Railway bolt loosening state monitoring and early warning method based on separated IMU

Abstract

The present application relates to the technical field of monitoring methods, in particular to a railway bolt loosening state monitoring and early warning method based on a split IMU, comprising the following steps: step one: based on the installation position and installation data of connecting bolts, a plurality of connecting bolts are screened out as key bolts; step two: a monitoring device is arranged on the key bolts, the monitoring device comprises a monitoring module and a communication module, the monitoring module collects three-axis acceleration data, and the communication module uploads data information; step three: the relative offset angle of the key bolts is analyzed, when the relative offset angle is greater than a preset angle offset threshold, a loosening early warning signal is generated, and the key bolt is marked as a problem bolt; step four: when a specific train passes, the early warning information is obtained and uploaded to the cloud. Through screening of key bolts and setting of monitoring devices, and through specific trains to recover the monitored abnormal data, the bolt split loosening monitoring along the track can be realized.

Description

Technical Field

[0001] This invention relates to the field of monitoring methods, specifically to a method for monitoring and early warning of loose railway bolts based on a split IMU. Background Technology

[0002] Rail connectors, commonly known as fishplates, are an important component of steel rails, primarily used to connect two independent rails, allowing multiple rails to be joined to form a complete track. If loose bolts on the rail connectors go undetected, it can lead to increased gaps and misalignment at the rail joints. When a train passes at high speed, this can range from exacerbating wheel-rail impact and reducing passenger comfort to potentially causing serious safety accidents such as rail breakage and train derailment.

[0003] The mainstream method for monitoring the loosening status of bolts on rail connecting plates in existing technologies is manual inspection. Although this method can check whether all connecting bolts are loose one by one, it takes a long time and is inefficient. Based on this, some tracks use monitoring equipment with triaxial accelerometers as the main component. Although this method can replace manual bolt loosening monitoring, it is limited by the large number of bolts and the length of the track, making it difficult to cover all applications. It can only be used on some tracks and has certain limitations. Summary of the Invention

[0004] In view of the above-mentioned shortcomings of the existing technology, the present invention provides a method for monitoring and early warning of loose railway bolts based on a split IMU, which can effectively solve the problem that the existing monitoring equipment is difficult to perform split full-line monitoring for long-distance track lines.

[0005] To achieve the above objectives, the present invention provides the following technical solution: This invention provides a method for monitoring and early warning of loose railway bolts based on a split IMU, comprising the following steps: Step 1: Record all the fastening bolts corresponding to the rail connecting plates on the railway track as connecting bolts. Based on the installation position and installation data of the connecting bolts, select multiple connecting bolts and record them as key bolts. Step 2: Install monitoring equipment on key bolts. The monitoring equipment includes a monitoring module and a communication module. The monitoring module collects triaxial acceleration data and uploads the data information based on the communication module. The monitoring cycle of the monitoring equipment and the multiple monitoring nodes within the monitoring cycle are determined based on the periodic passage time of multiple passing trains. The multiple monitoring nodes constitute a monitoring node set, and triaxial acceleration data is collected once at each monitoring node and recorded as the target acceleration set. Step 3: Obtain the target acceleration set collected by the monitoring module, analyze the target acceleration set to obtain the relative offset angle of the key bolt, and when the relative offset angle is greater than the preset angle offset threshold, generate a loosening warning signal and mark the key bolt as a problem bolt. The problem bolt packages its own information with the loosening warning signal into a warning message, stores the warning message and waits for it to be uploaded. Step 4: Select a specific train from multiple passing trains. The specific train is the one with the shortest departure cycle. Determine the upload time interval based on the specific train's route data and vehicle data. Instruct the monitoring equipment to upload the warning information within the upload time interval. When the specific train passes, the warning information is obtained and uploaded to the cloud.

[0006] Furthermore, the key bolt screening process is as follows: Obtain track data along the railway track, record multiple connecting bolts on the same rail connecting plate as a bolt combination, record the track position of the bolt combination as the combination coordinate, analyze and calculate the track impact assessment value based on the track data corresponding to the combination coordinate, and select multiple key bolt combinations based on the track impact assessment value and a preset distribution distance threshold. Based on the stress test data of the rail connecting plate, the stress data of different connecting bolts in the same critical bolt combination are analyzed. Combined with the data of the connecting bolts themselves and the installation data, the installation impact assessment value is calculated. The connecting bolt with the largest installation impact assessment value in the same critical bolt combination is selected and recorded as the critical bolt.

