A smart field control method and system suitable for urban rail transit

Abstract

The present invention discloses a smart field section control method and system suitable for urban rail transit, which belongs to the field of rail transit technology. Its technical points include: setting multiple types of sensors such as image, position, and status in key areas of the field section to collect vehicle, equipment, and infrastructure operation, status, and environmental data in real time; transmitting the collected data to a control center in real time through a wired and wireless communication network, so that the control center can analyze the received data in real time, identify anomalies, predict faults, and generate control instructions; automatically controlling intelligent execution equipment such as turnout machines, signal machines, and vehicle maintenance equipment according to the control instructions, so as to realize intelligent coordinated control of vehicle entry and exit, parking, maintenance, and equipment operation. The railway operation and maintenance management system is a highly complex and sophisticated system. The present invention ensures the safety, efficiency, and orderliness of railway transportation with the help of the railway operation and maintenance management system.

Description

Technical Field

[0001] The present invention relates to the field of rail transit technology, and more specifically to a smart site control method and system suitable for urban rail transit. Background Art

[0002] Train wheelsets are a key component of trains. To ensure train running safety, it is necessary to regularly inspect the surface of wheelsets for cracks.

[0003] A Chinese patent application with publication number CN116385543A provides a method for identifying, locating and measuring cracks in fluorescent magnetic particle inspection images of train wheelsets. The invention places the wheelset in a flaw detector for image acquisition and identifies cracks in the wheelset based on image recognition technology.

[0004] However, the above static detection method requires disassembly of the wheelset, which takes a long time to detect, so the existing technology has shortcomings. Summary of the invention

[0005] In view of the shortcomings of the prior art, the purpose of the present invention is to provide a smart field section control method and system suitable for urban rail transit. By collecting wheelset images during train operation, crack detection is performed without the need to disassemble the wheelset, thus saving detection time. At the same time, the present invention can identify the morphological changes of cracks under the actual stress conditions of the wheelset, and can more accurately identify fault cracks that affect the safety of train operation.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] The present invention provides a smart field section control method applicable to urban rail transit, comprising:

[0008] Image, location and status sensors are installed in key areas of the site to collect real-time operation, status and environmental data of vehicles, equipment and infrastructure;

[0009] The collected data is transmitted to the control center in real time through a wired and wireless communication network, so that the control center can analyze the received data in real time, identify anomalies, predict faults and generate control instructions;

[0010] Automatically control switch machines, signal machines and vehicle maintenance equipment according to management and control instructions to achieve intelligent and coordinated control of vehicle entry and exit, parking, maintenance and equipment operation.

[0011] As a further improvement of the present invention, the smart field segment management method further includes:

[0012] Acquire a first image and a second image, wherein the first image is an image of a wheelset when a train is traveling within a preset speed range, and the second image is an image of a wheelset when the train is departing;

[0013] According to the first image and the second image, a first crack and a second crack are identified, wherein the first crack is a wheelset crack when the train is in a high-speed running state, and the second crack is a wheelset crack when the train is departing;

[0014] Selecting a fault crack from among the first cracks according to the similarity between the first crack and the second crack;

[0015] Obtaining a fault detection result of a train wheelset according to the characteristic information of the fault crack;

[0016] Through the intelligent maintenance system, a maintenance plan is formulated according to the fault detection results of the train wheelset.

[0017] As a further improvement of the present invention, identifying a first crack according to the first image includes:

[0018] Performing image filtering and radial distortion correction on the first image to obtain a third image;

[0019] Identify the direction and length of the first crack according to the third image;

[0020] Performing smear correction on the third image according to the smear direction and the smear length to obtain a target image;

[0021] The first crack is identified according to the target image.

[0022] As a further improvement of the present invention, identifying the smear direction and the smear length of the first crack according to the third image includes:

[0023] Performing a second Fourier transform on the third image to obtain a second Fourier spectrum diagram;

[0024] The smear direction and the smear length are obtained according to the quadratic Fourier spectrum diagram.

[0025] As a further improvement of the present invention, obtaining the smear direction according to the quadratic Fourier spectrum diagram includes:

[0026] Obtaining a binary image according to the quadratic Fourier spectrum diagram and the contour detection algorithm;

[0027] Cropping the binary image to obtain a cropped image;

[0028] Performing Radon transform on the cropped image to obtain a transformation matrix;

[0029] The smear direction is calculated according to the transformation matrix.

[0030] As a further improvement of the present invention, a binary image is obtained according to the quadratic Fourier spectrum diagram and the contour detection algorithm, including:

[0031] Calculating the amplitude and adaptive threshold of each pixel in the quadratic Fourier spectrum, where the adaptive threshold is the grayscale average value within a preset neighborhood of the pixel;

[0032] Obtaining a plurality of contour points of the quadratic Fourier spectrum graph based on the amplitude and the adaptive threshold;

[0033] The plurality of contour points are connected to obtain the binary image.

