TECHNICAL FIELD
[0001] The subject disclosure relates to a railway monitoring sensor unit, more particular
to a railway monitoring sensor unit configured to measure vertical displacement of
a railway line, especially void (in Dutch "blinde vering").
BACKGROUND ART
[0002] Railway lines are usually supported by sleepers, which rest on a ballast bed consisting
of a packed bed of angular stones. Vibration caused by the passage of trains can lead
to the development of voids under the sleepers. This is especially the case at locations
susceptible to track subsidence, such as transition zones from ballasted track to
unballasted track, or at insulated rail joints. When a train passes along the railway
track, any such voids will allow vertical movement of the track and so of the train.
Where such voids occur under just one end of a sleeper, this may cause the vehicles
in the train to tilt or sway, and in extreme cases this can lead to derailment. Furthermore,
if the vertical movement is more than about 20 mm, this can impose excessive stresses
on the rails, particularly in the vicinity of welts. Accordingly, monitoring the presence
and size of such voids is desirable. Similar problems can also arise where the rails
are supported by a continuous support structure such as a concrete slab base rather
than by conventional ballast, and it should be appreciated that in this specification
the term ballast is to be interpreted as meaning the underlying medium that supports
the rails, and above which the rails extend.
[0003] GB 2420627 discloses an instrument for monitoring vertical movement of a railway line relative
to the supporting ballast, referred to as a void meter. The void meter consists of
a shaft connected at one end to a clamp such as a magnetic base so it can be clamped
onto a rail. An indicator pivots on the shaft, being displaced when a sensing member
attached to the shaft is turned, and being held by friction.
[0004] JP4338273B2 discloses a track bed subsidence detector is composed of a detector body attached
and fixed to the rail bottom portion of the rail and a base portion installed on the
track bed. A linear shaft of the detector body is pressed downward by a compression
coil spring 35 following the sinking of the base portion. Switches detect the amount
of downward movement of the detector body and consequently the amount of subsidence
of track bed.
[0005] CN110126877A discloses a railway track vibration monitoring system comprising a plurality of railway
track vibration monitoring nodes. A railway track vibration monitoring node comprises
a silicon MEMS based capacitive acceleration sensor and communicates with a backstage
monitoring center which analyses received track vibration data in real time.
[0006] WO2017105451A1 discloses a monitoring and warning system is provided that measures a track displacement
as an indication of an operational condition of railway tracks and a rail track structure.
The system comprises a sensor and a device coupled to the sensor. In response to a
physical measurement of a vertical displacement of a railway track in a down direction
by the sensor, the device is configured to provide a warning signal indicative of
a possible future failure of the railway track or the rail track structure for proactive
track maintenance purposes.
[0007] US2018/154914A1 discloses at least first and second attitude sensors fixed to the rail by glue at
respective positions offset along the axis of the rail; a processing circuit configured
to recover attitude measurements supplied by the first and second attitude sensors
and configured to calculate a deformation of the railway rail relative to the axis
as a function of the recovered attitude measurements. The attitude sensors are connected
one after the other by a wired connection including in particular power supply wiring
and communication wiring to the processing circuit.
SUMMARY OF INVENTION
[0008] It is an object of the present subject technology to provide an improved railway
monitoring sensor unit which comprises at least one of the following properties: more
accurate void measurement and which requires less service maintenance.
[0009] According to an aspect of the subject technology, this object is achieved by a railway
monitoring sensor unit having the features of claim 1. Advantageous embodiments and
further ways of carrying out the present technology may be attained by the measures
mentioned in the dependent claims.
[0010] A railway monitoring sensor unit according to the subject technology comprises an
MEMS acceleration sensor generating acceleration signals and, a wireless communication
module configured to transmit data to a server for further processing. The railway
monitoring sensor unit further comprises a signal processing unit configured for processing
the acceleration signals and generating vertical track displacement data and a power
supply to power the MEMS acceleration sensor, the signal processing unit and the wireless
communication module. The wireless communication module is configured to transmit
the vertical track displacement data to the server.
