Track state estimation method, track state estimation device, and vehicle

Abstract

To provide a track condition estimation method capable of easily inspecting a track using an operational vehicle.SOLUTION: A track condition estimation method includes an acceleration noise removal step, a tilt angle estimation step, and a track condition calculation step. The acceleration noise removal step eliminates the acceleration noise generated by the vehicle acceleration / deceleration from a three-dimensional acceleration measured by an inertia measuring device installed on a vehicle body or an axle box based on a vehicle travelling speed. The tilt angle estimation step estimates a horizontal tilt angle and a longitudinal tilt angle at the installation position of the inertia measuring device based on the three-dimensional acceleration with eliminated acceleration noise and the three-dimension acceleration measured by the inertia measuring device. The track condition calculation step calculates track condition information representing the track condition based on the estimated horizontal tilt angle and longitudinal tilt angle.SELECTED DRAWING: Figure 11

Description

[Technical Field]

[0001] The present invention relates to a track state estimation method, a track state estimation device, and a vehicle. [Background technology]

[0002] One of the causes of derailment of vehicles running on rails is irregularity of the rails relative to their original position (hereinafter referred to as "track irregularity").Since track irregularity has a direct impact on derailment, it is important to properly inspect and maintain track irregularity in order to reduce the risk of derailment.

[0003] Inspection of track irregularities (hereinafter also referred to as "track inspection") is carried out using, for example, a hand-pushed track inspection device. However, when the rail to be inspected becomes long, it becomes difficult to inspect it frequently. Furthermore, track inspection using a hand-pushed track inspection device is a static inspection performed under no-load conditions, which differs from the conditions at the time of a derailment. For this reason, it is desirable to be able to carry out dynamic inspections with vehicles running at high frequency.

[0004] One method for frequently performing dynamic inspections while a vehicle is in motion is to use an operating vehicle to perform track inspection. For example, Patent Document 1 discloses a track inspection device that uses a vehicle. This device attaches a detector unit made up of various detectors to the bogie frame of the vehicle, and uses the inertial sine wave method to measure five track irregularities while the vehicle is in motion. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] Patent No. 3411861 [Patent Document 2] Patent No. 6674544 Summary of the Invention [Problem to be solved by the invention]

[0006] However, while the technology described in Patent Document 1 can measure five track irregularity items using a detector unit, the detector unit's measurement mechanism is complex and expensive. Furthermore, because the displacement is calculated by integrating the acceleration signal twice, the accumulation of integration errors significantly reduces measurement accuracy when the vehicle is traveling at a slow speed, such as when it is used to transport heavy materials in a steelworks.

[0007] The present invention has been made in consideration of the above problems, and an object of the present invention is to provide a track condition estimation method, a track condition estimation device, and a vehicle that enable easy track inspection using an operating vehicle. [Means for solving the problem]

[0008] In order to solve the above problems, according to one aspect of the present invention, there is provided a track state estimation method including: an acceleration noise removal step of removing acceleration noise generated by vehicle acceleration and deceleration from triaxial acceleration measured by an inertial measurement unit (IMU) installed on the vehicle body or an axle box, based on the vehicle's running speed; an inclination angle estimation step of estimating the horizontal inclination angle and longitudinal inclination angle at the installation position of the IMU, based on the triaxial acceleration from which the acceleration noise has been removed and the triaxial angular velocities measured by the IMU; and a track state calculation step of calculating track state information representing the track state, based on the estimated horizontal inclination angle and longitudinal inclination angle.

[0009] In the tilt angle estimation step, a translational acceleration component during vehicle travel is extracted by filtering from the three-axis acceleration from which acceleration noise has been removed, an error covariance matrix of an observation model of a Kalman filter is set based on the extracted translational acceleration component, and a horizontal tilt angle and a longitudinal tilt angle at the installation position of the inertial measurement unit are estimated from the three-axis acceleration and three-axis angular velocity from which acceleration noise has been removed using a Kalman filter formulated from a state space model configured by the observation model with the set error covariance matrix. In the track state calculation step, elevation displacement may be calculated as track state information based on the estimated longitudinal tilt angle.

[0010] In the orbit state calculation step, the calculated elevation change may be subjected to high-pass filtering to remove estimation errors of low-frequency components.

[0011] The translational acceleration component may be extracted by band-pass filtering to remove frequency components in which track irregularity predominates from the lateral acceleration and longitudinal acceleration among the three-axis accelerations.

[0012] The error covariance matrix may be set by adding values ​​obtained by multiplying the extracted translational acceleration components by predetermined coefficients to the diagonal elements.

[0013] In the acceleration noise removal step, acceleration noise contained in the three-axis acceleration may be removed by subtracting an acceleration component calculated from the vehicle's traveling speed from the longitudinal acceleration component of the three-axis acceleration measured by the inertial measurement unit.

[0014] Furthermore, in order to solve the above-mentioned problems, according to another aspect of the present invention, there is provided a track state estimation device comprising: an acceleration noise removal processing unit that removes acceleration noise generated by vehicle acceleration and deceleration from triaxial acceleration measured by an inertial measurement unit (IMU) installed on the vehicle body or an axle box, based on the vehicle's running speed; an inclination angle estimation unit that estimates the horizontal inclination angle and longitudinal inclination angle at the installation position of the IMU, based on the triaxial acceleration from which the acceleration noise has been removed and the triaxial angular velocities measured by the IMU; and a track state calculation unit that calculates track state information representing the track state, based on the estimated horizontal inclination angle and longitudinal inclination angle.

