Pantograph state intelligent detection method

Abstract

The invention belongs to the field of railway vehicle pantograph overhauling, and discloses an intelligent pantograph state detection method. The method comprises the steps of carrying out fault detection on a pneumatic pantograph lifting mechanism of the pantograph, carrying out fault identification on a pantograph fastener and carrying out anomaly detection on a pantograph head of the pantograph. According to the pantograph state intelligent detection method provided by the invention, state detection and parameter measurement are carried out on the pantograph by using the existing vehicle equipment, planned repair is changed into state repair, automatic intelligent maintenance is realized, the labor cost is reduced, the efficiency is improved, the workload of maintenance and repair is reduced, the vehicle operation utilization rate is improved, and the vehicle operation cost is reduced. The vehicle real-time detection data is more efficient and accurate, faults can be found in time and early warned, problems can be handled in the first time, and the vehicle operation safety can be improved.

Description

Technical Field

[0001] The invention relates to the field of rail vehicle pantograph maintenance, and in particular to a pantograph state intelligent detection method. Background Art

[0002] As a high-voltage power supply device for locomotives and vehicles, pantographs play a key role in vehicle operation safety. Pantographs are usually installed on the roof of locomotives or EMUs. When the pantograph is raised, its slide contacts the contact wire, thereby introducing current into the vehicle and providing power for the locomotive. It is a key device for electric locomotives and urban rail vehicles to obtain current from the contact wire.

[0003] like Figure 1 As shown in the figure, the pantograph is mainly composed of modular structures such as chassis, lower arm, upper arm, bow head, carbon slide plate, pull rod, balance rod, drive airbag, solenoid valve and air path, damper, insulator, pneumatic control box, etc., which realize the up and down movement of the bow head and connect the contact network to provide high-voltage electricity for the vehicle. The maintenance of the pantograph has many points and a large workload. At present, the maintenance operation mode of the pantograph of rail transit vehicles is mainly manual maintenance. Since the pantograph is installed on the roof of the locomotive or EMU, the maintenance personnel need to work at a height, which has a high risk factor. Manual maintenance is time-consuming and labor-intensive and there is a risk of personal injury. In addition, manual maintenance can only be carried out at a set time point or when an obvious fault occurs. It is impossible to give early warning and intervention in time when the problem occurs, which poses certain safety hazards. Since the pantograph equipment itself does not have an automatic detection function, and the pantograph-network detection system currently configured on newly built vehicles mainly evaluates and warns the health status of the pantograph-network relationship, it does not give early warning for the maintenance of the pantograph itself.

[0004] Therefore, there is an urgent need in this field to study an intelligent pantograph status detection method to reduce the workload of maintenance personnel and improve the timeliness of maintenance. Summary of the invention

[0005] In view of the above problems, the present invention provides a method for intelligently detecting the state of a pantograph.

[0006] According to one aspect of the present invention, a method for intelligently detecting a pantograph state is provided, the method comprising performing fault detection on a pneumatic pantograph raising mechanism of the pantograph, performing fault identification on a pantograph fastener, and performing abnormality detection on a pantograph head, wherein all the detection items are mandatory inspection items of the pantograph maintenance work, and the fault detection of the pneumatic pantograph raising mechanism comprises: Collecting a total wind pressure value sent by an electronic brake control unit of a brake system, comparing the total wind pressure value with a set threshold, and generating a first valid signal when the total wind pressure value is greater than the set threshold, otherwise generating a first invalid signal; Collecting state data of the pantograph raising solenoid valve on the pantograph raising air circuit board, and generating a second valid signal when the state data indicates that the pantograph raising solenoid valve is in a closed state, otherwise generating a second invalid signal; collecting pantograph raising data from a train control and monitoring system regarding whether the pantograph is in place, and generating a third valid signal when the pantograph raising data indicates that the pantograph is in place, otherwise generating a third invalid signal; When the first valid signal, the second valid signal and the third valid signal are detected, it is determined that the air pipeline, the driving airbag, the bow lifting air circuit plate, the pneumatic bow lifting device and the auxiliary bow lifting device of the pneumatic bow lifting mechanism are fault-free. When at least one of the first invalid signal, the second invalid signal and the third invalid signal is detected, it is determined that the air pipeline, the driving airbag, the bow lifting air circuit plate, the pneumatic bow lifting device or the auxiliary bow lifting device is faulty.