[0034] As a further improvement of the present invention, the step of obtaining a plurality of contour points of the secondary Fourier spectrum graph based on the amplitude and the adaptive threshold comprises:

[0035] If the amplitude of the pixel point is greater than or equal to the adaptive threshold, the pixel point is determined to be a contour point;

[0036] If the amplitude of the pixel point is greater than half of the adaptive threshold and less than the adaptive threshold, it is determined whether the pixel point is a contour point according to the determination result of the previous pixel point adjacent to the pixel point.

[0037] As a further improvement of the present invention, obtaining the smear length according to the quadratic Fourier spectrum diagram includes:

[0038] Calculating the spacing between adjacent dark fringes in the quadratic Fourier spectrum;

[0039] The smear length is calculated according to the spacing and the size of the quadratic Fourier spectrum diagram.

[0040] As a further improvement of the present invention, selecting a fault crack from the first cracks according to the similarity between the first cracks and the second cracks includes:

[0041] The second crack edge is uniformly selected reference points, and obtain the reference point set ,in are the coordinates of the reference point;

[0042] Uniformly select the first crack edge edge points, and get the edge point set , the edge points correspond to the reference points one by one, where are the coordinates of the edge points;

[0043] The similarity is calculated based on the reference point and the edge point. ,in is the gradient vector of the reference point, is the gradient vector of the edge point;

[0044] According to the similarity level, a fault crack is selected from the first cracks.

[0045] As a further improvement of the present invention, if there are multiple fault cracks, obtaining the fault detection result of the train wheelset according to the characteristic information of the fault cracks includes:

[0046] Acquire characteristic information of each of the fault cracks, the characteristic information including the length, width and average distance of the fault crack to adjacent fault cracks;

[0047] Calculating the fault level of each fault crack according to the characteristic information and a preset possibility function;

[0048] According to the fault level, the fault detection result of the train wheelset is output.

[0049] The present invention provides a smart field section control system applicable to urban rail transit, comprising:

[0050] The acquisition module sets up image, position and status sensors in key areas of the field to collect the operation, status and environmental data of vehicles, equipment and infrastructure in real time;

[0051] The transmission module transmits the collected data to the control center in real time through a wired and wireless communication network, so that the control center can analyze the received data in real time, identify anomalies, predict faults and generate control instructions;

[0052] The control module automatically controls the switch machines, signal machines and vehicle maintenance equipment according to the control instructions, realizing intelligent coordinated control of vehicle entry and exit, parking, maintenance and equipment operation.

[0053] The present invention achieves the purpose of train wheel fault detection by accurately identifying fault cracks through the changes in crack morphology under real stress conditions, and eliminates the smear phenomenon in high-speed image shooting through image processing technology, further improving the recognition accuracy. Finally, the fault level is judged according to the characteristics of the fault crack, which facilitates subsequent maintenance work, thereby realizing the management and control of intelligent field sections. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 This is a schematic diagram of the management and control system architecture of the smart field segment of the present invention;

[0055] Figure 2 A schematic diagram of the control process of the control system of the intelligent field segment of the present invention;

[0056] Figure 3is a flow chart of the method steps of the present invention;

[0057] Figure 4 is a flow chart of the identification steps of the first crack in the present invention;

[0058] Figure 5 It is a flow chart of the steps of determining contour points in the present invention;

[0059] Figure 6 It is a schematic diagram of the system structure of the present invention. DETAILED DESCRIPTION

[0060] The technical solution of the present invention is described in detail below through the accompanying drawings and specific embodiments. It should be understood that the embodiments of the present invention and the specific features in the embodiments are detailed descriptions of the technical solution of the present invention, rather than limitations of the technical solution of the present invention.

[0061] The term "and / or" in the following text is only a description of the association relationship between the associated objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " generally indicates that the associated objects before and after are in an "or" relationship.

[0062] The smart section of urban rail transit is a new model that uses modern scientific and technological means such as information technology, automation technology, cloud computing and the Internet of Things to conduct efficient and intelligent management of urban rail transit vehicle sections. The smart section of this application realizes all-round control of train operation maintenance, fault diagnosis, spare parts management, etc. through an integrated railway operation and maintenance management system, improves operational efficiency, reduces maintenance costs, and supports data-driven decision-making, thereby providing safer and more reliable operation guarantees for urban rail transit systems.

[0063] In the present application embodiment, Figure 1 The railway operation and maintenance management system shown is a highly complex and sophisticated system that ensures the safety, efficiency and orderliness of railway transportation. The following is a detailed description of each link of the system:

[0064] Production information is the basis of the entire system. Construction information involves the arrangement of various types of construction along the railway, including new lines, line reconstruction and other projects. These constructions must be accurately planned to avoid interference with normal operations. Power outage information is related to the reasonable control of power supply, ensuring that power can be safely cut off during maintenance and other operations, and that power can be restored in time after the operation is completed. The power monitoring and data acquisition data provided by the PSCDA system can reflect the operating status of power equipment in real time, such as the voltage and current parameters of the substation, to help operators detect potential power failures in a timely manner. The operation diagram, vehicle location and track status information provided by the ATS train automatic monitoring system are even more critical. The operation diagram clarifies the departure time, route and arrival time of each train, and is the core basis for dispatching and commanding; real-time monitoring of vehicle location can prevent accidents such as train conflicts, and monitoring of track status can detect track diseases in advance to ensure the smooth and safe operation of trains.

