Turnout full-stroke multi-point analysis and early warning method and system

Abstract

The invention relates to a turnout full-stroke multi-point analysis and early warning method and system, and the method comprises the steps: firstly determining the number and action sequence of switch rails and / or point rails, then arranging displacement sensors at key parts, and monitoring the displacement and time in the whole process; by setting an action stage and a sampling number, correlation analysis is performed on displacement data, and a point rail / point rail path curve is drawn; actual data are verified according to the preset displacement and time threshold value, it is ensured that the action meets the standard, if the action does not reach the standard, a point switch fault is prompted, and accurate monitoring and maintenance are achieved. According to the invention, the multi-point traction turnout is used as a whole, the action process, the synchronism of a plurality of traction points, the influence of train turnout crossing and temperature change on the repulsion switch rail can be analyzed and diagnosed, and the whole-course monitoring, analysis and diagnosis of the operation condition of the whole switch rail in the turnout switching process are realized.

Description

Technical Field

[0001] The invention relates to the technical field of railway turnouts, in particular to a method and system for full-stroke multi-point analysis and early warning of a turnout. Background Art

[0002] In the prior art, turnouts are important basic equipment for railway operation support. They have complex working conditions and are prone to failure. They are the focus and difficulty of daily maintenance and repair, and require a lot of manpower and material resources. Turnouts are set on railway tracks to guide trains from one track to another. The turnouts usually include point rails or heart rails, basic rails, and switch machines; among them:

[0003] The point rail is the movable part of the turnout, used to change the direction of the train. When the switch machine or manual operation is performed on site, the point rail will change its relative position with the base rail, thereby achieving the purpose of changing the track of the train. In a complex turnout structure, the point rail also plays a key guiding role.

[0004] The stock rail is the fixed part of the turnout, used to form the path for the train to travel and to support and guide the train to travel;

[0005] The switch machine is a switching device for the turnout, which is used to switch the turnout, lock the turnout and reflect the position of the turnout point rail; the operating rod is a part of the switch machine, which is used to connect the switch machine and the turnout, and is responsible for transmitting the power of the switch machine to move the point rail or heart rail of the turnout, thereby realizing the switching of the turnout; the detection rod is another part of the switch machine, which is used to detect whether the point rail of the turnout has been correctly switched to the predetermined position and whether it is tightly attached. The position change of the detection rod will trigger the signal system to indicate the status of the turnout.

[0006] Among the signal failures that affect the safety of high-speed train operation, turnout failures account for a large proportion. Once a failure occurs, it will cause major accidents such as collisions and rollovers. Comprehensive turnout operating condition analysis can not only prevent the occurrence of turnout-related failures in advance, but also determine and analyze the use status of the turnout, which helps to guide on-site precise maintenance, improve maintenance efficiency, and reduce maintenance difficulty.

[0007] At present, the main switch monitoring method is to use a camera to collect the gap information of the switch. The gap information can reflect whether there is a gap in the close-fitting part (the fitting part of the point rail and the base rail (or heart rail) is called close-fitting). This switch monitoring method is difficult to realize the monitoring of the repulsion opening (the state where the point rail and the base rail (or heart rail) do not fit and maintain a certain gap is called repulsion) and the full stroke of the detection rod, and it is impossible to achieve full-process monitoring, analysis and diagnosis of the entire point rail operation status during the switch switching process.

[0008] Taking the dynamic external locking turnout as an example, when using a camera to collect the gap information of the turnout, usually only the state of the point rail in the static close position is monitored. However, in the initial stage of the dynamic external locking turnout conversion, the repulsive point rail is activated first, which leads to a lack of monitoring of the changes in the repulsive point rail. Therefore, it is impossible to monitor and analyze the initial stage of the dynamic external locking turnout conversion.

[0009] Most of the current high-speed turnouts use multi-point traction turnouts. During the turnout conversion process, multiple traction points are set to jointly drive the switch's point rail and heart rail to convert; each traction point is equipped with a corresponding switch machine or other drive device, and synchronous and coordinated movement is achieved through mechanical and electrical connections; due to the higher synchronization requirements of multi-point traction turnouts, the equipment status analysis of a single point alone is obviously not comprehensive enough, and it is impossible to obtain accurate and effective data to support complex working environment analysis. For example: In the prior art, it is impossible to monitor the full-stroke synchronous operation status of the point rail (heart rail) in real time during the turnout operation, and it is impossible to analyze the curve data formed during the turnout operation process, so it is impossible to judge whether the current point rail (heart rail) action is wrong based on the curve action form, and it is impossible to provide error information in real time. In addition to analyzing the synchronization status of the point rail (heart rail), the operation status of a certain point rail (heart rail) is also lacking in current turnout technology.

[0010] The curve data generated during a turnout operation is an important basis for reflecting the turnout performance and evaluating its health status. However, due to technical limitations, most turnout systems currently do not have such data analysis capabilities.

[0011] In a multi-point traction turnout, the operating status of each point rail (heart rail) may affect the performance of the entire turnout system. Therefore, in addition to paying attention to the overall synchronization, it is also necessary to independently analyze the operating status of a single point rail.

[0012] The information disclosed in this background technology section is only intended to deepen the understanding of the overall background technology of the present invention, and should not be regarded as acknowledging or suggesting in any form that the information constitutes the prior art already known to those skilled in the art. Summary of the invention

[0013] In view of the defects existing in the prior art, the purpose of the present invention is to provide a method and system for multi-point analysis and early warning of the entire turnout stroke, which regards the multi-point traction turnout as a whole, can analyze and diagnose the action process, the synchronization of multiple traction points, the influence of train crossing and temperature changes on the rejection point rail, and realize the full monitoring, analysis and diagnosis of the entire point rail operation status during the turnout conversion process.

