CN104477209A - Railway locomotive vehicle wheel online scratch automatic detection system and method - Google Patents

Abstract

The invention discloses an automatic detection system for on-line abrasion of railway locomotive and vehicle wheels, which includes an incoming line induction module, an offline induction module, n swing rod detection modules, a database storage module, a host computer communication module and an abrasion data analysis module; The rod detection module includes a left swing rod detection module, a right swing rod detection module and a scratch data acquisition module; the incoming line induction module is set at the entry end of the working rail, the offline sensing module is set at the leaving end of the and the right swing rod detection module are symmetrically arranged on the inner side of the two working rails, and the n swing rod detection modules are arranged on the inner side of the working rail with intervals and the total length is greater than the circumference of the wheel, and n is L ₁ Add 1 to the value of /L ₂ to get an integer, L ₁ is the circumference of the wheel, and L ₂ is the length of a single pendulum detection module. The invention also discloses a method based on the system. The invention can accurately judge information such as the depth, length, position and the like of the scratch, thereby improving the detection accuracy.

Description

A system and method for on-line automatic detection of wheel abrasions of railway locomotive vehicles

technical field

The invention relates to the technical field of wheel state detection of railway locomotive vehicles, in particular to a system and method for on-line automatic detection of wheel abrasions of railway locomotive vehicles.

Background technique

As one of the important parts of the running part of railway rolling stock, the wheel is running normally or not directly affects the safe driving. In recent years, my country's railway vehicles have experienced several major speed increases, from ordinary trains, express trains, EMUs to high-speed rail, showing an unprecedented good development trend. But the consequence is that the force between the wheel and the rail of the railway rolling stock is constantly increasing, and the dynamic load on the wheel is also increased. At the same time, the steady improvement of railway transportation speed and capacity means higher requirements for high-speed passenger transportation, especially the safety of railway rolling stock wheels is particularly important. Under numerous factors, the potential safety hazard caused by the scuffing of the tread roller circle has become an important source of failures and accidents of railway rolling stock. The safety of wheels has gradually become the key to ensure the safe operation of vehicles. Any type of abnormal defects of wheels will restrict the development of my country's rail transit industry.

At present, there are mainly two types of scratch detection methods for railway locomotive wheels: static detection and dynamic detection. Static detection is completed when the vehicle is stationary. The time is long and the process is relatively complicated; dynamic detection is carried out when the vehicle is running normally, so its detection efficiency is much higher than that of static detection, and its detection automation is high, it does not occupy the turnaround time of the vehicle, and it is convenient for storage The wheel information data has a high detection rate and has received attention in recent years. The most commonly used and relatively mature in dynamic detection is the scratch detection system constructed by vibration acceleration method and parallelogram method. Figure 1 is a schematic diagram of a scratch detection system using the vibration acceleration method. The acceleration measurement sensor for measuring the abnormal vibration caused by the scratch and the wheel counting sensor for detecting the passing of the wheel are installed on the rail. Sensors such as large and small, the data output by these sensors are transmitted to the data collection module next to the line, and then communicate with the data processing module to analyze and process the collected data. The vibration acceleration method has a high misjudgment rate, and it is easy to miss detection at low speeds, resulting in low detection accuracy. Figure 2a is a schematic diagram of the scratch detection system using the parallelogram method, and Figure 2b is a schematic diagram of the detection parts of the scratch detection system using the parallelogram method. The connecting rod mechanism is fixed on the steel rail in a parallelogram layout, and the displacement sensor is fixed on the steel rail on one side of the four-bar connecting mechanism through the support. The device can only run in one direction, and it will cause destructive impact when the train runs in reverse, which is inconvenient for maintenance.

In the invention patent with application publication number CN 102785679A, a wheel tread scratch and out-of-roundness online detection device is disclosed. Figure 3 is a cross-sectional structural schematic diagram of a wheel tread scratch and out-of-roundness online detection device. It mainly includes: a working rail 1 and a rail clamping block 2. A detection device 3 is arranged on the upper side of the rail clamping block 2. The detection device 3 includes a casing 4, and a rocker bracket 5 and a rocker swing are arranged in the cavity of the casing 4. Rod 6, the rocker arm swing bar 6 is embedded between the upper side of the rocker arm bracket 5 and the upper casing of the shell 4, the front section of the rocker arm swing bar 6 extends out of the shell 4 and extends to the working rail 1 and the rocker arm swing bar The front end of the 6 is provided with a measuring rod 7, and the rear end of the rocking arm swing rod 6 is provided with a driving gear 8 fixedly connected with it and arranged coaxially. The movable gear 9 is provided with a rotary encoder coaxially arranged with it. Other parts also have: back-moving spring 10, splint 11 and lifting device 12. In this patent application, the wheel tread is in contact with the measuring rod, so it cannot be guaranteed that the wheel rim can be pressed against the measuring rod; in this patent application, the swing rod detection module is placed outside the working track, so it needs to be installed inside the working track to ensure that the wheel does not move Serpentine guard rail device, increasing such guard rail device has increased cost.

