JPH07327302A - Automatic train operation system - Google Patents

Abstract

PURPOSE:To enhance control follow-up based on a desired travel pattern. CONSTITUTION:Since a train can run when a pattern of certain degree of accuracy can be reproduced from a desired travel pattern, a pattern calculator 3 splits a travel section finely to produce a desired travel pattern where the desired speed at each point is designated by a constant. A contraction unit 4 splits a travel section designated by the desired travel pattern into a plurality of sub-sections each generating a contraction pattern for approximating the speed with a specific function. A follow-up controller 6 controls a train 2 according to the contraction pattern. Consequently, the pattern data can be reduced by approximating the travel pattern to a function curve and the calculation of the speed can be simplified at the time of control follow-up by approximating with a low order function.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an automatic train operation system of target traveling pattern following type.

[0002]

2. Description of the Related Art Automatic Train Operati
on) system is to automatically control the running of the train. As one of the methods, there is a type in which a target traveling pattern is obtained in advance and a train is controlled by a follow-up control device so that the train follows the target traveling pattern. In this type of automatic train operation system, when the speed deviation between the current actual speed and the target travel pattern is fed back and speed follow-up control is performed, the calculated target travel pattern data is stored in the system to The target traveling pattern data is extracted according to the above, the target speed at the current traveling point is calculated, and the notch command is determined based on this target speed.

FIG. 6 shows the configuration of an automatic train operation system of that type. In this figure, the pattern calculation device 71 obtains a target traveling pattern defined as position-versus-speed data by taking into consideration conditions such as the traveling position of the train, the topography of the position, and the degree of curve of the track. Yes, 72 is the target travel pattern data. FIG. 7 is a graph showing this distance-versus-speed data. In this figure, (xi, vi) is the position-velocity data given to each position specified by the distance from the starting point. , For example, the position-to-velocity data can be given in units of 1 [sec]. This position-versus-speed data is given to the pattern following controller 74. The pattern tracking control device 74 has a control target value calculation unit 75 and a notch command calculation unit 76. Control target value calculation unit 75
When a target traveling pattern is given in the form of position-to-speed data, when calculating the target speed, two data before and after the current position are selected and interpolated. The interpolation will be performed based on the following equation, for example.

Vr = ((x−x2) / (x3−x2)) (v3−v2) + v2 (1) In this equation, x is the current position given by the distance measuring device 73, and x2 and x3 are shown in FIG. The distance values v2 and v3 between two adjacent points as shown in Fig. 2 are target velocities at points of distances x2 and x3, respectively.

The target speed vr obtained by the control target value calculation unit 75 is given to a notch command calculation unit 76, which is converted into a notch command which is a stepwise motor output torque command, and train control is performed. Device 77
Controls the speed of the electric motor according to this notch command.

[0006]

However, in the above-mentioned conventional automatic train traveling system, data must be given at sufficiently small intervals in order to reproduce the target traveling pattern in the form of a speed curve with respect to the position. However, there is a problem that the target traveling pattern data becomes large. For example, as described above, when the target travel pattern is given as position-to-speed data for every 1 [sec], the number of data elements reaches 110 in the travel section of the predetermined travel time 110 [sec]. I will end up. Therefore, if an attempt is made to store the target travel pattern for all travel sections of the route traveling on that day, most of the limited memory will be occupied only by the target travel pattern data.

Further, since the interpolation calculation is involved as described above, there is a problem that it takes time to calculate the target speed and the control tends to be delayed.

The present invention has been made in view of the above-mentioned prior art, and an object of the present invention is to provide an automatic train operation system capable of improving the followability of control based on a target traveling pattern. It is in.

In particular, an object of the present invention is to make it possible to reduce the size of the target traveling pattern data and the control target calculation time during the pattern follow-up control.

A further object of the present invention is to be able to find an approximate curve whose error falls within a predetermined allowable value.

[0011]

[Means for Solving the Problems] First, in actual driving,
Modeling error due to considering trains as mass points, fluctuations in vehicle conditions such as boarding rate and output torque, fluctuations in route conditions due to changes in weather, errors in signals from sensors such as distance measuring devices, wheel sliding / slip, Due to various factors, it is unavoidable that the train deviates from the target traveling pattern to some extent, and therefore control must be performed in consideration of those factors as well.

However, the target traveling pattern is a guideline for the running manner required for the train to run optimally under standard conditions, and the train needs to strictly follow the pattern under various conditions. There is no. Therefore, it suffices if the pattern can be reproduced to some extent accurately from the target traveling pattern data.

