and direction are input into said feature detectors.
21. A system according to any one of claims 1 to 14 wherein said installation is a pipeline.
22. A system according to claim 21 further comprising sensors distributed along said pipeline for providing said signal.
23. A system according to claim 21 or 22 wherein said signal is a vibration signal from the pipeline.
24. A system according to claim 21, 22 or 23 wherein data comprising at least one of pressure, temperature and flow rate are input into said feature detectors.
25. A system according to any one of claims 1 to 14 wherein said installation is an elevator.
26. A system according to any one of claims 1 to 14 wherein said installation is an escalator.
27. A system according to any one of claims 1 to 14 wherein said installation is a cable.
28. A system according to any one of claims 1 to 14 wherein said installation is a security device extending along a boundary.
29. A system constructed and arranged to operate substantially as hereinbefore described with reference to and as illustrated in the accompanying drawings.
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
GB0318339.9 | 2003-08-05 | ||
GB0318339A GB0318339D0 (en) | 2003-08-05 | 2003-08-05 | Installation condition monitoring system |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2005015326A1 true WO2005015326A1 (en) | 2005-02-17 |
Family
ID=27839673
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/GB2004/003322 WO2005015326A1 (en) | 2003-08-05 | 2004-08-02 | System for monitoring the working condition of an installation |
Country Status (2)
Country | Link |
---|---|
GB (1) | GB0318339D0 (en) |
WO (1) | WO2005015326A1 (en) |
Cited By (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2009067770A1 (en) * | 2007-06-12 | 2009-06-04 | Asel-Tech Technologia E Automação Ltda. | System for detecting leaks in single phase and multiphase fluid transport pipelines |
DE102008028264B3 (en) * | 2008-06-13 | 2009-12-17 | Knorr-Bremse Systeme für Schienenfahrzeuge GmbH | Method for monitoring at least one system parameter influencing the operating behavior of vehicles or vehicle trains |
WO2009153033A2 (en) * | 2008-06-20 | 2009-12-23 | Airbus Operations Gmbh | Aircraft conduit monitoring system and method |
WO2011160943A1 (en) * | 2010-06-21 | 2011-12-29 | Optimized Systems And Solutions Limited | Asset health monitoring using stable distributions for heavy-tailed data |
CN105518377A (en) * | 2014-10-31 | 2016-04-20 | 深圳市大疆创新科技有限公司 | Gas leakage processing method and apparatus, and air vehicle |
EP2064106B1 (en) | 2006-09-18 | 2016-06-15 | Bombardier Transportation GmbH | Diagnostic system and method for monitoring a rail system |
US10227215B2 (en) | 2016-08-24 | 2019-03-12 | Otis Elevator Company | Passenger conveyor and method for monitoring vibrations in a passenger conveyor |
CN110428109A (en) * | 2019-08-07 | 2019-11-08 | 佳都新太科技股份有限公司 | Subway shield door time between failures prediction model is established and prediction technique |
WO2020002018A1 (en) * | 2018-06-28 | 2020-01-02 | Konux Gmbh | System and method for traffic control in railways |
WO2020173636A1 (en) * | 2019-02-27 | 2020-09-03 | Siemens Mobility GmbH | Method for monitoring points of a railway track installation |
US10829344B2 (en) | 2017-07-06 | 2020-11-10 | Otis Elevator Company | Elevator sensor system calibration |
WO2021001219A1 (en) * | 2019-07-02 | 2021-01-07 | Konux Gmbh | Monitoring, predicting and maintaining the condition of railroad elements with digital twins |
CN112313139A (en) * | 2018-06-28 | 2021-02-02 | 科路实有限责任公司 | System and method for rail transit control |
US11014780B2 (en) | 2017-07-06 | 2021-05-25 | Otis Elevator Company | Elevator sensor calibration |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109033582A (en) * | 2018-07-12 | 2018-12-18 | 西安英特迈思信息科技有限公司 | A kind of bullet train multi parameter intallingent threshold value criterion |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5845230A (en) * | 1996-01-30 | 1998-12-01 | Skf Condition Monitoring | Apparatus and method for the remote monitoring of machine condition |