How to solve the deficiencies of the prior art is a major problem to be solved urgently in the field of rolling stock wheel state detection technology.

Contents of the invention

The technical problem to be solved by the present invention is to overcome the deficiencies of the prior art and provide an automatic detection system and method for the on-line abrasion of railway locomotive wheels. The cost of the system, in the present invention, the wheel rim contacts and the horizontally placed swing bar ensures that the wheel rim can always be pressed on the swing bar, and can accurately judge the current wheel position, speed and other information; the detection method in the present invention improves the detection precision and easy to use.

The present invention adopts the following technical solutions for solving the problems of the technologies described above:

According to the present invention, an on-line automatic detection system for railway locomotive and vehicle wheels includes an incoming line sensing module, an offline sensing module, n swing rod detection modules, a database storage module, a host computer communication module and a scratch data analysis module; The swing rod detection module includes a left swing rod detection module, a right swing rod detection module and a scratch data acquisition module; the left swing rod detection module and the right swing rod detection module are symmetrically arranged on the inner side of two working rails respectively, The n pendulum detection modules are arranged at intervals on the inner side of the working rail and the total length is greater than the circumference of the wheel, where n is the value of L ₁ /L ₂ plus 1 and rounded up to an integer, L ₁ is the circumference of the wheel, L ₂ is the length of a single pendulum detection module; where:

The incoming line sensing module is arranged at the entry end of the working rail, and is used to output the rolling stock incoming line sign signal to the database storage module when the railway rolling stock enters the detection area of the incoming line sensing module at a preset speed;

The offline sensing module is arranged at the departure end of the working rail, and is used to output the offline sign signal of the rolling stock to the database storage module when the railway rolling stock leaves the detection area of the offline sensing module at a preset speed;

The left swing rod detection module and the right swing rod detection module are used to collect the angle change data of the wheel flange when the wheel flange is pressed down, and output it to the scratch data acquisition module;

The scratch data acquisition module is used to receive the angle change data of wheel rim depression output by the left swing rod detection module and the right swing rod detection module, and output it to the upper computer communication module;

The database storage module is used to receive the incoming sign signal of the rolling stock and the off-line sign signal of the rolling stock;

The upper computer communication module is used to read the rolling stock incoming line sign signal and the rolling stock offline sign signal in the database storage module in real time; when the upper computer communication module reads a complete set of rolling stock incoming line sign signals from the database storage module After the signal and the off-line sign signal of the rolling stock, read the angle change data of the wheel rim pressing down from the scratch data acquisition module, and output it to the scratch data analysis module;

The scratch data analysis module is used to analyze and process the angle change data of wheel rim pressing down to obtain wheel scratch information.

As a further optimized scheme of the online automatic detection system for railway locomotive and vehicle wheels of the present invention, the left swing rod detection module and the right swing rod detection module both include swing rods and gear mechanisms that are in contact with the wheel rim and placed horizontally 1. A rotary encoder coaxial with the gear mechanism arranged on the swing rod; when the swing rod is pressed down by the rim of the wheel, the swing rod produces an angle change, and the gear mechanism is used to amplify the angle change of the swing rod before passing through The rotary encoder processes and outputs angle change data to the scratch data acquisition module.

As a further optimization scheme of the online automatic detection system for railway locomotive and vehicle wheels of the present invention, the swing rod detection module further includes two wheel positioning sensing modules arranged at both ends of the left swing rod detection module or the right swing rod detection module ; When the railway locomotive passes through the detection area of the wheel positioning sensing module, the trigger signal is output to the scratch data acquisition module, and the scratch data acquisition module obtains the time according to the received trigger signal, and calculates according to the pre-stored position parameters The speed of the wheel, when the speed is within the preset vehicle speed, the obtained wheel abrasion information is valid.

As a solution for further optimization of the on-line automatic detection system for railway locomotive wheel scratches of the present invention, the incoming line sensing module and the offline sensing module are both photoelectric switch sensors.

As a further optimized scheme of the online automatic detection system for rolling stock wheels of the present invention, the wheel positioning sensing module is an eddy-current proximity switch or a capacitive proximity switch or a Hall-type proximity switch or a photoelectric proximity switch or Ultrasonic proximity switches or microwave proximity switches.

As a solution for further optimization of the on-line automatic detection system for railway rolling stock wheels of the present invention, the scratch data acquisition module includes a controller unit, an encoder data receiving unit, an encoder data decoding unit, a storage unit, and a communication unit and the wheel alignment data receiving unit; the angle change data of wheel rim pressing is received by the encoder data receiving unit, and then sent to the controller unit after being decoded by the encoder data decoding unit; the wheel alignment data receiving unit receives the wheel alignment sensor The trigger signal output by the module is sent to the controller unit; the controller unit outputs the received angle change data after decoding and processing to the storage unit, and finally outputs it to the upper computer communication module through the communication unit.

