A system and method for detecting defects in turntable wheels

Abstract

This invention discloses a system and method for detecting defects in turntable wheels, relating to the field of non-destructive testing technology. The system comprises a field sensing layer, a network transmission layer, and a system platform layer. The field sensing layer employs a wheel defect detection device, which uses two sets of symmetrically arranged microswitches to detect the time interval between wheel passes, and determines defects based on a preset normal interval calculated from the wheel diameter and turntable speed. A crossarm detection device identifies the crossarm position using a photoelectric switch and a uniquely coded optical signal reflector. A periodic reset device determines the turntable's operating cycle and triggers data packetization. The network transmission layer provides power and data transmission based on PoE technology. The system platform layer enables real-time monitoring and alarm functions, data statistical analysis, and maintenance prediction. This invention overcomes the applicability issues of vision and sensor solutions in complex environments by using the time interval detection principle, achieving accurate and real-time detection and location of wheel defects.

Description

Technical Field

[0001] This invention relates to the field of nondestructive testing and intelligent equipment technology, specifically to a defect detection system and method for a turntable wheel. Background Technology

[0002] Airport baggage handling systems are a core component ensuring the efficiency of air transport, with baggage carousels (including arrival and departure carousels) playing a crucial role in baggage transport. The carousel wheels are the core load-bearing and transmission components of this system, their direct physical functions including: 1. Supporting the carousel's own weight and dynamic baggage load; 2. Rolling on the tracks to drive the carousel's operation. According to first principles, the normal working condition of the wheels depends on their intact geometric shape. However, under the fundamental physical effects of long-term cyclic stress, impact loads, and environmental aging, the wheel material inevitably develops defects such as wear, cracks, and plastic deformation, leading to a reduction in diameter or loss of roundness.

[0003] This fundamental geometric change triggers a series of chain reactions: First, the contact area and pressure distribution between the wheels and the track change, leading to an abnormally increased rolling resistance; second, the out-of-round wheels will produce periodic radial runout. These two effects ultimately manifest as increased vibration and abnormal noise on the carousel. The direct physical consequence is damage to the crossarms and tracks in the carousel's support structure; in severe cases, jamming can cause the carousel to stop operating, resulting in the interruption of the entire baggage system and economic losses.

[0004] Currently, in production practice, there is no automated inspection equipment specifically designed for the turntable's traveling wheel. The industry's commonly relied-upon manual visual inspection method suffers from the following fundamental flaws due to the limitations of human perception:

[0005] Inefficiency and subjectivity: Manual inspection cannot continuously and without omission observe every wheel running at high speed. The judgment results rely heavily on individual experience and attention, resulting in a high rate of false positives and false negatives.

[0006] Invisible defects cannot be identified: Many initial defects (such as microcracks and internal stress concentrations) occur underground in the material or are too small to be identified by the naked eye. By the time the fault appears, it has often developed into a serious problem.

[0007] Lack of real-time and predictive capabilities: Manual inspection is discrete and periodic, unable to achieve continuous physical condition monitoring, and unable to capture predictive changes in physical parameters (such as the trend of increased vibration) before a failure occurs. It is essentially a "post-failure maintenance" mode and cannot achieve "predictive maintenance".

[0008] Existing technologies for detecting defects in rail vehicle wheels, which function similarly to running wheels, attempt to solve human-related problems through technological means. However, their technical principles are fundamentally mismatched with the application scenarios and cannot be directly applied to airport carousel running wheels.

[0009] The vision-based approach: This approach works by capturing two-dimensional optical images of the wheels using an image sensor (camera) and then analyzing the image features using algorithms to identify defects. However, this approach faces insurmountable obstacles in the actual physical environment of an airport carousel: uneven lighting, complex backgrounds (reflections from metal structures), oil and dust contamination of the lens, and weak image features due to the small size of the wheels. These factors introduce a large amount of optical noise, resulting in a low signal-to-noise ratio and a sharp decline in the efficiency and accuracy of the detection algorithm.

[0010] Sensor-based solution: This solution's basic principle is to directly measure the geometric dimensions of the wheel's outer circumference using contact or non-contact sensors (such as laser displacement or eddy current sensors). This principle places extremely high demands on the sensor's absolute installation accuracy, measurement resolution, and relative positional stability with the measured surface. Airport turnout environments are characterized by high vibration and confined spaces, making it difficult to ensure the sensors maintain the required precise installation state over long periods. Even minute installation misalignments or vibrations can be misinterpreted by the system as wear on the wheels, resulting in unacceptable measurement errors.

