[TECHNICAL FIELD]
[0001] The present disclosure relates to a track circuit state determination apparatus that
determines the state of an AC track circuit.
[BACKGROUND ART]
[0002] Track circuits in railway traffic are apparatuses that detect the presence or absence
of an on-rail train by the use of rails as part of an electric circuit. The track
circuits are configured to transmit a signal from one end side of the rails and detect
the presence or absence of a received signal resulting from a short-circuit between
the rails on the axle of the on-rail train by a track relay provided on the other
end side of the rails. Since the track circuits are installed outdoors, there may
occur a problem that the track relay abnormally drops under the influence of a natural
environment such as rainfall and snow coverage. Thus, there are known various techniques
for detecting abnormal state of AC track circuits (for example, refer to Patent Documents
1 and 2).
[RELATED-ART DOCUMENTS]
[PATENT DOCUMENTS]
[0003]
[Patent Document 1] Japanese Unexamined Patent Application Publication No. 1992-113941
[Patent Document 2] Japanese Unexamined Patent Application Publication No. 1999-278269
[SUMMARY OF THE INVENTION]
[TECHNICAL PROBLEM]
[0004] The conventional state monitoring of track circuits is performed by comparing the
transitions of signals (for example, voltage value, current value, phase difference,
and the like) on the transmission side or the reception side with a predetermined
threshold to determine whether the track circuits are in the normal or abnormal state.
However, the track circuits differ in the circuit length (rail length), the distance
between the transmitter/receiver and the rails (cable length), and the parameters
of circuit elements including rails and ballast. This requires maintenance personnel
(user) to set a threshold appropriate for each track circuit depending on their experience
and knowledge. Thus, there has been demand for a new technique by which to mechanically
perform the state determination of track circuits without depending on the knowledge
the maintenance personnel (user) has.
[0005] An issue resolved by the present disclosure is to provide a new technique for determining
the state of AC track circuits.
[SOLUTION TO PROBLEM]
[0006] A first aspect to achieve the foregoing object is
a track circuit state determination apparatus that determines the state of an AC track
circuit, comprising:
a storage that stores a reference vector locus of a current vector to a voltage transmitted
to the AC track circuit;
a calculation section that, based on measurement values of a transmission voltage
and a transmission current of a transmitter in the AC track circuit measured by a
transmission-side measurer, calculates a current vector of the transmission current
to the transmission voltage; and
a determination section that determines the state of the AC track circuit by calculating
a vector locus from a locus of the current vector calculated by the calculation section
for a predetermined period of time and comparing the vector locus with the reference
vector locus.
[0007] As a result, in the first aspect, it is possible to determine the state of the AC
track circuit including at least a normal state or an abnormal state by a new method
by which to compare the current vector locus to the transmission voltage with the
reference vector locus.
[0008] A second aspect is the track circuit state determination apparatus in the first aspect,
in which
the storage stores an on-rail train-present reference vector locus that is the reference
vector locus in the presence of an on-rail train in the AC track circuit, and
the determination section calculates the vector locus from the current vector calculated
by the calculation section in the presence of an on-rail train in the AC track circuit
and compares the vector locus with the on-rail train-present reference vector locus.
[0009] As a result, in the second aspect, it is possible to determine the state of the track
circuit, targeting at the current vector locus in the presence of an on-rail train.
[0010] A third aspect is the track circuit state determination apparatus in the first or
second aspect, in which
the storage stores an on-rail train-absent reference vector locus that is the reference
vector locus in the absence of an on-rail train in the AC track circuit, and
the determination section calculates the vector locus from the current vector calculated
by the calculation section in the absence of an on-rail train in the AC track circuit
and compares the vector locus with the on-rail train-absent reference vector locus.
[0011] As a result, in the third aspect, it is possible to determine the state of the track
circuit, targeting at the current vector locus in the absence of an on-rail train.
[0012] A fourth aspect is the track circuit state determination apparatus in the second
or third aspect,
further comprising a segmentation section that, when the current vector calculated
by the calculation section satisfies a predetermined steep-change condition, divides
the current vector into a segment of the current vector in the presence of an on-rail
train from the satisfaction of the steep-change condition to return to the current
vector before the satisfaction and the other segment of the current vector in the
absence of an on-rail train.
[0013] The current vector hardly changes in the absence of an on-rail train. However, once
a train passes, the current vector steeply changes and then recovers to the original
current vector or its proximity. Thus, as in the fourth aspect, the current vector
can be divided into the segment in the presence of an on-rail train and the segment
in the absence of an on-rail train according to changes in the current vector at each
passage of a train through the track circuit.
