WO2013114135A2

WO2013114135A2 - Control of transport networks

Info

Publication number: WO2013114135A2
Application number: PCT/GB2013/050240
Authority: WO
Inventors: David John Hill
Original assignee: Optasense Holdings Limited
Priority date: 2012-02-01
Filing date: 2013-02-01
Publication date: 2013-08-08
Also published as: WO2013114135A3; GB201201768D0

Abstract

The application describes methods and apparatus for control of transport networks, especially rail networks. The method involves monitoring at least part of the rail network(202, 203,204) by performing distributed acoustic sensing on one or more optical fibres(104a, 104b, 104c) deployed along the path of the network to provide a plurality of acoustic sensing portions. The signals detected by the plurality of acoustic sensing portions are analysed to detect acoustic signals associated with trains(205) moving on the network. The signals are analysed to identifying the front and rear of said trains and to track the movement of trains on the network. Indentifying the front and rear of a train may involve identifying the start/end (302, 303) of a continuous acoustic signal generated by the train and/or identifying characteristic signatures generated by the wheelsets of the train. The method may be used to provide moving block signalling.

Description

Control of Transport Networks

The present invention relates to control and monitoring of transport networks, for example road or rail networks, and to active network control for such networks, in particularly using distributed acoustic sensing.

Various control, monitoring and protection systems for transport networks are known. For instance in rail networks there are various signalling and control systems and train protection systems that have been proposed to monitor the movement of trains on the network and provide control.

In general, train control systems tend to try to monitor the movement of trains on the network and control signals and/or issue movement authority to individual trains to move about the network. In some systems train protection systems may be fitted to at least some of the trains which may automatically impose speed limits in certain areas and prevent movement deemed to be unsafe or unauthorized.

The existing systems typically rely on a number of different sensors or disparate systems. For instance locating the position of trains relies on a number of different systems. Trains are typically provided with number of sensors to determine position, for instance speedometers, inertial sensors or other movement sensors to determine the trains position based on the movement of the train on the network. As an example wheel revolutions may be monitored and used as an indication of the distance travelled. However if any wheel skid occurs during the movement of the train the train will actually have travelled further than indicated by the wheel sensor. Likewise wheel slippage can lead to errors in recorded position. It is noted that GPS is not commonly used as a primary system as in some urban environments GPS may be insufficiently reliable, in terms of accuracy and/or coverage, and GPS doesn't work well in tunnels. GPS thus can't readily be used in underground transit systems.

In general errors can develop in the train's own internal measure of its position. There may therefore also be variety of active and/or passive track based position markers which can be used by the train to correct its idea of its own position, and/or feed information to a control centre. For example at intervals along at least some tracks some absolute position references (APRs) may be embedded into the track. These act as passive beacons that are interrogated by the train as it passes over them. Detection of a specific APR therefore allows the train to accurately know its position.

However with conventional trackside positional beacons and the like there is usually a relatively large spacing between the beacons, meaning that a relatively large uncertainty may exist in train position from time to time.

In terms of train control the need to integrate the readings from several diverse sensor systems can lead to complex systems and the inherent uncertainty means that relatively large safety margins need to be built into the control. Many control system for trains operate on the principles of block signalling, i.e. the sections of track are divided into blocks and, in normal circumstances, trains are only given movement authority to move into unoccupied blocks. Typically the blocks are fixed positions on the track and thus a train in one block must be found to clear a block before another train can be given movement authority to enter the block. The type of system described above can operate within a fixed block operation.

In a fixed block system a train must be able to stop within a single block if necessary if the expected movement authority to enter the next block is not received. Thus there is a link between the maximum speed of the train and the length of blocks. Allowing trains to operate faster requires longer blocks which may reduce network capacity.

It has been proposed to use moving block signalling, where the blocks effectively move with the train. The goal is to maintain a safe zone in front of the train. This however requires accurate indications of the trains position and speed. As mentioned above this can be difficult to achieve in conventional systems.

In addition, as mentioned above trains may be fitted with train protection systems that automatically reduce speed or prevent movement. Again however the reliance on multiple sensors can sometimes present a relatively confused picture of the current train status. Such systems generally fail-safe and thus in the event of contradictory information the system may default to the worst case scenario and act accordingly. Thus the system may act on the basis of the greatest error at any one time. This may result in unwanted and unnecessary slowing of the train or delays in moving.

Sometime the slowing may be maintained until better data is obtained, at which point the speed can be increased again, but the cycle of slowing and speeding may result in fuel wastage and/or delays.

In some instances therefore trains that have been fitted with automatic protection systems have had the protections systems disabled by the train operators due to undesirable performance and there have been instances of collisions involving trains that had protections systems available.

Similar issues may apply generally to road networks in terms of maximising efficiency of the network by setting speed limits and controlling signals, for instance controlling traffic flow onto a major highway. Highways may be provided with cameras or in-lane induction loops to detect traffic flow rates but any traffic conditions between successive monitoring locations must be inferred. It is therefore an object of the present invention to provide methods and apparatus for transport network control that mitigate at least some of the aforementioned issues.