[0011] Existing technologies either revert to inefficient and unreliable human perception or employ technical principles that are unsuitable, unstable, or overly complex in the specific physical environment of a turntable. Summary of the Invention

[0012] The purpose of this invention is to provide a defect detection system and method for airport baggage carousel wheels. Through a stable and reliable automated solution, defects in airport baggage carousel wheels can be detected in real time and accurately, and the fault location can be quickly located. This overcomes the fundamental shortcomings of manual inspection, such as inefficiency, subjectivity, and poor applicability of existing vision and sensor technologies in complex field environments.

[0013] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows:

[0014] A defect detection system for a turntable wheel includes a field sensing layer, a network transmission layer, and a system platform layer;

[0015] The field sensing layer includes a wheel defect detection device, a crossarm detection device, and a periodic reset device. The wheel defect detection device is used to detect the defect status of the wheel and includes a device housing, microswitches, and a connecting mechanism. The microswitches include two sets, which are symmetrically installed inside the device housing. When the wheel passes by, the microswitches are triggered by the surface of the wheel to generate a trigger signal. The trigger signal includes a trigger timestamp and a trigger status. The field sensing layer also includes a speed detection device. The speed detection device uses an incremental rotary encoder installed on the turntable drive shaft to measure the turntable running speed in real time, or calculates the turntable running speed by the time interval between adjacent crossarms passing through the crossarm detection device.

[0016] The crossarm detection device is used to identify the crossarm number and includes a photoelectric switch and an optical signal reflector. The optical signal reflector is installed on each crossarm and has a unique code. The photoelectric switch emits an optical signal and receives the reflected signal from the optical signal reflector to determine the crossarm position and number.

[0017] The cycle reset device is used to determine the turntable's operating cycle. It includes a photoelectric switch and an optical signal reflector. The optical signal reflector is installed on the first crossarm of the turntable. When the photoelectric switch detects the first crossarm, it triggers the cycle reset and packages the data.

[0018] The network transmission layer is based on PoE technology and includes a PoE fiber optic switch and network connection cable, used to provide DC power and data transmission to the field sensing layer; the system platform layer is used to receive data and includes a real-time monitoring and alarm module and a data statistics and analysis module. The real-time monitoring and alarm module determines whether the alarm threshold has been reached based on the trigger signal of the walking wheel defect detection device and issues an alarm. The data statistics and analysis module is used for historical data recording, visualization and maintenance reminders.

[0019] In one embodiment of the present invention, the micro switch of the wheel defect detection device is a high-precision mechanical micro switch with a triggering force range of 0.1N to 1.0N. The preset normal interval of the triggering time interval is calculated based on the wheel diameter and the turntable running speed. If the actual triggering time interval deviates from the preset value by more than 20%, the wheel defect is determined.

[0020] In one embodiment of the present invention, the optical signal reflector of the crossarm detection device is an infrared optical signal reflector, the photoelectric switch is an infrared photoelectric switch with an emission frequency of 38kHz, and the unique code is realized by a combination of different reflectivity regions set on the optical signal reflector. The photoelectric switch identifies the code value by detecting changes in the intensity of the reflected signal.

[0021] In one embodiment of the present invention, the data packaging of the periodic reset device includes encapsulating the trigger signals, crossarm numbers and timestamps of all walking wheels in the current period into a data packet, and uploading it to the system platform layer through the network transmission layer.

[0022] In one embodiment of the present invention, the real-time monitoring and alarm module of the system platform layer further includes a heartbeat detection mechanism for monitoring the online status of the field sensing layer device. If the device is offline for more than a set time, an offline alarm signal is issued.

[0023] In one embodiment of the present invention, the field sensing layer is based on an ESP32 microcontroller, which is configured to process the input signals of microswitches and photoelectric switches, calculate the trigger time interval, and perform data preprocessing.

[0024] In one embodiment of the present invention, the wheel defect detection device, the crossarm detection device, and the periodic reset device are fixed to the vicinity of the turntable track by bolts. The connecting mechanism includes a support bracket and fixing bolts, which are used to adjust the position of the device to adapt to the wheel trajectory.

[0025] In addition, the present invention also discloses a method for detecting defects in turntable wheels, applied to the aforementioned turntable wheel defect detection system, comprising the following steps:

[0026] PoE provides DC power to the wheel defect detection device, crossarm detection device, and periodic reset device;

[0027] Each turntable crossarm has a traveling wheel on its upper and lower sides. The crossarms and traveling wheels are numbered and bound together to ensure precise positioning of the traveling wheels.