[0014] A fifth aspect is the track circuit state determination apparatus in any of the first
to fourth aspects, in which
the storage stores a plurality of the reference vector loci in association with accompanying
information indicating at least a situation including one of season, time zone, and
weather condition in which the AC track circuit was operating, and
the determination section selects the reference vector locus satisfying a predetermined
condition in proximity with the situation in which the measurement was performed as
a comparative target and conducts the comparison.
[0015] Since the track circuit is installed outdoors, the transmission current varies depending
on an external environment involved in rain and temperature, which causes a change
in the current vector locus. Thus, as in the fifth aspect, selecting and using the
reference vector locus that is in proximity in the measurement situation of the transmission
voltage and the transmission current including season, time zone, and weather condition
for comparison with the current vector locus, allows higher-accuracy state determination
of the track circuit.
[0016] A sixth aspect is the track circuit state determination apparatus in any of the first
to fifth aspects, in which
the reference vector locus is data generated as an appearance probability distribution
at each locus position based on vector loci of past current vectors calculated by
the calculation section, and
the determination section calculates an evaluation value of a vector locus to be determined
based on an appearance probability on the appearance probability distribution traced
by the vector locus to be determined and determines the state of the AC track circuit
based on the evaluation value.
[0017] As a result, in the sixth aspect, the reference vector locus is the appearance probability
distribution at each locus position based on the past current vector loci. Thus, the
evaluation value of the current vector locus to be determined can be determined as
the appearance probability that is the degree of agreement with the past current vector
loci.
[BRIEF DESCRIPTION OF DRAWINGS]
[0018]
FIG. 1 is a diagram showing an application example of a track circuit state determination
apparatus.
FIG. 2 is a diagram describing a current vector.
FIG. 3A is a diagram describing a current vector locus.
FIG. 3B is a diagram describing a current vector locus.
FIG. 4 is a diagram describing segments of the current vector.
FIG. 5A is a diagram describing generation of an appearance probability distribution.
FIG. 5B is a diagram describing generation of an appearance probability distribution.
FIG. 6 is a diagram describing calculation of a degree of abnormality.
FIG. 7 is a functional configuration diagram of the track circuit state determination
apparatus.
FIG. 8 is a diagram showing an example of current vector locus data.
FIG. 9 is a diagram showing an example of determination result data.
FIG. 10 is a diagram showing an example of reference vector locus data.
FIG. 11 is a flowchart of a track circuit state determination process.
[DESCRIPTION OF EMBODIMENTS]
[0019] Preferred embodiments of the present disclosure are described below with reference
to the drawings. The present disclosure is not limited by the embodiments described
below, and embodiments to which the present disclosure is applicable are not limited
to the following embodiments. In the drawings, identical elements are denoted with
identical reference numerals.
[System configuration]
[0020] FIG. 1 shows an application example of a track circuit state determination apparatus
100 in the present embodiment. As shown in FIG. 1, there is a track provided with
track circuits 1T, 2T, 3T, ... in sections of left and right rails R with a predetermined
length. The track circuits are apparatuses that detect the presence of an on-rail
train, taking advantage of the fact that the left and right rails R are electrically
short-circuited by a wheelset of the train. In the present embodiment, double-rail
track circuits are provided with
rail joint insulation 1 on the left and right rails R at the section boundary between the track circuits.
Provided at the boundary between the track circuits are two sets of impedance bonds
3 with the
rail joint insulation 1 therebetween.
[0021] A transmission transformer 7 as a transmitter is connected across the rails R on
one end side (transmission side) of each track circuit via the impedance bonds 3 and
a current decreasing resistor 5. A track relay 11 is connected across the rails on
the other end side (reception side) via the impedance bonds 3 and a phase adjustor
9. The current decreasing resistor 5 is provided to limit current and prevent burning
of the devices.
[0022] The transmission transformer 7 transforms AC power supplied from a power source 21
such as a commercial power source to generate a track signal (train detection signal),
and transmits the signal across the rails R on the transmission side of each track
circuit. That is, the track circuits in the present embodiment are AC track circuits.
[0023] The track relay 11 is a binary track relay that has two coils, a track coil and a
local coil, and drives a contact by the voltages applied to the coils and a phase
difference between them. The track coil is connected across the rails R on the reception
side of the track circuit and to which the voltage of the track signal flowing through
the track circuit is applied. An AC voltage supplied from the power source 21 is applied
to the local coil. The voltage applied to the local coil (hereinafter called "local
voltage") is stable in phase (also referred to period). Thus, the phase of the local
voltage is used as a reference.
