RAILWAY TRACK QUALITY

Field of the Invention

[0001] The present invention relates to a computer system and computer implemented method of predicting track quality of a section of railway track.

Background of the Invention

[0002] In the railway industry, rail data measurements of track geometry are typically used to measure track quality. This track quality is normally measured in terms of the standard deviation of vertical and lateral alignment for each 1/8th mile (200 meters) section of railway track. Track maintenance engineers often use limited forecasting, for example by comparing the current standard deviation of each l/8th mile section with a previously obtained standard deviation. Unfortunately, this approach does not provide track maintenance engineers with a reliable prediction of the future track quality nor does it identify the durability of any tamping (i.e. whether any maintenance is effective or whether the track quickly returns to its pre- tamping geometry) . This tamping durability can vary considerably between otherwise similar track sections and depends, not only on track traffic type and track tonnage, but to a large extent on local conditions. Summary of the Invention

[0003] According to one embodiment of the invention there is provided a computer implemented method of predicting track quality of a section of railway track, the method comprising: accessing sets of track geometry quality data values for the section of railway track, wherein the sets of track geometry quality data values are obtained from one or more track geometry measuring devices traveling along the section at spaced time intervals and each of the sets of track geometry quality data values comprise at least one data channel; calculating, for the section of railway track, variance based values for at least one channel for each of the sets; calculating, for the section of railway track, degradation rates for each of the sets of the least one channel, wherein each of the degradation rates are based on a difference value between two of the variance based values over a time period therebetween; identifying if at least one of said degradation rates is a tamped degradation rate indicative an occurrence of tamping of the section of railway track; and determining a quality estimate for the section of railway track based on at least two selected said calculated degradation rates and a spaced time interval between dates associated with the selected calculated degradation rates.

[0004] According to another embodiment of the invention there is provided a computer system arranged to predicting track quality of a section of railway track, the computer system comprising: a memory storing sets of track geometry quality data values for the section of railway track, wherein the sets of track geometry quality data values are obtained from one or more track geometry measuring devices traveling along the section at spaced time intervals and each of the sets of track geometry quality data values comprise at least one data channel; a processor that in combination with the memory is arranged to: access the sets of track geometry quality data values; calculate, for the section of railway track, degradation rates for each of the sets of the least one channel, wherein each of the degradation rates are based on a difference value between two of the variance based values over a time period therebetween; identify if at least one of said degradation rates is a tamped degradation rate indicative an occurrence of tamping of the section of railway track; and determine a quality estimate for the section of railway track based on at least two selected said calculated degradation rates and a spaced time interval between dates associated with the selected calculated degradation rates.

[0005] According to another embodiment of the invention there is provided a non-transistory computer readable media which, when executed by a computer, is arranged to perform the method as recited above. Brief Description of the Drawings

[0006] For a better understanding of the invention and to show how the same may be carried into effect, there will now be described by way of example only, specific embodiments, methods and processes according to the present invention with reference to the accompanying drawings in which :

Figure 1 is a schematic block diagram of a computer system, for predicting degradation of a section of railway track, in accordance with a preferred embodiment of the present invention;

Figure 2 is a flow chart illustrating a computer implemented method of predicting track quality of a section of railway track, in accordance with a preferred embodiment of the present invention; and

Figure 3 is a flow chart illustrating a storing process of figure 2, in accordance with a preferred embodiment of the present invention.

Detailed Description of the Embodiments

[0007] The detailed description set forth below in connection with the appended drawings is intended as a description of presently preferred embodiments of the invention, and is not intended to represent the only forms in which the present invention may be practised. It is to be understood that the same or equivalent functions may be accomplished by different embodiments that are intended to be encompassed within the spirit and scope of the invention. In the drawings, like numerals are used to indicate like elements throughout. Furthermore, terms "comprises, " "comprising, " or any other variation thereof, are intended to cover a non-exclusive inclusion, such that module, circuit, device components, structures and method steps that comprises a list of elements or steps does not include only those elements but may include other elements or steps not expressly listed or inherent to such module, circuit, device components or steps. An element or step proceeded by "comprises ...a" does not, without more constraints, preclude the existence of additional identical elements or steps that comprises the element or step.

[0008] Referring now to Figure 1 there is illustrated a schematic block diagram of a computer system 100, for predicting degradation of a section of railway track, in accordance with a preferred embodiment of the present invention. The computer system 100 includes a processor 102 and a memory unit 104 that typically comprises both a Read Only Memory (ROM) 106 and a Random Access Memory (RAM) 108. There is also a transceiver 110 and a user interface 112 examples of which include one or more of a keypad, a conventional display screen, a touch screen and a mouse. The memory unit 104 is coupled by an address and data bus 120 to the processor 102. The transceiver 110 and a user interface 112 are also coupled to the processor 102 by respective interconnects 130, 140.

[0009] The processor 102 that in combination with the memory unit 104 is arranged to perform the processes or method described below with reference to figures 2 and 3.