[0015] Furthermore, in order to solve the above-mentioned problems, according to another aspect of the present invention, there is provided a vehicle comprising: a speed measurement device that measures the running speed of the vehicle; an inertial measurement unit (IMU) installed on the body or an axle box of the vehicle; an acceleration noise removal unit that removes acceleration noise generated by acceleration and deceleration of the vehicle from the triaxial acceleration measured by the IMU based on the running speed of the vehicle; an inclination angle estimation unit that estimates the horizontal inclination angle and longitudinal inclination angle at the installation position of the IMU based on the triaxial acceleration from which the acceleration noise has been removed and the triaxial angular velocities measured by the IMU; and a track condition estimation device that includes a track condition calculation unit that calculates track condition information representing the track condition based on the estimated horizontal inclination angle and longitudinal inclination angle. [Effects of the Invention]

[0016] As described above, according to the present invention, it is possible to easily perform track inspection using an operating vehicle. [Brief explanation of the drawings]

[0017] [Figure 1] 1 is a schematic diagram showing a general configuration of a vehicle according to an embodiment of the present invention; [Figure 2] FIG. 10 is a schematic diagram for explaining vertical displacement. [Figure 3] FIG. 10 is a schematic diagram showing elevation displacement calculated when the inertial measurement unit is installed in the vehicle body. [Figure 4] FIG. 10 is a schematic diagram showing elevation displacement calculated when an inertial measurement unit is installed in an axle box. [Figure 5] FIG. 2 is an explanatory diagram for explaining measurement values ​​obtained by an inertial measurement unit. [Figure 6] FIG. 2 is a schematic cross-sectional view of a wheelset for explaining values ​​measured by an inertial measurement unit. [Figure 7] 10 is a graph for explaining a process for removing acceleration noise from longitudinal acceleration. [Figure 8] 10 is a graph showing an example of calculation of an index Rx representing translational acceleration. [Figure 9] 10 is a graph showing an example of calculation of an index Ry representing translational acceleration. [Figure 10] FIG. 2 is a functional block diagram showing the configuration of a track state estimation device according to the embodiment. [Figure 11] 4 is a flowchart illustrating an example of a track state estimation method according to the embodiment. [Figure 12] 10 is a graph showing an example of elevation change before and after application of a high-pass filter. [Figure 13] 10 is a graph showing the estimated results of elevation displacement in the example. DETAILED DESCRIPTION OF THE INVENTION

[0018] Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. In this specification and drawings, components having substantially the same functional configurations are designated by the same reference numerals, and redundant explanations will be omitted.

[0019] [1. Vehicle configuration] First, a general configuration of a vehicle 1 according to one embodiment of the present invention will be described with reference to Fig. 1. Fig. 1 is a schematic diagram showing a general configuration of a vehicle 1 according to this embodiment.

[0020] A vehicle 1 according to this embodiment is configured to be able to run on a pair of rails 5 that are extended on sleepers (reference numeral 7 in FIG. 2). As shown in FIG. 1, the vehicle 1 is configured from a car body section 10 that is made up of a body structure that is the main structural part of the vehicle, and a bogie section 20 that supports the car body section 10 and has a running mechanism. In the following, the direction of travel of the vehicle (front-rear direction) is defined as the X direction, the width direction (left-right direction) as the Y direction, and the height direction (up-down direction) as the Z direction.

[0021] Four wheel pairs are arranged in the vehicle travel direction on the bogie section 20 of the vehicle 1 shown in Figure 1. Each wheel pair is formed by two wheels 23 that are paired in the width direction and connected by a wheel set 25. Axle boxes 27, which are bearing portions of the wheel set 25, are provided on both ends of each wheel set 25. The axle boxes 27 are connected to the bogie frame 21 by coil springs 29. Note that in Figure 1, the axle box 27 and coil spring 29 are shown for only one wheel 23 in order to show the wheel set 25, but all of the wheels 23 are equipped with axle boxes 27 and coil springs 29.

[0022] The vehicle 1 according to this embodiment is equipped with an inertial measurement unit (IMU) 30 in at least one of the carbody 10 or the axle box 27 of each wheel 23. In the vehicle 1 shown in FIG. 1, one IMU 30 is provided in each of the carbody 10 and the axle box 27 of one wheel 23, but it is sufficient to install only one of them. The IMU 30 is a device that detects three-dimensional inertial motion (translational motion and rotational motion in three orthogonal axial directions). The IMU 30 detects translational motion using an acceleration sensor and rotational motion using an angular velocity sensor. Measurement values ​​by the IMU 30 are output to a track condition estimation device (track condition estimation device 100 in FIG. 10) that estimates the track condition of the rails 5 on which the vehicle 1 travels.

[0023] The vehicle 1 according to this embodiment also includes a position detection device (not shown) that detects the absolute position of the vehicle 1 on the rails 5 and a speed measurement device (speed measurement device 40 in FIG. 10) that measures the traveling speed of the vehicle 1. The position detection device may be, for example, an RFID (Radio Frequency Identification) reader. In this case, the absolute position of the vehicle 1 can be detected by reading an RFID tag that is pre-installed at the middle position in the width direction of the pair of rails 5 with the RFID reader. Note that the method for detecting the absolute position of the vehicle 1 is not limited to this example, and well-known techniques may also be used. For example, position information obtained using radio waves from an artificial satellite may be used as the absolute position of the vehicle 1. The speed measurement device may be, for example, a laser Doppler speedometer installed in the vehicle body 10, and the relative speed from the ground surface calculated by the laser Doppler speedometer may be used as the traveling speed of the vehicle 1.

[0024] [2. Estimation of orbital conditions] Rail track conditions include planar deviation, alignment deviation, elevation deviation, level deviation, and gauge deviation. Below, we will explain how to estimate elevation deviation, which may indirectly cause a derailment, as a track condition. As shown in Figure 2, elevation deviation refers to the difference in height of the top surface of the rail 5 in the direction of vehicle travel (front-to-back direction). Elevation deviation h is expressed as the vertical distance between a line connecting two points on the top surface of the rail 5 that are a predetermined distance d apart in the direction of vehicle travel, and the top surface of the rail 5 at the midpoint between the two points.