[0007] Furthermore, the calculation process for the track impact assessment value is as follows: Obtain the speed limit value when the train passes through the combined coordinates. Simulate the train passing through the combined coordinates at the speed limit value and perform force analysis to obtain the comprehensive acceleration. Record the comprehensive acceleration as the track impact assessment value, where: Treating the track as a plane, with the plane angle depending on the gradient and superelevation, and treating the train as a point mass, when the track corresponding to the combined coordinates is a curved segment, analyze the acceleration components of gravitational acceleration and centrifugal acceleration in the direction perpendicular to the track plane; when the track corresponding to the combined coordinates is a straight segment, analyze the acceleration component of gravitational acceleration in the direction perpendicular to the track plane, and denote the acceleration component in the direction perpendicular to the track plane as the composite acceleration.

[0008] Furthermore, the screening process for key bolt combinations is as follows: Draw the track line and mark all bolt combinations on the track line. There is a preset track influence threshold. Bolt combinations that are greater than or equal to the track influence threshold are recorded as the first candidate combinations. When the distance between any two adjacent first candidate combinations is greater than or equal to the distribution distance threshold, multiple bolt combinations between these two first candidate combinations are selected as the second bolt combinations. The first bolt combinations and the second bolt combinations are collectively recorded as the key bolt combinations.

[0009] Furthermore, select the option that meets the following conditions: Condition 1: The spacing between multiple second candidate combinations is less than the distribution distance threshold; Condition 2: The number of the second candidate combinations is the minimum value that satisfies Condition 1.

[0010] Furthermore, the calculation process for the installation impact assessment value is as follows: There are multiple preset type numbers, each type number corresponds to a rail connection plate type. Each type of rail connection plate is assigned a set of screw hole numbers. Each screw hole number in the screw hole number set corresponds to a connecting bolt. The screw hole number and type number corresponding to any connecting bolt are recorded as the plate coordinates of that connecting bolt. Train passing tests were conducted on different types of rail connecting plates. During the test, the stress change around each bolt hole was measured and the alternating stress amplitude was calculated. A limit amplitude was preset, and the installation impact assessment value was obtained by dividing the alternating stress amplitude by the limit amplitude.

[0011] Furthermore, the process for determining monitoring nodes is as follows: The location of the key bolt on the rail is recorded as the target location. All trains that pass the target location during their journey are recorded as passing trains. The departure cycle of the passing trains is obtained, and the longest cycle in the departure cycle is recorded as the monitoring cycle. The time when each passing train passes the target location within the monitoring cycle is recorded as the passing time. Based on the passing time and the vehicle body parameters of the passing trains, the passing influence range is obtained. Each passing influence range is represented within the monitoring cycle. The portion of the non-passing influence range within the monitoring cycle is recorded as the target range. The midpoint of each target range is recorded as the monitoring node.

[0012] Furthermore, the process of obtaining the affected area is as follows: Any passing train is designated as the target train. The minimum speed of the target train when it passes the target position is designated as the target speed. The length of the target train is divided by the target speed to obtain the base time. A time interval is established with the passing time of the target train as the start point of the interval and the base time as the length of the interval, designated as the first influence interval. Obtain the full load weight of the target train, calculate the vibration influence distance using the vibration propagation formula, divide the vibration influence distance by the target speed to obtain the vibration influence duration, preset the departure deviation duration corresponding to the target train, calculate the sum of the vibration influence duration and the departure deviation duration to obtain the comprehensive influence duration, subtract the departure deviation duration from the starting point of the first influence interval, and add the departure deviation duration to the ending point of the first influence interval to obtain the traversed influence interval.

[0013] Furthermore, the relative offset angle analysis process is as follows: After tightening the key bolts, the monitoring equipment data is zeroed out. The first set of target accelerations collected after zeroing is recorded as the reference set, the most recently collected set of target accelerations is recorded as the comparison set, the vertical plane perpendicular to the bolt axis is recorded as the reference plane, and the rectangular coordinate system corresponding to the reference plane is recorded as the reference coordinate system. In the reference coordinate system, the Y-axis acceleration vector and Z-axis acceleration vector corresponding to the reference set are represented as the initial vector combination, and the Y-axis acceleration vector and Z-axis acceleration vector corresponding to the comparison set are represented as the comparison vector combination. The angle between the Y-axis acceleration vectors in the initial vector combination and the comparison vector combination is calculated and recorded as the relative offset angle.

[0014] Furthermore, the process for determining the upload time range is as follows: The passing time of a specific train is recorded as the anchor point time, the length of the specific train is recorded as the specific length, the minimum speed of the specific train passing the monitoring equipment is recorded as the specific speed, the specific length is divided by the specific speed to obtain the first duration, and a time interval is constructed with the anchor point time as the interval start point and the first duration as the interval length, recorded as the basic upload interval. The departure deviation duration is subtracted from the start point of the basic upload interval, and the departure deviation duration is added to the end point of the basic upload interval to obtain the upload time interval.