DE19852220A1 (en) * | 1998-11-12 | 2000-06-08 | Stn Atlas Elektronik Gmbh | Process for the detection of damage in rail traffic |
US20020077733A1 (en) * | 1999-06-15 | 2002-06-20 | Andian Technologies | Geometric track and track/vehicle analyzers and methods for controlling railroad systems |
WO2002047954A1 (en) * | 2000-12-12 | 2002-06-20 | Deutsche Bahn Ag | Method and device for monitoring the handling characteristics of rail vehicles and for diagnosing components of rail vehicles |
WO2004009422A1 (en) * | 2002-07-19 | 2004-01-29 | Aea Technology Plc | Assessment of railway track geometry |
-
2003
- 2003-08-05 GB GB0318339A patent/GB0318339D0/en not_active Ceased
-
2004
- 2004-08-02 WO PCT/GB2004/003322 patent/WO2005015326A1/en active Application Filing
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5845230A (en) * | 1996-01-30 | 1998-12-01 | Skf Condition Monitoring | Apparatus and method for the remote monitoring of machine condition |
DE19852220A1 (en) * | 1998-11-12 | 2000-06-08 | Stn Atlas Elektronik Gmbh | Process for the detection of damage in rail traffic |
US20020077733A1 (en) * | 1999-06-15 | 2002-06-20 | Andian Technologies | Geometric track and track/vehicle analyzers and methods for controlling railroad systems |
WO2002047954A1 (en) * | 2000-12-12 | 2002-06-20 | Deutsche Bahn Ag | Method and device for monitoring the handling characteristics of rail vehicles and for diagnosing components of rail vehicles |
WO2004009422A1 (en) * | 2002-07-19 | 2004-01-29 | Aea Technology Plc | Assessment of railway track geometry |
Cited By (23)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP2064106B1 (en) | 2006-09-18 | 2016-06-15 | Bombardier Transportation GmbH | Diagnostic system and method for monitoring a rail system |
WO2009067770A1 (en) * | 2007-06-12 | 2009-06-04 | Asel-Tech Technologia E Automação Ltda. | System for detecting leaks in single phase and multiphase fluid transport pipelines |
DE102008028264B3 (en) * | 2008-06-13 | 2009-12-17 | Knorr-Bremse Systeme für Schienenfahrzeuge GmbH | Method for monitoring at least one system parameter influencing the operating behavior of vehicles or vehicle trains |
US8577522B2 (en) | 2008-06-13 | 2013-11-05 | Knorr-Bremse Systeme Fur Schienenfahrzeuge Gmbh | Method for monitoring at least one system parameter which influences the operating behaviour of vehicles or trains of vehicles |
WO2009153033A2 (en) * | 2008-06-20 | 2009-12-23 | Airbus Operations Gmbh | Aircraft conduit monitoring system and method |
WO2009153033A3 (en) * | 2008-06-20 | 2010-04-29 | Airbus Operations Gmbh | Aircraft conduit monitoring system and method |
CN102066194A (en) * | 2008-06-20 | 2011-05-18 | 空中客车作业有限公司 | Aircraft conduit monitoring system and method |
US8511179B2 (en) | 2008-06-20 | 2013-08-20 | Airbus Operations Gmbh | Aircraft conduit monitoring system and method |
WO2011160943A1 (en) * | 2010-06-21 | 2011-12-29 | Optimized Systems And Solutions Limited | Asset health monitoring using stable distributions for heavy-tailed data |
US10520387B2 (en) | 2014-10-31 | 2019-12-31 | SZ DJI Technology Co., Ltd. | Gas leakage treatment method and aerial vehicle |
CN105518377A (en) * | 2014-10-31 | 2016-04-20 | 深圳市大疆创新科技有限公司 | Gas leakage processing method and apparatus, and air vehicle |
US10227215B2 (en) | 2016-08-24 | 2019-03-12 | Otis Elevator Company | Passenger conveyor and method for monitoring vibrations in a passenger conveyor |
US10829344B2 (en) | 2017-07-06 | 2020-11-10 | Otis Elevator Company | Elevator sensor system calibration |
US11014780B2 (en) | 2017-07-06 | 2021-05-25 | Otis Elevator Company | Elevator sensor calibration |
WO2020002018A1 (en) * | 2018-06-28 | 2020-01-02 | Konux Gmbh | System and method for traffic control in railways |
CN112313139A (en) * | 2018-06-28 | 2021-02-02 | 科路实有限责任公司 | System and method for rail transit control |
WO2020173636A1 (en) * | 2019-02-27 | 2020-09-03 | Siemens Mobility GmbH | Method for monitoring points of a railway track installation |
CN113574398A (en) * | 2019-02-27 | 2021-10-29 | 西门子交通有限公司 | Method for monitoring a switch of a railway track system |
AU2020227791B2 (en) * | 2019-02-27 | 2023-07-27 | Siemens Mobility GmbH | Method for monitoring points of a railway track installation |