[0025] Typically, elevation displacement is defined by a method called the 10m chord slash method, which involves stretching a chord at two points 10m apart on the rail and measuring the distance from the rail at the midpoint. While the measurement characteristics differ from the 10m chord slash method, the elevation difference between the front and rear supports can also be considered a type of elevation displacement that represents irregularities in the rail height direction. Therefore, in this embodiment, the elevation difference over a reference length L according to the carbody structure is calculated as elevation displacement h. If the front-to-rear tilt angle is φ and the distance between the supports is L, elevation displacement h can be expressed by the following formula (1):

[0026] h = L × tanφ (1)

[0027] In this embodiment, elevation displacement is measured using only the measurement values ​​of the inertial measurement unit 30 installed in the carbody 10 or the axle box 27. When the inertial measurement unit 30 is installed in the carbody 10 as shown in FIG. 3, the inertial measurement unit 30 measures the longitudinal tilt angle φ of the carbody 10 and calculates the difference in elevation between the front and rear bogie units 20 as elevation displacement. At this time, the fulcrum distance L is the distance between the front and rear bogie units 20 of the carbody 10. Also, when the inertial measurement unit 30 is installed in the bogie unit 20 (axle box 27) as shown in FIG. 4, the inertial measurement unit 30 measures the longitudinal tilt angle φ of the bogie unit 20 and calculates the difference in elevation between the front and rear wheel sets 25 as elevation displacement. At this time, the fulcrum distance L is the distance between the front and rear wheel sets 25 of the bogie unit 20.

[0028] Here, factors that cause errors in inclination angle estimation include noise caused by shock vibrations associated with vehicle travel and vehicle acceleration / deceleration noise. In particular, when the inertial measurement unit 30 is installed in the axle box 27, it is possible to detect track irregularity directly below the wheels, while also making it easier to measure the above-mentioned noise. Therefore, this embodiment presents a method for reducing the effects of these noises and estimating elevation irregularity, which is one index that represents the track condition of the rails.

[0029] As a method for estimating elevational displacement, for example, Patent Document 2 discloses an inspection device that inspects track irregularities, such as elevational irregularities, on a track on which a vehicle runs. In this inspection device, when a beam member of a predetermined length is moved on a rail by two rollers, the elevational irregularity at the midpoint between the two contact points is calculated by multiplying the angular velocity acting on the beam member by the velocity of the two rollers. However, Patent Document 2 uses a device that is different from the actual vehicle that runs on the rail, so it is also not possible to easily inspect the rail track condition. Furthermore, the inspection device in Patent Document 2 measures the response of the beam member, which is a rigid body, and is thought to have a different responsiveness from a vehicle that has an elastic body such as a spring.

[0030] Below, we will explain in detail a track state estimation technique that estimates elevation change as a track state, and a track state estimation device and track state estimation method that execute a track state estimation process based on the track state estimation technique.

[0031] [2-1. Orbit state estimation method] First, a track state estimation method for estimating track states according to this embodiment will be described with reference to Figs. 5 to 9. Fig. 5 is an explanatory diagram illustrating measured values ​​by the inertial measurement unit 30. Fig. 6 is a schematic cross-sectional view of a wheelset for explaining measured values ​​by the inertial measurement unit 30. Fig. 7 is a graph illustrating the process of removing acceleration noise from longitudinal acceleration. The upper graph shows the longitudinal acceleration measured by the inertial measurement unit 30 and the acceleration component calculated from the vehicle speed measured by the speed measurement device, and the lower graph shows the corrected longitudinal acceleration from which the acceleration noise has been removed. Fig. 8 shows an index R representing translational acceleration. x 9 is a graph showing an example of calculation of the index R representing the translational acceleration. y 10 is a graph showing an example of calculation of

[0032] In the following description, the measurement value of the three-axis acceleration measured by the inertial measurement unit 30 at time k is a xk , a yk , a zk , the measured values ​​of the three-axis angular velocity are ω xk , ω yk , ω zk As shown in Fig. 5, the roll angle at time k is θ k , the pitch angle is φ k , yaw angle is ψ k Let's say.

[0033] In this embodiment, a Kalman filter is used to estimate the roll angle θx of the wheelset 25 from the measurement values ​​of the inertial measurement unit 30. The Kalman filter is one of the data assimilation techniques, and is a technique that uses Bayesian estimation to sequentially estimate state quantities that minimize errors with observed quantities. The roll angle estimated value and pitch angle estimated value obtained from the measurement values ​​of the three-axis acceleration are used as observed quantities, and the Kalman filter can be formulated from a time update equation based on the integration of the roll angular velocity and pitch angular velocity in absolute coordinates.

[0034] State vector x k is defined by the following formula (2).

[0035]

number

[0036] Also, the observation vector y k is the measured value of the three-axis acceleration a shown in Fig. 6. xk , a yk , a zk The low-frequency components of the gravitational acceleration g are extracted and projected onto the three axes, and the observation vector y can be expressed as in the following equation (3). k In this case, θ 0k is the estimated roll angle at time k, φ 0k is the estimated pitch angle at time k.

[0037]

number

[0038] Here, the measured value of the three-axis acceleration a xk , a yk , a zk includes acceleration noise generated by acceleration and deceleration of the vehicle 1. In this embodiment, first, the measured value a of the three-axis acceleration is calculated based on the traveling speed of the vehicle 1. xk , a yk , a zk Of these, longitudinal acceleration a xkIn actual vehicle operation, the vehicle 1 maintains a constant speed while repeatedly accelerating and decelerating, and the effects of acceleration and deceleration in the longitudinal direction can be clearly observed from the longitudinal acceleration waveform. Here, since the vehicle speed is measured at a predetermined cycle (for example, once every 0.1 seconds), the longitudinal acceleration waveform has a stepped shape. Therefore, acceleration and deceleration noise cannot be removed by filtering a specific frequency.