[0028] Crossarm data is collected in real time using a crossarm detection device;

[0029] Data on the walking wheels is collected in real time using a walking wheel defect detection device;

[0030] The turntable's operating cycle is checked by a cycle reset device. After the turntable starts running, if the crossarm detection value matches the preset code value of the first crossarm, it is determined that the turntable has completed one cycle, and the crossarm number is reset. If they do not match, the operating cycle determination continues.

[0031] After the crossarm number is reset, the collected data such as the walking wheels will be uploaded to the host computer monitoring and alarm system via the lower-level microcontroller and PoE fiber optic switch.

[0032] The monitoring and alarm system analyzes the data and displays it through a visual human-computer interaction interface;

[0033] On the one hand, it determines whether the damage to the walking wheel has reached the set alarm threshold. If the alarm threshold is reached, the system issues a walking wheel fault alarm signal; if the alarm threshold is not reached, the walking wheel data continues to be analyzed in real time.

[0034] On the other hand, the online status of the field device is monitored in real time through a heartbeat detection mechanism. If the device is determined to be offline, an offline alarm signal is issued; if the device recovers from the offline state, the alarm signal is cleared; if the device is not offline, monitoring continues.

[0035] The monitoring and alarm system records and archives information such as data and alarm events. The data can be exported for in-depth analysis or reporting.

[0036] In one embodiment of the present invention, the formula for calculating the triggering time interval is: ,in For the trigger timestamp of the first group of microswitches, The preset normal interval is the trigger timestamp for the second group of microswitches. Calculations based on the diameter D of the traveling wheel and the linear velocity v of the turntable: Where L is the installation distance between the two sets of microswitches, and v is the tangential linear velocity at the contact point between the traveling wheel and the track.

[0037] In one embodiment of the present invention, the data statistics and analysis module of the system platform layer is further used to train a trend prediction model based on historical trigger data, predict the remaining life of the walking wheel and generate a maintenance plan. The trend prediction model is a linear regression model, and the input includes the trigger time interval change rate and crossarm load data. The trigger time interval change rate is calculated from historical Δt data. The crossarm load data is obtained in any of the following ways: (a) real-time measurement by a pressure sensor installed on the crossarm, the pressure sensor has a range of 0-100kg and an accuracy of ±1%; (b) obtained from the real-time weight data table of the baggage system control database of the airport baggage system, the data table contains the crossarm number, timestamp and load weight fields, and the data update frequency is not less than 1Hz.

[0038] Compared with the prior art, the present invention has the following beneficial effects:

[0039] This invention utilizes an innovative method of measuring the time interval of the traveling wheels using dual microswitches, transforming geometric changes into precise time parameters. This fundamentally overcomes the technical obstacles of visual solutions being affected by lighting interference and sensor solutions requiring high installation accuracy. The PoE-based power transmission architecture and heartbeat detection mechanism ensure long-term reliability in complex industrial environments, while the crossarm optical coding and periodic reset device enable precise fault location. The system platform layer trains a trend prediction model using historical data, upgrading maintenance from reactive to predictive. The bolt-fixed and adjustable bracket design eliminates the need to modify existing structures during installation, significantly reducing implementation costs. The overall solution, through simple hardware configuration and innovative algorithms, achieves accurate detection, rapid location, and intelligent early warning of traveling wheel defects in the complex operating environment of airport baggage carousels, effectively avoiding carousel downtime losses and significantly improving operational efficiency and economy. Attached Figure Description

[0040] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained from these drawings without creative effort.

[0041] Figure 1 This is a general framework diagram of the present invention.

[0042] Figure 2 This is a schematic diagram of the overall structure of the present invention.

[0043] Figure 3 This is a flowchart of the method for detecting defects in the walking wheel of the present invention.

[0044] Figure label:

[0045] 101 Walking wheel, 102 Turntable crossarm, 103 Optical signal reflector, 104 Walking wheel defect detection device, 105 Support bracket, 106 Photoelectric switch, 107 ESP32 microcontroller, 108 PoE fiber optic switch. Detailed Implementation

[0046] In the following description, only certain exemplary embodiments are briefly described. As those skilled in the art will recognize, the described embodiments can be modified in various ways without departing from the spirit or scope of the embodiments of the invention. Therefore, the drawings and description are considered to be exemplary in nature and not restrictive.