[0007] According to one embodiment of the present invention, the method also includes determining a specific component where a fault occurs based on a combination of one or more of the first valid signal, the second valid signal, and the third valid signal and one or more of the first invalid signal, the second invalid signal, and the third invalid signal.

[0008] According to an embodiment of the present invention, when the first valid signal, the second valid signal and the third invalid signal are detected, an alarm is issued that an air line or a driving airbag has an air leak.

[0009] According to an embodiment of the present invention, when the first valid signal, the second invalid signal and the third invalid signal are detected, an alarm of a lifting solenoid valve failure is output.

[0010] According to one embodiment of the present invention, when the second valid signal, the first invalid signal and the third invalid signal are detected, an alarm of failure of the auxiliary air supply unit and the pneumatic lifting device or leakage of the air pipeline is output.

[0011] According to one embodiment of the present invention, identifying a fault of the pantograph fastener includes: Using image recognition technology to determine whether a fastener is present at a set location; When it is determined that the fastener is not present at the set position, outputting a missing alarm; When it is determined that the fastener exists at the set position, identifying the size of the fastener; comparing the identified size to a size threshold; When the size is greater than the size threshold, using a first image recognition method to perform offset detection on the fastener, and determining whether to output a loosening warning based on the offset detection; When the size is smaller than the size threshold, a second image recognition method is used to perform position detection on the fastener, and based on the position detection, it is determined whether to output a looseness warning.

[0012] According to an embodiment of the present invention, performing offset detection on the fastener using a first image recognition method and determining whether to output a loosening warning based on the offset detection includes: capturing an image of the fastener in a normal installed state, analyzing the image and determining an initial position of a set feature of the fastener; capturing an image of the fastener in real time, analyzing the image and determining a current position of the set feature of the fastener; The offset between the current position and the initial position is determined, and when the offset exceeds a set angle range, a looseness warning is output.

[0013] According to an embodiment of the present invention, the method of detecting the position of the fastener using the second image recognition method and determining whether to output a loosening warning based on the position detection includes: Capturing two images at different angles when the fastener is in a normal installation state, and identifying the initial position of the fastener in the image at each angle based on the two images at different angles; Tracking the position of the fastener at two angles in real time to determine the current position; The changes between the current position of the fasteners and their respective initial positions at two angles are determined respectively, and when the change of at least one of them exceeds a set range, a loosening warning is output.

[0014] According to one embodiment of the present invention, performing abnormality detection on the pantograph head includes: Determining a normal posture of the bow head and a standard characteristic distribution vector of the bow head; Collecting image information of the bow head in real time, and identifying the current posture and current feature distribution vector of the bow head; Determine the angle deviation between the current posture of the bow head and the normal posture, when the deviation is greater than a set angle, it is a tilt state, monitor a first duration of the tilt state, and when the first duration exceeds a first time threshold, trigger a bow head abnormality alarm; and / or Determine the mismatch rate between the current feature distribution vector and the standard feature distribution vector, when the mismatch rate is higher than a first mismatch threshold, it is a mismatch state, monitor the second duration of the mismatch state, when the second duration exceeds a second time threshold, trigger a bow head abnormality alarm.

[0015] According to one embodiment of the present invention, the abnormality detection also includes: detecting the acceleration of the pantograph, and when the acceleration is greater than a set acceleration threshold, the mismatch rate is greater than a second mismatch threshold and the duration exceeds a third time threshold, a bow head abnormality alarm is triggered.

[0016] Due to the adoption of the above technical scheme, the pantograph state intelligent detection method provided by the present invention utilizes the existing vehicle equipment to perform state detection and parameter measurement on the pantograph, transforms the planned maintenance into state maintenance, realizes automatic intelligent maintenance, reduces labor costs and improves efficiency, reduces the workload of maintenance and repair, improves vehicle operation utilization, and reduces vehicle operation costs. The real-time vehicle detection data is more efficient and accurate, can detect faults in time and issue warnings, and handle problems as soon as possible, which can improve vehicle operation safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The accompanying drawings are used to provide a further understanding of the present disclosure and constitute a part of the specification. Together with the following specific embodiments, they are used to explain the present disclosure but do not constitute a limitation of the present disclosure. In the accompanying drawings: Figure 1 A schematic diagram of a pantograph structure according to the prior art; Figure 2 is a flow chart of a pantograph state intelligent detection method according to an embodiment of the present invention; Figure 3 is a flow chart of a pantograph state intelligent detection method according to another embodiment of the present invention; Figure 4 The present invention is a flowchart of a method for intelligently detecting a pantograph state according to another embodiment of the present invention. DETAILED DESCRIPTION