[0039] Therefore, in this embodiment, the traveling speed V of the vehicle 1 at time i measured by the speed measurement device is i Acceleration component a due to acceleration / deceleration of vehicle 1 Vi is calculated using the following formula (4).

[0040]

number

[0041] Here, in order to smooth the waveform, a low-pass filter may be applied to the traveling speed V of the vehicle 1 measured by the speed measurement device. The cutoff frequency of the low-pass filter is, for example, V ave / 2(V ave The acceleration component a calculated based on the above formula (4) may be used as the average velocity (m / s). V from the longitudinal acceleration a measured by the inertial measurement unit 30 xIMU is corrected as shown in the following equation (5), and the corrected longitudinal acceleration a x * IMU Ask for.

[0042]

number

[0043] The upper graph in FIG. 7 shows the longitudinal acceleration a measured by the inertial measurement unit 30. xIMU and the acceleration component a calculated based on the above equation (4) from the vehicle speed measured by the speed measurement device at this time. VThe longitudinal acceleration a measured by the inertial measurement unit 30 based on the above equation (5) is xIMU Acceleration component a V The graph in the bottom of Figure 7 shows the corrected longitudinal acceleration a x * IMU In the example in Figure 7, acceleration noise occurs around 120 m due to the acceleration of vehicle 1, and looking at the graph at the bottom of Figure 7, we can see that this effect has been removed.

[0044] In the subsequent processing, the measured value of the 3-axis acceleration a xk , a yk , a zk Of these, longitudinal acceleration a xk For the longitudinal acceleration a after the correction with the acceleration noise removed, x * IMU By using the above, the accuracy of estimating the orbital state can be improved. In the following explanation, the measured values ​​of the longitudinal acceleration of the three-axis acceleration are referred to as a xk However, in reality, the longitudinal acceleration a after correction by the above formula (5) x * IMU is used.

[0045] 3-axis acceleration measurement value a xk , a yk , a zk The gravitational acceleration projection component is a low-frequency acceleration component, for example, a component of 0 to 1 Hz. For example, a low-pass filter is used to filter the measured value of the three-axis acceleration a xk , a yk , a zk The gravitational acceleration projected component can be extracted from

[0046] The time update equation is derived from the integral of the angle in the absolute coordinate system, and is a nonlinear relational equation for the state vector. k-1 An extended Kalman filter is used to perform linearization around v. In this case, the system equation is expressed by the following equation (6). kis the process noise.

[0047]

number

[0048] Also, the observation noise w k , observation matrix H k When , the observation equation is expressed by the following equation (8).

[0049]

number

[0050] The measured value a of the three-axis acceleration is calculated by the Kalman filter based on the state space model expressed by the above equations (6) to (8). xk , a yk , a zk and the measured value of the three-axis angular velocity ω xk , ω yk , ω zk From the roll angle θ k and pitch angle φ k can be estimated.

[0051] In this embodiment, in order to reduce the influence of errors due to the translational acceleration component during vehicle travel included in the measurement values ​​of the inertial measurement unit 30, the measurement value a of the three-axis acceleration during vehicle travel, which is an error factor, is xk , a yk , a zk The translational acceleration component may be extracted by filtering processing, and the observation error covariance matrix of the Kalman filter may be set based on the extracted translational acceleration component.

[0052] The impact response that occurs when the wheel flange comes into contact with the rail in a sharp curve section is a factor that causes errors in the angle estimation value. In the above equation (3), which calculates the angle from the integral of the angular velocity, and the above equation (8), which calculates the angle from the low-frequency component of acceleration (i.e., the projected component of gravitational acceleration), the process noise v k and the observation noise w kis assumed to be Gaussian noise with zero mean. This noise setting represents the error component of each equation. The process noise v k means the measurement error and integration error of the angular velocity data. Observation noise w k means the measurement error of the acceleration data and the estimation error of the gravitational acceleration projection component extracted from the acceleration.

[0053] When the wheel flange comes into contact with the rail on a sharp curve, an impulsive response may be observed in both the angular velocity measurement value and the acceleration measurement value, but the impulsive response is characterized by being dominant across all frequency bands. Here, the above formula (3) is valid regardless of whether or not there is an impulsive response, but the observation vector y in the above formula (8) k is an angle calculated from the gravitational acceleration projection component, and becomes an error factor when translational acceleration acts on the inertial measurement unit 30. That is, in such a traveling section, the observation noise w k By setting to a large value, the reliability of the above formula (8) is reduced (in other words, the reliability of the above formula (6) is increased), thereby appropriately expressing the actual response.

[0054] Therefore, the observation noise w k is set to improve the accuracy of the angle estimation value. Specifically, the index R representing the translational acceleration is set. x , R y is introduced, and the observation noise w k The covariance matrix R of

[0055] Observation vector y k Roll angle θ 0k and pitch angle φ 0k The calculation accuracy of the left and right acceleration a yk , longitudinal acceleration a xk In both cases, errors can occur due to translational acceleration. yk Regarding the impact response when passing through a sharp curve, it appears as a translational acceleration, and the longitudinal acceleration a xkFor the vehicle speed, the acceleration / deceleration required to maintain a constant speed appears as translational acceleration. x , R y Define

[0056] Index R x is the measured value of longitudinal acceleration a x Only high frequency components are extracted from the index R using a band pass filter, and the extracted value is squared, and then the moving average over a predetermined interval (number of moving average samples) is taken. y is the measured value of the lateral acceleration a y Only high frequency components are extracted from the index R using a band pass filter, and the extracted value is squared, and then the moving average over a predetermined interval (number of moving average samples) is taken. x , R y Since is an index for correcting low-frequency responses (i.e., the tilt angle estimated from the gravitational acceleration projection component), the estimated value is stabilized by squaring it and taking a moving average to remove high frequencies. For example, the high-frequency components to be extracted may be frequency components between 1 and 7 Hz, and the number of moving average samples may be 250 sample times.