The present invention has been made paying attention to this point, and if the target traveling pattern is approximated by a low-order curve or the like, the target traveling pattern can be represented by an approximate parameter, so that the size of the pattern data is large. Therefore, the target speed can be easily obtained by substitution calculation.

Therefore, in the automatic train operation system of the present invention, first target travel pattern calculation means for obtaining first target travel pattern data given as position-to-speed data, and data of the first target travel pattern. Pattern data reducing means for generating second target travel pattern data which is position-to-speed data constituted by a function curve approximating a travel pattern of a predetermined section in the data of the first target travel pattern using , Pattern-following control means for determining a notch command based on the data of the second target travel pattern.

The pattern data reduction means is, for example, the first
All the traveling sections in the target traveling pattern data are divided as approximation target sections to the function curve according to their characteristics, and approximation section generating means for generating approximation section data indicating each section, and each approximation section indicated by the approximation section data. Is extracted as a reference for low-order curve approximation, an approximation curve expressed by a specific function that passes through the at least two reference points is obtained, and approximation curve generation means for generating data of the function parameter It can be realized by including and.

The pattern data reduction means determines the approximation accuracy based on the error between the approximate curve and the first target traveling pattern. As a result, when the approximation accuracy is larger than the allowable value, the approximation section concerned. It is also possible to have a configuration in which the section indicated by the data is divided again, and second approximation section generation means for generating the approximation section data indicating each section and giving it to the approximation curve generation means.

The second target travel pattern, that is, the contracted target travel pattern, may include approximate section data and function parameters of the approximate section.

[0018]

According to the present invention, the size of the target travel pattern data can be suppressed by dividing the given target travel pattern and approximating a specific function curve for each part. At this time, the control target calculation at the time of the pattern follow-up control can be simplified by approximating the function as low as possible.

Further, by determining the error between the approximate curve and the first target traveling pattern and adopting only the approximate curve in which the error falls within the allowable value, reliable control becomes possible.

[0020]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 shows the configuration of an automatic train operation system according to an embodiment of the present invention. In this figure,
The automatic train operation system of the present embodiment is roughly composed of a ground side device 1 and a train 2 installed on the ground side. The ground side device 1 includes a pattern calculation device 3 and a pattern data reduction device 4.
It has and. The pattern calculation device 3 is a standard condition (boarding rate, driving / braking device condition, weather, etc.) based on route data, train performance data, and operating condition data read from a data storage device or storage medium (not shown). Under the condition, the target travel pattern for the train to meet various requirements such as ride comfort and energy saving is calculated according to the train performance and operation schedule. This is the same as the target traveling pattern as shown in FIG. This target travel pattern data is given to the pattern data reduction device 4 and is compressed there. This compression process is a guideline for how the train is run so that it can run optimally under standard conditions, and the train does not have to run exactly according to the pattern under various conditions, and the target From the viewpoint that the pattern can be reproduced to some extent accurately from the traveling pattern data, two points that can approximate the original target traveling pattern by a low-order function are found, and the position-to-speed data between the two points is calculated as the low point. It is replaced with the following function to reduce the number of position-versus-speed data constituting the target traveling pattern. FIG. 2 shows distance-velocity data showing such a state, and corresponds to the data shown in FIG. Here, only n positions of the distances x1, x2, ..., Xn are used, and each section has a linear function vr = ai.x.
It is approximated by n + bi. In this formula, v
r is the target speed, i (= 1 to n) is the number of each section,
x is a travel position variable that represents the distance from the starting point and is Xi-1.
≤X≤Xi. Further, ai and bi are approximate parameters. FIG. 3 shows the internal configuration of the pattern data reduction device and the pattern tracking control device, which are the system components shown in FIG. As shown in this figure, the contracted target travel pattern data 5 is composed of a data string of values xi, ai, and bi.