[0024] At entry of a train into the track circuit, the rails R are short-circuited by the
axle of the train so that the voltage applied to the track coil in the track relay
11 (hereinafter called "reception voltage") and also referred to "incoming voltage")
decreases and the phase difference between the reception voltage and the local voltage
becomes smaller. Thus, the track relay 11 changes from a raised state to a dropped
state, whereby the entry of the train into the track circuit is detected. The phase
adjustor 9 is provided to adjust the phase of the reception voltage such that the
phase difference between the reception voltage and the local voltage in the absence
of an on-rail train becomes a value optimally suited for keeping the raised state
of the track relay 11.
[0025] The track circuit state determination apparatus 100 is configured such that a plurality
of measurement terminals 200 and a processing apparatus 300 are communicably connected
by a transmission line 102 to determine states of individual track circuits to be
determined.
[0026] Each measurement terminal 200 is provided at the section boundary between the track
circuits and into which a voltage of the track signal generated by the transmission
transformer 7 (transmission voltage) and current (transmission current) are input
as measurement values relating to one of the track circuits adjacent to each other
at the boundary, and into which a contact condition of the track relay 11 is input
as a measurement value relating to the other of the track circuits adjacent to each
other at the boundary. The measurement terminal 200 calculates a phase difference
of the transmission current from the transmission voltage (transmission current phase
difference), and outputs the calculated phase difference together with the input measurement
values to the processing apparatus 300 via the transmission line 102.
[0027] The transmission voltage is measured by a potential transformer (PT) 13 that is a
transmission-side measurer connected to a secondary side of the transmission transformer
7. The transmission current is measured by a current transformer (CT) 15 that is a
transmission-side measurer inserted into between the secondary side of the transmission
transformer 7 and the rail R. The transmission current may be calculated by detecting
a voltage across the current decreasing resistor 5.
[0028] The processing apparatus 300 is a kind of computer that includes an electronic circuit
performing arithmetic control. The processing apparatus 300 determines the state of
each track circuit including at least a normal state or an abnormal state based on
the measurement values input from the corresponding measurement terminal 200.
[Determination principle]
[0029] A principle of state determination on the track circuit by the processing apparatus
300 will be described. The processing apparatus 300 determines the state of the track
circuit to be determined by calculating a current vector of the track circuit from
the measurement values input from the measurement terminal 200 relating to the track
circuit to be determined and comparing a locus of the current vector with a reference
vector locus.
[0030] FIG. 2 is a diagram describing a current vector. As shown in FIG. 2, the current
vector is a vector (x, y) that has an origin O as the starting point in an XY orthogonal
coordinate system in which the voltage vector is oriented in an X-axis positive direction,
and has a phase difference θ from the X axis as transmission current phase difference
and has a magnitude as transmission current value. The measurement values including
the transmission current and the transmission current phase difference input from
the measurement terminal 200 are associated with measurement time. Thus, the current
vector at each measurement time can be calculated from the transmission current and
the transmission current phase difference at the measurement time. Time-series variations
in the current vector (x, y) at the continuous measurement times constitute the locus
of the current vector.
[0031] The shape of the locus of the current vector varies depending on whether a train
has entered in the corresponding track circuit (in the presence of an on-rail train)
or not (in the absence of an on-rail train). FIGS. 3A and 3B are diagrams schematically
showing examples of current vector loci. FIG. 3A shows a current vector locus in the
absence of an on-rail train for a certain period of time, and FIG. 3B shows a current
vector locus in the presence of an on-rail train for one cycle of the train from the
entry into to exit from the corresponding track circuit.
[0032] As shown in FIG. 3A, in the absence of an on-rail train, the transmission current
and the transmission voltage are almost constant, and thus the current vector hardly
changes and the current vector locus almost concentrates on one point.
[0033] On the other hand, in the presence of an on-rail train, as shown in FIG. 3B, the
transmission voltage is almost constant but the transmission current greatly changes
because the impedance varies depending on the short-circuit position of the rails
caused by the axle of the running train. That is, as the transmission current value
and the transmission current phase difference greatly change, the current vector greatly
changes and has a current vector locus that fluctuates more widely than that in the
absence of an on-rail train. In addition, the changes in the current vector are steeper
than those in the absence of an on-rail train. Specifically, the current vector locus
takes a crescent-like shape that makes a closed path at each passage of a train through
the corresponding track circuit. FIG. 3B shows a current vector locus at each passage
of a train from the entry into to exit from the corresponding track circuit. In the
current vector locus, the current vector in the absence of an on-rail train changes
due to the entry of the train such that the current phase difference becomes smaller,
and then changes again such that the current phase difference becomes larger, thereby
recovering to near the current vector before the entry (that is, the current vector
in the absence of an on-rail train). The current vector loci vary by the track circuit,
but form almost the same shape from track circuits of the same kind in the same state.