[0057] Impact responses or responses due to acceleration / deceleration are externally forced responses and can be assumed to be dominant across all frequency components. On the other hand, components due to angle changes are dominant in the low frequency band, for example, below 1 Hz, and measurement noise is dominant in the high frequency band, for example, above 7 Hz. Therefore, a bandpass filter is used to remove frequency components in these bands.

[0058] In detail, assuming that the vehicle speed is almost constant, the average vehicle speed in the measurement data to be filtered is V ave [m / s], maximum spatial frequency of track irregularity F space [cycle / m], the cutoff frequency F time [Hz] can be expressed by the following formula (9).

[0059] F time =F space ×Vave ···(9)

[0060] High frequency cutoff frequency F time_h can be expressed as an empirical formula with generality as in the following formula (10).

[0061] F time_h =7×F space ×V ave ···(10)

[0062] Index R x , R y The moving average sample number s can be expressed empirically by the following formula (11) using the sampling frequency fs [Hz] of the measurement values.

[0063] s=fs / (2×F time ) ···(11)

[0064] For example, the sampling frequency fs=500Hz, the cutoff frequency F time = 1Hz, the moving average sample number s is 250 sample times.

[0065] Figure 8 shows the index R representing the translational acceleration. x Fig. 9 shows an example of the translational acceleration index R y In Fig. 8, the indicator R x In Fig. 9, the indicator R indicated by the dashed line frame indicates that the vehicle is being affected by acceleration and deceleration to maintain a constant speed. y Locations with large values ​​indicate that the train is being affected by the impact of passing through a sharp curve.

[0066] Next, the calculated index R x , R y Using this, the observation noise w according to the magnitude of the translational acceleration component k The error covariance matrix R(k) may be set by adding a value obtained by multiplying the extracted translational acceleration component by a predetermined coefficient to the diagonal components. Specifically, the observation noise wk The variance-covariance matrix R(k) of y (k), R x (k) and index R y (k), R x Coefficient K for (k) y , K. x is used to set it as in the following equation (12).

[0067]

number

[0068] Furthermore, the process noise Q may be set constant regardless of time, for example, as shown in the following equation (13).

[0069]

number

[0070] In addition, the coefficient K y , K. x is the index R y (k), R x (k) is a value for setting the noise variance of the observation equation expressed by the above equation (8), and cannot be theoretically derived. y , K. x The value that maximizes the accuracy of estimating the elevation displacement is set as the optimum value for the process noise Q. The process noise Q is also set appropriately based on the value that maximizes the accuracy of estimating the elevation displacement.

[0071] In this manner, in this embodiment, the observed noise w k The variance-covariance matrix R(k) is set as shown in the above equation (12), and the observation noise w k is sequentially set according to the magnitude of the translational acceleration component. This reduces the influence of errors due to the translational acceleration component during vehicle travel, which is included in the measurement values ​​of the inertial measurement unit 30. As a result, the observation vector y at time k is k , i.e., the estimated value of the roll angle θ 0kand the estimated pitch angle φ 0k can be calculated with high accuracy. It is expressed by equations (6) to (8), and the observation noise w k The roll angle θ is calculated by a Kalman filter based on a state space model in which the variance-covariance matrix R(k) of k and pitch angle φ k Once is estimated, the elevation displacement h can be calculated using the above formula (1).

[0072] [2-2. Track State Estimation Device] The configuration of a track state estimation device 100 that estimates track states based on the above-mentioned track state estimation method will be described with reference to Fig. 10. Fig. 10 is a functional block diagram showing the configuration of the track state estimation device 100 according to this embodiment. As shown in Fig. 10, the track state estimation device 100 includes an acceleration noise removal unit 110, an inclination angle estimation unit 120, and a track state calculation unit 130.

[0073] The acceleration noise removal unit 110 removes the three-axis acceleration a measured by the inertial measurement unit 30 based on the traveling speed of the vehicle 1. xk , a yk , a zk Of these, longitudinal acceleration a xk (=a xIMU When the traveling speed V is acquired by the speed measurement device 40 installed in the vehicle 1, the acceleration noise elimination unit 110 calculates the acceleration component a due to the acceleration / deceleration of the vehicle 1 based on the above equation (4). V Then, the longitudinal acceleration a xk The acceleration noise elimination unit 110 corrects the longitudinal acceleration a xk (=a x * IMU ) to the tilt angle estimation unit 120.

[0074] The tilt angle estimation unit 120 estimates the three-axis acceleration a measured by the inertial measurement unit 30. xk , a yk , a zk and three-axis angular velocity ω xk , ωyk , ω zk Based on this, the horizontal inclination angle (roll angle θ k ) and the front-to-rear tilt angle (pitch angle φ k ) is estimated. xk is the longitudinal acceleration a corrected by the acceleration noise removal unit 110. xk (=a x * IMU ) is used.