Referring again to FIG. 1, the train 2 is equipped with a pattern following control device 6, and the pattern following control device 6 stores the given target traveling pattern data in a storage medium such as a memory (not shown). The target speed at the current location is determined from the target travel pattern data and the current position obtained from the monitor device (not shown), and the current speed and time obtained from the monitor device and the data storage device or storage medium (not shown) The notch command given to the train is determined based on the route data and the train performance data read from the above. Further, the pattern following control device 6 switches the speed following control to the position following control by determining the target position from the current speed instead of the target speed and setting it as the control target value, especially near the stop point. Therefore, the stopping accuracy can be secured, and as shown in FIG. 3, the distance measuring device 7, the train control device 8, the control target value calculation unit 9, and the notch command calculation unit 10 are provided. The distance measuring device 7 measures the distance traveled from the starting point of the train 2 and outputs the distance as the current position data x. The control target value calculation unit 9 refers to the contracted target travel pattern data 5 based on the current position data x from the distance measuring device 7 to calculate the target speed vr. At that time, the current position data x is xi- 1 ≤ x ≤ x
When i, the target speed is determined as vr = ai.x + bi by the function specified by the approximate parameters ai and bi corresponding to the position xi of the contracted target travel pattern data 5. For example, when the current position data x is x2 .ltoreq.x <x3, it is calculated as a3.x + b3. Target speed vr
Is supplied to the notch command calculation unit 10, and the notch command calculation unit 10 calculates the notch command based on the difference between Vr and the current speed V given by a speed measuring device (not shown). The train control device 8 controls the electric motor that drives the wheels by this notch command.

FIG. 4 shows the details of the pattern data reduction apparatus. In this figure, the pattern input device 10
Is for inputting the original target traveling pattern data from the pattern calculation device 3 and giving it to the pattern division device 11.
The pattern dividing device 11 divides the original target traveling pattern into some approximate sections at notch switching points, gradient change points, and the like, and stores and holds the approximate section data thus divided.

The pattern approximating device 12 includes a sample point extracting device 13, an approximate curve calculating device 14, and an approximate accuracy determining device 15.
And an approximate segment subdivision device 16. In response to the division end notification and the pattern approximation command from the pattern division device 11, the sample point extraction device 13 reads each of the approximation segment data held in the pattern division device 11 one by one, and the read approximation segment. Starting point, middle point,
The end point is extracted as a reference for linear approximation. The approximate curve calculation device 14 uses the least squares method or the like on the reference data so that the error at the sample points forming the reference data is minimized so that the approximate curve vr = ai.x.
Find + bi. When the approximation accuracy determination device 15 determines that the accuracy of the approximation straight line is insufficient, that is, the error at the sample point is larger than the reference error, it outputs a re-division command for the same section and is within the reference error. In this case, the end determination command is output together with the approximate parameter that specifies the approximate curve. The approximating section re-dividing device 15 subdivides the approximating section in response to the re-dividing command, and extracts the sample point re-extracting instruction and the approximating curve re-calculating instruction so as to re-calculate the approximating straight line for the new approximating section. It is given to the device 13. Then, the sample point extraction device 13 causes the start point, the intermediate point, and the
The end point is extracted as a reference for linear approximation, and then the approximate curve calculation device 14 obtains an approximate curve for the reference data.

The end judging device 17 is the approximation accuracy judging device 1.
5 in response to the end determination command from the pattern dividing device 11
It is determined whether or not the approximation processing has been completed for all the sections stored in, and if there are remaining sections, a pattern approximation command is given to the sample point extraction device 13. As a result, the sample point extraction device 13 reads the next section, the approximate curve calculation device 14 obtains the approximate curve, the approximate accuracy determination device 15 determines its accuracy, and the approximate interval subdivision device 16 determines if necessary. The process of being activated is repeated. When the pattern data reduction device 4 is connected to the pattern follow-up control device 6, the end determination device 17 reduces the approximate parameter sent together with the end determination command from the approximation accuracy determination device 15 to the reduced data output device 19. The pattern data reduction device 4 is transferred to the pattern following control device 6 via the, and stored in the reduction data storage device 18 when the pattern data reduction device 4 is not connected to the pattern following control device 6.

When the pattern follow-up controller 6 is connected, the contracted data output device 19 receives pattern data from the end determination device 17 each time one approximate curve data (approximate parameter) is received. In addition to having the function of transferring to the device 6, it also has the function of collectively transferring the approximate curve data stored in the contracted data storage device 18 in response to a transfer command from the pattern following control device 6.

According to the apparatus configured as described above, the number of divisions of the target traveling pattern is much smaller than that of the conventional method in which data is given in 1 [sec] steps. The size of the data given to the device 6 can be kept small. FIG. 5 shows a specific example of the target traveling pattern approximation processing. FIG. 5A shows distance-notch commands, FIG. 5B shows distance-gradients, and FIG. 5C shows distance-speed data. For example, if the target travel pattern shown in FIG. 5C is a target travel pattern with a travel time of 110 [sec] and the number of divisions is 24, compared to the case where position and speed data is given in 1 [sec] increments. The amount of data is 220 to 72 by simple calculation, and the size of the data is suppressed to about 1/3.