[0065] Maintenance procedures are formulated based on the long-term reliable operation requirements of equipment and facilities. The track demand plan needs to take into account factors such as track wear and service life, and regularly arrange track inspection, maintenance and replacement. The power supply demand plan should determine the maintenance cycle and projects of power equipment based on the degree of aging of power equipment and load size. The facility demand plan includes maintenance arrangements for infrastructure such as stations, platforms, and waiting rooms. The equipment demand plan comprehensively plans for trains, signal systems, communication equipment, etc., such as regular maintenance of trains and troubleshooting of signal systems. The personnel demand plan must ensure that there are sufficient numbers and qualifications of maintenance personnel participating in the operation, including different professionals such as electricians, fitters, and track workers. The material demand plan needs to prepare in advance the materials and parts required for various types of maintenance, such as rails, sleepers, power cables, electronic components, etc.

[0066] Vehicle monitoring is achieved through various sensors installed on the train. The sensors can monitor the status of the vehicle's running gear, such as wheel wear, bearing temperature, etc., and can promptly alarm if an abnormality occurs. At the same time, the operating parameters of key equipment such as the vehicle's braking system and traction system are also monitored in real time. Trackside inspection uses inspection equipment along the line to inspect the track. For example, the track flaw detection vehicle can detect whether there are defects such as cracks inside the track, and the track geometry inspection vehicle can measure the track gauge, superelevation, direction and other geometric parameters to ensure that the track meets the requirements of train operation.

[0067] The operation data records the details of each inspection and maintenance operation, including operation time, operator, operation content, etc. The fault management file records all the faults that have occurred in detail, including the time, location, phenomenon, handling process and results of the fault. By analyzing the fault data, the weak links of the equipment can be found, providing a basis for preventive maintenance.

[0068] Maintenance personnel need to strictly follow maintenance procedures and conduct quality inspections after the work is completed. In the fault management link, when equipment fails, the emergency handling procedures can be quickly initiated and personnel can be organized to carry out emergency repairs. Spare parts replacement requires a complete spare parts inventory management to ensure timely supply when parts need to be replaced. Event records record other important events besides failures, such as the impact of temporary construction on operations.

[0069] The yard operation includes shunting and receiving and dispatching operations, signal duty operations and mainline operations. Shunting and receiving and dispatching operations must ensure the orderly movement of trains in the yard and the smooth entry and exit of the mainline. The signal duty operations must ensure the normal operation of the signal system and accurately send operation signals to the trains. Mainline operations must be strictly carried out in accordance with the operation diagram, and at the same time, they must have the ability to respond to emergencies, such as train failures, natural disasters, etc., to ensure the safe transportation of passengers and goods.

[0070] like Figure 2 As shown in the figure, the workflow of a railway inspection and operation management system is shown. The system covers multiple complex and closely related subsystems to ensure the safety and efficiency of railway operations.

[0071] First, in terms of maintenance plan generation, the system integrates a variety of technical means. Among them, technologies such as 360° imaging and wheelset detection play an important role. 360° imaging can capture the appearance of railway equipment in all directions. Whether it is tracks, vehicles or facilities along the line, clear image information can be obtained through this technology to facilitate the discovery of potential damage or failures. Wheelset detection focuses on the wheelset part of the vehicle, which is a key component of vehicle operation. Accurate detection can avoid serious accidents such as derailment caused by wheelset problems. The data obtained through these detection technologies provides basic data support for the pre-generation of maintenance plans.

[0072] The perception intelligent analysis system plays a connecting role in the entire process. It analyzes and processes data from 360° imaging and wheelset detection. Through intelligent algorithms, it can accurately evaluate the status of the equipment, determine whether the equipment is in normal working condition, predict possible failures, and make production optimization suggestions based on the analysis results. For example, for the wear of the track, it can calculate the wear trend of the track in the future based on historical data and current detection data, thereby providing a scientific basis for the formulation of maintenance plans.

[0073] Manual adjustment of maintenance plans and dynamic adjustment of work items are the embodiment of system flexibility. Although the intelligent analysis system can provide scientific maintenance plans, in actual operation, it is still necessary to combine the actual situation on site and the judgment of experienced staff. Manual adjustments can be made to the maintenance plan according to special circumstances to ensure the feasibility and effectiveness of the plan. Dynamic adjustment of work items can flexibly change the focus and order of maintenance operations according to the real-time changes in the equipment status. For example, when a certain equipment suddenly has an abnormal alarm, the system can quickly increase its maintenance priority to ensure that the fault can be handled in a timely manner.