[0034] As described above, the current vector locus greatly differs depending on the presence
or absence of an on-rail train. Thus, the current vector locus is divided into segments
based on the difference to perform the state determination of the track circuit. FIG.
4 is a diagram schematically describing segments of the current vector locus. FIG.
4 outlines three-dimensionally time-series changes of a current vector of a certain
track circuit on an XY plane that is a vector plane along a depth direction and a
vertical direction and at time along a rightward direction. The XY axes are the same
as those shown in FIGS. 3A and 3B, and the X-axis positive direction is the direction
of the voltage vector.
[0035] As trains intermittently pass through the track circuit, an on-rail train-present
period and an on-rail train-absent period alternate repeatedly. Thus, the on-rail
train-present period and the on-rail train-absent period are each segmented as one
determination period. The segmentation boundary between the determination periods
can be determined depending on whether the current vector satisfies a predetermined
steep-change condition. The steep-change condition is a condition under which the
current vector can be regarded as having steeply changed, and for example, the steep-change
condition is set to one of the following: for a predetermined unit time, 1) the magnitude
of the current vector has changed by a first change amount or more and the orientation
of the current vector has changed by a first change angle or more; 2) the magnitude
of the current vector has changed by a second change amount or more; and 3) the orientation
of the current vector has changed by a second change angle or more.
[0036] That is, in the absence of an on-rail train, the current vector hardly changes (see
FIG. 3A) and thus does not satisfy the steep-change condition. When a train has entered
the track circuit, the current vector greatly changes within the unit time and thus
satisfies the steep-change condition. During the running of the train in the track
circuit, the current vector continues to satisfy the steep-change condition. Then,
when the train has exited the track circuit and the current vector has recovered to
near the state before the entry, the current vector no longer satisfies the steep-change
condition (see FIG. 3B). Therefore, assuming that a time point at which the current
vector has changed from the state not satisfying the steep-change condition to the
state satisfying the steep-change condition is a time point of the train's entry into
the track circuit, the on-rail train-present period is defined as a period from the
foregoing time point to a time point at which the current vector has recovered to
near the state immediately before the satisfaction of the steep-change condition.
The on-rail train-absent period is defined as the other period. The current vector's
"recover to near" the previous state means that the current vector has reached a coordinate
value that is almost equivalent to a coordinate value of the previous current vector.
The range of almost equivalent values can be set as appropriate. In short, the current
vector's "recover to near" the previous state can be said to the current vector's
"return" to the previous state, and thus the term "return" will be used as appropriate
in the present embodiment.
[0037] As described above, the state of the corresponding track circuit is determined by
dividing the current vector locus into the determination periods that are the on-rail
train-present period and the on-rail train-absent period and comparing each determination
period with the reference vector locus based on the past current vector loci. That
is, the current vector locus in the on-rail train-present period is compared with
the reference vector locus based on the current vector loci in the past on-rain train-present
periods, and the current vector locus in the on-rail train-absent period is compared
with the reference vector locus based on the current vector loci in the past on-rain
train-absent periods. In the present embodiment, comparison operations with the reference
vector locus are implemented in such a manner that the reference vector locus is used
as an appearance probability distribution that represents appearance probabilities
at locus positions.
[0038] FIGS. 5A and 5B are each diagrams describing a method for generating an appearance
probability distribution. FIG. 5A shows an appearance probability distribution relating
to the reference vector locus in the absence of an on-rail train, and FIG. 5B shows
an appearance probability distribution relating to the reference vector locus in the
presence of an on-rail train. The appearance probability distribution relating to
the reference vector locus is generated based on a plurality of past current vector
loci, where a vector locus in one determination period is treated as one vector locus.
The X and Y axes shown in FIGS. 5A and 5B are the same as the X and Y axes shown in
FIGS. 3A, 3B, and 4.
[0039] A current vector locus is actually a set of time-series data, that is, a set of current
vectors (values) as discrete data. FIGS. 5A and 5B each show the current vector locus
in a small number of plots for ease of understanding. In actuality, however, the current
vector loci are represented in a larger number of plots than shown in the drawings.
