Wheel slip prediction device

Abstract

To provide a wheel slip prediction device that can predict in advance slip of wheels due to fallen leaves.SOLUTION: A slip prediction device 9 is a device that predicts slip of wheels due to fallen leaves, and comprises: a fallen leaf coverage detection unit 13 that detects the state of coverage of a track with fallen leaves; a fallen leaf coating detection unit 16 that detects the state of formation of fallen leaf coating on rails of the track; and a slip prediction unit 18 that predicts slip of wheels on the basis of a result of detection performed by the fallen leaf coverage detection unit 13 and a result of detection performed by the fallen leaf coating detection unit 16. The slip prediction device 9 comprises: a fallen leaf coverage factor calculation unit 12 that calculates a fallen leaf coverage factor that is the ratio of an area occupied by color information of fallen leaves to a predetermined area of the track; and a fallen leaf coating factor calculation unit 15 that calculates a fallen leaf coating detection ratio that is the ratio of an area occupied by pixel values of fallen leave coating to a predetermined area of the rails on the track.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] The present invention relates to a wheel slippage prediction device that predicts wheel slippage caused by fallen leaves. [Background technology]

[0002] In railway vehicles, when wheel spin or skid is detected, damage to the rails and wheels is prevented and the necessary driving force is maintained by controlling the output of a prime mover such as a traction motor or engine. A conventional readhesion control method (hereinafter referred to as Prior Art 1) maintains the torque of the driving wheels in a reduced state when wheel spin or skid is detected, and restores the torque of the driving wheels when the numerical value of the acceleration of the rotation of the driving wheels changes in a predetermined change pattern while the torque of the driving wheels is reduced (see, for example, Patent Document 1). In Prior Art 1, the timing to start the recovery control of the torque of the driving wheels is appropriately determined from the change in acceleration.

[0003] In mountain railway sections, dead leaves from trees along the tracks are often seen falling around the tracks in autumn. As trains pass, the fallen leaves are crushed by the wheels and adhere to the rail surface. This deposit absorbs moisture from condensation and other sources, causing the tannins contained in the leaves to react with iron, a component of the rail, to form a black film. Over time, the black film thickens and, under humid conditions, becomes a soft paste-like substance, reducing adhesion and causing trains to skid or slide (see, for example, Non-Patent Document 1). Wheel skid and slide caused by fallen leaves in the autumn have been thought to be caused by the formation of a black film from fallen leaves on the rail head surface and the wetness of the rail head surface due to morning dew or light rain.

[0004] A conventional slip control device (hereinafter referred to as Prior Art 2) is equipped with an obstacle clearing mechanism that removes fallen leaves from the rail, a liquid injection mechanism that washes off resin from the top surface of the rail behind the obstacle clearing mechanism in the direction of travel, and an adhesion-increasing material spreading mechanism that spreads an adhesion-increasing material behind the liquid injection mechanism in the direction of travel (see, for example, Patent Document 2). Prior Art 2 cleans the rails from which fallen leaves have adhered, improving the adhesion of the wheels to the rails and appropriately controlling slip of the railway vehicle wheels. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] Japanese Patent Application Publication No. 2018-113850

[0006] [Non-Patent Document 1] Chen Hua and four others, "Understanding the mechanism of wheel slippage and sliding caused by fallen leaves," RRR, Vol. 76, Kenyusha, May 1, 2019, pp. 20-23

[0007] [Patent Document 2] Japanese Patent Application Publication No. 2017-013562 Summary of the Invention [Problem to be solved by the invention]

[0008] To combat wheel slippage and sliding caused by fallen leaves, railway operators have taken measures such as deploying sand-spreading vehicles and cleaning the rail top surface. The Railway Technical Research Institute, a public interest incorporated foundation, has also developed a method for removing the black film that is the main cause of wheel slippage by spraying citric acid on the rail top surface, which has proven to be somewhat effective. However, both Conventional Techniques 1 and 2 are related to controlling vehicle equipment and cleaning the rail top surface as countermeasures for wheel slippage and sliding, or for resin buildup on the rail top surface. Therefore, Conventional Techniques 1 and 2 have the problem of being unable to predict wheel slippage and sliding in advance. Conventional Technique 2 also has the problem of lacking indicators for removing fallen leaves from the rails or for determining the timing of removal.

[0009] An object of the present invention is to provide a wheel slippage prediction device that can predict wheel slippage caused by fallen leaves in advance. [Means for solving the problem]

[0010] The present invention solves the above problems by the means described below. Although the description will be given with reference numerals corresponding to the embodiments of the present invention, the present invention is not limited to these embodiments. The invention of claim 1 is a wheel slip prediction device (9) for predicting wheel (3) slippage caused by fallen leaves (L), as shown in Figures 1, 2 and 12, which includes a leaf cover detection unit (13) for detecting (S130) the state of coverage of a track (1) with fallen leaves, a leaf film detection unit (16) for detecting (S160) the state of formation of a leaf film (F) on a rail (1a) of the track, and a slip prediction unit (18) for predicting (S180) wheel slippage based on the detection results of the leaf cover detection unit and the leaf film detection unit.

[0011] The invention of claim 2 is a wheel slip prediction device as described in claim 1, characterized in that it comprises a color information extraction unit (11) that extracts (S110) color information of fallen leaves on the track based on a photographed image of the track, and a leaf coverage calculation unit (12) that calculates (S120) a leaf coverage rate (R1), which is the ratio of the area occupied by the color information of the fallen leaves to a specified area of ​​the track, based on the extraction result of the color information extraction unit, and the leaf coverage detection unit detects (S130) the coverage state of the fallen leaves based on the leaf coverage rate.