[0047] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0048] Example 1:

[0049] This embodiment discloses a defect detection system for a turntable wheel 101, including a field sensing layer, a network transmission layer, and a system platform layer. These layers work together to achieve defect detection, data transmission, and analysis monitoring of the wheel 101.

[0050] The system platform layer is developed based on PYSIDE6 software.

[0051] The field perception layer includes a wheel defect detection device 104, a crossarm detection device, and a periodic reset device, configured based on an ESP32 microcontroller 107. The ESP32 microcontroller 107 is used to process the input signals of the micro switch and the photoelectric switch 106, calculate the trigger time interval, and perform data preprocessing.

[0052] The wheel defect detection device 104 is used to detect the defect status of the wheel 101. It includes a device housing, microswitches, and a connecting mechanism. The microswitches are two sets of high-precision mechanical microswitches, symmetrically installed inside the device housing, with a triggering force range of 0.1N-1.0N. When the wheel 101 passes by, the outer circumference of the wheel contacts and presses down the triggering arms of the microswitches in sequence. The microswitches trigger to generate a triggering signal when the triggering force reaches the 0.1N-1.0N threshold. The center distance between the triggering points of the two sets of microswitches is L=5cm. The triggering signal includes a triggering timestamp and a triggering status. The wheel defect detection device 104 determines whether the wheel 101 is defective based on a comparison between the triggering time interval of the two sets of microswitches and a preset normal interval. The preset normal interval is determined based on the diameter D of the wheel 101, the turntable running speed, and the installation distance L of the two sets of microswitches. The connecting mechanism includes a support bracket 105 and fixing bolts, used to adjust the device position to adapt to the trajectory of the wheel 101.

[0053] Speed ​​detection device: An incremental rotary encoder (model E6B2-CWZ6C) is installed on the turntable drive shaft with a resolution of 1000 pulses / revolution. The linear speed of the turntable is calculated by measuring the drive shaft speed and combining it with the turntable transmission ratio. In the alternative solution, the turntable speed v=S / ΔT is calculated by measuring the time interval ΔT between adjacent crossarms passing through the crossarm detection device and the crossarm spacing S. The calculation period is 1 second.

[0054] Crossarm detection device: used to identify crossarm numbers, including photoelectric switch 106 and optical signal reflector 103. The optical signal reflector 103 is an infrared optical signal reflector, and the photoelectric switch 106 is an infrared photoelectric switch with an emission frequency of 38kHz. The optical signal reflector 103 is installed on each crossarm and has a unique code. The unique code is a binary code achieved by combining three different reflectivity regions (30%, 50%, and 70%) set on the optical signal reflector 103. The photoelectric switch 106 identifies the specific code value by detecting changes in the intensity of the reflected signal. The photoelectric switch 106 emits an optical signal and receives the reflected signal from the optical signal reflector to determine the position and number of the crossarm.

[0055] Periodic reset device: used to determine the turntable's operating cycle, including photoelectric switch 106 and optical signal reflector 103. The optical signal reflector 103 is installed on the first crossarm of the turntable. When the photoelectric switch 106 detects the first crossarm, it triggers a periodic reset and packages the data. The data packaging includes encapsulating the trigger signals of all traveling wheels 101 in the current cycle, the crossarm number, and the timestamp into a data packet.

[0056] The wheel defect detection device 104, the crossarm detection device, and the periodic reset device are fixed near the turntable track by bolts.

[0057] Network transport layer: Based on PoE technology, including PoE switches and network cables, it is used to provide DC power and TCP / IP data transmission to the field sensing layer.

[0058] System platform layer: used to receive and process data, including real-time monitoring and alarm modules and data statistics and analysis modules;

[0059] Real-time monitoring and alarm module: Based on the trigger signal of the walking wheel defect detection device 104, it determines whether the alarm threshold has been reached and issues an alarm. The preset normal interval of the trigger time interval is calculated based on the diameter D of the walking wheel 101 and the running speed of the turntable. If the actual trigger time interval deviates from the preset value by more than 20%, the walking wheel 101 is determined to be defective. The real-time monitoring and alarm module also includes a heartbeat detection mechanism to monitor the online status of the field sensing layer device. If the device is offline for more than a set time, an offline alarm signal is issued.