As a further optimization scheme of the online automatic detection system for railway rolling stock wheels of the present invention, the preset vehicle speed is uniform and ranges from 1 km/h to 20 km/h.

The detection method of a kind of railway locomotive vehicle wheel online abrasion automatic detection system based on the present invention comprises the following steps:

Step 1, pre-setting the range of the head area and the range of the tail area on the curve;

Step 2. Convert the angle change data of the wheel rim depression into the depth displacement data of the wheel rim depression, take the sampling point as the abscissa, and the depth displacement value as the ordinate, to obtain the original curve; for the head area of the original curve, The data curve in the tail area is fitted to obtain a new curve, and the depth displacement values corresponding to the same sampling points on the new curve and the original curve are respectively subtracted to obtain the first difference, and then the first difference is packaged. Line and smooth interpolation processing to form a scratch schematic curve; when there is a sample point on the scratch schematic curve whose ordinate meets the preset scratch depth threshold range, then the wheel is scratched at this time, and the scratch position is obtained , depth and length information;

Step 3: Analyze the middle area curve after removing the head area and tail area in the original curve, first find the highest point corresponding to the maximum depth displacement value in the middle area, start from the position of the first sampling point in the middle area, and set it as the starting point Sampling point;

A. Find the point with the same depth displacement value as the initial sampling point according to the sampling timing of the sampling point in turn, and subtract the abscissa of the point from the abscissa of the initial point to obtain the second difference. If the second difference meets the preset The threshold range of scratch length is set, then subtract the ordinate of the highest point from the ordinate of the starting point to obtain the third difference, when the third difference also meets the preset threshold range of scratch depth, then the wheel After being scratched, the position, depth and length of the scratch are obtained;

B. When the second difference does not meet the preset scratch length threshold range or the third difference does not meet the preset scratch depth threshold range, relocate to the sampling point at the next moment and set it as a new starting point Starting point, repeat step A.

Step 4: According to the position, depth and length information of the abrasion obtained in the above steps, output and alarm the abrasion situation of the railway locomotive vehicle wheel.

As a detection method of the online automatic detection system for railway locomotive and vehicle wheels of the present invention, the range of the head area on the curve is preset in the step 1. The sampling point is used as the sampling time sequence, and the first 200 sampling points on the curve are selected. ;The scope of the tail area is based on the sampling point as the sampling timing, and the last 100 sampling points on the curve are selected.

As a detection method of the online automatic detection system for railway locomotive wheel scratches of the present invention, the method of pre-setting the range of the head area and the range of the tail area on the curve is: the wheel rim is in contact with the two ends of the swing rod detection module When pressing and leaving the corresponding pendulum detection module, a jumping curve will be generated. The range of the jumping curve will be formed by comprehensive analysis of the actual passing data, and the range of the head area and the tail area will be obtained.

Compared with the prior art, the present invention adopts the above technical scheme and has the following technical effects:

(1) The fork detection module of the present invention includes a fork that is in contact with the wheel rim and placed horizontally, which ensures that the wheel rim can always be pressed on the fork, and can accurately collect the angle data of the wheel rim pressing down; The scratch automatic detection system can also accurately judge the current wheel position, speed and other information;

(2) The pendulum detection module of the present invention is arranged on the inner side of the detection work track, so improved, so that the guard rail device originally to be installed in the work track to ensure that the wheels do not go serpentine is no longer needed, saving The cost of the whole system can also guarantee the detection accuracy, and it is easy to use.

(3) By adopting the system analysis method of the present invention, the scratch data can be read and analyzed accurately, and information such as the depth, length, and position of the scratch can be judged, thereby improving the detection accuracy.

Description of drawings

FIG. 1 is a schematic diagram of a scratch detection system using a vibration acceleration method in the background art.

Fig. 2a is a schematic diagram of a scratch detection system using a parallelogram method in the background art.

Fig. 2b is a schematic diagram of detection parts of the scratch detection system using the parallelogram method in the background art.

Fig. 3 is a schematic cross-sectional structure diagram of an on-line detection device for wheel tread scratches and out-of-roundness.

Figure 4 is a schematic diagram of the composition of the on-line scuffing system for railway rolling stock wheels.

Fig. 5 is a trackside installation structure diagram of the on-line scratch automatic detection system for railway rolling stock wheels.

Figure 6 The measurement principle diagram of the pendulum detection module.

Fig. 7 is a structural diagram of the data acquisition module of the on-line scratch automatic detection system for railway rolling stock wheels.

Fig. 8 is a flow chart of scratch analysis of the on-line automatic detection system for wheel scratches of railway rolling stock.

Fig. 9 is a flow chart of the analysis method of the abrasion analysis module of the online abrasion automatic detection system for railway rolling stock wheels.

Fig. 10 is the local triangular wave scratch data of an example of the simulation data in the middle area processed by the scratch data analysis module.

Fig. 11 is the local rectangular wave scratch data of an example of the simulation data in the middle area processed by the scratch data analysis module.