Therefore according to the present invention there is provided a method of control of a rail network comprising: monitoring at least part of the rail network by performing distributed acoustic sensing on one or more optical fibres deployed along the path of the network to provide a plurality of acoustic sensing portions; analysing the signals detected by said plurality of acoustic sensing portions to detect acoustic signals associated with trains moving on the network; where analysing said signals comprises identifying the front and rear of said trains and tracking the movement of trains on the network.

The tracking the movement of trains on the network may be used to set one or more control conditions governing movement of traffic on the network. The method of this aspect of the present invention therefore uses fibre optic distributed acoustic sensing (DAS). Distributed acoustic sensing is a known type of sensing where an optical fibre is deployed as a sensing fibre and repeatedly interrogated with electromagnetic radiation to provide sensing of acoustic activity along its length.

Typically one or more input pulses of radiation are launched into the optical fibre. By analysing the radiation backscattered from within the fibre, the fibre can effectively be divided into a plurality of discrete sensing portions which may be (but do not have to be) contiguous. Within each discrete sensing portion mechanical disturbances of the fibre, for instance, strains due to incident acoustic waves, cause a variation in the properties of the radiation which is backscattered from that portion. This variation can be detected and analysed and used to give a measure of the intensity of disturbance of the fibre at that sensing portion. Thus the DAS sensor effectively acts as a linear sensing array of acoustic sensing portions of optical fibre. The length of the sensing portions of fibre is determined by the characteristics of the interrogating radiation and the processing applied to the backscatter signals but typically sensing portions of the order of a few meters to a few tens of meters or so may be used. As used in this specification the term "distributed acoustic sensing" will be taken to mean sensing by optically interrogating an optical fibre to provide a plurality of discrete acoustic sensing portions distributed longitudinally along the fibre and the term "distributed acoustic sensor" shall be interpreted accordingly. The term "acoustic" shall mean any type of pressure wave or mechanical disturbance that may result in a change of strain on an optical fibre and for the avoidance of doubt the term acoustic be taken to include ultrasonic and subsonic waves as well as seismic waves.

DAS can be operated to provide many sensing channels over a long length of fibre, for example DAS can be applied on fibre lengths of up to 40km or more with contiguous sensing channels of the order of 10m long. Thus a long stretch of the rail network can be monitored using a single DAS sensor. For lengths of more that 40km or so several DAS sensors units can be deployed at various intervals to provide continuous monitoring of any desired length of the rail network. The use of DAS for monitoring a rail network is particularly advantageous. As mentioned above a single DAS sensor can provide a contiguous series of sensing channels separated by 10m or so for a length of up to 40km or more and greater lengths can achieved by using more sensors. A single DAS interrogator unit may be multiplexed between two fibres to provide sensing over a distance of 80km (with the interrogator in the middle). This offers an unrivalled continuity of sensing along the network. The sensing fibre may be standard telecoms fibre and thus is relatively cheap. The fibre may be simply buried alongside the networks, e.g. along the sides or underneath the rail tracks in a narrow channel at any depth required. The optical fibre can be encased in a protective casing and can survive for a long time with no maintenance. Thus installation and maintenance costs are low. In many rail networks there may already be optic fibre deployed along at least the major routes and such existing communications infrastructure may comprise redundant optical fibres that can be used for DAS.

The optical fibre is interrogated by optical pulses generated by the interrogator unit (as will be explained in more detail later) and thus power is only needed for the interrogator units.

In a rail network setting the spatial resolution and coverage of the DAS system avoids all the issues described above.

Movement of a train on a train track adjacent a DAS sensing fibre will generate acoustic signals that can be detected and used to track the train as it moves, providing real time positional information to a resolution of a few tens of metres continuously along the entire length of the monitored section.

It will be appreciated that as a train move along the railway it produces a significant noise in the section of track that it travels along. In the present invention the acoustic signals produced by the train as it moves are detected and used to identify the exact location of the train. Thus the method may involve detecting an acoustic response from extended length of the fibre, i.e. from a plurality of contiguous sensor portions. By monitoring the detected acoustic signals over time the movement of the train can also be tracked. Thus DAS can inherently provide both the location and speed of the train. It will be noted that DAS provides an indication of the position of the full length of the train on the network. As the train moves acoustic signals are generated by the whole length of the train and detected by a plurality of sensing portions of the DAS sensing optical fibre. The acoustic signals detected by the DAS sensor can thus provide an indication of the length of the train and also a reliable indication of the position of both the front and the rear of the train. As mentioned the whole of the train will produce acoustic signals as it moves and thus the sensing portions alongside the whole of the train will show a strong signal.

Note that convention train positional systems will typically provide an estimate of the location of the front of the train only. The inertial guidance and any GPS receivers will be located at the front of the train. The control room will then use an assumed length of the train in any control operations. In many cases this will be sufficient but in some circumstances the train moving on the network may be a different length to the assumed length, for instance a freight train may be pulling more carriages than expected. Not knowing the exact length of the train could result in a following train keeping a greater distance than would actually be required for safety - resulting in network inefficiency - or conversely if the train is significantly longer than assumed the following train may not be kept at a sufficient distance.