[0060] Data statistics and analysis module: used for historical data recording, visualization and maintenance reminders, and also used to train a trend prediction model based on historical trigger data, predict the remaining life of the walking wheel 101 and generate a maintenance plan. The trend prediction model is a linear regression model. The input includes the trigger time interval change rate and crossarm load data. The trigger time interval change rate is calculated from historical Δt data. The crossarm load data is obtained in any of the following ways: (a) real-time measurement by a pressure sensor installed on the crossarm. The pressure sensor has a range of 0-100kg and an accuracy of ±1%; (b) obtained from the real-time baggage weight data table in the airport baggage system control database. The data table contains the crossarm number, timestamp and load weight fields, and the data update frequency is not less than 1Hz.

[0061] The load data of the crossarm is measured in real time by a pressure sensor installed on the crossarm 102, or obtained from the baggage weight data in the airport baggage system control database.

[0062] In addition, this embodiment also discloses a defect detection method for turntable traveling wheel 101, including the following steps:

[0063] PoE provides DC power to the wheel defect detection device 104, the crossarm detection device, and the periodic reset device.

[0064] Each turntable crossarm 102 has a traveling wheel 101 on its upper and lower sides. The crossarm and the traveling wheel 101 are numbered and bound together to accurately position the traveling wheel 101.

[0065] Crossarm data is collected in real time using a crossarm detection device;

[0066] The data of the walking wheel 101 is collected in real time by the walking wheel defect detection device 104;

[0067] The turntable's operating cycle is checked by a cycle reset device. After the turntable starts running, if the crossarm detection value matches the preset code value of the first crossarm, it is determined that the turntable has completed one cycle, and the crossarm number is reset. If they do not match, the operating cycle determination continues.

[0068] After the crossarm number is reset, the collected data such as the walking wheel 101 will be uploaded to the host computer monitoring and alarm system via the lower computer microcontroller and the PoE fiber optic switch 108.

[0069] The monitoring and alarm system analyzes the data and displays it through a visual human-computer interaction interface;

[0070] On the one hand, it determines whether the damage to the walking wheel 101 has reached the set alarm threshold. If the alarm threshold is reached, the system issues a fault alarm signal for the walking wheel 101; if the alarm threshold is not reached, the system continues to analyze the data of the walking wheel 101 in real time.

[0071] On the other hand, the online status of the field device is monitored in real time through a heartbeat detection mechanism. If the device is determined to be offline, an offline alarm signal is issued; if the device recovers from the offline state, the alarm signal is cleared; if the device is not offline, monitoring continues.

[0072] The monitoring and alarm system records and archives information such as data and alarm events. The data can be exported for in-depth analysis or reporting.

[0073] The trigger time interval is calculated to determine whether it deviates from the preset normal interval by more than a threshold. If so, a fault alarm for the walking wheel 101 is issued. The formula for calculating the trigger time interval is as follows: In the formula, For the trigger timestamp of the first group of microswitches, The second set of microswitches is triggered by a timestamp; the preset normal interval is... Calculations based on the diameter D of the traveling wheel 101 and the linear velocity v of the turntable: In the formula, L is the installation distance between the two sets of microswitches;

[0074] At the same time, the online status of the field device is monitored through a heartbeat detection mechanism, and an offline alarm is issued if the device is offline.

[0075] The data statistics and analysis module at the system platform layer trains a linear regression model based on historical trigger data. The inputs include the trigger time interval change rate and crossarm load data. The trigger time interval change rate is calculated from historical Δt data, and the crossarm load data is obtained through pressure sensors or the system database. The output is the remaining life prediction result and maintenance plan for the walking wheel 101.

[0076] The load data of the crossarm is measured in real time by a pressure sensor installed on the crossarm. The pressure sensor is a low-power wireless sensor that is powered by a battery or obtains power through an inductive power supply system on the turntable structure.

[0077] To facilitate a better understanding of the present invention by those skilled in the art, the present invention will be further described below in conjunction with specific embodiments.

[0078] In this embodiment, the system configuration is as follows:

[0079] On-site sensing layer: ESP32 microcontroller 107 (model ESP32-WROOM-32), the microswitch triggering force of the wheel defect detection device 104 is 0.1N, and the installation distance between the two sets of microswitches is L=5cm; the crossarm detection device adopts a 38kHz infrared photoelectric switch 106 (model E3Z-LT61), the optical signal reflector 103 is an infrared optical signal reflector, and its reflectivity pattern code is "30%-50%-70%"; the optical signal reflector 103 of the periodic reset device is installed on the side of the first crossarm, with a distance of 10cm from the photoelectric switch 106; the wheel defect detection device 104, the crossarm detection device and the periodic reset device are fixed to the outside of the turntable track by M8 bolts, and the height of the support bracket 105 of the connecting mechanism is adjustable from 5-15cm.