WO2021001219A1 (en) * | 2019-07-02 | 2021-01-07 | Konux Gmbh | Monitoring, predicting and maintaining the condition of railroad elements with digital twins |
JP2022537937A (en) * | 2019-07-02 | 2022-08-31 | コヌクス ゲーエムベーハー | Monitoring, Predicting, and Maintaining the Condition of Railroad Track Elements Using Digital Twins |
CN110428109A (en) * | 2019-08-07 | 2019-11-08 | 佳都新太科技股份有限公司 | Subway shield door time between failures prediction model is established and prediction technique |
CN110428109B (en) * | 2019-08-07 | 2020-11-03 | 佳都新太科技股份有限公司 | Subway shield door fault interval time prediction model establishing and predicting method |
Also Published As
Publication number | Publication date |
---|---|
GB0318339D0 (en) | 2003-09-10 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Wang et al. | Real-time defect detection of high-speed train wheels by using Bayesian forecasting and dynamic model | |
WO2005015326A1 (en) | System for monitoring the working condition of an installation | |
Hodge et al. | Wireless sensor networks for condition monitoring in the railway industry: A survey | |
US5529267A (en) | Railway structure hazard predictor | |
US20150009331A1 (en) | Real time railway disaster vulnerability assessment and rescue guidance system using multi-layered video computational analytics | |
WO2017070301A1 (en) | Operations monitoring for effect mitigation | |
AU2015264781B2 (en) | System for monitoring the operating conditions of a train | |
Vale et al. | Novel efficient technologies in Europe for axle bearing condition monitoring–the MAXBE project | |
Patra et al. | Availability analysis of railway track circuits | |
CN109353376A (en) | Rail vehicle monitors system and its monitoring method | |
CN111315630A (en) | Data fusion concept | |
Lingamanaik et al. | Using instrumented revenue vehicles to inspect track integrity and rolling stock performance in a passenger network during peak times | |
Oßberger et al. | Insights towards condition monitoring of fixed railway crossings | |
Pappaterra | A literature review for the application of artificial intelligence in the maintenance of railway operations with an emphasis on data | |
KR101663789B1 (en) | Monitoring system for vehicle running information by vibration analysis to predict danger | |
CN103852414A (en) | Method for monitoring corrosion and predicting service life of bridge | |
Ackroyd et al. | Remote ride quality monitoring of Acela train set performance | |
Velmurugan et al. | Advanced railway safety monitoring system based on wireless sensor networks | |
WO2020116031A1 (en) | Railroad monitoring system, railroad monitoring device, railroad monitoring method, and non-transitory computer-readable medium | |
Khoury et al. | Performance comparison of automatic vehicle identification and inductive loop traffic detectors for incident detection | |
Shadfar et al. | An index for rail weld health assessment in urban metro using in-service train | |
Nicks | Condition monitoring of the track/train interface | |
Aliev et al. | Intelligent system of noise control of the technical condition of railroad tracks | |
Chen et al. | Effective condition monitoring of line side assets | |
EP3896542A1 (en) | Method and device for diagnosing components of a technical system |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AK | Designated states |
Kind code of ref document: A1 Designated state(s): AE AG AL AM AT AU AZ BA BB BG BR BW BY BZ CA CH CN CO CR CU CZ DE DK DM DZ EC EE EG ES FI GB GD GE GH GM HR HU ID IL IN IS JP KE KG KP KR KZ LC LK LR LS LT LU LV MA MD MG MK MN MW MX MZ NA NI NO NZ OM PG PH PL PT RO RU SC SD SE SG SK SL SY TJ TM TN TR TT TZ UA UG US UZ VC VN YU ZA ZM ZW |
|
AL | Designated countries for regional patents |
Kind code of ref document: A1 Designated state(s): GM KE LS MW MZ NA SD SL SZ TZ UG ZM ZW AM AZ BY KG KZ MD RU TJ TM AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HU IE IT LU MC NL PL PT RO SE SI SK TR BF BJ CF CG CI CM GA GN GQ GW ML MR NE SN TD TG |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application | ||
122 | Ep: pct application non-entry in european phase |