Since the approximate curve is a linear function which is a straight line, the target speed can be obtained by a simple substitution calculation of the approximate parameter and the distance. Even when the position tracking control is performed near the stop point, the target position can be easily obtained from the current speed v and the approximate parameter as xr = (v-bi) / ai.

In the above embodiment, the pattern calculation device 3
The pattern data reduction device 4 is of the ground type, but may be of the on-vehicle type. When the pattern data reduction device is a ground-based type, the reduced target traveling pattern data is given to the pattern following control device 6 through a storage medium such as an IC card or a communication system, and the pattern data reduction device 4 is mounted on the vehicle. In the case of an on-board type, it can be configured to be provided to the pattern following control device 6 directly or through a storage medium such as a memory.

In the above embodiment, the case of approximating to a straight line has been described, but it is also possible to approximate to a quadratic curve or a spline curve. Further, it is possible to change the type of the approximate curve depending on the position in the target traveling pattern, the notch used, etc., such as selecting a curve having a good approximation accuracy near the stop point.

Furthermore, the pattern data reduction device and the pattern calculation device may be installed separately. In this case, the target travel pattern obtained by the pattern calculation device is directly
Alternatively, it can be given to the pattern data reducing apparatus through a storage medium such as an IC card or a memory or a communication system.

By approximating the target traveling pattern to a low-order curve for each part in this way, the size of the data of the target traveling pattern can be suppressed, so that a storage medium such as a memory can be saved. Further, the calculation of the control target value in the pattern following control device can be simplified, so that the calculation load during traveling can be reduced.

[0033]

As described above, according to the present invention, the size of the target traveling pattern data is suppressed by dividing the given target traveling pattern into parts and approximating to a low-order curve. It is possible to simplify the control target calculation calculation during the follow-up control.

[Brief description of drawings]

FIG. 1 is a block diagram showing the configuration of a pattern data reduction apparatus according to an embodiment of the present invention.

FIG. 2 is a conceptual diagram of approximation of a target traveling pattern of the device shown in FIG.

3 is a block diagram showing configurations of a pattern data reduction device and a pattern follow-up control device of the device shown in FIG.

4 is a block diagram showing a configuration of a pattern data reduction device of the device shown in FIG.

FIG. 5 is an explanatory diagram showing a specific example of target traveling pattern approximation shown in FIGS.

FIG. 6 is an explanatory diagram showing calculation of a target speed when position-to-speed data of an original target traveling pattern is given.

FIG. 7 is an explanatory diagram showing how to collect data by giving position-versus-speed data.

[Explanation of symbols]

 1 Ground-side device 2 Train 3 Pattern calculation device 4 Pattern data reduction device 5 Reduction target travel pattern data 6 Pattern follow-up control device 7 Distance measurement device 8 Train control device 9 Control target value calculation unit 10 Notch command calculation unit 11 pattern Dividing device 12 Pattern approximating device 13 Sample point extracting device 14 Approximating curve calculating device 15 Approximation accuracy determining device 16 Approximating section redivision device 17 End determining device 18 Reduced data storage device 19 Reduced data output device

Claims (4)

[Claims] 1. A first provided as position versus velocity data.
Target travel pattern calculation means for obtaining the target travel pattern data and the data of the first target travel pattern, and by a function curve that approximates the travel pattern of the predetermined section in the first target travel pattern data. Pattern data reduction means configured to generate second target travel pattern data which is position-to-speed data, and pattern follow-up control means for determining a notch command based on the second target travel pattern data. An automatic train operation system characterized by being equipped. 2. The pattern data reduction means divides all traveling sections in the first target traveling pattern data into sections to be approximated to a function curve according to their characteristics, and generates approximate section data indicating each section. Interval generation means and at least two points of each approximate interval indicated by the approximate interval data are extracted as a reference for low-order curve approximation, and an approximate curve that passes through the at least two reference points and is expressed by a specific function is generated. The automatic train operation system according to claim 1, further comprising: an approximate curve generation unit that obtains the data of the function parameter. 3. The pattern data reduction means determines an approximation accuracy based on an error between an approximate curve and a first target travel pattern, and as a result, when the approximation accuracy is larger than a permissible value, the approximation section concerned. The automatic train operation according to claim 2, further comprising: a second approximate section generation unit that divides the section indicated by the data again, generates approximate section data indicating each section, and gives the approximate curve generation unit the approximate section data. system. 4. The automatic train operation system according to claim 1, wherein the second target travel pattern includes approximate section data and a function parameter of the approximate section. .