[0080] Network transport layer: PoE switch (model TL-SF1005P), network connection cable is CAT5E network cable, transmission rate is 100Mbps.

[0081] System platform layer: The offline time threshold of the real-time monitoring and alarm module is set to 30 seconds, and the training sample size of the linear regression model of the data statistics and analysis module is 1000 sets of historical data.

[0082] Detection parameter definition:

[0083] The diameter D of the walking wheel 101 is 10cm. The linear velocity v of the turntable is measured in real time by a speed detection device. In this embodiment, the average speed is 0.2m / s, and the preset normal interval is... The defect determination threshold is 20% (i.e. (Deviating by 0.25s) exceeds (Defects are determined in time).

[0084] For the trigger timestamp of the first group of microswitches, This is the trigger timestamp for the second group of microswitches.

[0085] Testing process:

[0086] The PoE switch provides DC power to the field sensing layer to achieve PoE power supply, and the ESP32 microcontroller 107 initializes and completes the signal processing configuration.

[0087] Bind the crossarm to the number of the traveling wheel 101, and associate the unique code of each crossarm with the corresponding traveling wheel 101 (crossarm 1 corresponds to traveling wheel 1011-8, crossarm 2 corresponds to traveling wheel 1019-16).

[0088] The infrared photoelectric switch 106 of the crossarm detection device emits a 38kHz optical signal and receives the reflected signal from the optical signal reflector 103 to identify the number and position of the crossarm 3 (3m to the left of the track).

[0089] When the traveling wheel 10110 passes by, it triggers two sets of microswitches and records the data. =1620000001.23s, =1620000001.47s, calculate The deviation from the preset value by 0.01s (4%) is considered normal.

[0090] When the photoelectric switch 106 of the periodic reset device detects the optical signal reflector 103 of the first crossarm, it triggers the periodic reset and encapsulates the trigger signals, crossarm numbers and timestamps of all traveling wheels 101 in the current period (0-5min) into a data packet;

[0091] After the ESP32 microcontroller performs preprocessing on the packaged data, it uploads it to the system platform layer via a PoE switch;

[0092] The real-time monitoring and alarm module at the system platform layer determines that there are no defects based on trigger signals, and the heartbeat detection mechanism shows that all field sensing layer devices are online.

[0093] The data statistics and analysis module records and visualizes historical data. The linear regression model takes the trigger time interval change rate (0.002s / day) and crossarm load data (20kg) as input and outputs the remaining life prediction result of the 10110 walking wheel (180 days) and the monthly maintenance plan.

[0094] Example 2:

[0095] This embodiment discloses a turntable wheel vibration defect early warning system, applied to the turntable wheel defect detection system described in Embodiment 1, including:

[0096] The vibration sensing unit is independently encapsulated in a metal shielding shell and fixed to the side wall of the wheel defect detection device shell by a mounting bracket. It is mechanically isolated from the micro switch. The vibration sensing unit is connected to the ESP32 microcontroller through the SPI interface, sharing the same PoE power module but with independent signal processing. It includes a triaxial MEMS accelerometer. The sampling frequency of the triaxial MEMS accelerometer is set to 10kHz and the range is set to ±50g. It is used to collect the three-dimensional vibration acceleration signal when the wheel passes by.

[0097] The frequency domain analysis unit is configured to perform fast Fourier transform processing on the three-dimensional vibration acceleration signal to extract spectral feature parameters in the frequency range of 0.5kHz to 5kHz.

[0098] The defect identification unit is configured to establish a multi-dimensional defect identification model based on spectral feature parameters to identify the early material damage type of the walking wheel;

[0099] The decision fusion unit is configured to perform weighted fusion calculations on the vibration analysis results and the time interval detection results of the wheel defect detection device to generate a comprehensive defect score and trigger the corresponding early warning level.

[0100] The spectral characteristic parameters include:

[0101] The fundamental frequency amplitude parameter corresponds to the vibration amplitude of the fundamental frequency of the rotating wheel;

[0102] The first harmonic amplitude parameter and the second harmonic amplitude parameter correspond to the vibration amplitudes of the second and third harmonic frequencies, respectively.