[0074] Task scheduling is the command center of the entire maintenance process. It arranges manual maintenance and automated operations according to the maintenance plan and the real-time status of the equipment. In terms of manual maintenance, it ensures that maintenance personnel can arrive at the right place at the right time and are equipped with the right tools and equipment for maintenance. For automated operations, the system can control automated equipment to accurately perform maintenance tasks. For example, automated track grinding equipment can accurately grind the track according to preset parameters, improving operation efficiency and quality.

[0075] Automated operations have unique advantages in troubleshooting. They can quickly respond to fault alarms and automatically repair some common faults. For example, for some simple faults in the signal system, automated equipment can quickly locate and repair them, reducing the impact of the faults on railway operations. At the same time, manual maintenance operations play an important role in handling some complex field-level faults. Through on-site inspections and repairs by professionals, problems that automated equipment cannot handle can be solved.

[0076] In terms of operation and maintenance, the system involves multiple subsystems such as vehicle impoundment management, safety operation management, and material / tool ​​management. Vehicle impoundment management can reasonably arrange the parking and maintenance sequence of vehicles when a vehicle breaks down and needs to be repaired, so as to avoid affecting the normal operation of other vehicles. Safety operation management ensures the safety of personnel and equipment during the maintenance and operation process, and prevents safety accidents by formulating strict safety regulations and monitoring measures. Material / tool ​​management ensures that various materials and tools required for maintenance and operation can be supplied in a timely manner, and the quantity, location and usage of materials and tools are accurately managed through the intelligent inventory management system.

[0077] In addition, the system also includes peripheral security systems such as safety interlocking system, perimeter protection system and intelligent logistics system. The safety interlocking system can prevent misoperation during equipment operation and ensure safe and orderly operation between various equipment. The perimeter protection system ensures the physical safety of railway facilities and prevents external factors from interfering with and damaging railway operations. The intelligent logistics system is responsible for the transportation and distribution of materials required for maintenance and operation, ensuring that materials can arrive at the work site in a timely and accurate manner.

[0078] Finally, the DCC display system and the supervision business system realize visual management and supervision of the entire maintenance and operation process through the transmission of business dynamic information. They can display information such as equipment status, maintenance progress and fault handling in real time, which is convenient for managers to make overall arrangements and decisions, ensuring the efficient and stable operation of the entire railway maintenance and operation management system.

[0079] The railway inspection and operation management system, through close collaboration among various subsystems and making full use of modern detection technology, intelligent analysis technology and automation technology, has achieved full-process management from detection to maintenance, from fault handling to operation and maintenance guarantee, greatly improving the safety and reliability of railway operations.

[0080] The following is a detailed description of train fault detection.

[0081] The present application is applicable to smart sections of urban rail transit and is a smart management and control technology solution for smart sections of urban rail transit. Specifically, it involves train operation maintenance and fault detection.

[0082] As the part that is in direct contact with the rails, the main function of the train wheelset is to ensure the operation and turning of the locomotive on the rails, bear all static and dynamic loads from the locomotive, and transfer them to the rails. The main types of failure of the train wheelset include wear, cracks, etc. Among them, cracks may cause the wheelset to break during operation, causing serious safety accidents such as train derailment or overturning. However, if the wheelset is frequently inspected and maintained, the normal operation time of the train will be affected, thereby reducing the overall efficiency of railway transportation.

[0083] Therefore, in order to solve the above problems, Figure 3 As shown, the embodiment of the present application provides a smart field section control method applicable to urban rail transit, which can reduce the number of maintenance times while ensuring the accuracy of fault detection. The method includes:

[0084] Acquire a first image and a second image, the first image being an image of a wheelset when a train is traveling within a preset speed range, and the second image being an image of a wheelset when the train is departing;

[0085] According to the first image and the second image, a first crack and a second crack are identified, wherein the first crack is a wheelset crack when the train is in a high-speed running state, and the second crack is a wheelset crack when the train is departing;

[0086] According to the similarity between the first crack and the second crack, a fault crack is selected from the first crack;

[0087] According to the characteristic information of the fault crack, the fault detection result of the train wheelset is obtained;

[0088] Through the intelligent maintenance system, a maintenance plan is formulated according to the fault detection results of the train wheelset.

[0089] The preset speed range is 250km / h and above.

[0090] The intelligent maintenance system in the embodiment of the present application is an automated maintenance system suitable for smart sections of urban rail transit, which includes a large number of automated maintenance tools and configuration equipment, and can automatically perform train maintenance work by inputting maintenance plans or maintenance plans. When the train slowly enters the smart section, the intelligent maintenance system maintains the train according to the maintenance plans or maintenance plans received in advance, which involves detailed maintenance work such as personnel scheduling, and the present application does not limit or elaborate on this.

[0091] In an embodiment of the present application, the first image is an image taken when the train is traveling within a preset speed range, which is taken by a high-speed camera. The second image is an image taken when the train departs. At this time, the train is traveling at a slower speed and can be taken by an ordinary camera.