In one embodiment therefore the method comprises analysing the acoustic response from the acoustic sensor portions to locate an acoustic signal indicative of the front of the train and an acoustic signal associated with the rear of the train. The fact that a long section of track can be monitored with high spatial resolution with a DAS system, e.g. 40km or so of track with of the order of 10m sensing portions all along that 40km, is what enables such a method to be implemented in an effective manner.

As mentioned above the train will create a sound as it moves which will be detected by the sensing portions of fibre adjacent the relevant part of the track. All the sensing portions adjacent the train will detect a significant acoustic signal and thus the position of the train will show up as continuous area of acoustic noise. It will be appreciated of course that sound produced by the train will also travel ahead of the front of the train and backwards from the train and thus sections of fibre ahead of and behind the train will also detect noise due to the train movement. However although some sounds associated with train movement, such as ringing of the rails, may travel for a significant distance, such sounds will have a different characteristic to the sounds detected when the train is actually adjacent a sensing portion. Thus the acoustic response can be analysed to detect a generally continuous acoustic signal indicative of the train being adjacent or very near the sensing portions. The method may therefore involve identifying the beginning and end of a continuous acoustic disturbance indicative of the train.

Whilst the general acoustic noise may be used to determine the presence and location of a train, and used to track its progress, the identification may be improved by detecting characteristic signatures of the train. In one embodiment a train may be detected by detecting an acoustic signature resulting from the train passing track features. In particular the method may comprise identifying acoustic signals associated with the wheelsets of the train, for instance acoustic signals associated with the train passing track features. As will be appreciated as the train moves there will be a variety of sounds produced. In particular any track features that produce a noise as the wheelsets of the train pass over the feature will produce a characteristic repetitive pattern resulting from the arrangement of the wheelsets. Thus for instance for a jointed track there may be a noise produced when the wheels pass from one rail section to another. This will typically produce an acoustic signal. As the following wheelsets pass over the same joint they will also each produce a similar acoustic signal. Thus an acoustic sensing portion in the vicinity of the rail joint will detect a characteristic pattern. For instance consider a train having a locomotive with two front wheelsets and, after a larger gap, two rear wheelsets. If the train is travelling at a relatively constant speed the sensing portion adjacent the joint or other track feature may detect four distinct sounds corresponding to each of the four wheelsets crossing the feature in turn, with two sounds relatively close in time followed, after a larger gap by another two sounds relatively close in time. The distinctive repetitive pattern of the wheelsets crossing track features may therefore be used to detect the location of the train and identify the front and rear of the train by identifying the first and last sensing portions to reliably detect the characteristic sounds.

In general however the passage of the train on a track which is relatively featureless will still generate acoustic signals associated with the passage of the wheelsets/axels of the train which manifest as relatively intense broadband noise 'spikes' within the general acoustic signal due to passage of the train. The method may therefore involve identifying the general acoustic signals detected by a DAS sensor due to passage of a train on a monitored section of track. The detected acoustic signals due to the train may then be analysed to detect a series of relatively intense broadband signals, i.e. signals at a broad range of frequencies. Such signals have been found to correspond to signals generated by the wheelsets of the trains and thus can be used to identify the wheelsets/axels of the train. By identifying the signals due to the wheelsets of the train, as it moves along the track, the location of the train and the front and rear of the train can be accurately and reliably tracked over the whole of the monitored network to an accuracy of a few tens of metres.

As mentioned DAS can therefore be used to reliably detect and track the location of trains and can inherently provide information about the train length. The speed of the train can be determined by monitoring the change in position of the acoustic signals. The present invention therefore uses DAS to provide, in a single sensor arrangement, sufficient information to allow active train control. The acoustic signals from the one or more DAS sensors are therefore analysed to track the movement of various trains on a rail network. The tracking information may then used to set one or more control conditions governing movement of traffic on the network.

Setting control conditions may involve setting various signals appropriately, for instance the setting of speed limits or control of traffic signals. In a rail network setting the control conditions may involve controlling signals to allow trains to move or not. Setting control conditions may also involve issuing movement authority to trains and/or to controlling the positions of blocks in a moving block signalling system. It may also involving issuing speed limits to trains and/or activating emergency braking procedures. In a fully integrated system setting the control conditions may be providing the movement control for automatically operated rail vehicles. As all trains on the monitored network can be tracked, their exact location (including the location of the front and rear of the train) determined and speed and direction determined the control may maintain an optimal safe distance between trains. Thus network efficiency can be improved by achieving the optimal safe headroom that maximises network efficiency. Also, as DAS provides a continual update of the train status the control conditions can provide information to an automatic train protection unit that does not suffer from the drawbacks mentioned earlier. Thus the train protection system is more likely to operate reliably without inconveniencing the train operator. This increases the likelihood of train protection systems being installed and operated.

Indeed the improved information about the network status as a whole may allow assessment of how best to achieve fuel efficiency for a given train given network capability and any requirements on the train reaching a destination in a certain timeframe.