[0018] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with specific embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0019] The terms "including" and "having" and any variations thereof in the specification of the present invention and the above-mentioned drawings are intended to cover non-exclusive inclusions; the terms "first", "second", etc. in the specification of the present invention and the claims or the above-mentioned drawings are used to distinguish different objects rather than to describe a specific order. "Multiple" means two or more, unless otherwise clearly and specifically defined.

[0020] In the specification of the present invention and the above-mentioned drawings, when an element is referred to as being "fixed to" or "mounted on" or "disposed on" or "connected to" another element, it may be directly or indirectly located on the other element. For example, when an element is referred to as being "connected to" another element, it may be directly or indirectly connected to the other element.

[0021] In addition, reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present invention. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0022] In the following description of the present invention, the term "longitudinal" refers to the direction connecting the front and rear of the vehicle, which can also be understood as the length direction of the vehicle. The term "lateral" refers to the width direction of the vehicle. The terms "front", "rear", "up", "down", "top" and "bottom" are all based on the state of the vehicle when it is traveling forward.

[0023] The present invention collects image data and data information of various on-board systems through the vehicle pantograph detection system, analyzes and processes the data, and outputs warning data for pantograph status inspection, thereby realizing a method of transforming manual inspection into digital intelligent inspection, reducing manual inspection workload and lowering vehicle operation and maintenance costs.

[0024] One of the purposes of the present invention is to provide an intelligent method for detecting the state of a pantograph, the method comprising fault detection of a pneumatic pantograph lifting mechanism of the pantograph, fault identification of a pantograph fastener, and abnormality detection of a pantograph head.

[0025] Figure 2 The overall flow chart of the pantograph state intelligent detection method according to an embodiment of the present invention is shown. In this embodiment, the pneumatic pantograph lifting mechanism of the pantograph is fault-detected using existing vehicle equipment. The pneumatic pantograph lifting mechanism includes an air pipeline, a driving airbag, a pantograph lifting air circuit board, a pneumatic pantograph lifting device and an auxiliary pantograph lifting device.

[0026] Pantograph lifting is the process of filling the pantograph lifting airbag with compressed air through the pneumatic pantograph lifting device, thereby driving the pantograph mechanical device to lift the pantograph head and contact the power grid. Since the pneumatic pantograph needs to be raised by driving the airbag, the pneumatic pantograph needs to be checked for air tightness of the pneumatic pantograph lifting mechanism every two weeks to ensure the normal function of the pantograph lifting and lowering.

[0027] At present, the air tightness test uses soapy water to test all pipelines to see if there are bubbles. If no bubbles are generated, it is judged that there is no leakage. Since the pantograph air pipeline is connected from the ceiling pantograph to the air path control device on the end wall of the vehicle, and then to the vehicle chassis pantograph lifting device, the distance is long, the air pipeline is about 20 meters, the detection workload is large, and the pipeline installation position is hidden, there are many surrounding equipment, and the operation is difficult, so this method makes it difficult to find some leaks and easy to miss.