[0011] The concept of the present technology is based on the need of a railway monitoring
sensor unit that accurately measures the vertical movement of a rail and the void
below said rail due to passing trains, and that requires less service maintenance.
This object is reached by processing the acceleration data from a MEMS acceleration
sensor internally in the railway monitoring sensor unit and transmitting only the
measured vertical movement data wirelessly, rather than the raw acceleration data
from an acceleration sensor. In this way, the amount of data to be transmitted wirelessly
is significantly reduced and consequently less energy is needed to transmit the relevant
data wirelessly to a server. If the sensor unit is a stand-alone unit with internal
power source, e.g. a battery, the decrease of power consumption increases the lifetime
of the battery. As a result the interval between two service moments is increased,
which reduces the number of potential dangerous moments a service technician has to
work on the railway track.
[0012] Furthermore, by using a MEMS acceleration sensor, no mechanical moving parts are
needed to measure the vertical displacement of the rail. This makes the sensor unit
almost impervious to wear and tear and contamination. This also reduces the moments
of service maintenance of the sensor unit.
[0013] In an embodiment of the invention, the signal processing unit is configured to generate
a set of vertical track displacement data from acceleration signals corresponding
to a train passage. This feature makes it possible to further reduce energy consumption
because the sensor unit only sends data from periods with significant vertical displacement
of the rail and not from time periods without significant vertical movement.
[0014] In a further embodiment, the set of vertical track displacement data includes for
each train passage: a time stamp, and at least one of: maximal displacement downward,
maximal displacement upward, average displacement downward of wheelsets, average displacement
upward of the wheelsets, maximal swing caused by a wheelset, average swing caused
by the wheelsets. By only transmitting information that is used by the server and
maintenance software, the amount of data to be transmitted wirelessly can be further
reduced.
[0015] In an embodiment, the railway monitoring sensor unit is configured to repeatedly
collect a number of N sets of vertical track vibration data and to transmit the N
sets of vertical track displacement data in one transmission session to the server.
By transmitting sets of vertical track displacement data in a batch, the energy needed
to transmit the data can be further reduced.
[0016] In an embodiment, the railway monitoring unit further comprises a moving train detector
configured for detecting a moving train on the railway in vicinity of the sensor unit
and generating a presence detection signal indicative for the presence of a moving
train in vicinity of the sensor unit; and a controller to wake-up the signal processing
unit in response to the detection signal. By having the signal processing unit in
sleep mode when there is no train in the vicinity of the void measuring point which
comprises a railway monitoring sensor unit, the power consumption can be further reduced.
[0017] In a further embodiment, the moving train detector comprises a low-power MEMS acceleration
sensor generating acceleration signals, and a processor configured to derive the presence
detection signal from the acceleration signals generated by the low-power MEMS acceleration
sensor. By using a low-power MEMS acceleration sensor for detecting a moving train
in vicinity of the measuring point on the railway and a low-noise MEMS accelerometer
which consumes more power and generates acceleration signals to be processed by the
signal processing unit, the power consumption can be further reduced.
[0018] In an alternative embodiment, the presence detection signal indicates the presence
of a moving train in vicinity of the railway monitoring sensor unit when the acceleration
signals have an amplitude exceeding a first threshold value. By using only the acceleration
and not the displacement, no processing power is needed to determine whether a moving
train is in vicinity of the measuring point.
[0019] In an alternative embodiment, the signal processing unit is further configured for
determining a time instant corresponding to the moment that a train has passed, and
configured to derive vertical track displacement data from acceleration signals retrieved
after wake-up of the signal processing unit until the time instant. This feature allows
reducing the processing time of the signal processing unit and consequently the power
consumption of the railway monitoring sensor unit.
[0020] In a further embodiment, the signal processing unit is configured to detect that
a train has passed when the amplitude of the acceleration signals drops below a second
threshold value on times defined by a time profile. By using a simple algorithm to
detect the end of a train passage, less power is needed to determine corresponding
moment in time.
[0021] In an embodiment, the railway monitoring sensor is configured to switch to low-power
mode after transmission of vertical track displacement data of a train passage to
the server and optionally after generation of vertical track displacement data of
a train passage. This feature keeps energy consumption to a minimum.