[0015] The technical solution provided by this invention has the following advantages compared with the known prior art: 1. This invention, by setting up a battery-powered bolt loosening status monitoring device, enables bolt status monitoring to break free from reliance on manual inspection, achieving offline monitoring and proactive reporting. Furthermore, it filters a large number of connecting bolts to identify critical bolts significantly affected by train travel. By focusing on these critical bolts, a more flexible and accurate monitoring strategy can be provided while reducing the number of monitoring devices required, achieving separate monitoring across the entire track. Moreover, utilizing trains traveling on the track as data collection carriers covers the entire track. Compared to setting up fixed collection devices near the track, this method offers higher data acquisition efficiency and is more suitable for long-distance, multi-point, separate monitoring devices, significantly improving data utilization efficiency.

[0016] 2. The functions of the monitoring equipment of this invention are all timed to adapt to the rail transit behavior with obvious periodic patterns. This setting allows the monitoring equipment to be in standby and hibernation state most of the time, thereby obtaining a longer working time. The monitoring equipment is only awakened and performs the corresponding main functions when timed. Thus, the monitoring equipment only consumes power to perform its main functions when needed. Compared with the long-term online monitoring of similar monitoring equipment in the prior art, it saves more power and is more suitable for monitoring equipment that uses batteries as a power source. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0018] Figure 1 This is a diagram illustrating the method steps of the present invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0020] The present invention will be further described below with reference to embodiments.

[0021] See Figure 1 A method for monitoring and early warning of loosening of railway bolts based on a split IMU is applied to the monitoring of bolt status on rail connecting plates.

[0022] It should be noted that rail connecting plates, commonly known as fishplates, are long and slender with multiple bolt holes. In actual use, the connecting plate is placed tightly against both sides of the rail web, and bolts are then passed through the connecting plate and the rail to clamp the joint of the two rails, thus connecting them into a continuous section of railway track. When a train passes over the rail joint, it generates enormous shear forces and bending moments, causing vibrations and impacts to the connecting plate, which can lead to loosening of the bolts at that location.

[0023] At least the following steps are included: Step 1: Mark all the fastening bolts corresponding to the rail connecting plates on the railway track as connecting bolts. The connecting bolts are used to fix the rail connecting plates so that they clamp the ends of the two rail sections. Based on the installation position and installation data of the connecting bolts, select several connecting bolts as key bolts. Use the key bolts as the main monitoring objects for status detection, and thus indirectly monitor the locking status of all connecting bolts along the railway line by the loosening status of the key bolts.

[0024] It should be noted that the track laid along the track line consists of multiple individual rails, with adjacent rails connected by rail connecting plates. Each rail connecting plate is typically secured by 4 or 6 sets of bolts. Therefore, a section of track usually requires a large number of bolt connections. However, it is obviously unreasonable to install monitoring equipment for every bolt (this would require a large investment in equipment). In this case, it is necessary to select representative bolts for monitoring based on the actual working conditions to reflect the bolt loosening data at multiple key points along the entire track.

[0025] Specifically, the selection process for key bolts is as follows: The system acquires trackside data for railway tracks, including curve radius, gradient, and superelevation. Based on this data, track parameter information at any location on the railway track can be obtained. Multiple connecting bolts on the same rail connecting plate are denoted as a bolt combination, and the track position of the bolt combination is denoted as the combination coordinate. The track impact assessment value is calculated based on the trackside data corresponding to the combination coordinate to evaluate the impact of the trackside data on the fastening stability of the connecting bolts. Based on the track impact assessment value and a preset distribution distance threshold, multiple key bolt combinations are selected. Based on the stress test data of the rail connecting plate, the stress data of different connecting bolts in the same key bolt combination are analyzed. Combined with the data of the connecting bolts themselves and the installation data, the installation impact assessment value is calculated to evaluate the impact of the installation-related parameters of the connecting bolts on the fastening stability of the connecting bolts. Each connecting bolt corresponds to an impact assessment value. The connecting bolt with the largest installation impact assessment value in the same key bolt combination is selected and recorded as the key bolt. Each key bolt combination corresponds to one key bolt.

[0026] More specifically, the calculation process for the orbital impact assessment value is as follows: Obtain the speed limit value when the train passes through the combined coordinates. Simulate the train passing through the combined coordinates at the speed limit value and perform force analysis to obtain the comprehensive acceleration. Record the comprehensive acceleration as the track influence assessment value, which is the acceleration corresponding to the resultant force exerted by the train wheels on the rails. Where: Treating the track as a plane, with the plane angle depending on the gradient and superelevation, and treating the train as a point mass, when the track corresponding to the combined coordinates is a curved segment, analyze the acceleration components of gravitational acceleration and centrifugal acceleration in the direction perpendicular to the track plane; when the track corresponding to the combined coordinates is a straight segment, analyze the acceleration component of gravitational acceleration in the direction perpendicular to the track plane, and denote the acceleration component in the direction perpendicular to the track plane as the composite acceleration.