[0028] In order to reduce the workload of airtightness inspection and increase the accuracy and timeliness of inspection, the pantograph state intelligent detection method of the present invention includes fault detection of the pneumatic pantograph lifting mechanism of the pantograph. The pneumatic pantograph lifting mechanism fault detection method of this embodiment includes: Step S101: Collect the total wind pressure value sent by the electronic brake control unit of the brake system, compare the total wind pressure value with a set threshold, and generate a first valid signal when the total wind pressure value is greater than the set threshold, otherwise generate a first invalid signal; Step S102: collecting state data of the pantograph raising solenoid valve of the pantograph raising air circuit board, and generating a second valid signal when the state data indicates that the pantograph raising solenoid valve is in a closed state, otherwise generating a second invalid signal; Step S103: collecting pantograph raising data from a train control and monitoring system (TCMS system) on whether the pantograph is in place, and generating a third valid signal when the pantograph raising data indicates that the pantograph is in place, otherwise generating a third invalid signal; Step S104: When the first valid signal, the second valid signal and the third valid signal are detected, it is determined that the air pipeline, the driving airbag, the bow lifting air circuit plate, the pneumatic bow lifting device and the auxiliary bow lifting device of the pneumatic bow lifting mechanism are fault-free; when at least one of the first invalid signal, the second invalid signal and the third invalid signal is detected, it is determined that the air pipeline, the driving airbag, the bow lifting air circuit plate, the pneumatic bow lifting device or the auxiliary bow lifting device is faulty.

[0029] In step S101, the total wind pressure value (MR total wind pressure value) issued by the electronic brake control unit of the brake system is first collected, and the MR total wind pressure value is compared with the set threshold. For example, the train control and monitoring system (TCMS system) can collect the "MR total wind pressure value" data issued by the electronic brake control unit of the brake system, and compare the data with the set threshold. For example, the total wind pressure threshold can be set to 0.6MPa, and if the TCMS system determines that the MR total wind pressure value is greater than 0.6MPa, a first valid signal is generated. Optionally, if the TCMS system determines that the MR total wind pressure value is not greater than 0.6MPa, a first invalid signal is generated. The pressure threshold can be adjusted according to the specific vehicle model and vehicle operating conditions, and is not limited to the specific value given in this embodiment.

[0030] In step S102, the state data of the pantograph lifting solenoid valve of the pantograph lifting air circuit board is collected, and when the state data indicates that the pantograph lifting solenoid valve is in a closed state, a second valid signal is generated. For example, the TCMS system can collect the "pantograph lifting solenoid valve" data of the pantograph lifting air circuit board, and judge whether the solenoid valve is in a closed state or an open state based on the data. If the solenoid valve is in a closed state, a second valid signal is generated. Alternatively, if the solenoid valve is in an open state, a second invalid signal is generated.

[0031] In step S103, the pantograph raising data on whether the pantograph is in place is collected, and when the pantograph raising data indicates that the pantograph is in place, a third valid signal is generated. For example, the TCMS system can collect the "pantograph raising in place" signal sent by the pantograph control box to the TCMS system. If the TCMS system receives the "pantograph raising in place" signal, a third valid signal is generated. Alternatively, if the "pantograph raising in place" signal is not received, a third invalid signal is generated.

[0032] In step S104, when the first valid signal, the second valid signal and the third valid signal are detected, it is determined that the air pipeline, the driving airbag, the bow raising air circuit board, the pneumatic bow raising device and the auxiliary bow raising device of the pneumatic bow raising mechanism are fault-free; otherwise, it is determined that the air pipeline, the driving airbag, the bow raising air circuit board, the pneumatic bow raising device or the auxiliary bow raising device are faulty, and an early warning is output. For example, the TCMS system can confirm whether all the generated signals are valid signals. If all three signals are valid signals, it is determined that there is no fault and no early warning is output. If at least one signal is an invalid signal, an early warning is output.

[0033] This method can provide fault warning in time so that the fault can be checked and handled in time.

[0034] Optionally, in some embodiments, the specific faulty component may be determined based on a combination of one or more of the first valid signal, the second valid signal, and the third valid signal and one or more of the first invalid signal, the second invalid signal, and the third invalid signal.

[0035] For example, in some embodiments, when the first valid signal, the second valid signal and the third invalid signal are detected, an alarm is issued that the air pipeline or the driving airbag has air leakage. The air pipelines involved include the pantograph air pipeline, the roof air pipeline, the end wall air pipeline, the air pipeline connecting the pantograph air pipeline plate to the chassis auxiliary air supply unit, and the joints in the pipeline.

[0036] Optionally, in some embodiments, when the first invalid signal, the second valid signal, and the third invalid signal are detected, an alarm is outputted indicating that the auxiliary air supply unit (i.e., the auxiliary pantograph lifting device) and the pneumatic pantograph lifting device are faulty or the air pipeline is leaking. This detection result can eliminate the fault of the pantograph lifting solenoid valve of the pantograph lifting air circuit board and reduce the maintenance workload.