[0022] In an embodiment, the processing unit is configured to skip after the generation
of vertical track displacement data of a train passage the processing of acceleration
signals to derive vertical track displacement data for a predefined number of train
passages, a predefined time period or a combination thereof. As void below a sleeper
increases relative slowly, not every train passage has to be measured. This feature
reduces the power consumption of the sensor unit.
[0023] In an embodiment, the MEMS acceleration sensor is configured to generate a bandpass
filtered digital acceleration signals; the signal processing unit is configured to:
- noise filter the bandpass filtered digital acceleration signals to obtain a filtered
acceleration signal;
- integrating the filtered acceleration signal to obtain a raw velocity signal;
- noise filtering the raw velocity signal to obtain a filtered velocity signal;
- integrating the filtered velocity signal to obtain a displacement signal;
- analysing the displacement signal to obtain the vertical track displacement data.
With these process steps running on a signal processor, the vertical displacement
of the track can be accurately determined.
[0024] In a further embodiment, analysing comprises:
- determining for each wheelset of the train at least one of the following characteristics:
maximal displacement downward, maximal displacement upward and sum of maximal displacement
downward, maximal displacement upward to obtain wheelset data. In a further embodiment,
analysing further comprises:
- determining from the wheelset data maximal displacement downward, maximal displacement
upward, average displacement downward of wheelsets, average displacement upward of
the wheelsets, maximal swing caused by a wheelset, average swing caused by the wheelsets.
By determining internally a characteristic for each wheelset and not characteristics
for only whole train passage, a multitude of maintenance strategies can be applied
to determine when a ballast bed requires maintenance. From, the measured characteristics
of each wheelset a desired set of vertical track displacement data is assembled to
reduce the amount of data to be transmitted wirelessly to the server.
[0025] In an embodiment, the wireless communication module is a LPWAN transceiver. Such
a transceiver is very energy efficient.
[0026] In an embodiment, the railway monitoring sensor unit comprises a battery to power
the MEMS acceleration sensor, the signal processing unit and wireless communication
module. In this way damage to powerlines is eliminated.
[0027] In an embodiment, the railway monitoring sensor unit comprises a coupling structure
to affix the railway monitoring sensor unit to a rail. In a further embodiment, the
coupling structure includes a magnet. The coupling structure enables the sensor unit
to be firmly attached to a rail or sleeper. A magnet allows the sensor unit to be
easily and detachably fixed to an iron rail.
BRIEF DESCRIPTION OF THE DRAWINGS
[0028] These and other aspects, properties and advantages will be explained hereinafter
based on the following description with reference to the drawings, wherein like reference
numerals denote like or comparable parts, and in which:
Fig. 1 illustrates a railway monitoring system provided with displacement sensor units;
Fig. 2 illustrates the overall process of a railway monitoring system;
Fig. 3 shows a block diagram of an embodiment of railway monitoring sensor unit in
accordance with the present subject technology;
Fig. 4 illustrates the principle of void;
Fig. 5a is a graph showing a raw acceleration signal of a train passage as generated
in sensor unit in accordance with the subject technology over time;
Fig. 5b is a graph showing a filtered acceleration signal of a train passage as calculated
in a sensor unit in accordance with the subject technology over time;
Fig. 5c is a graph showing a raw velocity signal of a train passage as calculated
in a sensor unit in accordance with the subject technology over time;
Fig. 5d is a graph showing a filtered velocity signal as calculated in a sensor unit
in accordance with the subject technology over time; and,
Fig. 5e is a graph showing a raw deflection signal as calculated in a sensor unit
in accordance with the subject technology over time.