[0075] First, the tilt angle estimation unit 120 calculates the measured value a of the three-axis acceleration. xk , a yk , a zk The tilt angle estimation unit 120 extracts the gravitational acceleration projection component and the translational acceleration component from the measured value a of the three-axis acceleration. xk , a yk , a zk The translational acceleration component extracted from x , R y and calculate the observation noise w at time k using the above equation (12). k Then, the tilt angle estimation unit 120 calculates the observation vector y k (Estimated roll angle θ 0k and the estimated pitch angle φ 0k ) is calculated by the tilt angle estimation unit 120. The tilt angle estimation unit 120 calculates the observation noise w k The roll angle θ is calculated by a Kalman filter based on a state space model in which the variance-covariance matrix R(k) of k and pitch angle φ k The tilt angle estimation unit 120 calculates the calculated roll angle θ k and pitch angle φ k to the orbit state calculation unit 130.

[0076] The track condition calculation unit 130 calculates the horizontal tilt angle (roll angle θ k ) and the front-to-rear tilt angle (pitch angle φ k) and calculates track state information representing the track state. In this embodiment, elevation variation is calculated as the track state information. When elevation variation is calculated as the track state information, the track state calculation unit 130 calculates the pitch angle φ calculated by the inclination angle estimation unit 120. k Based on this, the elevation displacement h is calculated using the above formula (1).

[0077] Here, the calculated elevation displacement h has an estimation error of low frequency components. It is thought that this estimation error occurs because the horizontal tilt angle and longitudinal tilt angle calculated by the Kalman filter contain some integral errors of angular velocity. Therefore, a high-pass filter may be applied to the calculated elevation displacement h. The cutoff frequency of the high-pass filter may be determined according to the wavelength of the elevation displacement h to be obtained, for example, V ave / 30(V ave may be an average velocity [m / s]. By applying a high-pass filter, it is possible to suppress deviations in the elevation displacement h estimated each time the inertial measurement unit 30 measures the triaxial acceleration and triaxial angular velocity, thereby obtaining reproducible estimated values.

[0078] The track state calculation unit 130 outputs the calculated elevation change h to, for example, the output device 200. The output device 200 may be, for example, a display device such as a monitor.

[0079] The track state estimation device 100 according to this embodiment is configured with various processors such as a CPU (Central Processing Unit) and a DSP (Digital Signal Processor), and the above functions can be realized by the processors operating in accordance with predetermined programs.

[0080] The track state estimation device 100 according to this embodiment may be installed on the vehicle 1 or may be installed in a separate location. xk , a yk , a zk and the measured value of the three-axis angular velocity ω xk , ωyk , ω zk The travelling speed V of the vehicle 1 measured by the speed measuring device 40 is recorded in a storage device (not shown) in association with the absolute position of the vehicle 1 on the rail 5 detected by the position detecting device. The track state estimating device 100 uses the measured values ​​a of the three-axis acceleration recorded in the storage device. yk , a zk , the measured value of the three-axis angular velocity ω xk , ω yk , ω zk , the running speed V, and the absolute position of the vehicle 1, the track state can be estimated.

[0081] [2-3. Orbit state estimation method] A track state estimation method implemented by the track state estimation device 100 according to this embodiment will be described with reference to Fig. 11. Fig. 11 is a flowchart showing an example of the track state estimation method according to this embodiment.

[0082] (S10, S20: Acquisition of 3-axis acceleration and 3-axis angular velocity) First, the triaxial acceleration a is measured by the inertial measurement unit 30 installed in the body 10 or the axle box 27 of the vehicle 1. xk , a yk , a zk and three-axis angular velocity ω xk , ω yk , ω zk The three-axis acceleration and three-axis angular velocity measured by the inertial measurement unit 30 are measured (S10, S20). xk , a yk , a zk , ω xk , ω yk , ω zk is output to the acceleration noise removal unit 110 and the tilt angle estimation unit 120 of the track state estimation device 100.

[0083] (S30: Acquisition of vehicle speed) Furthermore, the running speed V is measured by the speed measurement device 40 installed in the vehicle 1 (S30). The running speed V of the vehicle 1 measured by the speed measurement device 40 is output to the acceleration noise removal unit 110 of the track state estimation device 100.

[0084] (S40: Acceleration noise removal) Next, the acceleration noise removal unit 110 calculates the three-axis acceleration a obtained in step S20 based on the traveling speed of the vehicle 1 obtained in step S30. xk , a yk , a zk Of these, longitudinal acceleration a xk (=a xIMU The acceleration noise removal unit 110 removes acceleration / deceleration noise from the traveling speed V of the vehicle 1 based on the above equation (4) by subtracting acceleration / deceleration component a V Then, the acceleration noise elimination unit 110 calculates the longitudinal acceleration a xk The acceleration noise elimination unit 110 corrects the longitudinal acceleration a xk (=a x * IMU ) to the tilt angle estimation unit 120.

[0085] (S50: Extraction of gravitational acceleration projection component) Next, the tilt angle estimation unit 120 calculates the measured value a of the three-axis acceleration. xk , a yk , a zk The gravitational acceleration projection component is extracted from (S50). xk In step S40, the longitudinal acceleration a xk (=a x * IMU ) is used. Measured value of 3-axis acceleration a xk , a yk , a zk The gravitational acceleration projection component is a low-frequency acceleration component, for example, a component of 0 to 1 Hz. The tilt angle estimation unit 120 uses, for example, a low-pass filter to estimate the measured value a of the three-axis acceleration. xk , a yk , a zk The gravitational acceleration projected component can be extracted from

[0086] (S60: Extraction of translational acceleration component) The tilt angle estimation unit 120 also estimates the three-axis acceleration measurement value a xk , a yk , a zk The translational acceleration component is extracted from the measured value a of the three-axis acceleration (S60). xk , a yk , a zk The translational acceleration component is a high-frequency acceleration component, for example, a component of 1 to 7 Hz. The tilt angle estimation unit 120 uses, for example, a band-pass filter to filter the measured value a of the three-axis acceleration. xk , a yk , a zk The tilt angle estimation unit 120 extracts the translational acceleration component and calculates an index R representing the translational acceleration. x , R y Calculate.