[0037] Optionally, in some embodiments, when the first valid signal, the second invalid signal and the third invalid signal are detected, an alarm of a failure of the lifting solenoid valve of the lifting gas circuit board is output.

[0038] By comprehensively analyzing the presence or absence of signals from the three subsystems or the types of signals, the specific component that has failed can be identified. When an alarm is triggered, operators can conduct further targeted inspections on related components based on the type of alarm, reducing the workload of manual maintenance and improving maintenance efficiency and timeliness.

[0039] Optionally, the pantograph status intelligent detection method of the present invention further provides a solution for fault identification of pantograph fasteners.

[0040] In the past, the fasteners installed on the pantograph were checked manually visually to see if they were firm, not loose, not rusted, with clear anti-loosening lines, and not offset. However, this inspection method cannot detect the loss and loosening of fasteners in time, and is prone to missed detection. To this end, the pantograph status intelligent detection method of the present invention also provides a method for identifying faults of pantograph fasteners. The pantograph fastener detection performs image analysis on the fasteners in the visible parts based on the pantograph video data collected by the pantograph-network detection system, identifies the missing or loose fasteners, and issues corresponding alarms. Figure 3 A specific example of a relevant detection scheme is shown.

[0041] In step S201, image recognition technology is used to determine whether there is a fastener at a set position. For example, the pantograph fasteners can be predefined by partitioning, and information such as the position and parameters of the pantograph fasteners can be stored in the system. The pantograph network detection system is used to collect the video data of the pantograph, and the area corresponding to the stored pantograph fastener position (also referred to as the target area) in the video data is identified as the fastener feature. In this process, the existing object classification and detection algorithms and models can be used to identify the identification features and environmental features related to the fasteners to be detected in the image to assist in determining the position of the fasteners. After the target area of ​​the fastener in the image is determined, the predetermined feature extraction is performed on the area. For example, when the fastener is a bolt, the outer contour of the bolt head can be used as a comparison feature. When the predetermined feature cannot be extracted in the target area, it means that there is no fastener at the position and the fastener has fallen off. When the predetermined feature can be extracted in the target area, it means that there is a fastener at the position.

[0042] In step S202, when it is determined that there is no fastener at the set position, a loss alarm is output, which may indicate the position and parameter information of the lost fastener.

[0043] In step S203, when it is determined that there is a fastener at the set position, the size of the fastener is identified based on image analysis. For example, the parameters on the fastener can be identified by using the image, and the size of the corresponding fastener can be determined according to the parameter information marked on the fastener, or the size of the fastener can be determined according to the fastener parameter information associated with the position stored in the system, or the size of the fastener can be determined according to the area ratio of the fastener in the captured image, etc. In short, any known method related to image recognition can be used to determine the size of the fastener based on the image.

[0044] In step S204, the identified size is compared with the size threshold. Since the smaller fasteners have a larger error when using images to locate the fastener identification features such as the anti-loosening wire. Therefore, different looseness detection methods are required for large fasteners and small fasteners. To this end, the size threshold is set to classify the fasteners first. For example, when the fastener is a bolt, the thread diameter of 8mm is used as the size threshold for classifying the fastener category. However, the specific threshold range can be set according to the type of fastener and the resolution of the image acquisition device.

[0045] In step S205, when the identified fastener size is greater than the size threshold, the first image recognition method is used to perform offset detection on the fastener, and based on the offset detection, it is determined whether to output a loosening warning. For example, the anti-loosening wire set on the fastener can be used as a comparison feature. When the fastener is in a normal installation state (that is, when the most recent maintenance has just been completed), an image of the fastener is captured to determine the initial position of the anti-loosening wire. In subsequent detections, the pantograph video data collected by the bow-net detection system is analyzed to determine the current position of the anti-loosening wire. The current position of the anti-loosening wire is compared with the initial position of the anti-loosening wire, and when the offset angle is greater than a set angle of, for example, 15 degrees, a loosening warning is output for early warning reminder.