[0092] This embodiment compares the cracks in the wheelset during high-speed driving and the cracks in the wheelset during train departure to obtain the fault cracks that are likely to affect the safety of train operation, and then accurately detects the faults in the train wheelset through the fault cracks. This embodiment only considers the cracks that deform under real stress conditions, namely the fault cracks, and the cracks that do not deform are called surface cracks. The deformation of surface cracks in high-speed driving is small, indicating that the crack amplitude is small and is not likely to affect the safety of train operation. Therefore, it is not used to detect train wheel set faults. Compared with the prior art that detects and repairs all cracks, the fault detection method provided in this embodiment can more effectively eliminate surface cracks and identify fault cracks. At the same time, since this embodiment only detects fault cracks, it can reduce the number of repairs and increase the service life of the train wheelset.

[0093] Furthermore, this embodiment provides a step of identifying a first crack according to the first image, including:

[0094] Performing image filtering and radial distortion correction on the first image to obtain a third image;

[0095] Identify the direction and length of the first crack according to the third image;

[0096] Performing smear correction on the third image according to the smear direction and the smear length to obtain a target image;

[0097] According to the target image, the first crack is identified.

[0098] Exemplarily, the formula for performing image filtering processing on the first image is:

[0099]

[0100] in, Indicates the coordinates of the pixel points that need to be filtered. is the pixel value after filtering. is the pixel value before filtering. is the normalized weight, represents the filter window, Indicates the length and width of the window. Represents the pixel value of the pixel within the filter window, is the kernel function in the spatial domain, The filtering method provided in this embodiment smoothes the image while retaining edge information by comprehensively considering the information in the spatial domain and the pixel value domain, and can effectively remove noise and maintain image details.

[0101] Exemplarily, the radial distortion correction method is:

[0102]

[0103]

[0104] in, Indicates the pixel coordinates that need to be corrected. represents the corrected pixel coordinates, , , and Represents the correction coefficient, which can be obtained through camera calibration.

[0105] The method for identifying the first crack provided in the present embodiment can eliminate the influence of noise and high-speed camera lens distortion on the image through image filtering and distortion correction. At the same time, since the train wheelset is located at the bottom of the train and the light is dim, in order to obtain the wheelset image in a high-speed and dim light environment, it is often necessary to set a lower frame rate and a longer exposure time. At this time, the ghosting phenomenon is prone to occur. The method provided in the present embodiment corrects the ghosting phenomenon by calculating the ghosting direction and the ghosting length, thereby alleviating the above problems and further improving the accuracy of crack identification.

[0106] Further, such as Figure 4 As shown, this embodiment provides a step of identifying the direction and length of the smear of the first crack according to the third image, including:

[0107] Performing a second Fourier transform on the third image to obtain a second Fourier spectrum diagram;

[0108] The smear direction and smear length are obtained according to the quadratic Fourier spectrum.

[0109] Specifically, the steps of performing a second Fourier transform on the third image are:

[0110] The third image is transformed by discrete Fourier transform Convert to spectrum ,in represents the pixel coordinates in the third image, Indicates the frequency domain coordinates corresponding to the pixel;

[0111] In order to enhance the clarity of the fringes in the spectrum, Perform logarithmic transformation to obtain ;

[0112] To further refine the stripes, Performing a second discrete Fourier transform, we get ,in for Frequency domain coordinates obtained by discrete Fourier transform;

[0113] Similarly, in order to further enhance the clarity of the fringes in the spectrum, Logarithmic transformation to obtain the quadratic Fourier spectrum .

[0114] Furthermore, the embodiment of the present application provides a step of obtaining the smear direction according to the quadratic Fourier spectrum diagram, including:

[0115] According to the quadratic Fourier spectrum and contour detection algorithm, a binary image is obtained;

[0116] Crop the binary image to obtain a cropped image;

[0117] Perform Radon transform on the cropped image to obtain the transformation matrix;

[0118] According to the transformation matrix, the smear direction is calculated.

[0119] Specifically, firstly, the quadratic Fourier spectrum Perform contour detection to obtain a binary image ;

[0120] Then the binary image is cropped, its center information is retained, and the redundant cross bright line part is cropped to obtain ;

[0121] For cropped images Perform Rando transform to get the transformation matrix ;

[0122] Then according to the matrix and rotation angle Draw an extreme value curve. The value range is 0°-180°, and the maximum value of the extreme value curve is obtained. According to the rotation angle corresponding to the maximum value The smear angle can be calculated , and The corresponding relationship is:

[0123]

[0124] In this embodiment, a specific smear angle is obtained by performing a Radon transform on the frequency domain image. The Rando transform is a transformation method for performing a line integral on an image at a specified angle, and the rotation angle can determine the projection direction. When smear occurs, the spectrum image will show regular light and dark stripes parallel to each other. The direction of the stripes is related to the smear angle. When the integrated straight line beam is parallel to the light and dark stripes in the spectrum, the smear angle can be determined by projection vectors in different directions.

[0125] Furthermore, this embodiment provides a step of obtaining the smear length according to the quadratic Fourier spectrum diagram, including:

[0126] Calculate the spacing between adjacent dark fringes in the quadratic Fourier spectrum;

[0127] The smear length is calculated based on the spacing and the size of the quadratic Fourier spectrum.