Fig. 12 is the local sine wave scratch data of an example of the simulated data in the middle area processed by the scratch data analysis module.

Fig. 13 is an example a of non-scratched data in the middle area of the actual on-site vehicle passing by the scratch data analysis module.

Figure 14 is an example b of non-scratched data in the middle area of the actual on-site vehicle passing by the scratch data analysis module.

Fig. 15 is an example c of non-scratched data in the middle area of the actual on-site vehicle passing by the scratch data analysis module.

Figure 16 is an example of scratch data in the middle area collected by the scratch data analysis module processing a set of actual passing vehicles.

Figure 17 is an example of the scratch data analysis module processing another set of data with scratches in the head area and tail area collected from actual passing vehicles.

The reference signs in the figure are interpreted as: 1-working rail, 2-track clamping block, 3-detection device, 4-housing, 5-rocker arm bracket, 6-rocker arm swing bar, 7-measuring rod, 8- Driving gear, 9-driven gear, 10-return spring, 11-splint, 12-lifting device.

Detailed ways

Below in conjunction with accompanying drawing, technical scheme of the present invention is described in further detail:

Figure 4 is a schematic diagram of the composition of the online scuffing system for railway rolling stock wheels. An automatic detection system for online abrasion of railway locomotive and vehicle wheels, including an incoming line induction module, an offline induction module, a swing rod detection module, a database storage module, a host computer communication module, an abrasion data analysis module, an abrasion data acquisition module, and a wheel Position sensing module. The general direction of the main data signals of the whole system is: the incoming line sensing module and the offline sensing module generate the incoming line and offline flag signals and store them in the database storage module, and the host computer communication module recognizes this group of flag signals in the database storage module; when the locomotive passes through When detecting the area, the pendulum detection module collects on-site measured data and sends it to the scratch data acquisition module, which then sends it to the upper computer communication module, and finally calls the scratch data analysis module to analyze and judge the collected data , if there is scratches, different levels of scratches will be alarmed and a report record file will be formed. The system can perform online automatic scratch detection on locomotive wheels, classify, analyze and process the collected data, and can accurately and quickly judge whether there is scratches on the wheel tread of the wheel.

As shown in Fig. 5, it is a trackside installation structure diagram of the on-line scratch automatic detection system for railway rolling stock wheels. An automatic detection system for on-line scratches of railway locomotive wheels, comprising an incoming line induction module, an offline induction module, n swing rod detection modules, a database storage module, a host computer communication module and a scratch data analysis module; the swing rod detection The modules include a left swing rod detection module, a right swing rod detection module and a scratch data acquisition module; wherein,

The installation and setting of the incoming line sensing module and the offline sensing module. The incoming line sensing module can be arranged at the entry end of the working rail, and when the railway rolling stock enters the detection area of the incoming line sensing module at a preset speed, a rolling stock incoming line sign signal is generated and output to the database storage module The offline sensing module can be arranged at the leaving end of the working rail, and when the railway rolling stock leaves the detection area of the offline sensing module at a preset speed, a rolling stock offline sign signal is generated and output to the database storage module;

The left swing rod detection module and the right swing rod detection module are symmetrically arranged on the inner sides of the two working rails respectively, and the n swing rod detection modules are arranged on the inner side of the working rails with intervals and the total length is greater than The circumference of the wheel, where n is the value of L ₁ /L ₂ plus 1 and rounded up to an integer, L ₁ is the circumference of the wheel, and L ₂ is the length of a single pendulum detection module, in view of the fact that the wheel diameters of the detected vehicle wheels are different , and this system is used to detect wheel tread abrasions of railway locomotives. Its wheel diameter is generally large. Taking an electric locomotive as an example, its wheel diameter is 1250mm, and the entire effective detection length is greater than 3925mm. At the same time, due to the limitation of production and processing and the limited length of the pendulum detection module, in general, the length of the pendulum detection module we use is 1200mm. After calculation, it can be known that four groups of pendulum detection modules are required to meet the length requirement of one round of the detected wheel diameter. Placing the swing rods next to each other will result in detection blind spots, so the four swing rod detection modules are placed at intervals to meet the design requirements, and the distance between the second swing rod detection module and the third swing rod detection module is greater than that of the first swing rod detection module. The distance between the first swing detection module and the second swing detection module. Its principle is: the first circle of the wheel starts from the first swing bar, and ends in the middle of leaving the second swing bar and the third swing bar detection module, that is, the first circle will not reach the third swing bar. The second circle begins to pass through the third sway detection module, and the position mapped on the wheel by the third sway detection module contains the distance between the first sway detection module and the second sway detection module and maps the position on the wheel . Continue to pass through an interval area to reach the fourth swing rod detection module, and the position mapped by the fourth swing rod detection module on the wheel includes the position mapped on the wheel by the interval between the second swing rod and the third swing rod. This way the entire circumference of the wheel tread will be mapped onto the sway detection module. In addition, taking into account the influence of different wheel diameters and wear, the requirements can be met by changing the distance between the swing bars;