[0014] In order to achieve the above purpose, the technical solution adopted by the present invention is:

[0015] A method for full-stroke multi-point analysis and early warning of a turnout, characterized in that it comprises the following steps:

[0016] According to the number of point rail groups and / or point rail groups included in the switch to be monitored, the specific number and action sequence of the point rails and / or point rails are obtained;

[0017] Displacement sensors are arranged on the point rail group and / or the center rail group, and several displacement sensors are set on the detection rod and the action rod of the switch machine to monitor the displacement and the acquisition time throughout the whole journey;

[0018] Set the specific stages of a turnout point rail or center rail movement process and the corresponding point rail or center rail sampling numbers for each stage;

[0019] Based on the action sequence, the path curve of the point rail and / or the center rail is obtained by associating and analyzing the displacement values ​​and acquisition time of each sampling number;

[0020] According to the displacement threshold and time threshold of the default action process, confirm whether the actual displacement value and collection time meet the threshold setting. If the expected displacement is not reached within the specified time, it indicates that the switch machine has a fault.

[0021] On the basis of the above technical solution, high-precision displacement sensors are arranged as required at the following key positions: close fitting notch, repelling opening, pointed rail, and detection rod.

[0022] On the basis of the above technical scheme, the high-precision displacement sensor uses a laser displacement sensor to record the static value and dynamic instantaneous value of the close contact of the forward and reverse positions of the pointed rail, and to record the static value and dynamic instantaneous value of the repulsion of the forward and reverse positions of the pointed rail; a wire displacement sensor or a non-contact magnetic displacement detection sensor is used to record the displacement change value of the detection rod.

[0023] On the basis of the above technical solution, the association analysis is implemented based on the fault diagnosis model, the corresponding sampling numbers are obtained in sequence according to the action order, the corresponding relationship between the displacement and the acquisition time is determined, and then according to the displacement threshold and time threshold of the default action process, it is determined whether the component movement distance reaches the expected displacement within the specified time;

[0024] The association analysis also includes the following data static analysis:

[0025] Receive a single-point detection value x from a sensor in real time; the single-point detection value x corresponds to a close contact distance or a repulsive distance;

[0026] Pre-set single-point static data thresholds x1, x2, x3, x4 to meet x1 <x2<x3<x4;

[0027] Make the following judgment:

[0028] If the single-point detection value x is within the range of (x2, x3), it is in a safe state;

[0029] If the single point detection value x is within the range of (x1,x2] or [x3,x4), it is in the warning state;

[0030] If the single point detection value x is within the range of (-∞,x1] or [x4,∞), it is in alarm state;

[0031] When in the warning state, further check the data status of the previous five times. If all the previous five data are in the warning state, send a warning message to the front end. If only a single data is in the warning state, no warning message is sent.

[0032] The association analysis also includes the following data dynamic analysis:

[0033] Receive the single-point detection value x of the sensors at all the point rails and the center rail of the same turnout;

[0034] Analyze the single-point detection value x to obtain the displacement value of the left detection rod and the displacement value of the right detection rod, which correspond to the displacement value from the close-fitting rail to the repelling rail and the displacement value from the repelling rail to the close-fitting rail respectively;

[0035] Pre-set dynamic data thresholds: x1, x2, x3, x4, x5, x6, satisfying x1 <x2<x3<x4<x5<x6;

[0036] Preset the time threshold t,

[0037] The displacement from the close orbit to the repelling orbit is divided into five stages: the first stage: (x1, x2], the second stage: (x2, x3], the third stage: (x3, x4], the fourth stage: (x4, x5], the fifth stage: (x5, x6];

[0038] After the dynamic process starts and the delay reaches the time threshold t, check whether the displacement value from the close rail to the repelling rail is in the last stage (x5, x6]. If the displacement value x is not in the last stage (x5, x6] after time t, it is considered that the switch machine has a fault.

[0039] The fault is analyzed as follows:

[0040] By detecting the rod displacement value and the displacement threshold, it is determined in which stage the fault occurs.

[0041] On the basis of the above technical solution, it also includes collecting the ambient temperature of the turnout through a temperature sensor, evaluating the impact of the ambient temperature on the action process based on a trend prediction model, generating early warnings and maintenance tasks and prompts.

[0042] On the basis of the above technical solution, the serial port is used to transmit the data collected by the sensor in real time, and WebSocket is used to transmit the data to the front-end interface for data visualization according to the collection time, presenting the path curve of the point rail and / or the center rail.

[0043] On the basis of the above technical solutions, based on big data analysis, the collected and saved historical data are analyzed to determine the fault diagnosis model and trend prediction model for identifying abnormal behavior, judging the fault type and fault location.

[0044] A system for multi-point analysis and early warning of turnout throughout its entire travel, used to implement any of the above methods, characterized in that it comprises:

[0045] A basic data setting module, used to set the specific number and action sequence of the point rails and / or the heart rails according to the number of the point rail groups and / or the heart rail groups included in the turnout to be monitored;

[0046] A sensor communication module, used to set the displacement sensors arranged on the point rail group and / or the center rail group, used to set the displacement sensors arranged on the detection rod and the action rod of the switch machine, and used to communicate with the displacement sensors to obtain real-time collected data;

[0047] An action process setting module is used to set the specific stages of a turnout point rail or center rail action process and the point rail or center rail sampling numbers corresponding to each stage;

[0048] A path curve generation module is used to obtain the path curve of the point rail and / or the center rail by associating and analyzing the displacement values ​​and acquisition time of each sampling number based on the action sequence;

[0049] The fault warning module is used to confirm whether the actual displacement value and collection time meet the threshold settings based on the displacement threshold and time threshold of the default action process. If the expected displacement is not reached within the specified time, it will prompt that the switch machine has a fault.