DESCRIPTION OF EMBODIMENTS
[0029] Fig. 1 illustrates a railway monitoring system provided with railway monitoring sensor
unit 110. The railway monitoring sensor unit is coupled to a railway 100 and is configured
to measure vertical displacement of the rails and/or when a train passes a location
where the sensor unit is firmly attached to the railway. The sensor unit 110 comprises
a housing with a coupling structure to affix the sensor unit to the rail 104 or to
a sleeper 102. A strong magnet, glue or mounting adhesive may be used to affix the
sensor unit to the rail, e.g. to the web of a rail. A rail comprises a head, a foot
and a web between the head and the foot. In an alternative embodiment the coupling
structure comprises through holes in the housing of the sensor unit. By means of a
screw coupling and/or bold and nut coupling the displacement sensor unit may be affixed
to the rail or sleeper. Fig. 4 illustrates the principle of void. As trains pass over
a given location, the wheelsets cause a movement of the rail 104. This is primarily
down to flexure of the rail 104, compression of the ballast 402, and voiding 404 between
the sleepers 102 and the ballast 402. Of primary concern to a railway engineer is
what the overall deflection of the rail is. The displacement sensor unit 110 according
to the subject technology is able to take also into account of any movement in the
ballast itself as a train passes.
[0030] The sensor unit determines vertical track displacement data, which may be logged
in the sensor unit 110. The vertical track displacement data may include any vertical
displacement data of the track that characterizes deterioration in the safety of the
railway and may give rise to maintenance. The sensor unit sends wirelessly the vertical
track vibration data to an antenna 112 which transfers to vertical track vibration
data to a secure server 114 to allow remote web based monitoring 116 of these logged
measurements. The monitoring of these operational characteristics, particularly the
rail displacement during the passage of rail traffic, will give a measure of voiding
and the maintenance requirements for the relevant section of rail. The sensor unit
110 also optionally measures and records rail and ambient temperature. The trending
of the vertical track displacement data with time may permit the prediction of maintenance
requirements or the imposing of speed limits on the relevant sections of the rail
network.
[0031] Fig. 2 illustrates the overall process of a railway monitoring system. First a rail
way monitoring is affixed to the railway at locations to be monitored (reference 201).
The track transition from a softer to a harder surface, and vice versa, at a railway
crossover or viaduct for example, can cause vibrations. Another cause may lie just
before/behind the track transition. The ballast (railway gravel) cannot support the
sleepers (concrete sleepers) as well as it should. This vertical displacement is called
in Dutch 'blinde vering'. The insulation rail joints before and after a track transition
may also affect vibrations. The sensor unit is now waiting for a train passage (reference
202). Reference 203 indicates that when a train approaches the sensor unit, the sensor
unit starts a measurement. During a train passage, indicated with 204, acceleration
data is recorded. Vertical track displacement data is calculated 205 from the acceleration
data by a signal processor. The calculated vertical track displacement data is wirelessly
transmitted 206 to the cloud 207 and distributed 208 to servers 209 for storage and
further processing. The stored vertical track displacement data may be inspected visually
on a screen by a graph showing 210 the vertical track displacement data as a function
of time. This inspection may also be performed automatically by algorithms that present
data for inspection to a user once the vertical track displacement data meets predefined
criteria. When the vertical track displacement data indicates that there is too much
vertical displacement or too much vibration to be expected in the near future, the
railway manager 211 is informed about said locations. If the vertical displacement
exceeds predefined threshold values, a notification 212 is given to start maintenance
and/or visual inspection or any other suitable action to avoid serious railway accidents.
[0032] Fig. 3 shows a block diagram of an embodiment of railway monitoring sensor unit 300
in accordance with the present subject technology. The railway monitoring sensor unit
300 comprises a first MEMS acceleration sensor 302 generating acceleration signals,
optionally a second MEMS acceleration sensor 304, a signal processing unit 306, data
storage 308, a wireless communication module 310, an antenna 312, a power supply 314
and a power regulator 316. All these parts are accommodated in the housing of the
unit.
[0033] The first MEMS acceleration sensor 302 comprises a 3-axis MEMS (Micro-Electro-Mechanical
Systems) acceleration sensing element. It should be noted that to measure only vertical
displacement a 1-axis MEMS acceleration sensing element may be used. The first MEMS
acceleration sensor 302 is configured to generate acceleration signals. The first
MEMS acceleration sensor converts the analogue acceleration signals to the digital
domain. Preferably, the first MEMS acceleration sensor comprises an analog, low-pass,
antialiasing filter to reduce out of band noise and to limit bandwidth in the analog
domain, an Σ-Δ ADC to digitize the filtered analog signal, a decimation filter with
a low-pass filter cutoff (3dB) at about 1 kHz and an interpolation filter after the
decimation filter to produce oversampled/upconverted raw acceleration signals to be
processed by the signal processing unit 306.