The left swing rod detection module and the right swing rod detection module are used to collect the angle change data of the wheel flange when the wheel flange is pressed down, and output it to the scratch data acquisition module;

The scratch data acquisition module is used to receive the angle change data of wheel rim depression output by the left swing rod detection module and the right swing rod detection module, and output it to the upper computer communication module;

The database storage module is used to receive the incoming sign signal of the rolling stock and the off-line sign signal of the rolling stock;

The upper computer communication module is used to read the rolling stock incoming line sign signal and the rolling stock offline sign signal in the database storage module in real time; when the upper computer communication module reads a complete set of rolling stock incoming line sign signals from the database storage module After the signal and the off-line sign signal of the rolling stock, the angle change data of the wheel rim pressing down is read from the scratch data acquisition module and output to the scratch data analysis module;

The scratch data analysis module is used to analyze and process the angle change data of wheel rim pressing down to obtain wheel scratch information.

The fork detection module also includes two wheel alignment sensing modules arranged on both sides of the left fork detection module or the right fork detection module; Scratch data acquisition module, the scrape data acquisition module obtains the time according to the received trigger signal, and calculates the speed of the wheel according to the pre-stored position parameters. When the speed is within the preset vehicle speed, the obtained wheel scratch information efficient. The setting of the wheel alignment sensing module shall be determined according to the number of scratch data acquisition modules. A plurality of proximity switches form a group of positioning induction sensors and correspond to a scratch data acquisition module.

The preset vehicle speed is constant and is 1km/h-20km/h. The pre-set vehicle speed is an important condition to ensure the detection accuracy of the system. Excessive vehicle speed during detection will lead to incomplete data collection of the online scratch automatic detection system and affect the detection accuracy of the system.

The left fork detection module and the right fork detection module all include a fork that is in contact with the wheel rim and placed horizontally, a gear mechanism, and a rotary encoder coaxial with the gear mechanism that is arranged on the fork; When the rim of the wheel is pressed down, the swing rod produces an angle change, and the gear mechanism is used to amplify the angle change of the swing rod, and then the rotary encoder processes and outputs the angle change data to the scratch data acquisition module.

Both the incoming line sensing module and the offline sensing module are photoelectric switch sensors.

The wheel alignment sensing module is an eddy current proximity switch or a capacitive proximity switch or a Hall proximity switch or a photoelectric proximity switch or an ultrasonic proximity switch or a microwave proximity switch.

Figure 6 The measurement principle diagram of the pendulum detection module. When the rim of the wheel without tread scratches is pressed against the detection module of the upper swing rod, it will drive the rocker swing rod to produce a certain downward pressure angle, and when the rim of the wheel with tread scratches is pressed against the detection module of the upper swing rod, it will Drive the rocker arm to produce an angle change, and then transmit it to the driven gear through the driving gear. At the same time, the angle change of the rocker arm swing is also amplified with the transmission ratio of the gear transmission mechanism. Then process the data, the angle change α obtained by the two measurements, and the swing radius r of the swing rod is known, and the wear condition of the tread surface, that is, the calculation formula of the scratch depth Δh can be obtained:

Δh=r*sinα;

The scratch data acquisition module includes a controller unit, an encoder data receiving unit, an encoder data decoding unit, a storage unit, a communication unit, and a wheel alignment data receiving unit; the angle change data of wheel rim depression is received by the encoder data The unit receives it, and then sends it to the controller unit after being decoded by the encoder data decoding unit; the wheel alignment data receiving unit receives the wheel alignment sensing module and outputs a trigger signal to the controller unit; the controller unit receives the angle change after decoding The data is output to the storage unit, and finally output to the upper computer communication module through the communication unit. When installing and setting, a scratch data acquisition module is responsible for receiving the data of the corresponding left and right swing bar detection modules. Fig. 7 is a structural diagram of the data acquisition module of the on-line scratch automatic detection system for railway rolling stock wheels. The basic working process of the scratch data acquisition module is: first, the wheel alignment data receiving unit receives the trigger signal of a wheel passing and notifies the controller unit, records the time of receiving the trigger signal, and calculates the wheel position according to the pre-stored position parameters. The encoder data receiving unit receives the scratch data actually collected on site, sends it to the encoder data decoding unit to decode the data, and then sends it to the controller unit. The controller unit controls the coordination of the entire system, stores data or communicates with the host computer, etc.

After the scratch data of the complete wheel is collected on site, it is necessary to analyze and process the data. While the railway rolling stock is passing, the host computer communication module reads a set of incoming and offline flag signals repeatedly from the database. If the complete flag signal is received, it reads the integrated data from the scratch data acquisition module , to generate a detection data file. Scratch data analysis software analyzes and judges the collected database storage modules. If scratches exist, different levels of scratches will be alarmed according to the regulations and a report record file will be formed.