[00010] Referring to Figure 2 there is a flow chart illustrating a computer implemented method 200 of predicting track quality of a section of railway track, in accordance with a preferred embodiment of the present invention. The method 200 commences at a start block 205 and includes at a storing block 210 that provides for storing sets of track geometry quality data values for the section of railway track. In this embodiment, the sets of track geometry quality data values are obtained from one or more track geometry measuring devices and are stored in the memory unit 104. The track geometry measuring devices obtain the data by traveling along the section at recorded dates, spaced by time intervals and collect raw data that is processed to provide the track geometry quality data values. In this embodiment, the track geometry measuring devices collected the data values in either an Unmanned Geometry Measuring System (UGMS) or Track Recording Vehicle (TRV) format.

[00011] Each of the sets of track geometry quality data values are obtained over a period of time on the recorded dates when a train with a geometry measuring kit travelled along the section. In this embodiment each set of the sets of geometry quality data values comprise four data channels that are known to a person skilled in the art as: AL35m; AL70m; Top35m; and Top70m.

[00012] At an accessing block 215 the sets of track geometry quality data values for the section of railway track are accessed from the memory unit 104. Next, a calculating block 220, performed by the processor 102 in combination with the memory unit 104, provides for calculating, variance based values that in this embodiment are standard deviation values SD for at least one data channel (AL35m, AL70m, Top35m or Top70m) for each of the sets. The variance based values are the square of the standard deviation values SD and the standard deviation values SD are calculated as follows:

where n is the number of data points along the section of railway track in the specific set, Xk is the value of a geometry parameter at point k in the section of railway track, and x is the average value of all the data points in the specific set. For the rest this specification standard deviation values SD will be described although the invention is not necessarily limited to use of standard deviation values used and it will be appreciated that variance values or other variance based values can be used. Thus for the section of track identified by a milepost m and index i along with time (date) k of the when the set for a channel was obtained, the standard deviation SD _imk represents the location in miles, index and time of when the set for a channel was obtained.

[00013] At another calculating block 225 there is performed a process of calculating, for the section of railway track, degradation rates D for each of the sets of for each channel. The calculating block 225 is performed by performed by the processor 102 in combination with the memory unit 104. More specifically, at the calculating block 225 each of the degradation rates D are based on a difference value between two of the standard deviation values SD over the time period between the obtaining of the sets. Thus each degradation rate D is defined as a variance in track quality value over time and is calculated as follows:

where T _k is an inspection date in days identifying the specific one of the sets j of track geometry quality data values. Thus, T _(k+1) - T _k is the difference in days (time period) between the obtaining of the raw data for the values of SD _im(k+1) and SD _imk .

[00014] Once the degradation rate for each channel of each set has been calculated the method 200, at an identifying block 230, provides for identifying if at least one of the degradation ratesD _imk is a tamped degradation rate indicative an occurrence of tamping of the section of railway track. When tamping data is available the degradation rates D _imk affected by known tamping are designated as tamped degradation rates N _imk which are set to 1 or 2 depending on the following:

[00015] tamping data may not be available for th section of railway track then threshold estimates are used to identify and set the tamped degradation rates N _imk . Thus, for the four channels AL35m; AL70m; Top35m; and Top70m the degradation rates N _imk are identified and set as follows:

[00016] By way of example, table 1 below illustrates ten degradation rates for eight sections of railway track (i=l to 8) for the 65 ^th mile of the track on a specific railway line. Each of the ten degradation rates Dl to D10 are listed in rows. As shown, if any of the ten degradation rates Dl to D10 for a section are identified as a tamped degradation rate then its calculated degradation rate D _imk is replaced with the tamped degradation rate N _imk of 1 or 2

Table 1 Degradation rates over time. [00017] At determining block 235, a quality estimate S _imk (or settlement rate) is determined for each section of railway track. The quality estimate S _imk is based on at least two selected calculated degradation rates and a spaced time interval between dates associated with the selected calculated degradation rates. For instance, consider the following four situations:

[00018] Case 1 where there are no tamped degradation rates N _imk in a section (i.e. in table 1 when i=3,4 and 8), then : -

k(N) = 0, V k

[00019] Case 2 where there is only one tamping degradation rate N _imk in a section and it is the most recent degradation rate (k=10, i.e. in table 1 when i=2), then:- k (N) =1, and K=max(k)

[00020] Hence, when the tamped degradation rate N _imk in a section is the most recent degradation rate for the period of time, the determining is based on all calculated degradation rates D(k=l) to D(k9) occurring before the occurrence of tamped degradation rate D(kl0) .

[00021] Case 3 where there is more than one tamping degradation rate N _imk in a section (i.e. in table 1 when i=l) , then: - k(N)=2, k(N)==l < k < k(N)=2

[00022] Hence, the determining of the quality estimate S _imk is characterised by the occurrence of the least two of the calculated degradation rates, and ideally every one of the calculated degradation rates D(k=4) to D(k9), are between occurrences of the two tamped degradation rates at D ( k=3 ) and D (klO) .