[0034] The signal processing unit 306 receives the acceleration signals from the first MEMS
acceleration sensor 302 and processes the acceleration signals to generate a set of
vertical track displacement data. The set of vertical track displacement data generated
by the signal processing unit 306 is transmitted by wireless communication module
310 to a server for further processing using antenna 312. Data storage 308 is used
by the signal processing unit to store intermediate signals and one or more sets of
vertical track displacement data. The power supply 314 is a battery and may be a rechargeable
battery to power the components of the railway monitoring sensor unit 300. In this
way, the railway monitoring sensor unit 300 is a stand-alone unit which does not require
a connection with a main voltage. A power regulator 316 is provided to generate a
fixed output voltage of a pre-set magnitude that remains constant regardless of changes
to its input voltage or load conditions.
[0035] The wireless communication module 310 may uses any suitable wireless communication
protocol. When the railway monitoring sensor unit is a battery powered stand-alone
unit, the wireless communication module 310 may be a LPWAN (low-power wide-area network)
transceiver, LoRaWAN (Long Range Wide Area Network) transceiver, or any other suitable
wireless network transceiver. The wireless communication module 310 is switched on
by the signal processing unit 306 when there is a set of vertical track displacement
data available for transmission. A set of vertical track displacement data is calculated
for a train passage and includes for each train passage: a time stamp, and at least
one of: maximal displacement downward, maximal displacement upward, average displacement
downward of wheelsets, average displacement upward of the wheelsets, maximal swing
caused by a wheelset, average swing caused by the wheelsets. After transmission of
the set of vertical track displacement data, the wireless communication module is
switched in a low-power mode to reduce power consumption. To reduce power consumption
further, the railway monitoring sensor unit may be configured to repeatedly collect
a number of N sets of vertical track displacement data and to transmit the N sets
of vertical track displacement data in one transmission session to the server.
[0036] As said above, a set of vertical track displacement data is calculated for a train
passage. This means that the signal processing unit 306 only has to process acceleration
signals corresponding to train passages. By switching the signal processing unit 306
in a low-power mode when there is no moving train in the vicinity of the location
of the railway monitoring sensor unit, the power consumption is reduced. The signal
processing unit 306 knows when it has finalized processing of acceleration data to
obtain a set of vertical track displacement data. Furthermore, the signal processing
unit 306 controls the wireless communication unit 310 to transmit one or more sets
of vertical track displacement data and thus knows when the wireless communication
unit 310 has transmitted the vertical track displacement data. This allows the signal
processing unit 306 to switch itself into an energy-saving mode. To wake-up the signal
processing unit 306 from energy-saving mode to active mode, the railway monitoring
unit further comprises a moving train detector configured for detecting a moving train
on the railway in vicinity of the sensor unit and generating a presence detection
signal indicative for the presence of a moving train in vicinity of the sensor unit;
and a controller to wake-up the signal processing unit in response to the detection
signal. The first MEMS acceleration sensor 302 may perform the function of moving
train detector. As soon as the amplitude of the acceleration measured by the acceleration
sensing element exceeds a predefined threshold value, the sensor 302 generates the
presence detection signal at an output pin which is coupled to an input pin of the
digital signal processing unit 306. After the presence detection signal changes from
no-moving train detected value to moving train detected value, the digital signal
processing unit 306 is waked-up and start capturing acceleration signals from the
first acceleration sensor unit 302.
[0037] Optionally the railway monitoring sensor unit 300 comprises a second MEMS acceleration
sensor 304 to perform the moving train detector function. The second MEMS acceleration
sensor 304 has a power consumption which is less than the power consumption than the
first MEMS acceleration unit 302. Normally, a MEMS acceleration sensor with high accuracy
has a higher power consumption than a MEMS acceleration sensor with a low accuracy.