[0012] The invention of claim 3 is a wheel slip prediction device according to claim 1, further comprising a pixel value calculation unit (14) that calculates (S140) the pixel values ​​of the leaf film based on the photographed image of the track, and a leaf film rate calculation unit (15) that calculates (S150) a leaf film detection rate (R2), which is the ratio of the area occupied by the pixel values ​​of the leaf film to a specified area on the rail, based on the calculation result of the pixel value calculation unit, and the leaf film detection unit detects (S170) the formation state of the leaf film based on the leaf film detection rate.

[0013] The invention of claim 4 is a wheel slip prediction device according to claim 1, further comprising a wet condition detection unit (17) that detects the wet condition of the rail (S170), and the slip prediction unit predicts wheel slip based on the detection result of the wet condition detection unit.

[0014] The invention of claim 5 is the wheel slip prediction device of claim 4, wherein the wet condition detection unit detects weather data around the track and a rail temperature (T P ) is a wheel slip prediction device that detects the wetness state of the rail based on the above.

[0015] The invention of claim 6 is a wheel slip prediction device (9) for predicting wheel (3) slippage caused by fallen leaves (L), as shown in Figures 1, 2 and 12, which includes a leaf film detection unit (16) for detecting (S160) the formation state of a leaf film (F) on the rail of a track (1), a wet condition detection unit (17) for detecting (S170) the wet condition of the rail, and a slip prediction unit (18) for predicting (S180) wheel slippage based on the detection results of the leaf film detection unit and the wet condition detection unit.

[0016] The invention of claim 7 is a wheel slip prediction device according to claim 6, further comprising a pixel value calculation unit (14) that calculates (S140) the pixel values ​​of the leaf film based on the photographed image of the track, and a leaf film rate calculation unit (15) that calculates (S150) a leaf film detection rate (R2), which is the ratio of the area occupied by the pixel values ​​of the leaf film to a specified area on the rail, based on the calculation result of the pixel value calculation unit, and the leaf film detection unit detects (S170) the formation state of the leaf film based on the leaf film detection rate.

[0017] The invention of claim 8 is the wheel slip prediction device of claim 7, wherein the wet condition detection unit detects weather data around the track and a rail temperature (T P ) is a wheel slip prediction device that detects the wetness state of the rail based on the above.

[0018] According to this invention, wheel slippage caused by fallen leaves can be predicted in advance. [Brief explanation of the drawings]

[0019] [Figure 1] 1 is a configuration diagram that schematically shows a wheel slip prediction system that includes a wheel slip prediction device according to an embodiment of the present invention. [Figure 2] 1 is a configuration diagram of a wheel slip prediction device according to an embodiment of the present invention; [Figure 3]1A and 1B are photographs showing examples of leaf litter that can cause wheel slippage predicted by a wheel slip prediction device according to an embodiment of the present invention. FIG. 1A is a photograph of wheel tracks on leaves on the top surface of a rail during the daytime in autumn, and FIG. 1B is a photograph of a black film formed on the top surface of a rail during the night or early morning in autumn. [Figure 4] 10A to 10C are photographs showing, as an example, in chronological order, the state of leaf fall within a prediction interval of a certain track predicted by a wheel slip prediction device according to an embodiment of the present invention. [Figure 5] 3 is a schematic diagram for explaining an extraction process of a color information extraction unit of the wheel slip prediction device according to the embodiment of the present invention; FIG. [Figure 6] 5A to 5C are photographs showing, in chronological order, examples of the results of extraction of color information by a color information extraction unit of the wheel slip prediction device according to the embodiment of the present invention. [Figure 7] 1 is a graph showing, as an example, the detection results of a leaf cover detection unit of a wheel slip prediction device according to an embodiment of the present invention and the occurrence of wheel slip; [Figure 8] 1 is a schematic diagram for explaining the detection process of the leaf film detection unit of a wheel slip prediction device according to an embodiment of the present invention, in which (A) is an image of the rail portion cut out from a photographed image, (B) is an enlarged image of the cut-out photographed image, and (C) is a graph showing an example of the leaf film area. [Figure 9] 1 is a graph showing, as an example, the detection results of a leaf film detection unit of a wheel slip prediction device according to an embodiment of the present invention and the occurrence of wheel slip; [Figure 10] 10 is a graph showing an example of the results of trial calculation of rail dew points by a wet condition detection unit of the wheel slip prediction device according to the embodiment of the present invention. [Figure 11] 1 is a graph showing an example of the detection results of the leaf cover detection unit and the leaf film detection unit of the wheel slip prediction device according to an embodiment of the present invention, and the occurrence of wheel slip. [Figure 12] 3 is a flowchart illustrating the operation of the wheel slip prediction device according to the embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0020] Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. The track 1 shown in Figure 1 is a passageway (railroad track) on which vehicles 2 travel. The track 1 has left and right rails 1a that support and guide wheels 3 of the vehicles 2 to allow the vehicles 2 to travel, and the rails 1a have head surfaces 1b that come into contact with the treads 3a of the wheels 3 and directly support the wheels 3. The track 1 may be a ballast track made up of ballast and sleepers, or a slab track made up of concrete slabs.

[0021] The vehicle 2 is a moving body that travels along the track 1. The vehicle 2 is, for example, a railway vehicle such as an electric train, a diesel railcar, a locomotive, a passenger car, or a freight car. The vehicle 2 is equipped with wheels 3 that are in rolling contact with the rail 1a. The wheels 3 are equipped with tread surfaces 3a that come into contact with the head surface 1b of the rail head and receive frictional resistance, and flange surfaces 3b that are formed continuously on the outer periphery of the wheels 3 to prevent them from coming off the track.

[0022] The fallen leaves L shown in Figure 1 are leaves that have fallen from plants. The fallen leaves L include leaves that have fallen from plants that shed their leaves periodically in a certain season, as well as some stems in addition to the leaves. The fallen leaves L are leaves that have fallen from deciduous broadleaf trees, deciduous coniferous trees, or deciduous tall trees that grow along railway lines, such as ginkgo, cedar, cherry, maple, and zelkova, which form black membranes.