[0103] The spectral entropy parameter characterizes the energy distribution complexity of a vibration signal within the frequency domain.

[0104] The characteristic frequency band energy ratio parameter is used to calculate the ratio of vibration energy to total vibration energy in the 1kHz to 3kHz frequency band.

[0105] Furthermore, the multi-dimensional defect identification model includes the following judgment rules:

[0106] Wear defect judgment rule: Based on the physical principle that the stiffness of the wheel material decreases due to wear, when the fundamental frequency amplitude parameter decreases by more than 30% and the ratio of the first harmonic amplitude parameter to the fundamental frequency amplitude parameter increases by more than 50%, it is judged as a wear defect;

[0107] Crack defect judgment rule: When the characteristic frequency band energy ratio parameter exceeds 0.15 and obvious resonance peaks appear in the frequency range of 2.3kHz to 2.7kHz, it is judged as a crack defect;

[0108] Bearing damage judgment rule: When a sideband modulation phenomenon centered on the fundamental frequency appears in the vibration spectrum, and the amplitude of the modulation sideband exceeds 20% of the fundamental frequency amplitude, it is judged as bearing damage.

[0109] Furthermore, the decision fusion unit executes a weighted fusion algorithm:

[0110] Overall defect score = 0.7 × vibration defect score + 0.3 × geometric defect score;

[0111] The vibration defect score is calculated based on the output of a multi-dimensional defect identification model, and the specific calculation rules are as follows:

[0112] When wear defects are identified, the vibration defect score = 0.3 + min(1, max(0, (percentage decrease in fundamental frequency amplitude - 30%) / 70%)) × 0.4;

[0113] When a crack defect is identified, the vibration defect score = 0.3 + (characteristic frequency band energy ratio parameter - 0.15) / 0.25 × 0.4;

[0114] When bearing damage is identified, the vibration defect score = 0.3 + (sideband modulation amplitude percentage - 20%) / 30% × 0.4;

[0115] When multiple defect types are identified simultaneously, the maximum score among the scores of each defect type is taken as the final vibration defect score.

[0116] The geometric defect score is calculated based on the degree of deviation between the trigger time interval of the wheel defect detection device and the preset normal interval;

[0117] An early warning signal is triggered when the vibration defect score is greater than 0.6 and the geometric defect score is less than 0.3.

[0118] An emergency alarm signal is triggered when the overall defect score is greater than 0.8.

[0119] The vibration sensing unit also includes a temperature compensation module. The temperature compensation module monitors the ambient temperature in real time through a temperature sensor and corrects the sensitivity drift of the triaxial MEMS accelerometer. The temperature compensation coefficient is set to -0.02% / ℃.

[0120] In addition, this embodiment also discloses a method for detecting defects in turntable wheels, including a vibration early warning function, applied to the aforementioned turntable wheel vibration defect early warning system, characterized by including the following steps:

[0121] The three-dimensional vibration acceleration signal when the walking wheel passes by is collected in real time using a triaxial MEMS accelerometer, with a sampling frequency of 10kHz.

[0122] The three-dimensional vibration acceleration signal is subjected to bandpass filtering and windowing function processing from 0.5kHz to 5kHz.

[0123] Perform a fast Fourier transform on the processed vibration signal and calculate the spectral characteristic parameters;

[0124] Based on the judgment rules of the multi-dimensional defect identification model, the defect type of the walking wheel is identified and the severity of the defect is calculated;

[0125] The vibration analysis results and the time interval detection results are weighted and fused together.

[0126] Based on the comprehensive defect score, the corresponding level of early warning or alarm signal will be triggered.

[0127] The following describes this embodiment in detail with reference to specific implementation details: Based on Embodiment 1, this embodiment further discloses the specific implementation method of the turntable traveling wheel vibration defect early warning system:

[0128] The hardware configuration is as follows:

[0129] A triaxial MEMS accelerometer, model ADXL1002, is integrated inside the housing of the wheel defect detection device 104. The sampling frequency is set to 10kHz, and the measurement range is set to ±50g. The vibration sensing unit is connected to the ESP32 microcontroller 107 via an SPI interface, with a data transmission rate set to 2Mbps. The vibration sensing unit also includes a DS18B20 temperature sensor for temperature compensation of the triaxial MEMS accelerometer.