The DAS system will provide tracking of any train movement on the monitored network. At some times the train may be stationary on the rail network, even if just for a short time. During the time the train is not moving the locomotive may still be producing an acoustic signal which may be detected by the DAS sensor but other parts of the train may be relatively quiet. If the train is powered down there may be no acoustic signals generated by the train at all. The method may therefore involve maintaining an indication of the position of the train on the network, including the front and rear of the train. For example consider a train moving on a monitored section of track. The acoustic signals generated by the train will be detected by the sensing fibre of the DAS sensor and the position of the front and rear of the train on the rail track can be determined and tracked. If the train slows to a halt and waits stationary for a while then as long as the DAS sensor or sensors remain active the absence of any acoustic signals is actually a positive indication that the train has not moved and thus the last recorded position for the train can be maintained. As soon as the train starts to move again acoustic signals will be generated along the length of the train and thus the movement will be clearly detected by the DAS sensor. A DAS sensor can be operated continually and is reasonably low power. Even when no train movement is scheduled on the network it may be wished to continue to operate the DAS sensors to detect any anomalous acoustic signals, for instance such as associated with intrusion onto the railway for copper theft or the like. With continuous DAS monitoring position information about all of the trains on the network that have moved at least once can be maintained.

It should be noted that DAS sensors are quite suited for continuous operation and that DAS has an inherent degree of fault reporting. If the sensing cable is damaged it is likely that an unexpected larger than normal reflection will be detected. Interference with the sensing cable will also likely lead to a detectable measurement signal which could be used to generate an alarm. Failure of the interrogator unit or optic coupling to the fibre would result in a loss of sensitivity which would manifest in a lack of expected measurement signals (as the system will always detect low levels of ambient background noise). DAS is thus particularly useful for monitoring systems where reliability is important.

Setting control conditions may also involving controlling other aspects of the rail network, for instance controlling the operation of rail crossing barriers. At level junctions there may be automated barriers that operate to prevent road vehicles from crossing the train track as a train approaches. Usually the barrier closure is activated when the train reaches a certain trigger point a set distance way from the crossing. Having a fixed trigger point however makes no allowance for the different speeds that different trains may be travelling and thus must be set for the fastest expected speed. Clearly this means a greater delay at the crossing for road vehicles for slower trains. Using DAS in an active train control system the barrier could be closed based on the time it will take for the train to reach the crossing. Thus activation will differ for different trains minimising waiting times for road users and the consequent risk that an impatient driver tries to cross when the barriers are down. The control system could also use the accurate location of trains to prompt

announcements at stations or more accurate updating of service schedules.

The method of the present invention thus provides acoustic information via DAS to provide feedback about the movement of traffic over a transport network and uses that feedback in a control loop. The information provide via DAS is sufficient to allow control of the network as described and thus the need for complex integration of systems and the potential for conflicting data is reduced. It will of course be

appreciated that in many instances it will be desirable to have additional

positional/additional locations device to provide at least one of general validity check and redundancy in case of a failure in DAS. The train may have inertial guidance and internal databases to fall back upon should communications fail or DAS data be unavailable for any reason.

In addition to enabling train control DAS can also be used to provide other information about the transport network and the traffic upon it which can be useful in many other ways.

As mentioned above as trains travel on a railway they typically produce characteristic sounds, such a repetitive patterns from track features such as track joints. In use the motion of the train will be monitored. Any departure from the expected sound profile may indicate that the train operation has altered or that something has changed about the track.

For instance skidding is to be avoided on trains as the skid can create a flat spot on the wheel which can result in track damage and reduced efficiency of the train. A skidding would produce a distinctive sound which varies from the normal operation of the train. Detection of a characteristic sound indicative of a skid could thus be used to flag a maintenance check for the train and/or relevant section of track at the next opportunity. Other types of unwanted train movement behaviour could also be detected. Likewise the acoustic response from a given section of track would be expected to be largely the same each time a train traverses that section. For instance any track features leading to a detectable acoustic response would be expected to be in the same position. The acoustic profile from a given section of track when a train is travelling on or near it may be compared to one or more previous profiles. Detection of a new track feature generating an acoustic response could potentially indicate damage to a section of track or be indicative of likely failure. Detection of a significant change could therefore be used to flag the need for a maintenance check. In addition DAS will detect not only the movement of traffic on the track but any other sounds will be detectable. For instance DAS has been used previously for intruder detection. DAS could therefore be used to detect unauthorised access onto the network. For instance in some rail networks, such an underground network, there are issues with unauthorised access into the network, for example by people seeking shelter or for applying graffiti. This can have serious safety implication. In some areas theft of copper from existing communication and control lines creates issues, not just in the fact of intruders on the line but also in terms of the resultant loss of capability. Thus people or vehicles on the line could be detected and safety warnings generated.

Security personnel could be dispatched to investigate.

It was mentioned above that DAS may be used to help control level crossing and the like more efficiently. In one embodiment DAS may also be used to make sure that any such crossings are clear. If the fibre optic cable is deployed across the road then as a road vehicle passes the fibre it will impart a strain on the fibre which is detectable. A length of optical fibre can thus act as an axle detector. By arranging optical fibre on both sides of the crossing a vehicle detector can be implemented to determine the number of axles entering the crossing area and the number of axles leaving. In other words it is possible to detect whether all vehicles have exited the crossing. If not then appropriate emergency measures may be implemented. Two or more separate fibres could be implemented to provide such crossing detection but in one embodiment a single fibre may be arranged in a looped formation on both sides of the crossing.