[0023] The leaf film F shown in Figures 1 and 3 is a film formed on the rail 1a due to fallen leaves. The leaf film F is a soft black film formed by leaves L crushed between the rail 1a and the wheels 3. Depending on environmental conditions such as the location where the black film forms and the type of leaves L, the leaf film F may form uniformly on the top surface 1b or sparsely on the top surface 1b with a diameter of approximately 1 to 10 cm. As shown in Figure 3(A), the leaf film F is formed as a black film on the top surface 1b of the rail 1a when leaves L are crushed between the rail 1a and the wheels 3 in a section of track where wheel slippage, such as wheel slip, occurs, as shown in Figure 3(B). The leaf film F is thought to be formed in a humid environment after leaves L accumulated on the track 1 are scattered by the passing train 2 and crushed between the rail 1a and the wheels 3. The leaf coating F can be predicted from two factors: the wet condition on the rail 1a, such as condensation or light rain, and the state of the track 1 covered with leaves L.

[0024] The slip prediction system 4 shown in Figures 1 and 2 is a system that predicts slippage of the wheel 3 due to fallen leaves. Here, slippage of the wheel 3 refers to a state in which a speed difference occurs at the contact surface between the wheel 3 and the rail 1a, resulting in a difference between the running speed of the wheel 3 and the rotational speed of the wheel 3. The slippage of the wheel 3 can be, for example, wheel slip or skid. Wheel slippage is a macroscopic slippage that occurs between the tread 3a of the wheel 3 and the top surface 1b of the rail 1a when the driving force transmitted to the wheel 3 during powering (driving) is greater than the adhesion force acting between the wheel 3 and the rail 1a. This type of wheel slippage is likely to occur when the rail 1a is wet and contaminated by snow, rain, fallen leaves, or the like. Skidpage is a macroscopic slippage that occurs between the tread 3a of the wheel 3 and the top surface 1b of the rail 1a when the rotational speed of the wheel 3 is lower than the running speed of the vehicle 2 during braking, causing the wheel 3 to slide on the rail 1a. Wheel slippage due to fallen leaves L is particularly likely to occur between 6:00 and 8:00 AM from October to December, and occurs when fallen leaves L and moisture get between the rail 1a and the wheel 3.

[0025] As shown in FIG. 1, the slippage prediction system 4 performs image analysis on photographed images of the track 1 to quantify the amount of fallen leaves L covering the track 1 and to quantify the leaf film F formed on the top surface 1b of the rail 1a. The slippage prediction system 4 acquires meteorological data around the track 1 and detects the temperature of the rail 1a to predict the occurrence of condensation on the rail 1a. Based on the quantification results of the coverage state of fallen leaves L, the detection results of the formation state of the leaf film F, and the prediction results of the occurrence of condensation, the slippage prediction system 4 comprehensively evaluates the occurrence of wheel slippage 3 and predicts the risk of wheel slippage 3 caused by fallen leaves L. As shown in FIGS. 1 and 2, the system is equipped with a photographing device 5, a rail temperature monitoring device 6, a meteorological observation device 7, a communication network 8, and a slippage prediction device 9.

[0026] The photographing device 5 shown in FIGS. 1 and 2 is a device that photographs the track 1. As shown in FIG. 1, the photographing device 5 is a photographing or imaging device that photographs a predetermined detection section S in the longitudinal direction of the track 1. Here, the detection section S is a predetermined area for detecting fallen leaves L and leaf litter F, and is a monitoring section for monitoring whether wheel 3 slippage occurs. The detection section S is designated in advance along the track 1 depending on the occurrence of fallen leaves L and the occurrence of wheel 3 slippage. The detection section S is, for example, a slope section where wheel 3 slippage is likely to occur, or a steep slope section with a relatively large gradient. As shown in FIG. 1, the photographing device 5 is positioned a predetermined distance away from the track 1 so that the detection section S of the track 1 is within the photographing range (photographing area) and does not interfere with the travel of the vehicle 2. The photographing device 5 is a fixed camera that can perform fixed-point observation, long-term photographing, and long-term recording using the same camera from the same position in the same direction. The photographing device 5 is, for example, a time-lapse camera that operates on a battery, photographs for a predetermined time at set time intervals, automatically edits the photographed images (still images) into a video (frame-by-frame video), stores the video as image data in an internal storage device, and transmits the image data at set time intervals. The photographing device 5 transmits the photographed images within the detection section S to the slip prediction device 9 as image data.

[0027] The rail temperature monitoring device 6 is a device for monitoring the temperature of the rail 1a. The rail temperature monitoring device 6 monitors the rail temperature T P The rail temperature monitoring device 6 is a temperature sensor such as a thermocouple that measures the rail temperature T at a detection point P within the detection section S. When the rail temperature monitoring device 6 is placed on the rail 1a or near the rail 1a, it is attached to the top side, bottom upper surface or bottom side of the rail 1a at a location that does not come into contact with the wheels 3. When the rail temperature monitoring device 6 is placed near the rail 1a, a member cut out from the rail 1a or a member made of the same material as the rail 1a is placed at a distance from the rail 1a. The rail temperature monitoring device 6 measures the rail temperature T at a detection point P within the detection section S. P and transmits it to the slip prediction device 9 as rail temperature data.

[0028] The weather observation device 7 is a device that observes various weather data. The weather observation device 7 automatically observes weather data such as precipitation, wind direction, wind speed, temperature, humidity, sunshine hours, and snow depth in order to closely monitor weather conditions such as temperature, humidity, and wind speed over time and in a particular area. The weather observation device 7 is, for example, the Japan Meteorological Agency's regional weather observation system, which observes weather data at meteorological stations, weather observation centers, and weather observatories throughout Japan. The weather observation device 7 transmits the actual measured values ​​of weather conditions around the detection point P within the detection section S to the slip prediction device 9 as weather data around the detection point P.