[0027] It should be noted that the track impact assessment values ​​indirectly reflect the influence on the connecting bolts at different track locations. This influence mainly stems from changes in the track's own parameters. In other words, the same train exerts different impacts on the track when passing through different locations. By analyzing the differences in the magnitude of the force exerted by the train on the track, characteristic points of track stress can be identified. This allows for the generation of early warnings or management and maintenance based on the bolt status at these characteristic points. In the process of analyzing and obtaining comprehensive acceleration, a simplified train-track model facilitates simpler stress analysis, thereby quickly obtaining the required track impact assessment values.

[0028] More specifically, the screening process for key bolt combinations is as follows: The track line is drawn and all bolt combinations are marked on the track line. A track influence threshold is preset (in a specific embodiment, the value is 1.1g, where g is the acceleration due to gravity). Bolt combinations greater than or equal to the track influence threshold are recorded as first candidate combinations. When the distance between any two adjacent first candidate combinations is greater than or equal to the distribution distance threshold, multiple bolt combinations between these two first candidate combinations are selected as second bolt combinations. The first bolt combinations and the second bolt combinations are collectively recorded as key bolt combinations, selected based on the following conditions: Condition 1: The spacing between multiple second candidate combinations is less than the distribution distance threshold; Condition 2: The number of the second candidate combinations is the minimum value that satisfies Condition 1.

[0029] More specifically, the calculation process for the installation impact assessment value is as follows: There are multiple preset type numbers, each type number corresponds to a rail connection plate type (the bolt distribution is different on different types of rail connection plates). Each type of rail connection plate is assigned a set of screw hole numbers, and each screw hole number in the screw hole number set corresponds to a connecting bolt. The screw hole number and type number corresponding to any connecting bolt are obtained and recorded as the plate coordinates of that connecting bolt. Train passing tests were conducted on different types of rail connecting plates. During the test, the stress change around each bolt hole was measured (by setting up a triaxial strain rosette around the bolt hole) and the alternating stress amplitude was calculated. A limit amplitude was preset (in a specific embodiment, the value was taken as the symmetrical cyclic fatigue limit of the connecting plate material, which can be found in the design manual according to the material grade). The alternating stress amplitude was calculated and divided by the limit amplitude to obtain the installation impact assessment value.

[0030] It should be noted that the installation impact assessment value directly reflects the degree of alternating stress amplitude generated near the bolt hole when the train passes, relative to the material fatigue limit. The larger the assessment value, the stronger the loosening driving force at the bolt location (i.e., the more obvious the degree of bolt loosening driven by periodic dynamic load), and the higher the priority of installing loosening monitoring equipment.

[0031] It is worth noting that the train passing test refers to simulating the process of a train wheel passing over a railway rail under laboratory conditions. The calculation of the alternating stress amplitude and the train passing test process are both conventional technical methods in the existing technology, and will not be elaborated on here.

[0032] Step 2: Install monitoring equipment on the critical bolts. The monitoring equipment includes a monitoring module and a communication module. The monitoring module uses a triaxial accelerometer to record the triaxial acceleration data of the critical bolts. The communication module uses wireless communication technology to upload the data collected by the monitoring module. The monitoring period of the monitoring equipment and multiple monitoring nodes within the monitoring period are determined based on the periodic passage time of multiple passing trains. The monitoring nodes correspond to the stable moment of the rail after the passing train (i.e., the rail is in a stable stationary state at this moment, and the bolt data measured at this time reflects the true state of the bolt). Multiple monitoring nodes constitute a monitoring node set. Triaxial acceleration data is collected once at each monitoring node and recorded as the target acceleration set.

[0033] It should be noted that each monitoring device corresponds to a set of monitoring nodes. That is, each monitoring device independently sets its own set of monitoring nodes to achieve independent control of each monitoring device. This allows the monitoring device to perform multiple active monitoring cycles with the monitoring nodes as time references, thereby obtaining the bolt loosening status of the critical bolts after each passing train.

[0034] Specifically, the process for determining monitoring nodes is as follows: The location of the key bolt on the rail is recorded as the target location. All trains that pass the target location during their journey are recorded as passing trains. The departure cycle of the passing trains is obtained, and the longest cycle is recorded as the monitoring cycle. The time when each passing train passes the target location within the monitoring cycle is recorded as the passing time. Each passing train passes the target location at least once within the monitoring cycle. Based on the passing time and the vehicle body parameters of the passing train, the passing influence range is obtained. Each passing influence range is represented within the monitoring cycle. The portion of the non-passing influence range within the monitoring cycle is recorded as the target range. The midpoint of each target range is recorded as the monitoring node.