[0128] Specifically, the smear length for:

[0129]

[0130] in, is the distance between adjacent dark fringes, are the length and width of the quadratic Fourier spectrum graph. In this embodiment, it is assumed that the length and width of the quadratic Fourier spectrum graph are equal. is the proportional correlation coefficient between the size of the quadratic Fourier spectrum and the real scene.

[0131] This implementation measures the length of the image smear by the spacing between adjacent dark stripes in the quadratic Fourier spectrum. When the spacing between adjacent dark stripes is smaller, the image smear phenomenon is more serious and the smear length is longer. This is because the bright stripes in the spectrum represent high-frequency information and the dark stripes represent low-frequency information. When the spacing between dark stripes is smaller, the width of the bright stripes is smaller, the image has less high-frequency information and the image has lower clarity, which means that the image smear phenomenon is more serious and the smear length is longer.

[0132] The method provided in this embodiment can accurately calculate the smear angle and smear length based on the stripe characteristics of the spectrum graph, and then the target image can be obtained by eliminating the smear portion in the third image according to the smear angle and smear length. If there are multiple cracks in the third image, that is, there are multiple smear portions, it is necessary to add an image segmentation step, and then repeat the above steps for each segmented crack image to eliminate the smear portion of each crack.

[0133] Furthermore, this embodiment provides a step of obtaining a binary image according to a quadratic Fourier spectrum diagram and a contour detection algorithm, including:

[0134] Calculate the amplitude and adaptive threshold of each pixel in the quadratic Fourier spectrum, where the adaptive threshold is the average grayscale value within a preset neighborhood of the pixel;

[0135] A plurality of contour points of a quadratic Fourier spectrum graph are obtained based on amplitude and adaptive threshold;

[0136] Connect multiple contour points to get a binary image.

[0137] Specifically, the pixel amplitude for:

[0138]

[0139] in, is the coordinate of the pixel point in the quadratic Fourier spectrum, , , and They are , , 45°, and 135° directions.

[0140] This embodiment uses multi-directional gradient calculation, compared with only and The calculation is performed in two directions. This method can capture the direction information of the edge more accurately, thereby improving the accuracy of edge detection.

[0141] Further, such as Figure 5 As shown, this embodiment provides a step of obtaining multiple contour points of a quadratic Fourier spectrum diagram based on amplitude and adaptive threshold, including:

[0142] If the amplitude of the pixel point is greater than or equal to the adaptive threshold, the pixel point is determined to be a contour point;

[0143] If the amplitude of the pixel point is greater than half of the adaptive threshold and less than the adaptive threshold, then whether the pixel point is a contour point is determined based on the determination result of the previous pixel point adjacent to the pixel point.

[0144] Specifically, first set the adaptive threshold for:

[0145]

[0146] in, is the side length of the preset neighborhood. In this embodiment, it is assumed that the preset neighborhood is a square neighborhood. Represents the coordinates of the pixel points in the preset neighborhood. Represents pixel Since the gray value in the neighborhood of each pixel is different, the adaptive threshold of each pixel is also different, achieving the purpose of adaptively adjusting the threshold according to each pixel.

[0147] Then according to the adaptive threshold Find contour points based on the relationship between the magnitude and amplitude: If the amplitude of the pixel point is greater than or equal to the adaptive threshold, the pixel point is determined to be a contour point, and the output value is 1; if the amplitude of the pixel point is less than or equal to half of the adaptive threshold, the pixel point is determined not to be a contour point, and the output value is 0. If the amplitude of the pixel point is greater than half of the adaptive threshold and less than the adaptive threshold, the pixel point is determined to be a contour point based on the determination result of the previous pixel point adjacent to the pixel point, that is, the output value is the same as the output value of the previous pixel point, which is specifically expressed as:

[0148]

[0149] in, is the output value of the current pixel, is the output value of the previous pixel, and finally connect the points with output value 1 to get the final image.

[0150] This embodiment adopts an adaptive threshold to filter contour points. Compared with a fixed threshold, the threshold provided by this embodiment can change with the change of the pixel position, and the neighborhood size can be set as needed, so that contour points can be selected more accurately.

[0151] Furthermore, this embodiment provides a step of selecting a fault crack from the first crack according to the similarity between the first crack and the second crack, including:

[0152] Uniformly select the second crack edge reference points, and obtain the reference point set ,in is the coordinate of the reference point;

[0153] Uniformly select the first crack edge edge points, and get the edge point set , the edge points correspond to the reference points one by one, where is the coordinate of the edge point;

[0154] Based on the reference points and edge points, the similarity is calculated ,in is the gradient vector of the reference point, is the gradient vector of the edge point;

[0155] According to the degree of similarity, the fault crack is selected from the first cracks.