[0029] The communication network 8 is a network for transmitting and receiving various data related to the slip prediction system 4. The communication network 8 transmits image data from the imaging device 5 to the slip prediction device 9, transmits temperature data from the rail temperature monitoring device 6 to the slip prediction device 9, and transmits weather data from the weather observation device 7 to the slip prediction device 9. The communication network 8 is a telecommunications line such as a telephone line or an internet line that connects the imaging device 5, rail temperature monitoring device 6, weather observation device 7, and slip prediction device 9 so that they can communicate with each other.

[0030] The slippage prediction device 9 is a device that predicts slippage of the wheels 3 due to fallen leaves L. The slippage prediction device 9 analyzes the image data transmitted by the photographing device 5 and comprehensively evaluates the quantitative results of the fallen leaves L on the track 1, the detection results of the leaf film F on the top surface 1b of the rail 1a, and the prediction results of the occurrence of condensation on the rail 1a, to predict slippage of the wheels 3 due to fallen leaves L within the detection section S. As shown in FIG. 2 , the slippage prediction device 9 includes a data input unit 10, a color information extraction unit 11, a leaf cover rate calculation unit 12, a leaf cover detection unit 13, a pixel value calculation unit 14, a leaf film rate calculation unit 15, a leaf film detection unit 16, a wetness detection unit 17, a slippage prediction unit 18, a data storage unit 19, a slippage prediction program storage unit 20, a display unit 21, and a control unit 22.

[0031] 2 is a means for inputting various data. The data input unit 10 inputs image data output by the imaging device 5, temperature data output by the rail temperature monitoring device 6, and weather data output by the weather observation device 7, and stores these data in a data storage unit 19. The data input unit 10 includes, for example, an interface (I / F) circuit that receives various data from the communication network 8 and inputs the data to the control unit 22.

[0032] The color information extraction unit 11 is a means for extracting color information of the fallen leaves L on the track 1 based on the captured image of the track 1. The color information extraction unit 11 extracts color information of the fallen leaves L on the track 1 by performing image analysis on the image data captured by the photographing device 5. The color information extraction unit 11 distinguishes the fallen leaves L in the detection section S by color based on the image data captured by the photographing device 5. The color information extraction unit 11 sets a threshold value according to the color of the fallen leaves L for each type of fallen leaves L so that the colors of the fallen leaves L shown in FIG. 6 can be extracted from the image data shown in FIG. 4. Here, the threshold value is set using an RGB color code that numerically represents colors using a combination of red, green, and blue. The threshold value is set, for example, within the range of (R, G, B) = (25 to 255, 125 to 255, 0) so that the fallen leaves L can be extracted. 4, when the detection section S is filled with mainly fallen cedar leaves, the color information extraction unit 11 sets the reference threshold values ​​to (R, G, B) = (155, 255, 0) and extracts green color information corresponding to fallen cedar leaves. The color information extraction unit 11 stores the color information of fallen leaves L extracted from the image data in the data storage unit 19 as color data.

[0033] The leaf coverage calculation unit 12 shown in Figure 2 is a means for calculating the leaf coverage rate R1 based on the extraction results of the color information extraction unit 11. The leaf coverage calculation unit 12 calculates the proportion of fallen leaves L within the detection section S based on the color data extracted by the color information extraction unit 11. The leaf coverage calculation unit 12 calculates the leaf coverage rate R1 (%), which is the proportion of the fallen leaf detection area to the entire area of ​​the detection section S, using the following equation 1.

[0034]

number

[0035] Here, the entire area of ​​the detection section S shown in Equation 1 is, for example, the entire area of ​​the sleeper section corresponding to the detection section S shown in Figure 1 when the track 1 is a bedded track. The leaf detection area is the area occupied by color information of fallen leaves L within the detection section S. The leaf coverage calculation unit 12 stores the leaf coverage R1 within the detection section S in the data storage unit 19 as leaf coverage data.

[0036] The graph shown in Figure 7 shows changes in the leaf coverage rate R1 within the detection section of an actual track on a certain line. The vertical axis in Figure 7 is the leaf coverage rate R1 (%) within the detection section, and the horizontal axis is the date. Wheel spins in the section in question and in the surrounding sections are reports of wheel spins reported by drivers of vehicles 2 traveling in or near the detection section S. As shown in Figure 7, wheel spins of wheels 3 occurred within or near the detection section S on days when the value or change in leaf coverage rate R1 was large.

[0037] The leaf cover detection unit 13 shown in FIG. 2 is a means for detecting the coverage state of the track 1 with fallen leaves L. The leaf cover detection unit 13 detects the coverage state of fallen leaves L within the detection section S based on the calculation result of the leaf cover rate calculation unit 12. The leaf cover detection unit 13 detects the coverage state of fallen leaves L accumulated on the track 1 based on the leaf cover rate R1 calculated by the leaf cover rate calculation unit 12. The leaf cover detection unit 13 quantifies, for example, the amount of fallen leaves or the amount of accumulation within the detection section S based on the leaf cover rate R1, and detects the coverage state of fallen leaves L within the detection section S. The leaf cover detection unit 13 stores the coverage state of fallen leaves L within the detection section S in the data storage unit 19 as leaf cover detection data.

[0038] The pixel value calculation unit 14 is a means for calculating the pixel values ​​of the leaf film F. Here, a pixel value is a value that numerically represents the color of a pixel, which is the smallest element that makes up a photographed image. The pixel value calculation unit 14 calculates the pixel values ​​of the leaf film F based on the photographed image of the track 1. The pixel value calculation unit 14 performs predetermined image processing based on the image data photographed by the photographing device 5. As shown in FIG. 8(A), the pixel value calculation unit 14 cuts out a photographed image corresponding to the top surface 1b of the rail 1a from the photographed image of the detection section S.