Plotting the current vectors constituting the plurality of current vector loci produces
a high-density plot group at the positions that
are likely to be taken as locus positions, and produces no or few plots at the positions that
are unlikely to be taken as locus positions. Consequently, superimposing the plots of the plurality
of current vector loci makes it possible to obtain a frequency distribution of the
locus positions that can be taken as vector locus. In the present embodiment, for
each of regions of the XY plane that is divided in a predetermined size, the ratio
of the number of the current vectors plotted in the region to the total number of
the plotted current vectors is defined as appearance probability p in the region.
[0040] However, the appearance probability p may be defined as described below without reference
to the number of plots. Specifically, the plots relating to one current vector locus
are binarized depending on whether they exist in the regions of the XY plane divided
in a predetermined size. If one or more plots exist in the region, the number of plots
in the region is set to one. Accordingly, the appearance probability distribution
obtained by superimposing the past current vector loci is based on the number of the
current vector loci that have passed through each region, and the appearance probability
in each region represents the rate at which the current vector loci pass through the
region.
[0041] In the present embodiment, the past current vector loci used for generation of the
reference vector locus are a predetermined number of current vector loci starting
from a time point to be determined (or a time point of measurement of measurement
data to be determined). Alternatively, the appearance probability p may be defined
as described below. Since the track circuits are installed outdoors, the transmission
current phase difference may differ from time to time under the influence of natural
environments such as rainfall, snow covering, and temperature. Thus, for example,
the current vector loci in the determination periods are classified by the measurement
situation including season, time zone, and weather conditions such as rain or clear
sky. Thus, the past current vector loci of which the measurement situation is in agreement
with or in proximity to that of the current vector locus to be determined may be used
to generate the appearance probability distribution relating to the reference vector
locus.
[0042] By comparison with the thus generated appearance probability distribution relating
to the reference vector locus, a degree of abnormality a is calculated as an evaluation
value of one current vector locus. FIG. 6 is a diagram describing calculation of the
degree of abnormality a. The X and Y axes are the same as those shown in the other
drawings. FIG. 6 shows an example of a current vector locus in the presence of an
on-rail train. The degree of abnormality a in the one current vector locus is calculated
by the following equation (1):

[0043] The term "p(i)" refers to the appearance probability in a region that contains a
current vector i representing the individual locus positions of the current vector
locus, and "n" refers to the number of the current vectors representing the individual
locus positions of the current vector locus.
[0044] That is, the degree of normality N is an average of the appearance probabilities
p(i) that correspond to the current vectors i constituting the current vector locus,
which indicates the degree of agreement with the reference vector locus. The degree
of normality N is an average of the appearance probabilities and thus takes a value
within a range of 0.0 ≤ N ≤ 1.0. The degree of abnormality a also takes a value within
a range of 0.0 ≤ a ≤ 1.0.
[0045] The calculated degree of abnormality a is compared with a predetermined threshold
to determine the state of the corresponding track circuit. For example, when the degree
of abnormality a is greater than the threshold, it is determined that the track circuit
is abnormal, and otherwise it is determined that the track circuit is normal. Stepwise
thresholds may be preset so that the abnormality levels can be determined stepwise.
In that case, when the abnormality level is low, it can be determined that there is
a sign of abnormality.
[Functional configuration]
[0046] FIG. 7 is a functional configuration diagram of the track circuit state determination
apparatus 100. The track circuit state determination apparatus 100 is configured such
that the plurality of measurement terminals 200 provided at the individual section
boundaries between the AC track circuits are communicably connected to the processing
apparatus 300.
[0047] Each of the measurement terminals 200 receives inputs of the transmission voltage
and the transmission current from the track circuit on the transmission side and an
input of a contact condition of the track relay 11 from the track circuit on the reception
side, at the section boundary between the track circuits where the measurement terminal
200 is provided. The measurement terminal 200 has a phase difference calculation section
202 and a transmission control section 204.
[0048] The phase difference calculation section 202 calculates the phase difference of the
transmission current from the transmission voltage, based on the measurement values
of the transmission voltage and the transmission current of the transmitter in the
track circuit obtained by the transmission-side measurer. That is, the phase difference
calculation section 202 calculates the phase difference of the input transmission
current from the input transmission voltage.
[0049] The transmission control section 204 associates the input transmission voltage and
transmission current as the measurement values relating to the track circuit on the
transmission side, the values of the phase difference calculated by the phase difference
calculation section 202, and the value of the input contact condition as the measurement
value relating to the reception-side track circuit with measurement date and time
and identification information of the track circuit, and transmits the same as measurement
data to the processing apparatus 300.
[0050] The processing apparatus 300 includes an input section 302, a display 304, a communication
section 306, a processing section 310, and a storage 330, and can be configured as
a sort of computer.