[0050] On the basis of the above technical solution, the sensor communication module is also used to set the temperature sensor, and is used to communicate with the temperature sensor to obtain real-time collected ambient temperature data.

[0051] On the basis of the above technical solutions, it also includes:

[0052] A big data processing module is used to analyze the collected and saved historical data based on big data analysis, and determine the fault diagnosis model and trend prediction model for identifying abnormal behavior, determining the fault type and fault location;

[0053] Based on the call of the fault warning module, the fault diagnosis model obtains the corresponding sampling number in sequence according to the action order, determines the corresponding relationship between the displacement and the acquisition time, and then determines whether the component movement distance reaches the expected displacement within the specified time according to the displacement threshold and time threshold of the default action process;

[0054] Based on the call of the fault warning module, the trend prediction model evaluates the impact of ambient temperature on the action process, generates warnings and maintenance tasks and prompts.

[0055] The method and system for full-stroke multi-point analysis and early warning of a turnout disclosed in the present invention realizes comprehensive monitoring of the full stroke of the detection rod, the repulsion exit, and the action process of multiple traction points by innovating on the basis of the displacement gap monitoring technology, and has the following beneficial effects:

[0056] 1. It can monitor the changes of the initial action of the repulsive point rail and analyze the initial factors of the turnout conversion;

[0057] 2. With high synchronization, it can conduct comprehensive analysis on single-point equipment and obtain accurate and effective data to support complex working environment analysis;

[0058] 3. It can accurately collect the key parameters such as displacement and time of the turnout during the process of close fitting gap, repelling opening and point rail movement;

[0059] By collecting multi-point monitoring data, the system can comprehensively analyze the adverse conditions that occur during the switch switching process, such as blocking and stopping, and accurately locate the specific fault point and its cause;

[0060] 4. It can provide guidance measures for troubleshooting, and realize early maintenance through trend analysis and early warning prompt functions, thereby effectively reducing the occurrence of turnout failures.

[0061] The method and system for full-stroke multi-point analysis and early warning of a turnout described in the present invention not only improve driving safety, but also enhance the efficiency of maintenance work, reduce the labor intensity of maintenance personnel, and have important economic and social benefits. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] The present invention has the following accompanying drawings:

[0063] The accompanying drawings are used to better understand the present invention and do not constitute an improper limitation of the present invention.

[0064] Figure 1 A flow chart of embodiment 1 of a method for full-stroke multi-point analysis and early warning of a turnout according to the present invention.

[0065] Figure 2 System architecture diagram of embodiment 1 of a system for full-stroke multi-point analysis and early warning of turnouts described in the present invention.

[0066] Figure 3 Specific application scenario example 1.

[0067] Figure 4 Specific application scenario example 2.

[0068] Figure 5 Specific application scenario example three.

[0069] Figure 6 Specific application scenario example 4.

[0070] Figure 7 Specific application scenario example five. DETAILED DESCRIPTION

[0071] The present invention is further described in detail below in conjunction with the accompanying drawings. The detailed description is an explanation made in conjunction with the exemplary embodiments of the present invention, including various details of the embodiments of the present invention to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those skilled in the art that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present invention. Similarly, for the sake of clarity and conciseness, the description of well-known functions and structures is omitted in the following description.

[0072] The present invention deeply expands and integrates the traditional displacement gap monitoring technology, not only retaining the ability to accurately monitor key parts (such as close-fitting gaps and repulsive gaps), but also introducing full-stroke monitoring, multi-point traction action process correlation analysis and comprehensive diagnosis and predictive maintenance mechanisms based on big data.

[0073] like Figure 1 As shown, the method for multi-point analysis and early warning of turnout full stroke of the present invention comprises the following steps:

[0074] According to the number of point rail groups and / or point rail groups included in the switch to be monitored, the specific number and action sequence of the point rails and / or point rails are obtained;

[0075] Displacement sensors are arranged on the point rail group and / or the heart rail group, and several displacement sensors are set on the detection rod and the action rod of the switch machine to monitor the displacement amount and the acquisition time (the acquisition time of the displacement information) throughout the entire journey; the displacement sensors are arranged in a multi-point layout to capture more subtle displacement changes, especially those transition states or abnormal conditions that may be ignored in traditional single-point monitoring, thereby greatly improving the comprehensiveness and accuracy of monitoring; the full journey refers to the entire journey from the beginning to the end of the rotation of the point rail or the heart rail when the turnout is moved, and the displacement values ​​collected during the full journey combined with visualization technology can generate an intuitive path curve, and the path curve shows the full journey path formed by moving the turnout once;

[0076] Set the specific stages of a turnout point rail or center rail movement process and the corresponding point rail or center rail sampling numbers for each stage;

[0077] Based on the action sequence, the path curve of the point rail and / or the center rail is obtained by correlating and analyzing the displacement values ​​and acquisition time of each sampling number. The correlation analysis method can accurately collect basic parameters such as displacement and time. These data are important bases for evaluating the operation status of the turnout and diagnosing potential faults.

[0078] According to the displacement threshold and time threshold of the default action process, confirm whether the actual displacement value and collection time meet the threshold setting. If the expected displacement is not reached within the specified time, it will prompt that the switch machine has a fault. The comprehensive judgment of the displacement value and collection time can more accurately judge the synchronization and coordination of the turnout conversion, providing strong support for troubleshooting.