Figure 8 is a flow chart of the scratch analysis of the online scratch automatic detection system for railway rolling stock wheels. The core idea of the analysis algorithm is to segment the scratch data. Regions use an analysis method similar to image horizontal fill search. Then summarize the results of the two sections of analysis, and comprehensively judge whether there is a scratch phenomenon within the given scratch length and scratch depth threshold range, and output the position, size, curve, etc. of the scratch data.

The detection method of a kind of railway locomotive vehicle wheel online abrasion automatic detection system based on the present invention comprises the following steps:

Step 1, pre-setting the range of the head area and the range of the tail area on the curve;

Step 2. Convert the angle change data of the wheel rim depression into the depth displacement data of the wheel rim depression, take the sampling point as the abscissa, and the depth displacement value as the ordinate, to obtain the original curve; for the head area of the original curve, The data curve in the tail area is fitted to obtain a new curve, and the depth displacement values corresponding to the same sampling points on the new curve and the original curve are respectively subtracted to obtain the first difference, and then the first difference is packaged. Line and smooth interpolation processing to form a scratch schematic curve; when there is a sample point on the scratch schematic curve whose ordinate meets the preset scratch depth threshold range, then the wheel is scratched at this time, and the scratch position is obtained , depth, and length information; according to the actual passing data at the scene and the past inspection experience of the current staff, the existence of abrasions is characterized by length and depth. For this reason, another set of general parameters for inspection is obtained through comprehensive analysis and induction: the preset abrasion The threshold range of the scratch length and the threshold range of the scratch depth, Fig. 9 is a flow chart of the analysis method of the scratch analysis module of the on-line scratch automatic detection system for railway rolling stock wheels. ;

Step 3: Analyze the middle area curve after removing the head area and tail area in the original curve, first find the highest point corresponding to the maximum depth displacement value in the middle area, start from the position of the first sampling point in the middle area, and set it as the starting point Sampling point;;

A. Find the point with the same depth displacement value as the initial sampling point according to the sampling timing of the sampling point in turn, and subtract the abscissa of the point from the abscissa of the initial point to obtain the second difference. If the second difference meets the preset The threshold range of scratch length is set, then subtract the ordinate of the highest point from the ordinate of the starting point to obtain the third difference, when the third difference also meets the preset threshold range of scratch depth, then the wheel After being scratched, the position, depth and length of the scratch are obtained;

B. When the second difference does not meet the preset scratch length threshold range or the third difference does not meet the preset scratch depth threshold range, relocate to the sampling point at the next moment and set it as a new starting point Repeat step A from the beginning.

Step 4: According to the position, depth and length information of the abrasion obtained in the above steps, output and alarm the abrasion situation of the railway locomotive vehicle wheel.

The first 200 sampling points on the curve are selected as the sampling point as the sampling timing for the head region range on the curve preset in the step 1; the tail region range is based on the sampling point as the sampling timing sequence and the last 100 sampling points on the curve are selected point.

The method of pre-setting the range of the head area and the range of the tail area on the curve is: when the wheel rim touches the two ends of the swing bar detection module, that is, the jump curve that will be generated when pressing down and leaving the corresponding swing bar detection module, Comprehensively analyze the actual passing data to form the range of jumping curves, and obtain the range of the head area and the tail area.

Figures 10, 11, and 12 are data examples of the middle area processed by the scratch data analysis module, which are used to verify the accuracy of the middle area detection method. Figure 10 is the local triangular wave scratch data of the simulated data example of the middle area processed by the scratch data analysis module, Figure 11 is the local rectangular wave scratch data of the simulated data example of the middle area processed by the scratch data analysis module, and Figure 12 is the scratch data The data analysis module processes the local sine wave abrasion data of the simulated data instance in the middle region. Three sets of scratch data were simulated by scratch data simulation software, and their local scratch forms showed triangular wave, rectangular wave, and sine wave respectively, which are the curves shown in Figures 10, 11, and 12. The length of their data is 1200mm, the starting position of the scratch is 600mm, the length of the scratch is 40mm, and the depth of the scratch is 0.5mm. In the figure, in order to facilitate intuitive understanding, the head and tail areas of the original data curve are processed, the middle area is reserved, and filled with a dark color until a local scratch that meets the preset scratch length threshold range and scratch depth threshold range is found. injury information. The results obtained by the scratch analysis module are: the first shaft, the left wheel, and the No. 1 pendulum. The analysis curve represents triangular wave data, the No. 2 pendulum is rectangular wave data, and the No. 3 pendulum is sine wave data. The three sets of scratch information obtained by the scratch analysis software are: triangle wave data, position 608.0mm, length 58.0mm, depth 0.48mm; rectangular wave data, position 596.0mm, length 50.0mm, depth 0.58mm; sine wave data , the position is 602.0, the length is 46.0mm, and the depth is 0.51mm; compared with the actual given starting position of 600.0mm, the length of the scratch of 40.0mm, and the depth of the scratch of 0.50mm, the depth and length of the scratch can be accurately identified , to meet the detection accuracy requirements of the system.