The invention also relates to a rail traffic control system comprising at least one distributed acoustic sensor configured to provide distributed acoustic sensing on one or more optical fibres deployed along the path of a rail network to provide a plurality of acoustic sensing portions; and processor configured to analyse the signals detected by said plurality of acoustic sensing portions to detect acoustic signals associated with trains moving on the network; wherein said processor is configured to identify the front and rear of said trains and to track the movement of trains on the network. The system may also comprise a control unit configured to use said movement tracking to set one or more control conditions governing movement of traffic on the network.

The apparatus of this aspect of the invention may be configured to implement any of the features of the methods described above. The invention has been mainly described in relation to rail network but many of the same concepts apply to road networks.

In general aspects of the invention provide a method of control of a transport network comprising: monitoring at least part of the transport network by performing distributed acoustic sensing on one or more optical fibres deployed along the path of the network to provide a plurality of acoustic sensing portions; analysing the signals detected by said plurality of acoustic sensing portions to track the movement of traffic on the network; and using said movement tracking to set one or more control conditions governing movement of traffic on the network.

The invention will now be described by way of example only with respect to the following drawings; of which:

Figure 1 shows a convention DAS sensor arrangement;

Figure 2 illustrates a transport network been provided with DAS sensors;

Figure 3 illustrates the acoustic signals detected by a DAS sensor of a train travelling on a track;

Figure 4 illustrates the principles of detecting characteristics associated with trains;

Figure 5 shows how DAS may be integrated into a control system; and Figure 6 shows the acoustic signals detected by a DAS sensor deployed along a rail track as a train moves along the track.

Figure 1 shows a schematic of a distributed fibre optic sensing arrangement. A length of sensing fibre 104 is removably connected at one end to an interrogator 106. The output from interrogator 106 is passed to a signal processor 108, which may be co- located with the interrogator or may be remote therefrom, and optionally a user interface/graphical display 110, which in practice may be realised by an appropriately specified PC. The user interface may be co-located with the signal processor or may be remote therefrom.

The sensing fibre 104 can be many kilometres in length and can be, for instance 40km or more in length. The sensing fibre may be a standard, unmodified single mode optic fibre such as is routinely used in telecommunications applications without the need for deliberately introduced reflection sites such a fibre Bragg grating or the like. The ability to use an unmodified length of standard optical fibre to provide sensing means that low cost readily available fibre may be used. However in some embodiments the fibre may comprise a fibre which has been fabricated to be especially sensitive to incident vibrations. The fibre will be protected by containing it with a cable structure. In use the fibre 104 is deployed in an area of interest to be monitored which, in the present invention may be along the path of a railway as will be described. In operation the interrogator 106 launches interrogating electromagnetic radiation, which may for example comprise a series of optical pulses having a selected frequency pattern, into the sensing fibre. The optical pulses may have a frequency pattern as described in GB patent publication GB2,442,745 the contents of which are hereby incorporated by reference thereto, although DAS sensors relying on a single

interrogating pulse are also known and may be used. Note that as used herein the term "optical" is not restricted to the visible spectrum and optical radiation includes infrared radiation and ultraviolet radiation. As described in GB2,442,745 the

phenomenon of Rayleigh backscattering results in some fraction of the light input into the fibre being reflected back to the interrogator, where it is detected to provide an output signal which is representative of acoustic disturbances in the vicinity of the fibre. The interrogator therefore conveniently comprises at least one laser 112 and at least one optical modulator 114 for producing a plurality of optical pulses separated by a known optical frequency difference. The interrogator also comprises at least one photodetector 1 16 arranged to detect radiation which is Rayleigh backscattered from the intrinsic scattering sites within the fibre 104. A Rayleigh backscatter DAS sensor is very useful in embodiments of the present invention but systems based on Brillouin or Raman scattering are also known and could be used in embodiments of the invention.

The signal from the photodetector is processed by signal processor 108. The signal processor conveniently demodulates the returned signal based on the frequency difference between the optical pulses, for example as described in GB2,442,745. The signal processor may also apply a phase unwrap algorithm as described in

GB2,442,745. The phase of the backscattered light from various sections of the optical fibre can therefore be monitored. Any changes in the effective optical path length within a given section of fibre, such as would be due to incident pressure waves causing strain on the fibre, can therefore be detected. The form of the optical input and the method of detection allow a single continuous fibre to be spatially resolved into discrete longitudinal sensing portions. That is, the acoustic signal sensed at one sensing portion can be provided substantially

independently of the sensed signal at an adjacent portion. Such a sensor may be seen as a fully distributed or intrinsic sensor, as it uses the intrinsic scattering processed inherent in an optical fibre and thus distributes the sensing function throughout the whole of the optical fibre. The spatial resolution of the sensing portions of optical fibre may, for example, be approximately 10m, which for a continuous length of fibre of the order of 40km say provides 4000 independent acoustic channels or so deployed along a 40km section of transport network, such as a section of a rail network. This can provide effectively simultaneous monitoring of the entire 40km section of track. In an application to train monitoring the individual sensing portions may each be of the order of 10m in length or less.

As the sensing optical fibre is relatively inexpensive the sensing fibre may be deployed in a location in a permanent fashion as the costs of leaving the fibre in situ are not significant. The fibre may be deployed alongside or under the track (or road) and may for instance be buried alongside a section of track.