[0051] The input section 302 is implemented by input devices such as button switches, touch
panel, and keyboard, to output an operation signal according to a performed operation
to the processing section 310. The display 304 is implemented by a display device
such as a liquid crystal display (LCD) or touch panel, to provide various indications
according to a display signal from the processing section 310. The communication section
306 is implemented by a wired or wireless communication device, for example, to communicate
with the measurement terminals 200 via a transmission line.
[0052] The processing section 310 is implemented by an arithmetic device such as a central
processing unit (CPU), for example, to transmit instructions or data to the components
of the processing apparatus 300 based on programs and data stored in the storage 330,
thereby to perform overall control of the processing apparatus 300. The processing
section 310 also executes a track circuit state determination program 332 stored in
the storage 330 to serve as functional blocks such as a current vector calculation
section 312, a current vector segmentation section 314, a state determination section
316, a notification section 318, and a reference vector locus generation section 320.
However, these functional blocks can be each configured as independent arithmetic
circuits by application specific integrated circuits (ASICs) or field programmable
gate arrays (FPGAs).
[0053] The current vector calculation section 312 calculates a current vector of the transmission
current to the transmission voltage based on the measurement values input from the
measurement terminal 200. Specifically, the current vector calculation section 312
calculates a current vector (x, y) in the XY orthogonal coordinate system in which
the phase difference θ from an X axis with the voltage vector along the X-axis positive
direction is a transmission current phase difference and the magnitude of the current
vector is a transmission current value. Since the measurement values including the
transmission current and the transmission current phase difference input from the
measurement terminal 200 are associated with the measurement times, the current vector
at each measurement time can be calculated from the transmission current and the transmission
current phase difference at the corresponding measurement time (see FIG. 2). The current
vectors at the individual measurement times can be set in time-series manner to obtain
a current vector locus.
[0054] The current vector segmentation section 314 divides the current vector calculated
by the current vector calculation section 312 into the segments in the presence of
an on-rail train and the segments in the absence of an on-rail train. Specifically,
assuming that a time point at which the time-series current vector according to the
measurement time satisfies the steep-change condition is a time point of a train's
entry of into the corresponding track circuit, the on-rail train-present period for
the train to pass through the track circuit one time is defined as a period from the
foregoing time point to a time point at which the current vector has recovered to
near the state immediately before the satisfaction of the steep-change condition.
Then, the on-rail train-absent period is defined as the period other than the on-rail
train-present period (see FIGS. 3A and 3B).
[0055] The state determination section 316 sets each of the on-rail train-present period
and the on-rail train-absent period divided by the current vector segmentation section
314 as one determination period, and determines the state of the corresponding track
circuit including the normal state and the abnormal state, in each determination period
from the current vector locus. Specifically, the state determination section 316 calculates
the degree of abnormality a by comparing the current vector locus in each determination
period with the reference vector locus, and compares the degree of abnormality a with
a predetermined threshold to determine the state of the corresponding track circuit.
When the corresponding determination period is the on-rail train-present period, the
state determination section 316 compares the current vector locus with the reference
vector locus in the presence of an on-rail train, and when the corresponding determination
period is the on-rail train-absent period, the state determination section 316 compares
the current vector locus with the reference vector locus in the absence of an on-rail
train. In the present embodiment, the reference vector locus is the distribution data
of the appearance probabilities p at the individual positions. Thus, the average of
the appearance probabilities p at the positions corresponding to the current vectors
constituting the current vector locus is calculated as the degree of normality N,
and the degree of normality N is subtracted from "1.0" to calculate the degree of
abnormality a (see FIG. 6).
[0056] The notification section 318 performs a predetermined notification according to the
result of determination by the state determination section 316. For example, when
the state determination section 316 determines that a track circuit is in the abnormal
state, the notification section 318 may perform a notification such as displaying
on the display 304 a message indicating the abnormality in the corresponding track
circuit, outputting the message from a sound output section, or turning on a lamp
associated with the corresponding track circuit. Further, if a plurality of stepwise
thresholds are set for the state determination, the notification section 318 may notify
the abnormality level depending on any of the thresholds has been exceeded, or may
notify the occurrence of a sign of abnormality when the threshold condition for a
sign of abnormality has been satisfied.