[0035] More specifically, the process of obtaining the affected area is as follows: Any passing train is designated as the target train. The minimum speed of the target train when it passes the target location (based on historical data) is designated as the target speed. The length of the target train is divided by the target speed to obtain the base time. A time interval is established with the passing time of the target train as the interval start point and the base time as the interval length, designated as the first influence interval. The full-load weight of the target train is obtained. The vibration influence distance is calculated by combining the full-load weight with the vibration propagation formula. The vibration influence distance is divided by the target speed to obtain the vibration influence duration. The departure deviation duration corresponding to the target train is preset (based on empirical values, set to 10 minutes in a specific embodiment). The sum of the vibration influence duration and the departure deviation duration is calculated to obtain the comprehensive influence duration. The departure deviation duration is subtracted from the starting point of the first influence interval, and the departure deviation duration is added to the ending point of the first influence interval to obtain the traversed influence interval.

[0036] Furthermore, the vibration propagation formula is: Where 'a' represents the vibration amplitude at a distance of x meters from the target train, and 'm' is the full load weight of the target train. The preset reference vehicle weight, The preset wheel vibration reference value (i.e., the rail vibration amplitude near the wheels when a train with a weight equal to the reference weight passes by, obtained through actual measurement), D is the preset track attenuation rate (i.e. the attenuation rate of vibration propagating through the track), and a preset vibration influence threshold is set. The vibration influence threshold is substituted into this formula and solved. The distance value obtained by the solution is recorded as the vibration influence distance.

[0037] It should be noted that since the monitoring equipment uses a triaxial accelerometer to monitor the bolt's condition, it is necessary to avoid other interferences that could cause changes in the triaxial acceleration values. In other words, it is crucial to minimize vibration interference caused by passing trains (for rail transit systems, the only vibration source is usually the vibration from passing trains). Unlike ordinary ground vibration propagation, because rails consist of multiple fixedly connected rail sections (which can be considered a continuous medium), vibrations generated by trains can propagate much further through the continuous rails, thus affecting distant monitoring equipment. Therefore, calculating the vibration impact distance can be used to determine the furthest impact distance. This distance value can then be used to set the monitoring equipment's acquisition time to avoid interference from passing trains, thereby obtaining a more accurate representation of the bolt's rotation state under static conditions.

[0038] Step 3: Obtain the target acceleration set collected by the monitoring module, analyze the target acceleration set to obtain the relative offset angle of the key bolt, and when the relative offset angle is greater than the preset angle offset threshold, generate a loosening warning signal and mark the key bolt as a problem bolt. The problem bolt packages its own information (i.e., coordinates on the board) with the loosening warning signal into a warning message, stores the warning message and waits for it to be uploaded.

[0039] Specifically, the relative offset angle analysis process is as follows: After tightening the key bolts, the monitoring equipment data is zeroed out. The first set of target accelerations collected after zeroing is recorded as the reference set, and the most recently collected set of target accelerations is recorded as the comparison set. The vertical plane perpendicular to the bolt axis is recorded as the reference plane. A rectangular coordinate system corresponding to the reference plane is established and recorded as the reference coordinate system. The Y-axis acceleration vector and Z-axis acceleration vector corresponding to the reference set are represented in the reference coordinate system and recorded as the initial vector combination. The Y-axis acceleration vector and Z-axis acceleration vector corresponding to the comparison set are represented in the reference coordinate system and recorded as the comparison vector combination. The angle between the Y-axis acceleration vectors in the initial vector combination and the comparison vector combination is calculated and recorded as the relative offset angle.

[0040] It should be noted that since the monitoring equipment is fitted onto a critical bolt, and the critical bolt is perpendicular to the vertical plane, when the critical bolt loosens (i.e., rotates by a certain angle), it will inevitably cause changes in the triaxial acceleration measured by the monitoring equipment, especially the acceleration changes on the Y-axis and Z-axis (corresponding to the vertical plane). Since the vector sum of the vertical acceleration in the static state is known to be the acceleration due to gravity, the direction can be determined by the specific values ​​of the Y-axis acceleration and Z-axis acceleration, thereby constructing the corresponding vector. Then, by analyzing the angle of change of direction of the same measuring axis of the same monitoring equipment, the rotation angle can be determined, so that even slight loosening of the critical bolt can be accurately identified, effectively improving the efficiency of loosening monitoring.