[0087] (S70: Estimation of tilt angle) Next, the inclination angle estimation unit 120 estimates the inclination angle of the vehicle body 10 or the wheel set 25 (S70). The inclination angle can be calculated using a Kalman filter based on the state space model expressed by the above equations (6) to (8). Here, the inclination angle estimation unit 120 estimates the inclination angle of the vehicle body 10 or the wheel set 25 (S70). x , R y Using the above equation (12), the observation noise w k The variance-covariance matrix R(k) of the observation equation shown in the above equation (8) is set. k is sequentially set according to the magnitude of the translational acceleration component. As a result, the influence of errors due to the translational acceleration component during vehicle travel, which is included in the measurement value, is reduced, and the observation vector y k (Estimated roll angle θ 0k and the estimated pitch angle φ 0k ) can be calculated with high accuracy.

[0088] The tilt angle estimation unit 120 is expressed by equations (6) to (8), and the observation noise w k The roll angle θ is calculated by a Kalman filter based on a state space model in which the variance-covariance matrix R(k) of k and pitch angle φ kThen, the tilt angle estimation unit 120 calculates the calculated roll angle θ of the vehicle body 10 or the wheel set 25. k and pitch angle φ k to the orbit state calculation unit 130.

[0089] (S80: Estimation of track state information) Thereafter, the track state calculation unit 130 calculates the horizontal tilt angle (roll angle θ k ) and the front-to-rear tilt angle (pitch angle φ k ) and calculates track condition information representing the track condition (S80). In this embodiment, elevational deviation is calculated as the track condition information. When elevational deviation is calculated as the track condition information, the track condition calculation unit 130 calculates the track condition information based on the pitch angle φ of the car body 10 or the wheel set 25 calculated in step S70. k Based on this, the elevation displacement h is calculated using the above formula (1).

[0090] Here, a high-pass filter may be applied to the elevation displacement h calculated in step S80. As described above, the calculated elevation displacement h generates an estimation error of low-frequency components. By applying a high-pass filter, it is possible to suppress deviations in the elevation displacement h estimated each time the inertial measurement unit 30 measures the triaxial acceleration and triaxial angular velocity, and to obtain reproducible estimated values. The cutoff frequency of the high-pass filter may be determined according to the wavelength of the elevation displacement h to be obtained, for example, V ave / 30(V ave may be the average velocity [m / s]).

[0091] Figure 12 shows an example of the elevation displacement h calculated in step S80 before and after high-pass filtering. Figure 12 shows the true value of the elevation displacement measured by the 10m chord sine wave method using a hand-operated track inspection device. Here, V is used as the cutoff frequency of the high-pass filter. ave12, it can be seen that the deviation of the calculated elevation deviation h from the true value can be reduced by applying high-pass filtering. The track state calculation unit 130 outputs the calculated elevation deviation h to the output device 200.

[0092] [3. Summary] The above describes a vehicle according to one embodiment of the present invention, as well as a track condition estimation device and a track condition estimation method for estimating elevational displacement as a track condition. According to this embodiment, an inertial measurement unit (IMU) is installed on the vehicle's carbody or axle box, and the horizontal tilt angle and longitudinal tilt angle at the IMU installation location are estimated based on the triaxial acceleration and triaxial angular velocity measured by the IMU. The horizontal tilt angle and longitudinal tilt angle can be estimated using a Kalman filter. In this case, acceleration noise is removed from the longitudinal acceleration of the triaxial acceleration before estimating the horizontal tilt angle and longitudinal tilt angle. To reduce the influence of errors due to translational acceleration components contained in the IMU measurements while the vehicle is traveling, the translational acceleration components, which are an error factor, are extracted from the triaxial acceleration measurements while the vehicle is traveling by filtering, and the observation error covariance matrix of the Kalman filter is sequentially set based on the extracted translational acceleration components. This enables accurate estimation of the horizontal tilt angle and longitudinal tilt angle of the wheelset. High-precision estimation of the horizontal tilt angle and longitudinal tilt angle of the wheelset improves the accuracy of the elevation displacement estimation. Furthermore, the three-axis acceleration and three-axis angular velocity can be measured by simply running the vehicle, making it possible to carry out the measurements easily. [Example]

[0093] To verify the effectiveness of the track state estimation method according to the above embodiment, the elevation change estimated by the track state estimation method was evaluated. The vehicle used for the verification was the one shown in Fig. 1, and an inertial measurement unit was installed near the connection point between the car body and the bogie, as shown in Fig. 3. The distance between supports (reference length) L was 6.3 m.

[0094] In this verification, the true value was the elevation deviation measured by the 10m chord sine wave method using a hand-operated track inspection device. The elevation deviation was also estimated using the track condition estimation method according to the above embodiment. At this time, the coefficient K y , K. x are both 1.0 × 10 -6 The variance-covariance matrix R of the above equation (12) was set as follows. The process noise Q was set as given by the above equation (13). The elevation displacement was estimated using the measurement values ​​of the inertial measurement unit as is (i.e., using the longitudinal acceleration before correction) and using the longitudinal acceleration after correction, in which the acceleration noise was removed from the longitudinal acceleration.

[0095]

number

[0096] FIG. 13 shows elevation displacement estimated using the track state estimation method according to the above embodiment and elevation displacement using the 10-meter-chord sine wave method. Note that FIG. 13 shows the elevation displacements using the 10-meter-chord sine wave method for two paired rails. As shown in FIG. 13, the elevation displacements estimated using the track state estimation method according to the above embodiment closely track the true value. Furthermore, when the corrected longitudinal acceleration was used in estimating elevation displacement, the deviation from the true value was smaller than when the uncorrected longitudinal acceleration was used, with the error reduced by approximately 5 to 10 mm at maximum. This demonstrates that applying the track state estimation method according to the above embodiment reduces the influence of errors due to translational acceleration, including the influence of impact responses during vehicle travel, and also reduces the influence of acceleration noise due to vehicle acceleration and deceleration, thereby improving the estimation accuracy of elevation displacement.