[0039] The pixel value calculation unit 14 corrects the extracted photographed image using a color reference. The pixel value calculation unit 14 corrects the color of the extracted photographed image of the top surface 1b of the rail 1a using the color target T shown in FIG. 1 so that the photographed image has the same color as the actual color of the subject. Here, the color target T is a reference standard color sample used to correct the color of the photographed image to a standardized color. As shown in FIG. 8(A), the color target T is placed near the top surface 1b of the rail 1a and is photographed by the photographing device 5. The pixel value calculation unit 14 corrects the brightness and color of the extracted photographed image so that, for example, the numerical value of the color of the leaf litter F and the color of the patch of the color target T photographed together with the leaf litter F become similar.

[0040] The pixel value calculation unit 14 averages the pixel values ​​of the corrected captured image along the length of the rail 1a. As shown in FIG. 8(C), the pixel value calculation unit 14 calculates the average pixel value along the rail length at each position in the width direction of the rail 1a (the direction perpendicular to the length direction). Here, the vertical axis shown in FIG. 8(C) represents the average pixel value along the length of the rail 1a, and the horizontal axis represents the position along the rail width. The pixel value calculation unit 14 identifies, as a leaf film region, an area in the width direction of the rail 1a where the average pixel value is equal to or less than a threshold value. Here, the threshold value is set to the pixel value of a typical leaf film F. The pixel value calculation unit 14 stores the pixel values ​​of the leaf film F as pixel value data in the data storage unit 19.

[0041] The leaf film rate calculation unit 15 is a means for calculating the leaf film detection rate R2 based on the calculation results of the pixel value calculation unit 14. The leaf film rate calculation unit 15 calculates the proportion of the leaf film F in the detection section S based on the pixel values ​​of the leaf film F calculated by the pixel value calculation unit 14. The leaf film rate calculation unit 15 calculates the leaf film detection rate (area rate) R2 (%), which is the proportion of the area occupied by the pixel values ​​of the leaf film F to the entire area of ​​the detection section S, using the following equation 2.

[0042]

number

[0043] Here, the leaf film region shown in Equation 2 is a region where the average pixel value in the longitudinal direction of the rail 1a is equal to or less than a threshold value. The leaf film rate calculation unit 15 stores the leaf film detection rate R2 calculated in the detection section S in the data storage unit 19 as leaf film rate data.

[0044] The graph shown in Figure 9 shows the change in the leaf litter detection rate R2 within the detection section of an actual track on a certain line. The vertical axis in Figure 9 is the leaf litter detection rate R2 (%) within the detection section, and the horizontal axis is the date. Wheel spins in the detection section S and surrounding sections are reports of wheel spins reported by drivers of vehicles 2 traveling in or near the detection section S. As shown in Figure 9, wheel spins of wheels 3 occurred within or near the detection section S on days when the value or change in the leaf litter detection rate R2 was large.

[0045] The leaf film detection unit 16 shown in FIG. 2 is a means for detecting the state of formation of leaf film F on the rails 1a of the track 1. The leaf film detection unit 16 detects the state of formation of leaf film F within the detection section S based on the calculation result of the leaf film rate calculation unit 15. The leaf film detection unit 16 detects the state of formation of leaf film F within the detection section S based on the leaf film detection rate R2 calculated by the leaf film rate calculation unit 15. The leaf film detection unit 16 quantifies, for example, the area of ​​leaf film F formed on the rails 1a within the detection section S based on the leaf film detection rate R2, and detects the state of formation of leaf film F within the detection section S. The leaf cover detection unit 13 stores the state of cover by fallen leaves L within the detection section S in the data storage unit 19 as leaf cover detection data.

[0046] The wetness detection unit 17 is a means for detecting the wetness of the rail 1a. The wetness detection unit 17 is a means for detecting the wetness of the rail 1a based on meteorological data around the track 1 and the rail temperature T PThe wetness detection unit 17 detects the wetness of the rail 1a based on the weather data around the track 1 received from the weather observation device 7 and the rail temperature data of the rail 1a received from the rail temperature monitoring device 6. The wetness detection unit 17 acquires the saturated water vapor amount, water vapor pressure, ambient temperature, and ambient relative humidity in the vicinity of a detection point P on the track 1 from the weather observation device 7, and also calculates the rail temperature T P is acquired from the rail temperature monitoring device 6. The wet state detection unit 17 calculates the dew point Td of the rail 1a using the following equation 3.

[0047]

number

[0048] Here, the dew point Td is the temperature of the rail 1a when the water vapor in the atmosphere cools and condensation begins to form on the rail 1a. In Equation 3, es is the saturated water vapor amount [hPa], e is the water vapor pressure [hPa], T is the ambient temperature [°C], and U is the ambient relative humidity [%RH]. The wetness detection unit 17 detects the rail temperature T P is equal to or lower than the dew point Td calculated by Equation 3, it is determined that the surface of the rail 1a is wet due to condensation.

[0049] FIG. 10 is a graph showing an example of the results of trial calculations of rail dew points. Here, the X axis in FIG. 10 is ambient temperature T [°C], the Y axis is ambient relative humidity U [5RH], and the Z axis is rail dew point Td [°C]. The wetness condition detection unit 17 calculates the dew point Td of the rail 1a, for example, as shown in FIG. 10. The wetness condition detection unit 17 detects the wetness condition of the rail 1a within the detection section S and stores the wetness condition within the detection section S in the data storage unit 19 as wetness data.