[0041] In another specific embodiment, the relative offset angle is calculated based on the separate IMU detection nodes fixedly installed on each railway bolt, and a loosening warning signal is generated, wherein: Establish a three-dimensional coordinate system corresponding to the bolt mounting reference plane, and calibrate the X, Y, and Z axes of the accelerometer with this three-dimensional coordinate system (when the detection node is stationary or in a low-dynamic state, the triaxial output measured by the accelerometer mainly consists of the gravitational acceleration component). The microcontroller performs zero-bias calibration, temperature compensation, and low-pass filtering on the raw triaxial data to obtain ax, ay, and az (representing the acceleration values ​​measured by the accelerometer in the x, y, and z directions, respectively), and calculates the tilt angle of the detection node relative to the mounting reference plane according to the following formula: θx = atan2(ay, az), where θx represents the tilt angle calculated around the X-axis, which can usually be understood as the tilt angle in one direction. θy = atan2(-ax, sqrt(ay × ay + az × az)), where θy represents the tilt angle calculated around the Y-axis, which can usually be understood as the tilt angle in another direction. When it is necessary to obtain the composite tilt angle of the plane, calculate θ = sqrt(θx × θx + θy × θy), where θ represents the composite total tilt angle or the composite tilt angle of the plane.

[0042] Where sqrt represents the square root function, atan2 is the two-parameter arctangent function, which can determine the quadrant of the angle based on the sign of the components, and sqrt(ay×ay + az×az) represents the magnitude of the combined acceleration components in the Y and Z directions.

[0043] After the equipment is installed, the initial angles θx0, θy0, or θ0 are recorded as reference values ​​(θx0 represents the initial tilt angle in the X direction recorded when the equipment is first installed and not yet loosened, θy0 represents the initial tilt angle in the Y direction recorded when the equipment is first installed and not yet loosened, and θ0 represents the initial total tilt angle recorded when the equipment is first installed). During periodic data acquisition, the angle change Δθx = θx - θx0, Δθy = θy - θy0, or Δθ = θ - θ0 is calculated, and false positives caused by train vibration and transient impacts are suppressed through moving average, median filtering, or multiple sampling voting. When the angle change continuously exceeds a preset threshold, or accumulates to exceed a preset number of times within a specified time window, it is determined that the bolt or nut has become loose and a loosening warning signal is generated. The angle change, sampling time, node number, and battery status are packaged and sent to the acquisition box.

[0044] It should be noted that the accelerometer in the monitoring equipment is a low-power accelerometer with a range of ±2g, which can improve the ability to resolve slow tilt changes and avoid long-term drift caused by gyroscope integration.

[0045] More specifically, the monitoring equipment is powered by solar energy, including: The monitoring equipment (data acquisition box) is equipped with a solar panel, charging management circuit, energy storage battery, input reverse connection protection and overcharge / overdischarge protection circuit, and system voltage regulator circuit. The solar panel converts ambient light along the railway line into DC power, which is then used by the reverse connection protection device and charging management circuit to charge the energy storage battery using constant current, constant voltage, or pulse charging. The charging management circuit monitors the solar input voltage and current, as well as the voltage and temperature of the energy storage battery in real time. When the battery reaches its charging limit, it reduces or stops charging; when the battery voltage falls below a set threshold, it limits the load or enters a low-power protection state. The output of the energy storage battery, through the voltage regulator circuit, powers the main control unit of the data acquisition box, the LoRa communication module, the 4G or NB-IoT communication module, and the sensors. The main control unit dynamically adjusts the acquisition cycle, wireless transmission power, and communication timing based on the battery voltage and solar input status: it performs normal data acquisition and upload when there is sufficient sunlight; during continuous cloudy or rainy weather or when the battery is low on power, it extends the acquisition cycle and shuts down unnecessary modules, retaining only the timed wake-up, bolt node polling, and low battery alarm functions. The data acquisition box is powered by a combination of solar panels and energy storage batteries, reducing the frequency of battery replacements. This makes the equipment suitable for environments without mains power along railway lines and ensures that it can still complete data acquisition and communication at night and under short-term light-free conditions.

[0046] Step 4: Select a specific train from multiple passing trains. The specific train is the one with the shortest departure cycle. Install wireless interaction equipment (such as UHF RFID technology equipment) on the specific train. The wireless interaction equipment is used to receive data uploaded by the monitoring equipment. Based on the passing data of the specific train and the vehicle data, determine the upload time interval, and instruct the monitoring equipment to upload warning information within the upload time interval (continuously sending warning information outward during the upload period). When the specific train passes, obtain the warning information and upload it to the cloud for scheduling and management. Inspect and tighten the problematic bolts and the nearby connecting bolts.

[0047] It should be noted that a specific train is bound to a monitoring device on a one-to-one basis. This means that each monitoring device is pre-determined to a specific train during the initial installation phase, ensuring that the monitoring device has a clear data output target. Using specific trains to collect data can cover the entire track. Compared to setting up fixed collection devices near the track, this method has higher data collection efficiency and is more suitable for monitoring devices that are set up at multiple points in a long distance. In addition, by setting the specific train to the train with the shortest departure cycle, it is ensured that early warning information can be received in a timely manner, so as to facilitate subsequent maintenance and handling of loose bolts.