Figure 2 illustrates a section of traffic network, in this instance, a rail network 201 , having optical fibre buried alongside the tracks. In this example the track has three braches 202, 203 and 204. As mentioned above fibre optic sensing can be performed on fibre lengths of the order of 40 - 50km. However for some DAS sensors it can be difficult to reliably sense beyond 50km or so along a fibre. A length of 40-50km may be sufficient to monitor a desired section of track, say between main stations, and other fibres could be deployed to monitor other sections of track. For very long tracks it may be necessary to chain several DAS sensors together. Figure 2 illustrates one interrogator unit 106 arranged to monitor one optical fibre 104a deployed along one part of the track (including part of braches 202 and 204) and another optical fibre 104b deployed along another length of track (branch 202). The interrogator unit could house two lasers and detectors etc., i.e. dedicated components for each fibre or the laser and possibly detector could be multiplexed between the two fibres. After 40km say of fibre 104b another fibre could be deployed which is monitored by another interrogator unit. Thus there could be 80km or so between interrogator units. In this example branch 203 is also monitored by a DAS sensor using a different sensing fibre 104c which is connected to a different interrogator unit (not shown).

In use the interrogator operates as described above to provide a series of contiguous acoustic sensing channels along the path of the track branches. In use the acoustic signals generated by a train 205 in motion along the track 204 may be detected and analysed to determine the exact train location and the speed. The DAS sensor thus provides a monitoring system that can monitor long lengths of track with a high spatial resolution. As mentioned the sensing portions may be the order of metres in length. Deploying the sensor however simply involves laying a fibre optic cable along the path of the track - and in some instance suitable fibre optics may already be in place.

As a significant length of track can be monitored by contiguous sensing portions of fibre it can relatively straightforward to detect train movement along the track. Clearly movement of the train will create a range of noises, from the engine noise of the locomotive, noises from the train cars and the couplings and noise from the wheels on the track. The acoustic signals will be greatest in the vicinity of the train and thus be looking at the intensity of the signals detected by the sensor the returns from the sensing portions of fibre adjacent the current position of the train will exhibit a relatively high acoustic intensity. As illustrated in figure 3, which illustrates detected acoustic intensity against channel of the DAS sensor, the position of the train can thus be generally determined by detecting a continuous acoustic disturbance of relatively high intensity.

It is therefore possible to try to estimate the position of the front and rear of the train by detecting where the continuous disturbance starts 302 and ends 303. By monitoring the position of the front and rear of the train as it moves the exact location of the train is inherently provided. Note that no previous knowledge about the length of the train or the arrangement of cars that make up to the train is required although this information can be used to provide a data check if available.

As illustrated in Figure 3 it may be possible to detect the position of the front or rear of the train simply by looking at the intensity profile of the sensing portions and identify the beginning and end of a generally continuous acoustic disturbance above a certain intensity. However general noise associated with train movement may travel away from the train and the speed of sound, especially along the rails, will be much faster than the train speed. Thus in one embodiment a distinctive sound associated with the train passing track features may be used to detect where parts of the train are. For example in jointed track where there are gaps between rail sections there may be a noise generated at the wheel passes from one rail section to the next. This will create a relatively short duration relatively high intensity noise. A repeated pattern of such noises will then occur due to the passage of the various wheel sets over the same joint. Note that any track feature that leads to an acoustic response from a wheelset will generate such a repeated pattern of acoustic signals. This could be a rail junction, rail joint, rail weld or the like or a defect on the rail. It will be understood that rail joints tend to occur every few tens of metres of track. Thus typically there will be at most one rail joint per 10m sensing portion of fibre allowing the response of a single track feature to be detected by an individual sensing portion.

Figure 4a illustrates this principle. Figure 4a shows part of a train comprising three similar coupled cars 401 , for example boxcars, on a rail track 201. Each car 401 has a front bogie supporting two wheelsets and a rear bogie supporting two wheelsets.

Figure 4a shows this section of the train moving towards a feature 402 on the track, e.g. a rail joint, which will result in an acoustic signature as the wheelset moves over it. If the train moves at a constant speed this will result in a series of distinct acoustic signals as shown in Figure 4b. Figure 4b shows intensity against time (ignoring background noise of the general train motion for clarity).

It can therefore be seen that sensing portion in the vicinity of feature 402 will detect a repeated pattern of acoustic signals as the wheels pass the feature at that location. This allows the determination of the fact that wheels of the train are passing that location.

It is of course possible that the acoustic signal from the wheels the feature may travel to the next sensing portion. Given the level of general noise the signals from a given track feature may only be detectable by sensing portions close to that feature. If any signals do propagate relatively long distances and thus are detected by several sensing portions the time of arrival at the different portions can be determined to detect the earliest time of arrival - which obviously corresponds to the closest sensing portion to that feature. Thus by analysing the signals detected by the sensing portions at the front and rear of the continuous series of disturbances illustrated in Figure 3 sensing portion which detects the sounds of the wheelsets of the front or rear of the train can be determined. Thus the position of the train can be accurately determined.