[0050] The graph in Figure 11 shows the changes in leaf coverage rate R1 and leaf litter detection rate R2 within the detection section of an actual track on a certain line. The vertical axis on the left side of Figure 11 is the leaf coverage rate R1 (%) within the detection section, the vertical axis on the right side is the leaf litter detection rate R2 (%), and the horizontal axis is the date. Wheel spins in the detection section S and surrounding sections are reports of wheel spins reported by drivers of vehicle 2 traveling in or near the detection section S. As shown in Figure 11, wheel spins of wheel 3 occurred within or near the detection section S on days when the leaf coverage rate R1 and leaf litter detection rate R2 were high.

[0051] The slippage prediction unit 18 shown in FIG. 2 is a means for predicting slippage of the wheel 3. The slippage prediction unit 18 predicts slippage of the wheel 3 by comprehensively evaluating the detection results of the leaf cover detection unit 13, the leaf film detection unit 16, and the wet condition detection unit 17. For example, the slippage prediction unit 18 weights the detection results of the leaf cover detection unit 13, the leaf film detection unit 16, and the wet condition detection unit 17 to predict the occurrence of slippage of the wheel 3; predicts the occurrence of slippage of the wheel 3 when the leaf cover rate R1 and the leaf film detection rate R2 exceed predetermined values; predicts the occurrence of slippage of the wheel 3 when changes in the leaf cover rate R1 and the leaf film detection rate R2 exceed predetermined values; or predicts the occurrence of slippage of the wheel 3 when the amount of fallen leaves L, the amount of accumulated leaves, or the area of ​​the leaf film F exceeds predetermined values.

[0052] The slippage prediction unit 18 predicts slippage of the wheel 3 based on the detection results of the leaf cover detection unit 13 and the leaf film detection unit 16. The slippage prediction unit 18 also predicts slippage of the wheel 3 based on the detection results of the leaf cover detection unit 13, the detection results of the leaf film detection unit 16, and the detection results of the wet condition detection unit 17. The slippage prediction unit 18 also predicts slippage of the wheel 3 based on the detection results of the leaf film detection unit 16 and the detection results of the wet condition detection unit 17. The slippage prediction unit 18 stores the prediction result of the occurrence of slippage of the wheel 3 in the data storage unit 19 as slippage prediction data.

[0053] The data storage unit 19 is a means for storing various data related to the slip prediction system 4. The data storage unit 19 is a storage device that stores, for example, image data, rail temperature data, and weather data, as well as color data, leaf cover rate data, pixel value data, Takuyo film rate data, leaf cover detection data, leaf film detection data, wetness data, and slip prediction data, etc., in association with each detection section S.

[0054] The slip prediction program storage unit 20 is a means for storing a slip prediction program for predicting slippage of the wheels 3 caused by fallen leaves L. The slip prediction program storage unit 20 is a storage device or the like that stores a waveform prediction program read from an information recording medium or a waveform prediction program downloaded via an electric communication line.

[0055] The display unit 21 is a means for displaying various data related to the slip prediction device 9. The display unit 21 is a display device that displays, for example, the time change of the leaf cover rate R1 as shown in Fig. 7, the time change of the leaf litter detection rate R2 as shown in Fig. 9, and the time change of the leaf cover rate R1 and the leaf litter detection rate R2 as shown in Fig. 11 on a screen for each detection section S.

[0056] The control unit 22 is a central processing unit (CPU) that controls various operations related to the slip prediction device 9. The control unit 22 reads out a slip prediction program from the slip prediction program storage unit 20 and executes slip prediction processing in accordance with this slip prediction program. The control unit 22, for example, instructs the data storage unit 19 to store image data, rail temperature data, and weather data input from the data input unit 10, outputs image data read out from the data storage unit 19 to the color information extraction unit 11, instructs the color information extraction unit 11 to extract color information of fallen leaves L, outputs color data read out from the data storage unit 19 to the leaf coverage calculation unit 12, instructs the leaf coverage calculation unit 12 to calculate the leaf coverage R1, outputs leaf coverage data read out from the data storage unit 19 to the leaf coverage detection unit 13, instructs the leaf coverage detection unit 13 to detect the coverage state of fallen leaves L, and outputs image data read out from the data storage unit 19 to the leaf coverage detection unit 13. The image data is output to the pixel value calculation unit 14, the leaf film rate calculation unit 15 is instructed to calculate the leaf film detection rate R2, the leaf film rate data read from the data storage unit 19 is output to the leaf film detection unit 16, the leaf film detection unit 16 is instructed to detect the formation state of the leaf film F, the weather data read from the data storage unit 19 is output to the wetness condition detection unit 17, the wetness condition detection unit 17 is instructed to detect the wetness state of the rail 1a, the leaf coverage rate data, the leaf film rate data and the wetness data are output to the slippage prediction unit 18, the slippage prediction unit 18 is instructed to predict the slippage of the wheels 3, and the display unit 21 is instructed to display various data. The control unit 22 is connected to a data input unit 10, a color information extraction unit 11, a leaf cover rate calculation unit 12, a leaf cover detection unit 13, a pixel value calculation unit 14, a leaf film rate calculation unit 15, a leaf film detection unit 16, a wetness detection unit 17, a slippage prediction unit 18, a data memory unit 19, a slippage prediction program memory unit 20, and a display unit 21 so that they can communicate with each other.

[0057] Next, the operation of the wheel slip prediction device according to the embodiment of the present invention will be described. The following description will focus on the operation of the control unit 22 shown in FIG. In step (hereinafter referred to as S) 100, the control unit 22 reads the slip prediction program from the slip prediction program storage unit 20. When the slip prediction device 9 receives the photographic data photographed by the photographing device 5 at a predetermined time and at a predetermined interval, the control unit 22 reads the slip prediction program and starts a series of slip prediction processes.

[0058] In S110, the control unit 22 commands the color information extraction unit 11 to extract color information. As a result, the color information extraction unit 11 extracts color information such as that shown in Fig. 6 from the captured image such as that shown in Fig. 4 captured by the imaging device 5 within the detection section S shown in Fig. 1.