By using a low-power MEMS acceleration sensor with relative low accuracy for performing
the moving train detector and controller functions for generating the presence detection
signal as second MEMS acceleration sensor 304, the power consumption can be further
reduced. In that case, second MEMS acceleration sensor 304 continuously measures vibration
and as soon as the acceleration corresponding to the vibration caused by a moving
train exceeds a predefined threshold value, the sensor changes an output signal from
no-moving train value to moving train value. This output signal is used by both the
signal processing unit 306 and first MEMS acceleration sensor 302 to wake-up from
low-power mode. The first MEMS acceleration sensor 302 starts sensing the vibration
of the railway and generates acceleration signals and the signal processing unit 306
starts processing the acceleration signals to obtain a set of vertical track displacement
data from acceleration signals corresponding to a train passage. The signal processing
unit 306 monitors the amplitude of the acceleration signals to determine a time instant
corresponding to the moment a train has passed. The signal processing unit detects
that a train has passed when the amplitude of the acceleration signals drops below
a second threshold value on times defined by a time profile. An exemplar embodiment
of a time profile is: detecting that the amplitude of the acceleration signal is for
a period of 3 seconds below a threshold value. Another exemplar embodiment of a time
profile is: detecting the time instant the amplitude of the acceleration signal drops
for a period of X ms below the second threshold value, check Y ms after said time
instant whether the amplitude of the acceleration signal is again for a period of
X' ms below the second threshold value. If this is the case, the decision is made
that the train has passed the location of the railway monitoring sensor unit. The
signal processing unit 306 processes the acceleration signal up to the moment a train
has passed. In the exemplar embodiment of a time profile, this is the moment in time
corresponding to the time instant the amplitude of the acceleration signal drops for
the first period of X ms below the second threshold value. Possible values for X,
Y and X' are 100, 3000 and 100 respectively. After processing the acceleration signals
corresponding to a train passage and transmitting the set of vertical track displacement
data, the signal processing unit 306 switches the first MEMS acceleration sensor and
itself into low-power mode. The second MEMS acceleration sensor 304 monitors the vibration
of the railway and as soon as the amplitude exceeds the predefined threshold value,
the sensor generates again a control signal to wake-up the signal processing unit
306 and first MEMS acceleration sensor 302.
[0038] The processing unit may further be configured to skip after the generation of vertical
track displacement data of a train passage the processing of acceleration signals
to derive vertical track displacement data for a predefined number of train passages,
a predefined time period or a combination thereof. The void displacement increases
slowly in time and also due to each train passage. By skipping the processing of train
passages and generating and transmitting sets of vertical track displacement data,
the power consumption is reduced. In a first exemplar embodiment, the railway monitoring
sensor unit 300 determines one set of vertical track displacement data for a train
passage each two hours. In a second exemplar embodiment, the railway monitoring sensor
unit 300 determines a set of vertical track displacement data for a train passage
each tenth train passage. In a second exemplar embodiment, the railway monitoring
sensor unit 300 determines a set of vertical track displacement data for a train passage
each tenth train passage and if a tenth train passage is not within a two hours period,
a set of vertical track displacement data for a train passage is generated for the
first train passage after the 2 hour period.
[0039] The railway monitoring sensor unit according to the subject technology processes
acceleration signals from a 3-axis MEMS acceleration sensor to obtain a set of vertical
track displacement data as follows. A characteristic of the vertical displacement
of the rail is that the relevant information is in the frequency range from 1 Hz to
24 Hz. The first MEMS acceleration sensor 302 generates a bandpass filtered digital
acceleration signal which is the raw acceleration signal that is supplied to the signal
processing unit 306. Fig. 5a is a graph showing a raw acceleration signal of a train
passage as generated by the MEMS acceleration sensor 302 over time. The train comprises
four train units. Each unit comprises a front side wheelset and a rear side wheelset.