[0156] In this implementation, the gradient vector is used to measure the similarity between the first crack and the second crack. The direction and magnitude of the gradient vector reflect the trend and degree of grayscale change of the image at that point, so it can be used as an indicator to describe the contour features. When the value is greater than or equal to the preset value, it can be considered that the first crack and the second crack are basically the same, that is, the deformation degree of the second crack after being subjected to pressure is small, and it is a surface crack. This type of crack has little impact on the safety of train operation. When the second crack is smaller than the preset value, it is considered that the deformation degree of the second crack after being subjected to pressure is large, and it is a fault crack and needs to be repaired in time. If there are multiple first cracks and multiple second cracks, the above steps need to be repeated multiple times to determine the similarity of each group of cracks.

[0157] Furthermore, this embodiment provides a step of obtaining a fault detection result of a train wheelset according to characteristic information of a fault crack, including:

[0158] Obtain characteristic information of each fault crack;

[0159] Calculate the fault level of each fault crack according to the characteristic information and the preset possibility function;

[0160] According to the fault level, the fault detection results of the train wheelset are output.

[0161] Specifically, the characteristic information may include the length and width of each fault crack and the average distance between each fault crack and its adjacent fault cracks. Assuming that the total number of fault cracks is , is a positive integer greater than 1. The total number of indicators contained in the feature information is First, set the number of fault levels according to the detection needs , and divide the value range of each indicator into For example, The value range of the indicator Divide into The small intervals are , ;

[0162] Then confirm and The corresponding first and Level turning point and , and calculate the geometric midpoint of each cell , ,in and Indicates the endpoints of each interval;

[0163] Then construct the first-level lower limit measure possibility function and Upper limit measure possibility function of rank , assuming For the 1 observation value of an indicator, when hour:

[0164]

[0165] when hour,

[0166]

[0167] For Levels, simultaneous connections With The geometric midpoint of the level (or the turning point of the first order )as well as With The geometric midpoint of the level ( or Level turning point ), get the Indicators about The trigonometric possibility function of the level , assuming For the 1 observation of an indicator, which belongs to The possibility function of each level is expressed as:

[0168]

[0169] Then determine the weight of each feature information , and calculate the The fault crack is about The comprehensive clustering coefficient of ,in Indicates The indicator belongs to The possibility function of the level, Indicates The first fault crack The observed value of an indicator;

[0170] like , , then The fault crack belongs to Failure level.

[0171] This embodiment calculates the fault level of each fault crack according to the characteristic information of each fault crack and a preset possibility function, so as to facilitate targeted maintenance at a later stage.

[0172] The smart field section control method applicable to urban rail transit provided in the embodiment of the present application first eliminates the influence of noise and smearing through image processing technology, can accurately identify fault cracks, and obtain the fault level of the cracks through the characteristic information of the fault cracks, thereby realizing fault detection of train wheelsets.

[0173] Further, such as Figure 6 As shown, the embodiment of the present application provides a smart field section control system applicable to urban rail transit, including:

[0174] An acquisition module, used to acquire a first image and a second image, the first image being an image of a wheelset when the train is in a high-speed running state, and the second image being an image of a wheelset when the train is departing;

[0175] an identification module, configured to identify a first crack and a second crack according to the first image and the second image, wherein the first crack is a wheelset crack when the train is in a high-speed running state, and the second crack is a wheelset crack when the train is departing;

[0176] A selection module, for selecting a fault crack from the first cracks according to the similarity between the first crack and the second crack;

[0177] The detection module is used to obtain the fault detection result of the train wheelset according to the characteristic information of the fault crack.

[0178] The embodiments of the present application provide a smart field section control method and system suitable for urban rail transit. By identifying the morphological changes of cracks under the actual stress conditions of the wheelset, fault cracks that affect the safety of train operation can be accurately identified, and the fault level of the cracks can be determined based on the characteristic information of the fault cracks, thereby realizing the detection of train wheel set faults. At the same time, the embodiments of the present application effectively eliminate the ghosting phenomenon of images taken by high-speed cameras based on the characteristics of the spectrum diagram, further improving the accuracy of detection.

[0179] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0180] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0181] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0182] The above is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technicians in this technical field, some improvements and modifications without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.

Claims

1. A smart field section control method applicable to urban rail transit, characterized in that: include: Image, location and status sensors are installed in key areas of the site to collect real-time operation, status and environmental data of vehicles, equipment and infrastructure; The collected data is transmitted to the control center in real time through a wired and wireless communication network, so that the control center can analyze the received data in real time, identify anomalies, predict faults and generate control instructions; Automatically control turnout machines, signal machines and vehicle maintenance equipment according to control instructions to achieve intelligent coordinated control of vehicle entry, exit, parking, maintenance and equipment operation; Acquire a first image and a second image, wherein the first image is an image of a wheelset when a train is traveling within a preset speed range, and the second image is an image of a wheelset when the train is departing; According to the first image and the second image, a first crack and a second crack are identified, wherein the first crack is a wheelset crack when the train is in a high-speed running state, and the second crack is a wheelset crack when the train is departing; Selecting a fault crack from among the first cracks according to the similarity between the first crack and the second crack; Obtaining a fault detection result of a train wheelset according to the characteristic information of the fault crack; Formulate a maintenance plan based on the fault detection result of the train wheelset through an intelligent maintenance system; The step of identifying a first crack according to the first image includes: Performing image filtering and radial distortion correction on the first image to obtain a third image; Performing a second Fourier transform on the third image to obtain a second Fourier spectrum diagram; Obtaining a smear direction and a smear length according to the quadratic Fourier spectrum diagram; Performing smear correction on the third image according to the smear direction and the smear length to obtain a target image; The first crack is identified according to the target image.