[0059] In S120, the control unit 22 commands the leaf coverage calculation unit 12 to calculate the leaf coverage R1. As a result, the color information corresponding to the fallen leaves L is used as a threshold value, and the detection area of ​​the fallen leaves L is identified from the color information extracted by the color information extraction unit 11, and the leaf coverage calculation unit 12 calculates the leaf coverage R1 using Equation 1.

[0060] In S130, the control unit 22 commands the leaf cover detection unit 13 to detect the coverage state of the fallen leaves L. As a result, based on the leaf cover rate R1, the leaf cover detection unit 13 quantifies the amount of fallen leaves or the amount of accumulated leaves within the detection section S, and detects the coverage state of the fallen leaves L within the detection section S.

[0061] In S140, the control unit 22 commands the pixel value calculation unit 14 to calculate the pixel values ​​of the leaf litter F. The pixel value calculation unit 14 cuts out an image corresponding to the rail 1a from a photographed image such as that shown in Fig. 8(A) taken by the photographing device 5 within the detection section S shown in Fig. 1, and corrects the cut-out image using a color reference. As a result, the cut-out photographed image is corrected so that the color of the leaf litter F in the photographed image becomes closer to the actual color of the leaf litter F.

[0062] In S150, the control unit 22 instructs the leaf film rate calculation unit 15 to calculate the leaf film detection rate R2. As a result, the leaf film rate calculation unit 15 identifies the detection area of ​​the leaf film F on the rail 1a, and calculates the leaf film detection rate R2 using Equation 2.

[0063] In S160, the control unit 22 commands the leaf film detection unit 16 to detect the formation state of the leaf film F. As a result, based on the leaf film detection rate R2, the leaf film detection unit 16 quantifies the formation area of ​​the leaf film F within the detection section S, and detects the formation state of the leaf film F within the detection section S.

[0064] In S170, the control unit 22 commands the wet condition detection unit 17 to detect a wet condition. Based on the temperature of the rail 1a measured by the rail temperature monitoring device 6 at the detection point P in the detection section S shown in Fig. 1 and the weather data measured by the weather observation device 7 around the detection section S shown in Fig. 1, the wet condition detection unit 17 calculates the dew point Td using Equation 3. The rail temperature T at the detection point P P When the wet condition detection unit 17 determines that the temperature is equal to or lower than the dew point Td, the wet condition detection unit 17 determines that the surface of the rail 1a is wet due to condensation.

[0065] In S180, the control unit 22 commands the slippage prediction unit 18 to predict slippage of the wheel 3. As a result, the slippage prediction unit 18 comprehensively evaluates the detection results of the leaf cover detection unit 13, the leaf film detection unit 16, and the wet condition detection unit 17, and predicts slippage of the wheel 3. For example, when the leaf cover rate R1 exceeds a predetermined value and the leaf film detection rate R2 exceeds a predetermined value, the slippage prediction unit 18 predicts that there is a sign that the wheel 3 is about to slip.

[0066] In S190, the control unit 22 commands the display unit 21 to display various data. As a result, for example, the display unit 21 displays the detection section S where the occurrence of wheel slippage of the wheels 3 is predicted.

[0067] The wheel slip prediction device according to this embodiment has the following advantages. (1) In this embodiment, the leaf covering detection unit 13 detects the state of coverage of the track 1 with fallen leaves L, the leaf covering detection unit 16 detects the state of formation of leaf film F on the rails 1a of the track 1, and the slippage prediction unit 18 predicts wheel slippage of the wheels 3 based on the detection results of the leaf covering detection unit 13 and the leaf film detection unit 16. Also in this embodiment, the leaf film detection unit 13 detects the state of formation of leaf film F on the rails 1a of the track 1, the wetness detection unit 17 detects the wetness of the rails 1a, and the slippage prediction unit 18 predicts wheel slippage of the wheels 3 based on the detection results of the leaf film detection unit 13 and the wetness detection unit 17. Therefore, signs of wheel skidding or sliding of the wheels 3 can be predicted from the leaves L and condensation on the rails 1a. As a result, the risk of wheel slippage of the wheels 3 can be predicted, and wheel slippage caused by fallen leaves can be suppressed. Furthermore, it is possible to establish the timing for railway operators to take measures to deal with unexpected wheel slippage of the wheels 3, which can contribute to the determination of the timing of such measures. For example, for sections where wheel slippage of the wheels 3 due to fallen leaves is predicted, it is possible to determine the timing for taking measures to prevent wheel slippage, such as selectively scattering sand from the vehicle 2 onto the top surface 1b of the rail 1a, cleaning the top surface 1b, or spraying citric acid onto the top surface 1b.

[0068] (2) In this embodiment, the color information extraction unit 11 extracts color information of the leaves L on the track 1 based on the captured image of the track 1, and the leaf coverage calculation unit 12 calculates the leaf coverage rate R1, which is the ratio of the area occupied by the color information of the leaves L to a predetermined area of ​​the track 1, based on the extraction result of the color information extraction unit 11, and the leaf coverage detection unit 13 detects the coverage state of the leaves L based on the leaf coverage rate R1. Therefore, by quantifying the leaves L accumulated on the track 1 within the detection section S, the coverage state of the leaves L within the detection section S can be easily detected with high accuracy.

[0069] (3) In this embodiment, the pixel value calculation unit 14 calculates the pixel values ​​of the leaf film F based on the photographed image of the track 1, and the leaf film detection rate calculation unit 15 calculates the leaf film detection rate R2, which is the ratio of the area occupied by the pixel values ​​of the leaf film F to a predetermined area on the rail 1a, based on the calculation result of the pixel value calculation unit 14. The leaf film detection unit 16 detects the formation state of the leaf film F based on the leaf film detection rate R2. Therefore, by quantifying the formation area of ​​the leaf film F formed on the rail 1a within the detection section S, the formation state of the leaf film F within the detection section S can be detected easily and with high accuracy.