Each wheelset comprises two axles. This raw acceleration signal is processed by the
signal processing unit 306 to obtain a set of vertical track displacement data. The
raw acceleration signal is first applied to a first noise filter to attenuate frequencies
below 1,3 Hz and above 24 Hz to obtain a filtered acceleration signal. The first noise
filter may be a 2
nd order Butterworth filter. Fig. 5b is a graph showing the filtered acceleration signal
of a train passage. Subsequently, the filtered acceleration signal is integrated to
obtain a raw velocity signal. Fig. 5c is a graph showing the raw velocity signal of
a train passage. A second noise filter is applied on the raw velocity signal to obtain
a filtered velocity signal. Frequency components above 24 Hz are attenuated. The second
noise filter may be a 2
nd order Butterworth filter. Fig. 5d is a graph showing a filtered velocity signal as
calculated in a sensor unit in accordance with the subject technology over time. The
filtered velocity signal is integrated to obtain a displacement signal. Fig. 5e is
a graph showing a raw deflection signal after integrating the filtered velocity signal.
From this graph can be seen that the train comprises four train wagons and each wagon
comprises two wheel sets, each with two axles. The distance between the wheelsets
of a wagon is larger than the distance between the back side wheelset of a previous
wagon and the front side wheelset of a subsequent wagon coupled to the back side of
the previous wagon.
[0040] The raw deflection signal represents a displacement signal. The raw deflection signal
is finally analysed to reduce the amount of digital data and to obtain the set of
vertical track displacement data. The raw deflection signal shows the displacement
of each wheelset over time. Also the deflection caused by each axle can be seen in
the graph. From this raw deflection signal the displacement downward and upwards cause
by a wheelset can be easily determined. It can be seen that the displacement for each
wheelset is unique. Before a wheelset passes the sensor unit, the rail is lifted slightly
before the rail is displaced downward. Similarly, the rail is lifted slightly after
the passage of the wheelset. This allows determining for each wheelset the maximal
displacement swing downward, maximal displacement swing upward. The maximal displacement
swing is the displacement between a top 502 of raising of the rail and bottom 504
of downward displacement of the rail by an axle next to the top of lift-up of the
rail. Furthermore, the data derived for each wheelset can be used to determine for
the train passage the average of maximal displacement downward of the wheelsets, average
of maximal displacement upward of the wheelsets, maximal displacement downward, the
maximal displacement upward relative to a reference value, the maximal displacement
downward relative to the reference value. It should be noted that in Fig. 5E the raising
of the track is more than in reality and the lowering in less than in reality. However,
the distance between the top of the raising and the bottom of the lowering, defining
the displacement swing, corresponds to the physical displacement. Furthermore, if
the change over time is monitored, the known error in measured tops of raisings and
tops of lowering can be eliminated. Through the analysis of the raw deflection signal,
predefined characteristics are determined from the digital samples of the raw deflection
signal, allowing the sensor unit to reduce the amount of data to be sent wirelessly
to a server. Hence, a battery in the sensor unit does not need to be replaced as quickly
as if the raw deflection signal were sent to a server and thus less maintenance is
required on the sensor unit.
[0041] Next to vertical track displacement data, data such as a time stamp, ambient temperature,
rail temperature may be added to the set of vertical track displacement data corresponding
to a train passage.
[0042] While the invention has been described in terms of several embodiments, it is contemplated
that alternatives, modifications, permutations and equivalents thereof will become
apparent to those skilled in the art upon reading the specification and upon study
of the drawings. It must be understood that this description is given solely by way
of example and not as limitation to the scope of protection, which is defined by the
appended claims.
1. A railway monitoring sensor unit (300) comprising an MEMS acceleration sensor (302)
generating acceleration signals and, a wireless communication module (310) configured
to transmit data to a server for further processing;
characterized in that
the railway monitoring sensor unit further comprises a signal processing unit (306)
configured for processing the acceleration signals and generating vertical track displacement
data and a power supply (314) to power the MEMS acceleration sensor, the signal processing
unit and the wireless communication module; and, the wireless communication module
is configured to transmit the vertical track displacement data to the server.
2. The railway monitoring sensor unit according to claim 1, wherein the signal processing
unit (306) is configured to generate a set of vertical track displacement data from
acceleration signals corresponding to a train passage.