In general the acoustic signal detected from a train by a DAS sensor will therefore produce a relatively intense acoustic signal which will be detected by the sensing portions of the DAS sensor in the vicinity of the train. Within this general signal there will be a series of identifiable acoustic signals corresponding to the wheelsets/axels of the train. These signals will typically be relatively intense broadband signals. Figure 6 shows the acoustic signals recorded by multiples channels of a DAS sensor acquired from a sensing fibre alongside a rail track when a train passed by. It can be seen that there is a clear acoustic signal detected by multiple contiguous channels of the DAS sensor which corresponds to the train. It can be seen that this signal is relatively well defined in terms of a leading and trailing edge of the high intensity signals. This can be used generally to determine the position of the front and rear of the train as discussed above. The presence of a repetitive series of broadband noise spikes is also readily apparent. These signals are due to the passage of the wheelsets/axles of the train along the track. These signals are detectable due to the high spatial resolution of the DAS sensor and the ability of the DAS sensor to detect acoustic signals at a range of frequencies from a long length of track.

A DAS sensor according to embodiments of the present invention may therefore be arranged to analyse the acoustic signals to detect such broadband noise spikes. From a detection of such broadband noise spikes the train is effectively monitored along its whole length as it moves along the track. The leading and trailing spike can be used to identify the front and rear of the train.

As the positions of trains and the front and the rear of the train can be accurately determined this information can be used for train control as illustrated in Figure 5. Figure 5 shows two trains 501 , 502 travelling along the same section of track 503 which is monitored by DAS interrogator 106 and sensing fibre 104. The acoustic signals from both trains can be detected and transmitted to control centre 504.

Alternatively some local processing may be done in the vicinity of the interrogator and the processed results transmitted to the control centre. The control centre analyses the received data and detects the characteristic signature of two trains. From the locations of both trains and by monitoring the change in location the speed of the trains is determined. The control centre knows where the rear of train 502 is and thus can control the speed of both trains to maintain a safe distance. This may be achieved in various ways. The control centre may transmit control signals to a trackside signal unit 505 to stop or slow train 501 if necessary. Alternatively or additionally the control centre may provide direct movement authority and speed limits to trains 501 and 502. The control centre may also control other network infrastructure such as level crossing barriers 506 across road 507 at an appropriate time given the speed of train 502.

In some parts of a rail network there may be several tracks which are relatively close together and which may be monitored by a single DAS sensor. For example there may be a single optic fibre laid along the path of tracks and used to monitor movement of traffic on each of the various tracks. In such case there may be a need to determine, when acoustic signals are detected, which track a train is moving on to generate the signals.

In some instances this may be apparent from prior motion of the train, for instance two tracks which run substantially alongside one another in a first part of the network, e.g. in the vicinity of the station, may diverge in a second part of the network. In the second part of the network the two racks may be monitored using different DAS sensors or by different parts of the same DAS sensor. If a train arrives from the second part of the network the track it is travelling on will be apparent. However if two trains arrive on the parallel tracks and both are stationary whilst in the station and then one train starts moving the DAS system should ideally be able to determine which train is moving. Identifying which track a train is on may be done in various ways. In one embodiment there may be an acoustic signature associated with each track. As mentioned above track features such as joins and the like may lead to a detectable signal in the acoustic noise generated by a train moving on the track. The distribution of such feature may be unique to each track and thus the acoustic signals may be analysed to determine the track by the patterns of the acoustic signals detected. As a simple example if one track has a feature that generates a relatively intense sound in a give acoustic channel when a train passes and the other track has no such feature at the same place so the same acoustic channel is not stimulated in the same way the acoustic signals from the relevant channel may be analysed to look for presence of a repetitive signal indicating train wheelsets passing the track feature. In some instances track features that lead to characteristics sounds when a train passes could be deliberately positioned so as to all unique identification of tracks which are close together.

Doppler shift techniques could also be applied to determine the lateral offset of the acoustic source from the sensing fibre and hence identify which track a train is moving on. As will be understood by one skilled in the art a signal at a given frequency will be Doppler shifted as the train moves relative to a sensing portion. The amount of Doppler shift experienced will depend on the offset of the sensing portion form the moving acoustic source - along with the speed of the train (and obviously the original frequency). For some acoustic signals the true frequency (e.g. of a train locomotive) may be know. The general speed of the train can be determined for the rate at which the acoustic signals are detected to move along the sensing fibre. Thus the amount of Doppler shift may be calculated and used to determine which track a train is moving on. In some applications the path of optical fibre used as the sensing fibre for DAS may be arranged to resolve any positional ambiguity. For instance the fibre could have a meandering arrangement that crosses under the various tracks allowing the relevant track to be identified by looking at the relative intensity of the signals detected and the time of detection at different sensing portions.