[0057] The reference vector locus generation section 320 generates the reference vector
locus to be compared with the current vector locus. Specifically, in the present embodiment,
since the reference vector is represented as an appearance probability distribution,
the reference vector locus generation section 320 generates the appearance probability
distribution relating to the reference vector. The reference vector locus generation
section 320 classifies the past current vector loci into the current vector loci in
the presence of an on-rail train and the current vector loci in the absence of an
on-rail train, and uses the current vector loci in the presence of an on-rail train
to generate the appearance probability distribution relating to the reference vector
locus in the presence of an on-rail train (see FIG. 5B). The reference vector locus
generation section 320 also uses the current vector loci in the absence of an on-rail
train to generate the appearance probability distribution relating to the reference
vector locus in the absence of an on-rail train (see FIG. 5A).
[0058] At this time, the reference vector locus generation section 320 uses a predetermined
number of past current vector loci regarded as latest from the measurement dates and
times to generate the appearance probability distribution relating to the reference
vector locus. A plurality of classification conditions may be set from combinations
of measurement situations including season, time zone, and weather. Thus, the reference
vector locus generation section 320 uses the past current vector loci satisfying each
of the classification conditions to generate the appearance probability distribution
relating to the reference vector locus. The current vector loci may vary between before
and after the maintenance work of the track circuit. Thus, the reference vector locus
generation section 320 may use the past current vector loci of which the measurement
dates and times are the same as or later than the latest date and time of the maintenance
work to generate the appearance probability distribution relating to the reference
vector locus.
[0059] The storage 330 is implemented by a storage device such as a hard disk, read only
memory (ROM), or random access memory (RAM) to store programs, data, and the like
for the processing section 310 to comprehensively control the processing apparatus
300. The storage 330 also serves as a work area for the processing section 310 to
temporarily store results of arithmetic operations executed by the processing section
310 according to the programs, data input via the input section 302 and the communication
section 306, and the like. In the present embodiment, the storage 330 stores the track
circuit state determination program 332 and track circuit data 340.
[0060] The track circuit data 340 is generated for each track circuit and contains measurement
data 344, current vector locus data 346, determination result data 348, reference
vector locus data 350, threshold data 352, and maintenance history data 354, in association
with a track circuit ID 342 for identifying the corresponding track circuit.
[0061] The measurement data 344 is data of measurement values input from the corresponding
measurement terminal 200. Specifically, the measurement data 344 is data of measurement
values of transmission voltage, transmission current, the contact condition of the
track relay 11, the phase difference between the transmission voltage and the transmission
current (transmission current phase difference), which are associated with the measurement
times.
[0062] The current vector locus data 346 is data of the current vector locus in the determination
period that is the on-rail train-present period or the on-rail train-absent period.
As in an example shown in FIG. 8, the current vector locus data 346 includes on-rail
train-present data 346a relating to the current vector locus in the on-rail train-present
period and on-rail train-absent data 346b relating to the current vector locus in
the on-rail train-absent period. Either of the data contains the measurement date
and time, measurement time zone equivalent to the determination period, weather, and
the current vector locus, as accompanying information indicating the measurement situation,
in association with locus No. for identifying the current vector locus. The current
vector locus is time-series data of the current vectors at the measurement times in
the measurement time zone.
[0063] The determination result data 348 is data relating to results of state determination
on the current vector locus in each determination period. As in an example shown in
FIG. 9, the determination result data 348 contains on-rail train-present data 348a
relating to the current vector locus in the on-rail train-present period and on-rail
train-absent data 348b relating to the current vector locus in the on-rail train-absent
period. Either of the data contains reference No. of the reference vector locus used
for the state determination, the degree of abnormality, and the determination result
such as the normal state or the abnormal state, in association with locus No. of the
corresponding current vector locus.
[0064] The reference vector locus data 350 is data of the reference vector loci used for
state determination. As in an example shown in FIG. 10, the reference vector locus
data 350 contains on-rail train-present data 350a relating to the on-rail train-present
current vector locus and on-rail train-absent data 350b relating to the on-rail train-absent
current vector locus. Either of the data contains classification conditions, an applied
current vector locus list, and appearance probability distribution data, in association
with reference No. for identifying the corresponding reference vector locus. The classification
conditions are conditions for the current vector loci used for generation of the corresponding
reference vector locus, which constitute a combination of measurement situations including
season such as spring, summer, fall, or winter, time zone such as daytime or nighttime,
weather conditions such as clear sky, rain, or snow. The applied current vector locus
list is a list of locus Nos. of the past current vector loci used for generation of
the corresponding reference vector locus, which are selected from among the past current
vector loci satisfying the classification conditions described above. The appearance
probability distribution data is data indicating the corresponding reference vector
locus, which contains distribution data of appearance probabilities p (0.0 ≤ p ≤ 1.0)
at the individual positions (the individual regions in the present embodiment) on
the XY plane.