[0070] (4) In this embodiment, the wetness detection unit 17 detects the wetness of the rail 1a, and the slippage prediction unit 18 predicts slippage of the wheel 3 based on the detection result of the wetness detection unit 17. Therefore, slippage of the wheel 3 can be predicted in advance in cases where a leaf litter film F has formed on the rail 1a or where fallen leaves L cover the track 1 and condensation has formed on the rail 1a.

[0071] (5) In this embodiment, the weather data around the track 1 and the rail temperature T P Therefore, for example, when the temperature and humidity around the detection section S are relatively high and the rail temperature T P When the temperature is relatively low, it is possible to detect signs of condensation on the rail 1a and predict wheel slippage of the wheels 3.

[0072] The present invention is not limited to the above-described embodiment, and various modifications and alterations are possible as described below, and these are also within the scope of the present invention. (1) In this embodiment, the prediction of wheel slippage of a railway vehicle wheel 3 has been described as an example, but the present invention can also be applied to the prediction of wheel slippage of an automobile or the like. Furthermore, in this embodiment, spinning or sliding has been described as an example of wheel slippage of the wheel 3 caused by fallen leaves, but the present invention can also be applied to sticking, which is a state in which the wheel 3 does not rotate due to a brake not being released, or a state in which the wheel 3 starts to slide due to sudden braking while traveling, and the wheel 3 stops rotating completely, causing it to slide on the rail 1a. Furthermore, in this embodiment, the leaf coating F has been described as a black coating as an example, but the present invention can also be applied to leaf coating F other than a black coating.

[0073] (2) In this embodiment, the rail temperature monitoring device 6 is a contact-type temperature sensor. However, the rail temperature T P The present invention can also be applied to a case where the weather observation device 7 is a non-contact temperature sensor that measures temperature. In addition, in this embodiment, an example has been described in which the weather observation device 7 is a regional weather observation system of the Japan Meteorological Agency, but the present invention can also be applied to a case in which the weather observation device 7 is installed near the detection section S to acquire weather data for the detection section S. [Explanation of symbols]

[0074] 1 orbit 1a Rail 1b Parietal surface 2 vehicles 3 wheels 4. Slip prediction system 5. Imaging equipment 6. Rail temperature monitoring device 7. Weather Observation Equipment 8. Communication Networks 9. Slip prediction device 11 Color information extraction section 12 Leaf cover calculation unit 13 Leaf cover detection unit 14 Pixel value calculation unit 15 Leaf litter coating rate calculation unit 16 Leaf litter film detection unit 17 Wetness detection unit 18 Slip Prediction Section 20 Control Unit L Fallen leaves F. Leaf litter membrane (black membrane) P detection point T Color target section R1 leaf litter coverage R2 leaf litter film detection rate T P Rail Temperature Td dew point

Claims

1. A wheel slippage prediction device that predicts wheel slippage caused by fallen leaves, a leaf cover detection unit that detects a state of coverage of the track by fallen leaves; a leaf litter detection unit for detecting a state of leaf litter formation on the rails of the track; a slip prediction unit that predicts slip of the wheel based on the detection results of the leaf cover detection unit and the leaf film detection unit; A wheel slip prediction device comprising:

2. 2. The wheel slip prediction device according to claim 1, a color information extraction unit that extracts color information of fallen leaves on the track based on the captured image of the track; a leaf coverage calculation unit that calculates a leaf coverage rate, which is the rate of an area occupied by the color information of the fallen leaves to a predetermined area of ​​the track, based on the extraction result of the color information extraction unit, the fallen leaf coverage detection unit detects the leaf coverage state based on the fallen leaf coverage rate; A wheel slip prediction device characterized by the above.

3. 2. The wheel slip prediction device according to claim 1, a pixel value calculation unit that calculates pixel values ​​of the leaf litter film based on the captured image of the trajectory; a leaf film detection rate calculation unit that calculates a leaf film detection rate, which is the rate of an area occupied by pixel values ​​of the leaf film to a predetermined area on the rail, based on the calculation result of the pixel value calculation unit; the leaf litter film detection unit detects the state of formation of the leaf litter film based on the leaf litter film detection rate; A wheel slip prediction device characterized by the above.

4. 2. The wheel slip prediction device according to claim 1, a wetness detection unit for detecting the wetness of the rail; the slippage prediction unit predicts slippage of the wheel based on a detection result of the wet condition detection unit; A wheel slip prediction device characterized by the above.

5. 5. The wheel slip prediction device according to claim 4, the wetness detection unit detects the wetness of the rail based on meteorological data around the track and a temperature of the rail of the track; A wheel slip prediction device characterized by the above.

6. A wheel slippage prediction device that predicts wheel slippage caused by fallen leaves, a leaf film detection unit that detects the state of leaf film formation on the rail of the track; a wetness detection unit that detects the wetness of the rail; a slippage prediction unit that predicts slippage of the wheel based on the detection results of the leaf film detection unit and the wet state detection unit; A wheel slip prediction device comprising:

7. 7. The wheel slip prediction device according to claim 6, a pixel value calculation unit that calculates pixel values ​​of the leaf litter film based on the captured image of the trajectory; a leaf film detection rate calculation unit that calculates a leaf film detection rate, which is the rate of an area occupied by pixel values ​​of the leaf film to a predetermined area on the rail, based on the calculation result of the pixel value calculation unit; the leaf litter film detection unit detects the state of formation of the leaf litter film based on the leaf litter film detection rate; A wheel slip prediction device characterized by the above.

8. 8. The wheel slip prediction device according to claim 7, the wetness detection unit detects the wetness of the rail based on meteorological data around the track and a temperature of the rail of the track; A wheel slip prediction device characterized by the above.

Images