Thus although the sensing fibre may generally be arranged to run along the path of the tracks in some instances it may be desired to vary the path of the fibre to provide some additional sensing functionality. For instance in some instances a certain length of fibre may be arranged in one or more fibre loops to provide increased sensitivity and/or better spatial resolution in one or more area. As will be understood by on skilled in the art the sensing portions of optical fibre in DAS sensor are defined by the properties of the interrogating light pulse(s) and the processing of the backscatter returns. If the sensing portions are 10m long say and arranged to run longitudinally alongside the track then any acoustic signal acting on that 10m section of fibre may be detectable. If however a relatively low intensity stimulus is only affecting part of the 10m section then the effect on the backscattered light may not be large and the detected signal will be faint. If however the 10m length were coiled into a i m diameter coil say then any stimulus affecting that 1 m would excite the whole 10m section. This would improve sensitivity (at the cost of requiring a greater length of fibre and the need to arrange it in a coil). In some areas however extra sensitivity may be required, for instance near points or track

intersections/branches. The optical fibres used for the DAS sensor(s) may therefore, in some locations, be arranged to provide one or more areas of increased sensitivity. In some cases there may even be one or more fibre optic point sensors integrated with the distributed acoustic sensor portions. When fibre optic cable is being deliberately laid to provide monitoring the fibre may be laid in a desired arrangement. In some instances however monitoring may be arranged on fibre optic cables which were previously installed for other reasons, for example for communications. In many such installations along linear structures the fibre optic cable may generally run along the path of the structure, e.g. the rail track, but there may be some areas where additional cable length was provided. In other words in some sections the rail track there may be x metres of fibre optic cable (or just slightly more) for a section of track which is x metres. In another section however a length of x metres of track may be provided with x+y meters of cables where y may be several metres or tens of metres of cable. In these areas therefore part of the cable may be looped in one or more fibre loops. Such fibre loops may accidentally result when laying the cable or they may be deliberately introduced to provide 'spare' cable in the event of the need to reposition the cable at a later date or to remove a damaged section of cable.

The presence of such loops will of course create a mismatch in certain areas between the length of the sensing fibre and the length of the rail track being monitored. The presence of such loops may therefore be detected in an initial calibration/set-up phase for a DAS monitoring system and the signal returns from such loops may therefore be discounted for determining train motion. However such loops may provide areas of relatively high sensitivity of greater spatial resolution.

Claims

1. A method of control of a rail network comprising:

monitoring at least part of the rail network by performing distributed acoustic sensing on one or more optical fibres deployed along the path of the network to provide a plurality of acoustic sensing portions; and analysing the signals detected by said plurality of acoustic sensing portions to detect acoustic signals associated with trains moving on the network;

where analysing said signals comprises identifying the front and rear of said

trains and tracking the movement of trains on the network.

2. A method as claimed in claim 1 comprising analysing the acoustic response from the acoustic sensor portions to locate an acoustic signal indicative of the front of the train and an acoustic signal associated with the rear of the train.

3. A method as claimed in claim 2 comprising identifying the beginning and end of a continuous acoustic disturbance indicative of the train.

4. A method as claimed in any preceding claim comprise identifying acoustic signals associated with at least one wheelset of the train.

5. A method as claimed in claim 4 wherein identifying acoustic signals associated with a wheelset comprises identifying a relatively intense broadband noise component in the acoustic signals detected by said acoustic sensing portions.

6. A method as claimed in claim 4 or claim 5 comprising identifying acoustic signals associated with the first and last wheelsets of a train.

7. A method as claimed in any of claims 4 to 6 comprising identifying acoustic

signals associated with the wheelsets of the train passing track features.

8. A method as claimed in any preceding claim comprising determining the speed of a train by monitoring the change in position of the acoustic signals associated with the train.

9. A method as claimed in any preceding claim wherein tracking the movement of trains on the network is used to set one or more control conditions governing movement of traffic on the network.

10. A method as claimed in claim 9 wherein setting one or more control conditions comprises controlling signals to allow trains to move or not and/or issuing movement authority to trains.

1 1. A method as claimed in claim 9 or claim 10 wherein setting one or more control conditions comprises controlling the positions of blocks in a moving block signalling system.

12. A method as claimed in claim 9 or claim 10 wherein setting one or more control conditions comprises at least one of: controlling speed limits for at least part of the rail network; activating emergency braking procedures of a train; providing movement control for automatically operated rail vehicles; and controlling the operation of rail crossing barriers..

13. A method as claimed in any preceding claim comprising maintaining an indication of the position of the front and rear of a non-moving train on the network.

14. A method as claimed in any preceding claim further comprising analysing the acoustic signals generated by movement of the train to detect any departure from an expected sound profile.

15. A method as claimed in any preceding claim further comprising analysing the acoustic signals generated by movement of the train to detect any acoustic signals characteristic of undesired train movement behaviour.

16. A method as claimed in any preceding claim further comprising comparing the acoustic profile from a given section of track when a train is travelling on or near it to one or more previous profiles to detect any change in profile.

17. A method as claimed in any preceding claim further comprising analysing the signals detected by said plurality of acoustic sensing portions to detect unauthorised access onto the network. A rail traffic control system comprising:

at least one distributed acoustic sensor configured to provide distributed acoustic sensing on one or more optical fibres deployed along the path of a rail network to provide a plurality of acoustic sensing portions; and

a processor configured to analyse the signals detected by said plurality of

acoustic sensing portions to detect acoustic signals associated with trains moving on the network;

wherein said processor is configured to identify the front and rear of said trains and to track the movement of trains on the network.

A rail traffic control system as claimed in claim 18 further comprising a control unit configured to use said movement tracking to set one or more control conditions governing movement of traffic on the network.