[0046] In step S206, when the size of the fastener is less than the size threshold, the position of the fastener is detected by a second image recognition method, and it is determined whether to output a loosening warning based on the position detection. In some embodiments, images of two different angles when the fastener is in a normal installation state (i.e., just completed in the most recent maintenance) are collected. Taking the bolt as an example, the image can be collected from a first angle facing the bolt head bearing surface (i.e., the upper surface of the bolt head) and a second angle along the bolt radial direction. The first angle photographs the bolt head bearing surface, and the bolt head bearing surface is used as the position tracking object. The second angle photographs the side of the bolt, which can monitor the depth of the bolt rod screwed into the screw hole. The change in depth can indicate the looseness of the bolt, and the bolt head bearing surface can be used as the position tracking object. The first initial position is identified based on the initial image of the bolt at the first angle, and the second initial position is identified based on the initial image of the bolt at the second angle. The position of the bolt is tracked in real time at two angles, for example, the image or video frame of the bolt is continuously captured, and the current position of the bolt in each frame is determined. The current position identified at the first angle is referred to as the first current position, and the current position identified at the second angle is referred to as the second current position. Compare the first current position with the first initial position. If the position change exceeds 30% of the outer diameter of the bolt thread, which is about two thread pitches, it indicates that the bolt is loose or broken, and a loosening warning is output. Compare the second current position with the second initial position. If the position change exceeds two thread pitch heights, it indicates that the bolt is loose or broken, and a loosening warning is output. In addition, a comprehensive judgment can be made based on the results of the two angles to improve the accuracy of the detection. For example, if the position change between the first current position and the first initial position exceeds 20% of the outer diameter of the bolt thread, and the bolt head height changes by more than two thread pitches, a loosening warning is issued.

[0047] By adopting the pantograph state intelligent detection method of the present invention to detect fasteners, loosening and missing of fasteners can be discovered in time, thereby improving the safety of train operation.

[0048] Optionally, the pantograph status intelligent detection method of the present invention also provides a method for detecting abnormalities in the bow head of the pantograph, which is used to determine the dynamic abnormality of the bow head, and trigger an alarm if any condition of tilting, flipping, or deformation occurs. Figure 4 A flow chart for detecting abnormality of a pantograph head is shown.

[0049] In step S301, the normal posture of the bow head and the standard characteristic distribution vector of the bow head are determined; In step S302, image information of the bow head is collected in real time to identify the current posture of the bow head and the current characteristic distribution vector of the bow head; In step S303, the angle deviation between the current posture of the bow head and the normal posture is determined. When the deviation is greater than a set angle, it is a tilt state. A first duration of the tilt state is monitored. When the first duration exceeds a first time threshold, an abnormal bow head alarm is triggered; and / or In step S304, the mismatch rate between the current feature distribution vector and the standard feature distribution vector is determined. When the mismatch rate is higher than a first mismatch threshold, it is a mismatch state. The second duration of the mismatch state is monitored. When the second duration exceeds a second time threshold, a bow head abnormality alarm is triggered.

[0050] In step S301, the normal posture of the pantograph and the standard feature distribution vector of the pantograph are determined. For example, the license plate number image of the vehicle to be inspected can be obtained, and the optical character recognition method can be used to recognize the license plate number of the detected vehicle, and the reference contour corresponding to the vehicle to be inspected in the pantograph reference database is retrieved as the reference data according to the license plate number. Specifically, the YOLOv3 algorithm is first used to train the license plate number target, and after obtaining the target characters of the license plate number, the support machine is used to recognize the specific characters to obtain the license plate number of the vehicle to be inspected. Carry out license plate number recognition, obtain the standard pantograph model and reference data, thereby obtaining the normal posture data of the pantograph and the standard feature distribution vector representing the normal shape of the pantograph head.

[0051] In step S302, the image information of the pantograph is collected in real time, and the current posture of the pantograph and the current characteristic distribution vector of the pantograph are identified. For example, the three-dimensional point cloud data of the pantograph to be detected can be obtained through a 3D imaging device, and then the regional positioning of the pantograph head is performed to determine the position of the pantograph area in the data, and the position of the pantograph head can be determined by the spatial coordinate height. After the regional positioning of the pantograph head is determined, the position of each part of the pantograph head in the three-dimensional world coordinate system of the data collected by the 3D imaging device can be obtained. Based on the position data of the pantograph head in the three-dimensional world coordinate system, its current posture and current characteristic distribution vector can be determined.