[0079] On the basis of the above technical solution, high-precision displacement sensors are arranged as required at the following key positions: close fitting notch, repelling opening, pointed rail, and detection rod.

[0080] High-precision displacement sensors can be used to continuously track and quantify turnout performance indicators, providing detailed data support for subsequent fault analysis. The high-precision displacement sensor uses a laser displacement sensor to record the static value and dynamic instantaneous value of the close contact of the point rail in the forward and reverse positions, and to record the static value and dynamic instantaneous value of the separation of the point rail in the forward and reverse positions; a wire displacement sensor or a non-contact magnetic displacement detection sensor is used to record the displacement change value of the detection rod.

[0081] In terms of data collection, in order to obtain more and more accurate data for comprehensive analysis, another optional implementation scheme is that the outdoor equipment uses high-precision and high-reliability magnetic displacement sensors, current sensors and temperature and humidity sensors. Among them:

[0082] The most important part is the displacement sensor, which is used to collect the gap value of the turnout in static and dynamic conditions. The displacement sensor adopts a non-contact magnetic scale displacement detection chip. By reading the displacement data of the magnetic scale on the detection rod, it can accurately and reliably realize the data collection of the gap gap. The detection chip reads the change of the N / S pole magnetic field when the magnetic scale moves, and converts the change of the magnetic field into displacement data through amplification and conversion, thereby realizing the detection of the required displacement.

[0083] Temperature and humidity sensors mainly rely on changes in the physical properties of materials. They usually use thermocouples or thermistors to determine the temperature by measuring changes in resistance or voltage. Humidity sensors use the property that the dielectric constant changes with humidity and obtain the humidity value by measuring changes in capacitance.

[0084] One of the signs of the switch's transition from static to dynamic is determined by the current signal generated by the current sensor, and each transition corresponds to a current change.

[0085] After collecting the notch displacement, temperature, humidity and current signals, these data are packaged into frames and transmitted to the data analysis center through the serial port in the form of data frames. After receiving the data frames, the data analysis center parses the data frames, extracts the data and performs the final analysis.

[0086] On the basis of the above technical solution, the association analysis is implemented based on the fault diagnosis model, and the corresponding sampling numbers are obtained in sequence according to the action order to determine the correspondence between the displacement and the acquisition time. Then, based on the displacement threshold and time threshold of the default action process, it is determined whether the component movement distance reaches the expected displacement within the specified time.

[0087] Exemplarily, the association analysis also includes the following data static analysis, which preferably uses the displacement values ​​of the left detection rod and the right detection rod as analysis objects:

[0088] Receive a single-point detection value x from a sensor in real time; the single-point detection value x corresponds to a close contact distance or a repulsive distance;

[0089] Pre-set single-point static data thresholds x1, x2, x3, x4 to meet x1 <x2<x3<x4;

[0090] Make the following judgment:

[0091] If the single-point detection value x is within the range of (x2, x3), it is in a safe state;

[0092] If the single point detection value x is within the range of (x1,x2] or [x3,x4), it is in the warning state;

[0093] If the single point detection value x is within the range of (-∞,x1] or [x4,∞), it is in alarm state;

[0094] When in the warning state, the data status of the previous five times is further checked. If the data of the previous five times are all in the warning state, a warning message is sent to the front end. If only a single data is in the warning state, no warning message is sent.

[0095] Exemplarily, the association analysis also includes the following data dynamic analysis, which preferably uses the displacement value of the left detection rod and the displacement value of the right detection rod as the analysis object:

[0096] Receive the single-point detection value x of the sensors at all the point rails and the center rail of the same turnout;

[0097] Analyze the single-point detection value x to obtain the displacement value of the left detection rod and the displacement value of the right detection rod, which correspond to the displacement value from the close-fitting rail to the repelling rail and the displacement value from the repelling rail to the close-fitting rail respectively;

[0098] Pre-set dynamic data thresholds: x1, x2, x3, x4, x5, x6, satisfying x1 <x2<x3<x4<x5<x6;

[0099] Preset the time threshold t,

[0100] The displacement from the close orbit to the repelling orbit is divided into five stages: the first stage: (x1, x2], the second stage: (x2, x3], the third stage: (x3, x4], the fourth stage: (x4, x5], the fifth stage: (x5, x6];

[0101] After the dynamic process starts and the delay reaches the time threshold t, check whether the displacement value from the close rail to the repelling rail is in the last stage (x5, x6]. If the displacement value x is not in the last stage (x5, x6) after time t, it is considered that the switch machine has a fault; the delay reaching the time threshold t can be regarded as the end of the component movement and entering a stable state;

[0102] The fault is analyzed as follows:

[0103] By detecting the rod displacement value and the displacement threshold, it is determined in which stage the fault occurs; for example, if x∈(x4,x5], a fault occurs in the fourth stage.

[0104] As one of the optional implementation plans, for the four switch machines that complete an action synchronously, the corresponding eight displacement values ​​are subjected to static data analysis respectively, and then the alarm information is counted, and then the data is dynamically analyzed, and the most serious point is selected to be confirmed as the fault point. That is: for the four switch machines involved in an action, for each switch machine (including two displacement values), if any detection rod triggers an alarm, the switch machine is marked as a potential fault point. If all switch machines do not trigger an alarm, multi-point analysis is not performed. If there is a potential fault point, by comparing the alarm severity of each fault point (for example, judging by the fault stage, the closer to the initial stage, the more serious the fault), the most serious point is selected as the final fault point.