[0048] Specifically, the process for determining the upload time range is as follows: The passing time of a specific train is recorded as the anchor point time, the length of the specific train is recorded as the specific length, the minimum speed of the specific train passing the monitoring equipment is recorded as the specific speed, the specific length is divided by the specific speed to obtain the first duration, and a time interval is constructed with the anchor point time as the interval start point and the first duration as the interval length, recorded as the basic upload interval. The departure deviation duration is subtracted from the start point of the basic upload interval, and the departure deviation duration is added to the end point of the basic upload interval to obtain the upload time interval.

[0049] It should be noted that the upload time interval corresponds to the time range during which a specific train may pass through the location of the monitoring equipment. By determining the upload time interval, the early warning information uploaded by the monitoring equipment can be received at the moment the train passes by, thereby avoiding missed warnings, greatly improving the robustness of the bolt loosening status early warning, and thus improving the stability and safety of track operation.

[0050] It should be noted that the main functions of the monitoring equipment include data monitoring and data uploading. These main functions are triggered at specific times, meaning that the monitoring equipment is in standby or hibernation mode most of the time, thus maximizing its working time. The monitoring equipment is only woken up and performs the corresponding main functions when triggered at specific times. This allows the monitoring equipment to consume power only when needed to perform its main functions. Compared with the long-term online monitoring of similar monitoring equipment in the existing technology, it saves more power and is more suitable for monitoring equipment powered by batteries (since power supply is generally not available along railway lines, only battery-powered equipment can be used).

[0051] For those skilled in the art, any modifications, equivalent substitutions, or improvements can be made to the specific implementation methods and application scope based on the ideas of this invention, and all such modifications, equivalent substitutions, or improvements should be included within the scope of protection of this invention.

Claims

1. A method for monitoring and early warning of loose railway bolts based on a split IMU, characterized in that, Includes the following steps: Step 1: Record all the fastening bolts corresponding to the rail connecting plates on the railway track as connecting bolts. Based on the installation position and installation data of the connecting bolts, select multiple connecting bolts and record them as key bolts. Step 2: Install monitoring equipment on key bolts. The monitoring equipment includes a monitoring module and a communication module. The monitoring module collects triaxial acceleration data and uploads the data information based on the communication module. The monitoring cycle of the monitoring equipment and the multiple monitoring nodes within the monitoring cycle are determined based on the periodic passage time of multiple passing trains. The multiple monitoring nodes constitute a monitoring node set, and triaxial acceleration data is collected once at each monitoring node and recorded as the target acceleration set. Step 3: Obtain the target acceleration set collected by the monitoring module, analyze the target acceleration set to obtain the relative offset angle of the key bolt, and when the relative offset angle is greater than the preset angle offset threshold, generate a loosening warning signal and mark the key bolt as a problem bolt. The problem bolt packages its own information with the loosening warning signal into a warning message, stores the warning message and waits for it to be uploaded. Step 4: Select a specific train from multiple passing trains. The specific train is the one with the shortest departure cycle. Determine the upload time interval based on the specific train's route data and vehicle data. Instruct the monitoring equipment to upload the warning information within the upload time interval. When the specific train passes, the warning information is obtained and uploaded to the cloud.

2. The method for monitoring and early warning of loose railway bolts based on a split IMU according to claim 1, characterized in that, The selection process for critical bolts is as follows: Obtain track data along the railway track, record multiple connecting bolts on the same rail connecting plate as a bolt combination, record the track position of the bolt combination as the combination coordinate, analyze and calculate the track impact assessment value based on the track data corresponding to the combination coordinate, and select multiple key bolt combinations based on the track impact assessment value and a preset distribution distance threshold. Based on the stress test data of the rail connecting plate, the stress data of different connecting bolts in the same critical bolt combination are analyzed. Combined with the data of the connecting bolts themselves and the installation data, the installation impact assessment value is calculated. The connecting bolt with the largest installation impact assessment value in the same critical bolt combination is selected and recorded as the critical bolt.

3. The method for monitoring and early warning of loose railway bolts based on a split IMU according to claim 2, characterized in that, The calculation process for the track impact assessment value is as follows: Obtain the speed limit value when the train passes through the combined coordinates. Simulate the train passing through the combined coordinates at the speed limit value and perform force analysis to obtain the comprehensive acceleration. Record the comprehensive acceleration as the track impact assessment value, where: Treating the track as a plane, with the plane angle depending on the gradient and superelevation, and treating the train as a point mass, when the track corresponding to the combined coordinates is a curved segment, analyze the acceleration components of gravitational acceleration and centrifugal acceleration in the direction perpendicular to the track plane; when the track corresponding to the combined coordinates is a straight segment, analyze the acceleration component of gravitational acceleration in the direction perpendicular to the track plane, and denote the acceleration component in the direction perpendicular to the track plane as the composite acceleration.