[0052] In step S303, the angle deviation between the current posture of the bow head and the normal posture is determined. When the deviation is greater than a set angle, it is a tilted state. The first duration of the tilted state is monitored. When the first duration exceeds a first time threshold, an abnormal alarm of the bow head is triggered. For example, the current posture of the bow head determined in step S302 can be compared with the normal posture of the bow head determined in step 301 to determine the angle between the two. If the angle is greater than 5°, the duration of the bow head in this tilted state is further recorded to avoid a short tilt caused by vibration or the like being misreported as an abnormality. When the duration is greater than a time threshold of, for example, 3 seconds, it is determined that the bow head is abnormal and an abnormal alarm of the bow head is triggered.

[0053] In step S304, the mismatch rate between the current feature distribution vector and the standard feature distribution vector is determined. When the mismatch rate is higher than a first mismatch threshold (e.g., 25%), it is a mismatch state. The second duration of the mismatch state is monitored. When the second duration exceeds a second time threshold (e.g., 3 seconds), a bow head abnormality alarm is triggered.

[0054] Optionally, in some embodiments, the acceleration detection data provided by the acceleration sensor configured by the pantograph detection system can also be combined to analyze and judge the abnormality of the bow head. The acceleration sensor is installed on the outside of the pantograph bow head. The impact vibration and abnormal collision of the pantograph during vehicle operation can be reflected in the acceleration fluctuation value monitored by the acceleration sensor. When the acceleration value suddenly increases, the pantograph or vehicle may have a large speed change, and may have a collision or bow scraping, ram horn collision and other faults, so this value plays an important role in judging the abnormality of the pantograph bow head. At the same time, combined with the real-time video data collected by the pantograph detection system, the real-time feature vector of the bow head is analyzed with the standard feature distribution vector. When the mismatch rate and the tilt angle exceed a certain threshold, the integrity does not match the standard model, and it can be concluded that the pantograph morphology is abnormal and cannot work normally. There are safety hazards and timely warnings should be issued. Therefore, the method can also include detecting the acceleration of the pantograph. When the acceleration is greater than the set acceleration threshold (for example, 60g), the mismatch rate is higher than the second mismatch threshold (for example, 50%) and the duration exceeds the third time threshold (for example, 3 seconds), the bow head abnormal alarm is triggered.

[0055] Optionally, in some embodiments, the method may further include detecting the integrity of the pantograph bow head. For example, the current shape of the bow head may be obtained based on the real-time collected image information analysis, and the shape may be compared with the standard shape in the pantograph model. If the integrity similarity is less than 98% and the duration exceeds a set value such as 3 seconds, a bow head abnormality alarm is triggered.

[0056] The pantograph status intelligent detection method of the present invention transforms manual maintenance into intelligent maintenance, reduces labor costs, improves efficiency, and improves accuracy; transforms planned maintenance into status maintenance, reduces the pantograph maintenance workload by 25%, and simplifies maintenance work; reduces maintenance time, improves vehicle operation utilization, and reduces vehicle operation costs; can perform real-time fault warnings, detect faults in a timely manner and handle them, and improve vehicle operation safety.

[0057] The above-mentioned embodiments only express the implementation methods of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the present invention. It should be pointed out that for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention.

Claims

1. A pantograph state intelligent detection method, characterized in that: The method comprises performing fault detection on a pneumatic pantograph raising mechanism of the pantograph, performing fault identification on a pantograph fastener, and performing abnormality detection on a pantograph head, wherein the fault detection on the pneumatic pantograph raising mechanism comprises: Collecting a total wind pressure value sent by an electronic brake control unit of a brake system, comparing the total wind pressure value with a set threshold, and generating a first valid signal when the total wind pressure value is greater than the set threshold, otherwise generating a first invalid signal; Collecting state data of the pantograph raising solenoid valve on the pantograph raising air circuit board, and generating a second valid signal when the state data indicates that the pantograph raising solenoid valve is in a closed state, otherwise generating a second invalid signal; collecting pantograph raising data from a train control and monitoring system regarding whether the pantograph is in place, and generating a third valid signal when the pantograph raising data indicates that the pantograph is in place, otherwise generating a third invalid signal; When the first valid signal, the second valid signal and the third valid signal are detected, it is determined that the air pipeline, the driving airbag, the bow lifting air circuit plate, the pneumatic bow lifting device and the auxiliary bow lifting device of the pneumatic bow lifting mechanism are fault-free. When at least one of the first invalid signal, the second invalid signal and the third invalid signal is detected, it is determined that the air pipeline, the driving airbag, the bow lifting air circuit plate, the pneumatic bow lifting device or the auxiliary bow lifting device is faulty.