[0097] Although the preferred embodiments of the present invention have been described in detail above with reference to the accompanying drawings, the present invention is not limited to these examples. It is clear that a person skilled in the art to which the present invention pertains can conceive of various modifications and alterations within the scope of the technical ideas set forth in the claims, and it is understood that these also naturally fall within the technical scope of the present invention.

[0098] For example, in the above embodiment, the measured value of the three-axis acceleration a xk , a yk , a zk The low-frequency acceleration component, which is the gravitational acceleration projection component of the acceleration, is taken as the component of 0 to 1 Hz, and the measured value of the three-axis acceleration a xk , a yk , a zk The high-frequency acceleration components, which are translational acceleration components, are set to 1 to 7 Hz, but the present invention is not limited to this example. The frequency bands of the low-frequency acceleration components and high-frequency acceleration components are set assuming a low-speed vehicle used to transport heavy materials in a steelworks, for example. The frequency bands of the low-frequency acceleration components and high-frequency acceleration components may be set appropriately according to the running speed of the vehicle, for example, based on the above formulas (9) and (10). The index R x , R y The number of moving average samples s may also be set based on the above equation (11). [Explanation of symbols]

[0099] 1 vehicle 5 Rail 10 Body 20 Bogie section 21 Bogie frame 23 wheels 25 wheel set 27 Axle box 29 Coil spring 30 Inertial Measurement Unit 40 Speed ​​measuring device 100 Orbital state estimation device 110 Acceleration noise removal section 120 Tilt angle estimation part 130 Track state calculation unit 200 Output Device

Claims

1. an acceleration noise removal step of removing acceleration noise generated due to acceleration and deceleration of the vehicle from three-axis acceleration measured by an inertial measurement unit installed on a body or an axle box of the vehicle based on the running speed of the vehicle; an inclination angle estimation step of estimating a horizontal inclination angle and a longitudinal inclination angle at an installation position of the inertial measurement unit based on the triaxial acceleration from which the acceleration noise has been removed and the triaxial angular velocity measured by the inertial measurement unit; a track state calculation step of calculating track state information representing a track state based on the estimated horizontal inclination angle and the estimated longitudinal inclination angle; Including, a longitudinal acceleration component of the three-axis acceleration measured by the inertial measurement unit, the longitudinal acceleration component being calculated from the vehicle's traveling speed, to remove acceleration noise contained in the three-axis acceleration;

2. In the tilt angle estimation step, extracting a translational acceleration component during vehicle travel from the three-axis acceleration from which the acceleration noise has been removed by filtering; setting an error covariance matrix of an observation model of a Kalman filter based on the extracted translational acceleration component; using a Kalman filter formulated from a state space model configured by an observation model in which the error covariance matrix is ​​set, to estimate a horizontal tilt angle and a longitudinal tilt angle at an installation position of the inertial measurement unit from the three-axis acceleration and the three-axis angular velocity from which the acceleration noise has been removed; 2. The track state estimation method according to claim 1, wherein the track state calculation step calculates elevational deviation as the track state information based on the estimated longitudinal tilt angle.

3. 3. The track state estimation method according to claim 2, wherein the track state calculation step performs high-pass filtering on the calculated elevation deviation to remove estimation errors of low-frequency components.

4. 4. The orbit state estimation method according to claim 2, wherein the translational acceleration components are extracted by band-pass filtering to remove dominant frequency components of track irregularity from the lateral acceleration and longitudinal acceleration among the three-axis accelerations.

5. 5. The orbit state estimation method according to claim 2, wherein the error covariance matrix is ​​set by adding, to diagonal elements, values ​​obtained by multiplying the extracted translational acceleration components by predetermined coefficients.

6. an acceleration noise removal unit that removes acceleration noise generated due to acceleration and deceleration of the vehicle from three-axis acceleration measured by an inertial measurement unit installed on a body or an axle box of the vehicle based on the running speed of the vehicle; an inclination angle estimation unit that estimates a horizontal inclination angle and a longitudinal inclination angle at an installation position of the inertial measurement unit based on the triaxial acceleration from which the acceleration noise has been removed and the triaxial angular velocity measured by the inertial measurement unit; a track state calculation unit that calculates track state information representing a track state based on the estimated horizontal inclination angle and the estimated longitudinal inclination angle; Equipped with the acceleration noise removal unit removes acceleration noise contained in the three-axis acceleration by subtracting an acceleration component calculated from the vehicle's traveling speed from a longitudinal acceleration component of the three-axis acceleration measured by the inertial measurement unit.

7. a speed measuring device for measuring the traveling speed of a vehicle; an inertial measurement unit installed on a body or an axle box of the vehicle; an acceleration noise removal unit that removes acceleration noise generated due to acceleration and deceleration of the vehicle from the three-axis acceleration measured by the inertial measurement unit based on the traveling speed of the vehicle; an inclination angle estimation unit that estimates a horizontal inclination angle and a longitudinal inclination angle at an installation position of the inertial measurement unit based on the triaxial acceleration from which the acceleration noise has been removed and the triaxial angular velocity measured by the inertial measurement unit; a track state estimation device including a track state calculation unit that calculates track state information representing a track state based on the estimated horizontal inclination angle and the estimated longitudinal inclination angle; Equipped with the acceleration noise removal unit removes acceleration noise contained in the three-axis acceleration by subtracting an acceleration component calculated from the vehicle's traveling speed from a longitudinal acceleration component of the three-axis acceleration measured by the inertial measurement unit.