3. The railway monitoring sensor unit according to claim 2, wherein the set of vertical
track displacement data includes for each train passage: a time stamp, and at least
one of: maximal displacement downward, maximal displacement upward, average displacement
downward of wheelsets, average displacement upward of the wheelsets, maximal swing
caused by a wheelset, average swing caused by the wheelsets.
4. The railway monitoring sensor unit according to any of the claims 2 - 3, wherein the
railway monitoring sensor unit is configured to repeatedly collect a number of N sets
of vertical track displacement data and to transmit the N sets of vertical track displacement
data in one transmission session to the server.
5. The railway monitoring sensor unit according to any of the claims 1 - 4, wherein the
railway monitoring unit further comprises a moving train detector configured for detecting
a moving train on the railway in vicinity of the sensor unit and generating a presence
detection signal indicative for the presence of a moving train in vicinity of the
sensor unit; and a controller to wake-up the signal processing unit in response to
the detection signal.
6. The railway monitoring sensor unit according to claim 5, wherein the moving train
detector comprises a low-power MEMS acceleration sensor generating acceleration signals,
and a processor configured to derive the presence detection signal from the acceleration
signals generated by the low-power MEMS acceleration sensor.
7. The railway monitoring sensor unit according to any of the claims 5 - 6, wherein the
presence detection signal indicates the presence of a moving train in vicinity of
the railway monitoring sensor unit when the acceleration signals have an amplitude
exceeding a first threshold value.
8. The railway monitoring sensor unit according to any of the claims 5 - 7, wherein the
signal processing unit is further configured for determining a time instant corresponding
to the moment that a train has passed, and configured to derive vertical track displacement
data from acceleration signals retrieved after wake-up of the signal processing unit
until the time instant.
9. The railway monitoring sensor unit according to claim 8, wherein the signal processing
unit is configured to detect that a train has passed when the amplitude of the acceleration
signals drops below a second threshold value on times defined by a time profile.
10. The railway monitoring sensor unit according to any of the claims 5 - 9, wherein the
railway monitoring sensor is configured to switch to low-power mode after transmission
of vertical track displacement data of a train passage to the server and optionally
after generation of vertical track displacement data of a train passage.
11. The railway monitoring sensor unit according to any of the claims 5 - 10, wherein
the processing unit is configured to skip after the generation of vertical track displacement
data of a train passage the processing of acceleration signals to derive vertical
track displacement data for a predefined number of train passages, a predefined time
period or a combination thereof.
12. The railway monitoring sensor unit according to any of the claims 1 - 11, wherein
the MEMS acceleration sensor is configured to generate a bandpass filtered digital
acceleration signals; the signal processing unit is configured to:
- noise filter the bandpass filtered digital acceleration signals to obtain a filtered
acceleration signal;
- integrating the filtered acceleration signal to obtain a raw velocity signal;
- noise filtering the raw velocity signal to obtain a filtered velocity signal;
- integrating the filtered velocity signal to obtain a displacement signal;
- analysing the displacement signal to obtain the vertical track displacement data,
preferably wherein analysing comprises:
- determining for each wheelset of the train at least one of the following characteristics:
maximal displacement downward, maximal displacement upward and sum of maximal displacement
downward, maximal displacement upward to obtain wheelset data,
more preferably wherein analysing further comprises:
- determining from the wheelset data maximal displacement downward, maximal displacement
upward, average displacement downward of wheelsets, average displacement upward of
the wheelsets, maximal swing caused by a wheelset, average swing caused by the wheelsets.
13. The railway monitoring sensor unit according to any of the claims 1 - 12, wherein
the wireless communication module is a LPWAN transceiver.
14. The railway monitoring sensor unit according to any of the claims 1 - 13, wherein
the power supply comprises a battery to power the MEMS acceleration sensor, the signal
processing unit and wireless communication module.
15. The railway monitoring sensor unit according to any of the claims 1 - 14, wherein
the railway monitoring sensor unit comprises a coupling structure to affix the railway
monitoring sensor unit to a rail,
preferably wherein the coupling structure includes a magnet.