2. The pantograph state intelligent detection method according to claim 1, characterized in that: The method further includes determining a specific component that has failed based on a combination of one or more of the first valid signal, the second valid signal, and the third valid signal and one or more of the first invalid signal, the second invalid signal, and the third invalid signal.

3. The pantograph state intelligent detection method according to claim 2, characterized in that: When the first valid signal, the second valid signal and the third invalid signal are detected, an alarm is issued indicating that there is air leakage in the air pipeline or the driving airbag.

4. The pantograph state intelligent detection method according to claim 2, characterized in that: When the first valid signal, the second invalid signal, and the third invalid signal are detected, an alarm of a lift solenoid valve failure is output.

5. The pantograph state intelligent detection method according to claim 2, characterized in that: When the second valid signal, the first invalid signal and the third invalid signal are detected, an alarm is outputted indicating a failure of the auxiliary air supply unit and the pneumatic lifting device or an air pipeline leakage.

6. The pantograph state intelligent detection method according to claim 1, characterized in that: Fault identification of the pantograph fastener includes: Using image recognition technology to determine whether a fastener is present at a set location; When it is determined that the fastener is not present at the set position, outputting a missing alarm; When it is determined that the fastener exists at the set position, identifying the size of the fastener; comparing the identified size to a size threshold; When the size is greater than the size threshold, using a first image recognition method to perform offset detection on the fastener, and determining whether to output a loosening warning based on the offset detection; When the size is smaller than the size threshold, a second image recognition method is used to perform position detection on the fastener, and based on the position detection, it is determined whether to output a looseness warning.

7. The pantograph state intelligent detection method according to claim 6, characterized in that: The method of performing offset detection on the fastener using the first image recognition method and determining whether to output a loosening warning based on the offset detection includes: capturing an image of the fastener in a normal installed state, analyzing the image and determining an initial position of a set feature of the fastener; capturing an image of the fastener in real time, analyzing the image and determining a current position of the set feature of the fastener; The offset between the current position and the initial position is determined, and when the offset exceeds a set angle range, a looseness warning is output.

8. The pantograph state intelligent detection method according to claim 6, characterized in that: The method of detecting the position of the fastener using the second image recognition method and determining whether to output a loosening warning based on the position detection includes: Capturing two images at different angles when the fastener is in a normal installation state, and identifying the initial position of the fastener in the image at each angle based on the two images at different angles; Tracking the position of the fastener at two angles in real time to determine the current position; The changes between the current position of the fasteners and their respective initial positions at two angles are determined respectively, and when the change of at least one of them exceeds a set range, a loosening warning is output.

9. The pantograph state intelligent detection method according to claim 1, characterized in that: The abnormality detection of the pantograph head includes: Determining a normal posture of the bow head and a standard characteristic distribution vector of the bow head; Collecting image information of the bow head in real time, and identifying the current posture and current feature distribution vector of the bow head; Determine the angle deviation between the current posture of the bow head and the normal posture, when the deviation is greater than a set angle, it is a tilt state, monitor a first duration of the tilt state, and when the first duration exceeds a first time threshold, trigger a bow head abnormality alarm; and / or Determine the mismatch rate between the current feature distribution vector and the standard feature distribution vector, when the mismatch rate is higher than a first mismatch threshold, it is a mismatch state, monitor the second duration of the mismatch state, when the second duration exceeds a second time threshold, trigger a bow head abnormality alarm.

10. The pantograph state intelligent detection method according to claim 9, characterized in that: The abnormality detection also includes: detecting the acceleration of the pantograph, and when the acceleration is greater than a set acceleration threshold, the mismatch rate is greater than a second mismatch threshold and the duration exceeds a third time threshold, a pantograph abnormality alarm is triggered.

Images