[0105] On the basis of the above technical solution, it also includes collecting the ambient temperature of the turnout through a temperature sensor, evaluating the impact of the ambient temperature on the action process based on a trend prediction model, generating early warnings and maintenance tasks and prompts.

[0106] On the basis of the above technical solution, the serial port is used to transmit the data collected by the sensor in real time, and WebSocket is used to transmit the data to the front-end interface for data visualization according to the collection time, presenting the path curve of the point rail and / or the center rail.

[0107] The data collected by the sensors deployed on site are transmitted to the analysis system in the form of data frames through the serial port, and then the data frames are extracted and analyzed. WebSocket is a network communication protocol that provides the ability to perform full-duplex communication on a single TCP connection.

[0108] On the basis of the above technical solutions, based on big data analysis, the collected and saved historical data are analyzed to determine the fault diagnosis model and trend prediction model for identifying abnormal behavior, judging the fault type and fault location.

[0109] By mining and pattern recognition of historical data, fault diagnosis models and trend prediction models can automatically identify abnormal behavior patterns and predict possible fault trends. This predictive maintenance capability makes maintenance work more proactive and efficient, effectively reducing the probability of failure and maintenance costs.

[0110] Through trend analysis and prompt functions, the system can issue early warning signals before a fault occurs, reminding maintenance personnel to take timely measures to intervene. This proactive maintenance strategy not only reduces the impact of faults on transportation efficiency, but also extends the service life of turnout equipment and improves the reliability and safety of the overall system.

[0111] like Figure 2 As shown, this embodiment provides a system for multi-point analysis and early warning of a turnout full stroke based on the aforementioned method for multi-point analysis and early warning of a turnout full stroke, including:

[0112] A basic data setting module, used to set the specific number and action sequence of the point rails and / or the heart rails according to the number of the point rail groups and / or the heart rail groups included in the turnout to be monitored;

[0113] A sensor communication module, used to set the displacement sensors arranged on the point rail group and / or the center rail group, used to set the displacement sensors arranged on the detection rod and the action rod of the switch machine, and used to communicate with the displacement sensors to obtain real-time collected data;

[0114] An action process setting module is used to set the specific stages of a turnout point rail or center rail action process and the point rail or center rail sampling numbers corresponding to each stage;

[0115] A path curve generation module is used to obtain the path curve of the point rail and / or the center rail by associating and analyzing the displacement values ​​and acquisition time of each sampling number based on the action sequence;

[0116] The fault warning module is used to confirm whether the actual displacement value and collection time meet the threshold settings based on the displacement threshold and time threshold of the default action process. If the expected displacement is not reached within the specified time, it will prompt that the switch machine has a fault.

[0117] On the basis of the above technical solution, the sensor communication module is also used to set the temperature sensor, and is used to communicate with the temperature sensor to obtain real-time collected ambient temperature data.

[0118] On the basis of the above technical solutions, it also includes:

[0119] A big data processing module is used to analyze the collected and saved historical data based on big data analysis, and determine the fault diagnosis model and trend prediction model for identifying abnormal behavior, determining the fault type and fault location;

[0120] Based on the call of the fault warning module, the fault diagnosis model obtains the corresponding sampling number in sequence according to the action order, determines the corresponding relationship between the displacement and the acquisition time, and then determines whether the component movement distance reaches the expected displacement within the specified time according to the displacement threshold and time threshold of the default action process;

[0121] Based on the call of the fault warning module, the trend prediction model evaluates the impact of ambient temperature on the action process, generates warnings and maintenance tasks and prompts.

[0122] Taking the turnout point rail as an example, the implementation details of the present invention are described in detail below in combination with specific application scenarios.

[0123] 1. Fault scenario simulation design:

[0124] The turnout is mainly divided into a point rail group and a heart rail group. In this embodiment, the point rail group includes point rail No. 1 (marked as J1 in the figure), point rail No. 2 (marked as J2 in the figure) and point rail No. 3 (marked as J3 in the figure), and the heart rail group includes heart rail No. 1. In general, the action process of the point rail is more complicated than that of the heart rail. The point rail involves the linkage problem of different rails. Therefore, the simulated fault scenario is mainly designed and analyzed for the point rail group.

[0125] 2. Five processes and failure scenarios of turnout point rail movement:

[0126] A turnout point rail action can be divided into five processes, namely, the unlocking process inside the switch machine, the unlocking process of the rejection point rail, the unlocking process of the close contact point rail, the synchronous conversion process, the original rejection point rail close in place and the original close contact point rail continue to move to complete the locking process.

[0127] Through actual operation on site, five fault scenarios of switch point rail were simulated. Figure 3-Figure 7These five fault scenarios are shown, and data analysis is also performed on these five fault scenarios. According to the data at the end of the action, it is inferred which specific process has an error. Specific faults include: the repulsion rail is not unlocked or the conversion is poor, the close rail is not unlocked or the conversion is poor, the synchronous conversion is poor and the locking is abnormal. After the analysis is completed, the results are transmitted to the front end for alarm display.

[0128] 3. Specific process of simulating fault scenarios:

[0129] 3.1 Data Collection:

[0130] The displacement sensors and time acquisition devices arranged on site collect the displacement and time of the turnout point rail group (heart rail), and the collected data are sent to the data center through the serial port for processing.

[0131] It mainly involves displacement sensors and time acquisition devices, which are used to monitor the close contact state between the point rail, the heart rail, etc. and the changes in the repulsion opening. This is crucial to ensure that the point rail can be accurately positioned and close to the base rail during the switchover process, and can be smoothly repelled when needed. It is precisely arranged at the key positions of the close contact gap, repulsion opening, point rail and the entire travel path of the turnout.