[0065] The threshold data 352 is data of thresholds used for the state determination on
the corresponding track circuit.
[0066] The maintenance history data 354 is a history of maintenance work performed on the
corresponding track circuit. For example, the maintenance history data 354 contains
the dates and times of the maintenance work, the track circuit ID of the track circuit
having undergone the maintenance work, and the contents of the maintenance work performed,
in association with one another.
[Process flow]
[0067] FIG. 11 is a flowchart of a track circuit state determination process. This process
is executed by the processing section 310 on each track circuit as a target.
[0068] First, the current vector calculation section 312 calculates the current vector whenever
necessary based on the measurement values input from the measurement terminal 200
(step S1). The current vector segmentation section 314 determines the boundary between
the segments in the presence of an on-rail train and in the absence of an on-rail
train, depending on whether changes in the current vector satisfy the steep-change
condition. When the boundary between the segments is determined (YES in step S3),
the state determination section 316 sets one determination period from the previous
segment to the current segment, and calculates the current vector locus from the current
vectors in the determination period (step S5). The state determination section 316
also specifies whether the determination period is in the presence of an on-rail train
or in the absence of an on-rail train (step S7). The state determination section 316
stores the calculated current vector locus in association with the measurement situations
including the measurement date and time, time zone, and weather (step S9). The reference
vector locus generation section 320 specifies the classification conditions of the
current vector locus from the measurement situation, and generates the reference vector
locus by the use of the past current vector loci satisfying the classification conditions
(step S11).
[0069] The state determination section 316 compares the current vector locus with the generated
reference vector locus to calculate the degree of abnormality a (step S13). The state
determination section 316 compares the calculated degree of abnormality a with a threshold
to determine the state of the track circuit (step S15). Then, the notification section
318 performs a predetermined notification such as displaying and outputting the track
circuit and the determination results (step S17). Upon completion of the foregoing
series of steps, the process returns to step S1 to repeat the same processing.
[Advantageous effects]
[0070] As described above, according to the present embodiment, it is possible to determine
the state of the AC track circuit including at least the normal state or the abnormal
state by the new method by which the current vector locus relative to the transmission
voltage is compared with the reference vector locus. The current vector locus changes
differently between in the presence of an on-rail train and in the absence of an on-rail
train, and thus discriminating these changes allows high-accuracy determination. The
current vector locus differs among the track circuits, and thus generating the reference
vector locus by the use of the past current vector loci of the corresponding track
circuit makes it possible to generate data indicating features specific to the corresponding
track circuit.
[0071] Since the track circuits are installed outdoors, the measurement values of the transmission
current and others are vulnerable to external environments. Thus, classifying the
past current vector loci depending on the measurement situation, generating the reference
vector loci under the classifications, and comparing the current vector locus to be
determined with the reference vector locus belonging in the classification according
to the measurement situation allows higher-accuracy determination.
[0072] Note that applicable embodiments of the present disclosure are not limited to the
embodiment described above, and that the foregoing embodiments can be modified as
appropriate without deviating from the scope of the present disclosure.
(A) Setting of the thresholds
[0073] For the state determination of the track circuit, the thresholds to be compared with
the degree of abnormality a based on the current vector loci may be set according
to time-series transition of the past degrees of abnormality a, for example. In this
case, the thresholds are set separately for the on-rail train-present section and
the on-rail train-absent section of the past current vector loci. The setting of the
thresholds may be performed in such a manner that the transition of the past degrees
of abnormality a are presented to the user by displaying on the display 304 or the
like so that the user provides an operation instruction for setting the thresholds
by the input section 302. Alternatively, the past current vector loci may be classified
by the classification conditions such as season, time zone, and weather condition
so that the thresholds are set from the transition of the degree of abnormality a
in the corresponding current vector locus.
(B) Reference vector locus
[0074] Instead of generating the reference vector locus for each determination, the reference
vector loci are generated in advance in correspondence with the plurality of classification
conditions. The current vector locus to be determined may be compared with the reference
vector locus selected from among the reference vector loci under the classification
conditions satisfying the measurement situation.
[REFERENCE SIGNS LIST]
[0075]
100 track circuit state determination apparatus
200 measurement terminal
202 phase difference calculation section, 204 transmission control section
300 processing apparatus
310 processing section
312 current vector calculation section, 314 current vector segmentation section
316 state determination section, 318 notification section
320 reference vector locus generation section
330 storage
332 track circuit state determination program
340 track circuit data
342 track circuit ID, 344 measurement data
346 current vector locus data, 348 determination result data
350 reference vector locus data, 352 threshold data
354 maintenance history data