Figures 13, 14, and 15 are examples of scratch-free data in the middle area collected by the scratch data analysis module for processing three groups of actual passing vehicles on the spot. Data example a, Fig. 14 is the actual on-site passing car without scratch data example b in the middle area processed by the scratch data analysis module, and Fig. 15 is the actual on-site passing car without scratch data example c in the middle area processed by the scratch data analysis module. The result of the scratch analysis module is: no scratch information was found. It is in line with the actual results of the on-site wheel manual review.

Figure 16 is an example of the scratch data analysis module processing a set of data with scratches in the middle area collected by the actual passing vehicles on the spot, which is used to verify the accuracy of the middle area detection method. The result obtained by the scratch analysis module is: the initial position is 766.0mm, the length is 32.0mm, and the depth is 0.35mm; the result of the on-site wheel manual review is that the scratch length is 35mm, and the scratch depth is 0.4mm, which meets the detection accuracy requirements of the system , while the starting position is 1121.0mm, the length is 86.0mm, and the depth is 0.16mm, which does not meet the scratch length threshold range, so it is not considered as scratch here.

Figure 17 is an example of the scratch data analysis module processing another set of scratch data in the head area and tail area collected from the actual passing vehicles on the spot, which is used to verify the accuracy of the head area and tail area detection method. The e in the figure represents: the curve of the original data, the f in the figure represents: the curve after the original data is fitted and translated downward, and the g in the figure represents: the threshold judgment is made after the difference between the fitting and the original data, The curve of smooth interpolation and zoom-in processing. The h in the figure represents: the scratch schematic curve after the envelope and threshold judgment processing. The results obtained by the scratch analysis module are: starting position 1174.0mm, length 28.0mm, depth 0.25mm, meeting the scratch length threshold range, the result of the on-site wheel manual review is that the scratch length is 35mm, and the scratch depth is 0.4mm, which basically meets the detection accuracy requirements of the system.

In actual engineering, this technical condition is applicable to the online automatic detection of scratches on the wheel tread roller circles of railway locomotives and vehicles before they enter the warehouse. The speed of detection passing: 1-20km/h constant speed; the wheel set position for effective detection of scratches: roller circles scratch depth detection accuracy: ±0.2mm; system power supply voltage: AC220V (China), other regions can be customized; power supply voltage of equipment next to the track: DC24V; power supply of equipment next to the track: ≤50W; system operating temperature range: -40℃～70℃; Waterproof grade of equipment beside the track: IP67.

To sum up, the system structure of the present invention is simple in design, easy to use and low in production cost, and can be applied to on-site environments with different conditions, and online detection of tread scratches and out-of-roundness of different car models. In addition, the above embodiments are only to illustrate the technical ideas of the present invention, and cannot limit the scope of protection of the present invention. Any changes made on the basis of the technical solutions according to the technical ideas proposed in the present invention fall within the scope of the present invention. within the scope of protection. Technologies not involved in the present invention can be realized by existing technologies.

Claims (10)

1. An online automatic detection system for wheel abrasion of a railway rolling stock is characterized by comprising an incoming line induction module, an offline induction module, n swing rod detection modules, a database storage module, an upper computer communication module and an abrasion data analysis module; the swing rod detection module comprises a left swing rod detection module, a right swing rod detection module and a scratch data acquisition module; the left swing rod detection module and the right swing rod detection module are mutually symmetrical and are respectively arranged at the inner sides of the two working rails, the n swing rod detection modules are arranged at the inner sides of the working rails at intervals, and the total length is more than that of the swing rod detection modulesWheel circumference, where n is L₁/L₂Adding 1 to the value of (A) and then taking an integer, L₁Is the wheel circumference, L₂Detecting the length of the module for a single swing link; wherein:

the system comprises an incoming line induction module, a database storage module and a data processing module, wherein the incoming line induction module is arranged at an incoming end of a working rail and used for outputting a locomotive incoming line mark signal to the database storage module when a railway locomotive drives into a detection area of the incoming line induction module at a preset speed;

the off-line sensing module is arranged at the leaving end of the working rail and used for outputting a locomotive vehicle off-line mark signal to the database storage module when the railway locomotive vehicle drives away from the detection area of the off-line sensing module at a preset speed;

the left swing rod detection module and the right swing rod detection module are used for collecting angle change data of wheel rim pressing when the wheel rim is pressed down and outputting the angle change data to the scratch data collection module;

the scratch data acquisition module is used for receiving the angle change data of the wheel rim pressing output by the left swing rod detection module and the right swing rod detection module and outputting the angle change data to the upper computer communication module;

the database storage module is used for receiving locomotive incoming line marking signals and locomotive off-line marking signals;

the upper computer communication module is used for reading the locomotive incoming line marking signals and the locomotive off-line marking signals in the database storage module in real time; after the upper computer communication module reads a group of complete locomotive vehicle incoming line mark signals and locomotive vehicle off-line mark signals from the database storage module, reading angle change data of wheel rim pressing from the scratch data acquisition module and outputting the angle change data to the scratch data analysis module;

and the scratch data analysis module is used for analyzing and processing the angle change data of the wheel rim pressing to obtain the scratch information of the wheel.