[0132] 3.2 Data analysis:

[0133] After receiving the data, the data center processes it immediately, for example:

[0134] The serial port data received by the data center is binary data of 0 / 1, which is received by bytes and converted into hexadecimal storage. After receiving a frame of data, the frame header and frame tail are checked first, and the CRC polynomial is used to verify the security of the frame. Then the hexadecimal data is converted according to the protocol to obtain the following information: frame type, frame content length, switch machine number, fault bit, sampling number, close gap gap, repulsive gap gap, left detection rod displacement value and right detection rod displacement value.

[0135] The parameters are passed to the data transmission thread through the signals and slots owned by QT. The data transmission thread analyzes the values, assigns color and other features to the adhesion value and repulsion value of this frame, and then passes them to the database and to the front-end interface through WebSocket. After receiving the data, the front-end interface adds it to the array and then refreshes the interface at intervals of 1 second to display the data.

[0136] 3.3 Data Analysis

[0137] Analyze and check the data values of the preset time threshold for this action process. By comparing and analyzing with previous normal data and experience, if the data at this time node is found to be abnormal, infer forward which process of this switch action the value should normally appear in, and finally obtain the fault point. Finally, mark this action process as an abnormal state, and send the data and analysis results to the front-end interface through WebSocket for curve and alarm display. For example:

[0138] The system is provided with displacement thresholds and time thresholds for the default action process, and the default values can be manually corrected. The displacement thresholds are arranged in ascending (descending) order from small (large) to large (small) in five stages, and the time threshold is the time point when a dynamic process movement should have ended;

[0139] After each dynamic process starts and reaches the time threshold, monitor the data in the received dynamic data frame. If it is not within the range of the last stage, the switch machine fails. At this time, judge which stage interval the fault detection rod displacement value is in by comparing the displacement value of the detection rod with the displacement threshold, and then determine the fault in this interval.

[0140] Exemplarily, for static data, the data is divided into single-point analysis and multi-point linkage analysis.

[0141] After obtaining the contact distance and repulsive force distance from the received data through processing, compare them with the pre-set static frame data thresholds. There is a set of thresholds x1, x2, x3, x4 (x1 < x2 < x3 < x4) for the repulsive force distance and the contact distance respectively. Taking the contact distance as an example, when the contact value x belongs to (x2, x3), the contact position is in a safe state; when the contact value belongs to (x1, x2] or [x3, x4), the contact position is in a warning state; when the contact value belongs to (-∞, x1] or [x4, ∞), the contact position is in an alarm state.

[0142] Perform such single-point analysis on each received data. If it is in the alarm state, directly send it to the front-end for alarm. If it is in the warning state, perform multi-point linkage to check whether the states of the previous five times are all in the warning state. If they are all in the warning state, then send a warning message to the front-end. If only this time is in the warning state, it may be slightly interfered, and the next measurement may be a normal value. Therefore, do not send a warning message to the front-end to reduce unnecessary push.

[0143] Exemplarily, for dynamic data, single-point analysis, multi-point linkage analysis and fault diagnosis models are adopted.

[0144] After processing the received data, all of it is sent to the front-end interface for data visualization. For a single switch operation, there are data for three switch points and one stock rail. The data for each rail is further divided into the displacement value of the left detection rod and the displacement value of the right detection rod, which respectively correspond to the displacement value from the closed rail to the separated rail and the displacement value from the separated rail to the closed rail. Taking the displacement value from the closed rail to the separated rail, since it is from the closed rail to the separated rail, the displacement value changes from small to large. The closed rail is divided into 5 stages, and the fifth stage is the locking stage, that is, the stage where the operation is completed. There are six thresholds x1, x2, x3, x4, x5, x6 (x1 < x2 < x3 < x4 < x5 < x6) for these five stages, and the five stages are respectively (x1, x2], (x2, x3], (x3, x4], (x4, x5], (x5, x6], and there is also a time threshold t.

[0145] After t seconds from the start of the dynamic process, under normal conditions, the switch machine moves into place, that is, the displacement value from the closed rail to the separated rail is within the range of (x5, x6]. The fault diagnosis model monitors the data in the received dynamic data frame after the time threshold t is reached for each dynamic process. If the displacement value x is not within the range of the last stage, that is, (x5, x6], then the switch machine has a fault. At this time, it is judged in which stage interval the displacement value of the detection rod is by comparing the displacement value of the detection rod with the displacement threshold, and then it is determined that it is a fault in that interval. For example, if x ∈ (x4, x5], it means that there is a fault in the fourth stage. For the 4 switch machines in one operation, single-point analysis is performed on each of the 8 displacement values respectively for alarm, and then the alarm information is statistically analyzed. Through multi-point analysis, the most serious point is selected as the fault point, because a fault point will slightly affect other non-fault points, resulting in non-fault points not reaching the position and generating alarm information.

[0146] 3.4. Example of the path curve corresponding to the full stroke path

[0147] Figure 3 、 Figure 4 、 Figure 5 、 Figure 6 、 Figure 7 They respectively correspond to five different application scenarios of the switch machine (switch point group) for the switch in actual operation on-site, and show five different types of fault alarms of the switch machine, where:

[0148] Figures 3 to 6 They respectively represent four different errors simulated for switch point No. 1, and the simulations are respectively: the separated rail cannot be released or has poor conversion, the closed rail cannot be released or has poor conversion, synchronous conversion is poor, and locking is abnormal.

[0149] Figure 7 Represents the starting curve of the fault position, simulating the conversion of the fault position to the normal position. The orange pop-up window in the figure indicates that J1 and J2 are converted from the previous fault position to the normal position.