2. According to claim 1, a smart section control method applicable to urban rail transit is characterized in that: Obtaining the smear direction according to the quadratic Fourier spectrum diagram includes: Obtaining a binary image according to the quadratic Fourier spectrum diagram and the contour detection algorithm; Cropping the binary image to obtain a cropped image; Performing Radon transform on the cropped image to obtain a transformation matrix; The smear direction is calculated according to the transformation matrix.

3. The smart section control method applicable to urban rail transit according to claim 2 is characterized in that: According to the quadratic Fourier spectrum diagram and the contour detection algorithm, a binary image is obtained, including: Calculating the amplitude and adaptive threshold of each pixel in the quadratic Fourier spectrum, where the adaptive threshold is the grayscale average value within a preset neighborhood of the pixel; Obtaining a plurality of contour points of the quadratic Fourier spectrum graph based on the amplitude and the adaptive threshold; The plurality of contour points are connected to obtain the binary image.

4. The smart section control method applicable to urban rail transit according to claim 3 is characterized in that: The step of obtaining a plurality of contour points of the secondary Fourier spectrum diagram based on the amplitude and the adaptive threshold comprises: If the amplitude of the pixel point is greater than or equal to the adaptive threshold, the pixel point is determined to be a contour point; If the amplitude of the pixel point is greater than half of the adaptive threshold and less than the adaptive threshold, it is determined whether the pixel point is a contour point according to the determination result of the previous pixel point adjacent to the pixel point.

5. The smart section control method applicable to urban rail transit according to claim 1 is characterized in that: Obtaining the smear length according to the quadratic Fourier spectrum diagram includes: Calculating the spacing between adjacent dark fringes in the quadratic Fourier spectrum; The smear length is calculated according to the spacing and the size of the quadratic Fourier spectrum diagram.

6. The smart section control method applicable to urban rail transit according to claim 1 is characterized in that: According to the similarity between the first crack and the second crack, selecting a fault crack from the first crack comprises: The second crack edge is uniformly selected reference points, and obtain the reference point set ,in are the coordinates of the reference point; Uniformly select the first crack edge edge points, and get the edge point set , the edge points correspond to the reference points one by one, where are the coordinates of the edge points; The similarity is calculated based on the reference point and the edge point. ,in is the gradient vector of the reference point, is the gradient vector of the edge point; According to the similarity level, a fault crack is selected from the first cracks.

7. The smart field section control method applicable to urban rail transit according to claim 1 is characterized in that: If there are multiple fault cracks, obtaining a fault detection result of the train wheelset according to characteristic information of the fault cracks includes: Acquire characteristic information of each of the fault cracks, the characteristic information including the length, width and average distance of the fault crack to adjacent fault cracks; Calculating the fault level of each fault crack according to the characteristic information and a preset possibility function; According to the fault level, the fault detection result of the train wheelset is output.

8. A smart field control system suitable for urban rail transit, characterized in that: include: The acquisition module sets up image, position and status sensors in key areas of the field to collect the operation, status and environmental data of vehicles, equipment and infrastructure in real time; The transmission module transmits the collected data to the control center in real time through a wired and wireless communication network, so that the control center can analyze the received data in real time, identify anomalies, predict faults and generate control instructions; The control module automatically controls the turnout machine, signal machine and vehicle maintenance equipment according to the control instructions, realizing the intelligent coordinated control of vehicle entry and exit, parking, maintenance and equipment operation; An acquisition module, used to acquire a first image and a second image, wherein the first image is an image of a wheelset when the train is in a high-speed running state, and the second image is an image of a wheelset when the train is departing; an identification module, configured to identify a first crack and a second crack according to the first image and the second image, wherein the first crack is a wheel pair crack when the train is in a high-speed running state, and the second crack is a wheel pair crack when the train is departing; wherein, identifying the first crack according to the first image comprises: performing image filtering and radial distortion correction on the first image to obtain a third image; performing a quadratic Fourier transform on the third image to obtain a quadratic Fourier spectrum diagram; obtaining a smear direction and a smear length according to the quadratic Fourier spectrum diagram; performing smear correction on the third image according to the smear direction and the smear length to obtain a target image; and identifying the first crack according to the target image; A selection module, configured to select a fault crack from among the first cracks according to a degree of similarity between the first crack and the second crack; The detection module is used to obtain the fault detection result of the train wheelset according to the characteristic information of the fault crack.

Images