2. The automatic online detecting system for the scratches on the railway rolling stock wheel of claim 1, wherein the left swing link detecting module and the right swing link detecting module each comprise a swing link which is in contact with the wheel rim and is horizontally arranged, a gear mechanism, and a rotary encoder which is arranged on the swing link and is coaxial with the gear mechanism; when the swing rod is pressed down by the wheel rim of the wheel, the swing rod generates angle change, and the gear mechanism is used for amplifying the angle change of the swing rod and then processing and outputting angle change data to the scratch data acquisition module through the rotary encoder.

3. The system of claim 1, wherein the pendulum bar detection module further comprises two wheel alignment sensing modules disposed at opposite ends of the left pendulum bar detection module or the right pendulum bar detection module; when the railway rolling stock passes through the detection area of the wheel positioning sensing module, a trigger signal is output to the scratch data acquisition module, the scratch data acquisition module obtains time according to the received trigger signal and calculates the speed of the wheel according to the position parameter stored in advance, and when the speed is within the preset speed, the obtained information of the scratch of the wheel is effective.

4. The system of claim 1, wherein the incoming line sensing module and the outgoing line sensing module are both photoelectric switch type sensors.

5. The system of claim 3, wherein the wheel alignment sensing module is an eddy current type proximity switch, a capacitive type proximity switch, a Hall type proximity switch, a photoelectric type proximity switch, an ultrasonic type proximity switch, or a microwave type proximity switch.

6. The system of claim 3, wherein said scratch data collection module comprises a controller unit, an encoder data receiving unit, an encoder data decoding unit, a storage unit, a communication unit and a wheel positioning data receiving unit; the angle change data of the wheel rim pressing is received by the encoder data receiving unit, decoded by the encoder data decoding unit and then sent to the controller unit; the wheel positioning data receiving unit receives the trigger signal output by the wheel positioning sensing module and then sends the trigger signal to the controller unit; the controller unit outputs the received angle change data after decoding processing to the storage unit and finally outputs the angle change data to the upper computer communication module through the communication unit.

7. An automatic railway rolling stock wheel scratch detection system as claimed in claim 1, wherein said preset speed is constant and is 1km/h-20 km/h.

8. The method for detecting the automatic online rail vehicle wheel scratch detection system according to claim 1, comprising the following steps:

step one, presetting a head area range and a tail area range on a curve;

step two, converting the angle change data of wheel rim pressing into depth displacement data of wheel rim pressing, and obtaining an original curve by taking a sampling point as an abscissa and a depth displacement value as an ordinate; fitting data curves in the head region and the tail region of the original curve to obtain a new curve, subtracting depth displacement values corresponding to the same sampling points on the new curve and the original curve respectively to obtain a first difference value, and performing envelope curve and smooth interpolation processing on the first difference value to form a scratch schematic curve; when the ordinate of a certain sampling point on the scratch schematic curve meets a preset scratch depth threshold range, the wheel is scratched at the moment, and information of the position, the depth and the length of the scratch is obtained;

step three, analyzing a middle area curve of the original curve except the head area and the tail area, firstly finding a highest point corresponding to the maximum depth displacement value in the middle area, and setting the highest point as an initial sampling point from the first sampling point position of the middle area;

A. sequentially finding out a point with the same depth displacement value as the initial sampling point according to the sampling time sequence of the sampling point, respectively subtracting the abscissa of the point from the abscissa of the initial point to obtain a second difference value, if the second difference value meets the preset scratch length threshold range, subtracting the ordinate of the highest point from the ordinate of the initial point to obtain a third difference value, and if the third difference value also meets the preset scratch depth threshold range, scratching the wheel at the moment to obtain the information of the position, the depth and the length of the scratch;

B. and B, when the second difference value does not meet the preset scratch length threshold range or the third difference value does not meet the preset scratch depth threshold range, repositioning the sampling point at the next moment, setting the sampling point as a new starting point, and repeating the step A.

And step four, outputting and alarming the abrasion condition of the railway locomotive vehicle wheels according to the abrasion position, depth and length information obtained in the step.

9. The method as claimed in claim 8, wherein the preset head area range on the curve in the first step is a sampling sequence of sampling points, and the first 200 sampling points on the curve are selected; the tail area range takes the sampling points as a sampling time sequence, and the last 100 sampling points on the curve are selected.

10. The method for detecting the automatic online detecting system for the railway locomotive wheel scratches as claimed in claim 8, wherein the method for presetting the head area range and the tail area range on the curve is as follows: when the wheel rim is in contact with the two ends of the swing rod detection modules, namely the wheel rim is pressed down and leaves the corresponding swing rod detection modules, a jumping curve can be generated, the actual vehicle passing data is comprehensively analyzed to form an interval range where the jumping curve occurs, and a head area range and a tail area range are obtained.