[0150] The horizontal axis in the figure represents time, and the vertical axis represents the magnitude of the collected point rail repulsion exit and close contact value, that is, the displacement value. It can be understood that in order to clearly distinguish which point rail has an error, the fault simulation is performed using point rail No. 1 as an example in the figure, and the fault simulation can also be performed on point rail No. 2 and point rail No. 3.

[0151] Figure 3 The simulated fault is that the No. 1 point rail is not disconnected or the conversion is poor. The staff simulates the fault of the point rail group at the turnout site. After setting the fault point, they move the turnout and observe whether the front-end interface issues a corresponding alarm.

[0152] It can be seen from the end of the curve that the action process of this point rail is abnormal. The simulated fault is that the point rail No. 1 is not unlocked or the conversion is poor. The alarm result on the interface shows that the unlocking conversion of the original repelling point rail No. 1 is abnormal, and the unlocking conversion of the original repelling point rail No. 1 causes the abnormal unlocking conversion of the original close-fitting point rail No. 2, the abnormal synchronization process of the two point rails of point rail No. 2, and the abnormal unlocking conversion of the original close-fitting point rail No. 3. Synchronization errors of the point rail group. Since the point rail group is synchronously operated, they will affect each other. The curve shown in the figure is also in an abnormal state, which is basically consistent with the results of the previous simulation of the unsolvable or poor conversion of the point rail No. 1.

[0153] Figure 4 The simulated fault is that the No. 1 close-fitting rail of the switch rail is not released or the conversion is poor. The staff simulates the fault of the switch rail group at the switch site. After setting the fault point, they move the switch and observe whether the front-end interface issues a corresponding alarm.

[0154] By observing the end of the curve, it can be seen that the actions of the No. 1, No. 2 and No. 3 point rails are abnormal. The simulated fault is that the No. 1 close-fitting rail is not unlocked or the conversion is poor. Through the alarm interface, it can be seen that the unlocking conversion of the original close-fitting point rail of the No. 1 point rail is abnormal. The abnormal unlocking conversion of the original close-fitting point rail of the No. 1 point rail causes the abnormal synchronization process of the No. 2 point rails and the abnormal locking process of the original close-fitting point rail of the No. 3 point rail. Since the point rail group is operated as a whole, there is a high probability that a problem with one point rail will cause abnormalities in other point rails. The green pop-up window indicates that the previous fault has been restored to normal.

[0155] Figure 5 The simulated fault was poor synchronous conversion of point rail No. 1. The staff simulated the fault of the point rail group at the turnout site. After setting the fault point, they moved the turnout and observed whether the front-end interface issued a corresponding alarm.

[0156] By observing the end of the curve, it can be seen that the actions of point rails No. 1 and No. 2 are abnormal. The simulated fault is the poor synchronization conversion of point rail No. 1. Through the alarm interface, it can be seen that the synchronization process of the two point rails of point rail No. 1 is abnormal. The abnormal synchronization process of the two point rails of point rail No. 1 causes the abnormal locking process of the original repulsion point rail of point rail No. 2. The fault position of point rail No. 1 is basically consistent with the simulated fault. The error prompt of point rail No. 2 may be caused by the fault of point rail No. 1.

[0157] Figure 6 The simulated fault was abnormal locking of point rail No. 1. The staff simulated the fault of the point rail group at the turnout site. After setting the fault point, they moved the turnout and observed whether the front-end interface issued a corresponding alarm.

[0158] By observing the end of the curve, it can be seen that the action of Point Rail No. 1 is abnormal. The simulated fault is that the locking of Point Rail No. 1 is abnormal. Through the alarm interface, it can be seen that the process of Point Rail No. 1 repelling the point rail and continuing to complete the locking is abnormal. The process of Point Rail No. 1 repelling the point rail and continuing to complete the locking is abnormal. There is a probability that the synchronization process of the two point rails of Point Rail No. 1 will be abnormal. The fault position of Point Rail No. 1 is basically consistent with the simulated fault.

[0159] The switch needs to be moved twice during one switch movement. If a fault occurs during the first movement, the system will need to give a prompt when moving the switch for the second time because the switch machine movement curve is not the standard curve, the switch machine movement starting position is abnormal, and the movement time is abnormal. This means that a fault occurred during the last switch movement, which is indicated by an orange pop-up window.

[0160] The contents not described in detail in this specification belong to the prior art known to professional and technical personnel in this field.

[0161] The above description is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiment. Any equivalent modifications or changes made by those skilled in the art based on the contents disclosed by the present invention should be included in the protection scope recorded in the claims.

Claims

1. A method for multi-point analysis and early warning of turnout full stroke, characterized in that: The steps include: According to the number of point rail groups and / or point rail groups included in the switch to be monitored, the specific number and action sequence of the point rails and / or point rails are obtained; Displacement sensors are arranged on the point rail group and / or the center rail group, and several displacement sensors are set on the detection rod and the action rod of the switch machine to monitor the displacement and the acquisition time throughout the whole journey; Set the specific stages of a turnout point rail or center rail movement process and the corresponding point rail or center rail sampling numbers for each stage; Based on the action sequence, the path curve of the point rail and / or the center rail is obtained by associating and analyzing the displacement values ​​and acquisition time of each sampling number; According to the displacement threshold and time threshold of the default action process, confirm whether the actual displacement value and collection time meet the threshold setting. If the expected displacement is not reached within the specified time, it indicates that the switch machine has a fault.

2. A method for full-stroke multi-point analysis and early warning of a turnout as claimed in claim 1, characterized in that: High-precision displacement sensors are arranged as required at the following key positions: close fitting gap, repelling opening, point rail, and detection rod.