4. The method for monitoring and early warning of loose railway bolts based on a split IMU according to claim 3, characterized in that, The screening process for key bolt combinations is as follows: Draw the track line and mark all bolt combinations on the track line. There is a preset track influence threshold. Bolt combinations that are greater than or equal to the track influence threshold are recorded as the first candidate combinations. When the distance between any two adjacent first candidate combinations is greater than or equal to the distribution distance threshold, multiple bolt combinations between these two first candidate combinations are selected as the second bolt combinations. The first bolt combinations and the second bolt combinations are collectively recorded as the key bolt combinations.

5. The method for monitoring and early warning of loose railway bolts based on a split IMU according to claim 4, characterized in that, Select the option that meets the following conditions: Condition 1: The spacing between multiple second candidate combinations is less than the distribution distance threshold; Condition 2: The number of the second candidate combinations is the minimum value that satisfies Condition 1.

6. The method for monitoring and early warning of loose railway bolts based on a split IMU according to claim 4, characterized in that, The calculation process for the installation impact assessment value is as follows: There are multiple preset type numbers, each type number corresponds to a rail connection plate type. Each type of rail connection plate is assigned a set of screw hole numbers. Each screw hole number in the screw hole number set corresponds to a connecting bolt. The screw hole number and type number corresponding to any connecting bolt are recorded as the plate coordinates of that connecting bolt. Train passing tests were conducted on different types of rail connecting plates. During the test, the stress change around each bolt hole was measured and the alternating stress amplitude was calculated. A limit amplitude was preset, and the installation impact assessment value was obtained by dividing the alternating stress amplitude by the limit amplitude.

7. The method for monitoring and early warning of loose railway bolts based on a split IMU according to claim 1, characterized in that, The process for determining monitoring nodes is as follows: The location of the key bolt on the rail is recorded as the target location. All trains that pass the target location during their journey are recorded as passing trains. The departure cycle of the passing trains is obtained, and the longest cycle in the departure cycle is recorded as the monitoring cycle. The time when each passing train passes the target location within the monitoring cycle is recorded as the passing time. Based on the passing time and the vehicle body parameters of the passing trains, the passing influence range is obtained. Each passing influence range is represented within the monitoring cycle. The portion of the non-passing influence range within the monitoring cycle is recorded as the target range. The midpoint of each target range is recorded as the monitoring node.

8. The method for monitoring and early warning of loose railway bolts based on a split IMU according to claim 7, characterized in that, The process of obtaining the affected area through the route is as follows: Any passing train is designated as the target train. The minimum speed of the target train when it passes the target position is designated as the target speed. The length of the target train is divided by the target speed to obtain the base time. A time interval is established with the passing time of the target train as the start point of the interval and the base time as the length of the interval, designated as the first influence interval. Obtain the full load weight of the target train, calculate the vibration influence distance using the vibration propagation formula, divide the vibration influence distance by the target speed to obtain the vibration influence duration, preset the departure deviation duration corresponding to the target train, calculate the sum of the vibration influence duration and the departure deviation duration to obtain the comprehensive influence duration, subtract the departure deviation duration from the starting point of the first influence interval, and add the departure deviation duration to the ending point of the first influence interval to obtain the traversed influence interval.

9. The method for monitoring and early warning of loose railway bolts based on a split IMU according to claim 1, characterized in that, The relative offset angle analysis process is as follows: After tightening the key bolts, the monitoring equipment data is zeroed out. The first set of target accelerations collected after zeroing is recorded as the reference set, the most recently collected set of target accelerations is recorded as the comparison set, the vertical plane perpendicular to the bolt axis is recorded as the reference plane, and the rectangular coordinate system corresponding to the reference plane is recorded as the reference coordinate system. In the reference coordinate system, the Y-axis acceleration vector and Z-axis acceleration vector corresponding to the reference set are represented as the initial vector combination, and the Y-axis acceleration vector and Z-axis acceleration vector corresponding to the comparison set are represented as the comparison vector combination. The angle between the Y-axis acceleration vectors in the initial vector combination and the comparison vector combination is calculated and recorded as the relative offset angle.

10. The method for monitoring and early warning of loose railway bolts based on a split IMU according to claim 1, characterized in that, The process for determining the upload time range is as follows: The passing time of a specific train is recorded as the anchor point time, the length of the specific train is recorded as the specific length, the minimum speed of the specific train passing the monitoring equipment is recorded as the specific speed, the specific length is divided by the specific speed to obtain the first duration, and a time interval is constructed with the anchor point time as the interval start point and the first duration as the interval length, recorded as the basic upload interval. The departure deviation duration is subtracted from the start point of the basic upload interval, and the departure deviation duration is added to the end point of the basic upload interval to obtain the upload time interval.

Images