[00023] Case 4 where there is only one tamping degradation rate N _imk in a section and there are two or more degradation rates after the tamping degradation rate N _imk (i.e. in table 1 when i=6 or 7), then:- k(N)=l and K(kmax - 1 )≥2

[00024] Hence, determining of the quality estimate S _imk is characterised by the occurrence of the least two of the calculated degradation rates D(k=7) to D(kl0), being after the occurrence of the tamped degradation rate D(k=6) .

[00025] Case 5 where there is only one tamping degradation rate N _imk in a section and there are less than two degradation rates after the tamping degradation rate N _imk (i.e. in table 1 when i=5) , then:- k(N)=1 and K(kmax - 1)<2

[00026] Hence, the determining of the quality estimate S _imk is characterised by the occurrence of the calculated degradation rates D(k=l) to D(k=8) is before the occurrence of the tamped degradation rate D(k=8) of the quality estimate S _imk .

[00027] At a predicting block 240, the method 200 performs predicting a track quality at a future date Fdate. The track quality predicted includes a degradation rate D _imk of the section of track based on the quality estimate S _imk , a number of days T _x from a reference date Ref until the future date Fdate and a degradation rate D _im(Kmax) of the track at the reference date Ref. This can be expressed in the following equation.

[00028] After the predicting block 240, a block 245 outputs the results at the user interface 112 and/or sends the results to a remote location via the transceiver 110. The method 200 then ends at an end block 245.

[00029] Referring to figure 3 there is a flow chart illustrating the storing process of block 210, in accordance with a preferred embodiment of the present invention. At a block 305 the process receives raw data from obtained from one or more track geometry measuring devices that have travelled along the section of railway track on different dates. In this embodiment the raw data is in either the Unmanned Geometry Measuring System (UGMS) or Track Recording Vehicle (TRV) format as will be apparent to a person skilled in the art. At a block 310, the raw data is sorted by use of an Engineering Line Reference (ELR) code and the necessary geometry channels are extracted which in this embodiment includes 12 such channels namely: Location Miles, Location Yards, AL35m, AL70m, Top35m, Top70m, Gauge, Cant, Curvature, Velocity, Twist .

[00030] At a block 315 the method cleans the data by checking for various anomalies including: spikes which are massively over standard thresholds, gradient anomalies and drop outs that are flagged and stored as a separate data future analysis if required. The data associated with the spikes (relatively fast transients that rise above or fall below set threshold values) is removed which results in a cleansed data set.

[00031] Because of alignment issues with UGMS data formats a decision block 320 determines if alignment is required. If the cleansed data set was obtained from a UGMS data format then data alignment is required, otherwise if the data set was obtained from a TRV data format then no aligning is required. When aligning is required a process of aligning the data in the data set is performed at an aligning block 325. However when aligning is not required the process 300 goes to a storing block 330.

[00032] At the aligning block 325 an aligning process aligns nonaligned sets of track geometry quality data values are data points that are aligned by use of a reference set of data points. The aligning process filters the data in the sets to reduce noise. The data is filtered using a median filter that runs through the data point by point, replacing each point with the median of neighbouring points. Next, splitting points on both reference data set and nonaligned inspection data set are identified. The splitting points are used to partition the data sets into sections and the reference set is also partitioned in the same manner. The alignment process then employs one of three methods to carry out the alignment for a nonaligned section to its corresponding reference section. These three methods are an average, a scale and an interpolation method and are described below.

[00033] The average method calculates the average value of all signal points on X coordinates of both sections, separately. Then the difference between two average values will be regarded as the shift between two signal sections. This method can be described as follows:

Where S _average is the shift; is the average value of all signal points of X coordinates of reference inspection; is the average value of all signal points of X

coordinates of nonaligned inspection. [00034] The scale method can be described by the following equations:

Where is the X value of the last point in the reference

section; is the X value of the first point in the same

section with is the X value of the last point in the

nonaligned section; is the X value of the first point in

the same section; is the reference section length; and is the length of a corresponding section to be aligned.

[00035] The interpolation method generates the same number of points in the corresponding the nonaligned section and its corresponding reference section. Firstly, the number of points of the two sections are compared and the smaller number of sections is selected as follows:

Where n _r is the number of points of the reference section and n _n is the number of points of nonaligned section .

The new steps of X data coordinates can be obtained by:

Once setp _x and setp ₂ are obtained they are used to interpolate and generate new coordinates of points of reference and nonaligned sections, separately.

[00036] After alignment, or when alignment is not required, the cleansed data set or cleansed data and aligned data set is stored, at the storing block 330, as one of the sets of track geometry quality data values in the RAM 108.

[00037] Advantageously the present invention alleviates the problem associated with reliable predictions of the future track quality or track durability identification resulting from tamping.

[00038] The description of the preferred embodiments of the present invention has been presented for purposes of illustration and description, but is not intended to be exhaustive or to limit the invention to the forms disclosed. It will be appreciated by those skilled in the art that changes could be made to the embodiments described above without departing from the broad inventive concept thereof. It is understood, therefore, that this invention is not limited to the particular embodiment disclosed, but covers modifications within the spirit and scope of the present invention as defined by the appended claims.