3. A method for full-stroke multi-point analysis and early warning of a turnout as claimed in claim 2, characterized in that: The high-precision displacement sensor uses a laser displacement sensor to record the static value and dynamic instantaneous value of the close contact of the forward and reverse positions of the pointed rail, and to record the static value and dynamic instantaneous value of the repulsion of the forward and reverse positions of the pointed rail; a wire displacement sensor or a non-contact magnetic displacement detection sensor is used to record the displacement change value of the detection rod.

4. A method for full-stroke multi-point analysis and early warning of a turnout as claimed in claim 1, characterized in that: The correlation analysis is implemented based on the fault diagnosis model, and the corresponding sampling numbers are obtained in sequence according to the action sequence, the corresponding relationship between the displacement and the acquisition time is determined, and then according to the displacement threshold and time threshold of the default action process, it is determined whether the component movement distance reaches the expected displacement within the specified time; The association analysis also includes the following data static analysis: Receive a single-point detection value x from a sensor in real time; the single-point detection value x corresponds to a close contact distance or a repulsive distance; Pre-set single-point static data thresholds x1, x2, x3, x4 to meet x1 <x2<x3<x4; Make the following judgment: If the single-point detection value x is within the range of (x2, x3), it is in a safe state; If the single point detection value x is within the range of (x1,x2] or [x3,x4), it is in the warning state; If the single point detection value x is within the range of (-∞,x1] or [x4,∞), it is in alarm state; When in the warning state, further check the data status of the previous five times. If all the previous five data are in the warning state, send a warning message to the front end. If only a single data is in the warning state, no warning message is sent. The association analysis also includes the following data dynamic analysis: Receive the single-point detection value x of the sensors at all the point rails and the center rail of the same turnout; Analyze the single-point detection value x to obtain the displacement value of the left detection rod and the displacement value of the right detection rod, which correspond to the displacement value from the close-fitting rail to the repelling rail and the displacement value from the repelling rail to the close-fitting rail respectively; Pre-set dynamic data thresholds: x1, x2, x3, x4, x5, x6, satisfying x1 <x2<x3<x4<x5<x6; Preset the time threshold t, The displacement from the close orbit to the repelling orbit is divided into five stages: the first stage: (x1, x2], the second stage: (x2, x3], the third stage: (x3, x4], the fourth stage: (x4, x5], the fifth stage: (x5, x6]; After the dynamic process starts and the delay reaches the time threshold t, check whether the displacement value from the close rail to the repelling rail is in the last stage (x5, x6]. If the displacement value x is not in the last stage (x5, x6] after time t, it is considered that the switch machine has a fault. The fault is analyzed as follows: By detecting the rod displacement value and the displacement threshold, it is determined in which stage interval the fault occurs.

5. A method for full-stroke multi-point analysis and early warning of a turnout as claimed in claim 1, characterized in that: It also includes collecting the ambient temperature of the turnout through a temperature sensor, evaluating the impact of the ambient temperature on the movement process based on a trend prediction model, and generating early warnings and maintenance tasks and prompts.

6. A method for full-stroke multi-point analysis and early warning of a turnout as claimed in claim 1, characterized in that: The serial port is used to transmit the data collected by the sensor in real time, and WebSocket is used to transmit the data to the front-end interface according to the collection time for data visualization, showing the path curve of the point rail and / or the center rail.

7. A method for full-stroke multi-point analysis and early warning of a turnout as claimed in claim 1, characterized in that: Based on big data analysis, the collected and saved historical data are analyzed to determine the fault diagnosis model and trend prediction model for identifying abnormal behavior, judging the fault type and fault location.

8. A system for multi-point analysis and early warning of turnout throughout its entire travel, used to implement the method described in any one of claims 1 to 7, characterized in that: include: A basic data setting module, used to set the specific number and action sequence of the point rails and / or the heart rails according to the number of the point rail groups and / or the heart rail groups included in the switch to be monitored; A sensor communication module, used to set the displacement sensors arranged on the point rail group and / or the center rail group, used to set the displacement sensors arranged on the detection rod and the action rod of the switch machine, and used to communicate with the displacement sensors to obtain real-time collected data; An action process setting module is used to set the specific stages of a turnout point rail or center rail action process and the point rail or center rail sampling numbers corresponding to each stage; A path curve generation module is used to obtain the path curve of the point rail and / or the center rail by associating and analyzing the displacement values ​​and acquisition time of each sampling number based on the action sequence; The fault warning module is used to confirm whether the actual displacement value and collection time meet the threshold settings based on the displacement threshold and time threshold of the default action process. If the expected displacement is not reached within the specified time, it will prompt that the switch machine has a fault.

9. A turnout full-stroke multi-point analysis and early warning system as claimed in claim 8, characterized in that: The sensor communication module is also used to set the temperature sensor and to communicate with the temperature sensor to obtain real-time collected ambient temperature data.

10. A turnout full-stroke multi-point analysis and early warning system as claimed in claim 8, characterized in that: Also includes: A big data processing module is used to analyze the collected and saved historical data based on big data analysis, and determine the fault diagnosis model and trend prediction model for identifying abnormal behavior, determining the fault type and fault location; Based on the call of the fault warning module, the fault diagnosis model obtains the corresponding sampling number in sequence according to the action order, determines the corresponding relationship between the displacement and the acquisition time, and then determines whether the component movement distance reaches the expected displacement within the specified time according to the displacement threshold and time threshold of the default action process; Based on the call of the fault warning module, the trend prediction model evaluates the impact of ambient temperature on the action process, generates warnings and maintenance tasks and prompts.

Images