[0130] The method for extracting spectral feature parameters is as follows:

[0131] Vibration signal processing includes the following steps:

[0132] First, a Hanning window function is applied to the acquired three-dimensional vibration acceleration signal to reduce spectral leakage. Then, a fast Fourier transform is performed on the windowed vibration signal to convert the time-domain signal into a frequency-domain signal. Finally, the following spectral feature parameters are extracted from the frequency-domain signal:

[0133] Fundamental frequency amplitude parameter: Find the frequency corresponding to the maximum amplitude in the frequency range of 0-500Hz as the fundamental frequency, and record the amplitude at that frequency as the fundamental frequency amplitude parameter;

[0134] Harmonic amplitude parameters: Calculate the vibration amplitude at the second and third harmonic frequencies, and use them as the first harmonic amplitude parameters and the second harmonic amplitude parameters, respectively;

[0135] Spectral entropy parameter: calculates the energy distribution probability of the frequency domain signal and calculates the spectral complexity based on the information entropy formula;

[0136] Characteristic frequency band energy ratio parameter: Calculate the ratio of vibration energy in the 1kHz to 3kHz frequency band to vibration energy across the entire frequency band.

[0137] The defect identification model is implemented as follows:

[0138] The multi-dimensional defect identification model is implemented based on the following rules:

[0139] Wear defect identification: When the fundamental frequency amplitude parameter drops by more than 30% relative to the normal reference value, and the ratio of the first harmonic amplitude parameter to the fundamental frequency amplitude parameter increases by more than 50%, the wear defect of the walking wheel is determined.

[0140] Crack defect identification: When the characteristic frequency band energy ratio parameter exceeds 0.15 and a significant resonance peak appears in the frequency range of 2.3kHz to 2.7kHz, the wheel is determined to have a crack defect.

[0141] Bearing damage identification: When a sideband modulation phenomenon centered on the fundamental frequency appears in the vibration spectrum, and the amplitude of the modulation sideband exceeds 20% of the fundamental frequency amplitude, it is determined that the traveling wheel bearing is damaged.

[0142] Enhanced testing process:

[0143] Based on the detection process in Example 1, the following vibration analysis steps are added:

[0144] When the walking wheel 101 passes the walking wheel defect detection device 104, the triaxial MEMS accelerometer synchronously acquires a three-dimensional vibration acceleration signal for 100ms, collecting a total of 1000 data points. The ESP32 microcontroller 107 performs spectrum analysis on the vibration signal and extracts the following spectral characteristic parameters: fundamental frequency amplitude parameter is 0.85, first harmonic amplitude parameter is 0.12, second harmonic amplitude parameter is 0.08, spectral entropy parameter is 2.1, and characteristic frequency band energy ratio parameter is 0.09.

[0145] The defect identification unit, based on a multi-dimensional defect identification model, determined that all spectral characteristic parameters were within the normal range and that no early defect characteristics were present. The decision fusion unit calculated a vibration defect score of 0.15, a geometric defect score of 0.04, and a comprehensive defect score of 0.7 × 0.15 + 0.3 × 0.04 = 0.117. Since the comprehensive defect score was less than the warning threshold of 0.6, the system determined that the walking wheel 10110 was in normal condition and did not trigger a warning signal.

[0146] Example of anomaly detection:

[0147] When testing the walking wheel 101, the vibration analysis results showed that the fundamental frequency amplitude parameter was 0.45 (a decrease of 47% compared to the normal value), the ratio of the first harmonic amplitude parameter to the fundamental frequency amplitude parameter was 0.85 (an increase of 210%), and the characteristic frequency band energy ratio parameter was 0.18. Time interval analysis showed that the trigger time interval deviated from the preset value by 4%, while the geometric dimensions were normal.

[0148] Based on the wear defect judgment rules, the defect identification unit determined that the traveling wheel 101 exhibits early wear characteristics. The decision fusion unit calculated a vibration defect score of 0.75, a geometric defect score of 0.04, and a comprehensive defect score of 0.537. Since the vibration defect score is greater than 0.6 and the geometric defect score is less than 0.3, the system triggered an early warning signal, indicating that the traveling wheel 10125 exhibits early wear characteristics and recommends enhanced monitoring.

[0149] Through 30 days of continuous testing of 50 walking wheel samples, including 10 walking wheels with pre-existing defects, the following technical performance data were obtained:

[0150] Early warning accuracy: 92.3% (48 / 52 warning events); False alarm rate: 4.7% (2 / 43 no-defect alarms); Average warning lead time: 21.5 days earlier than single geometric detection method (range 15-28 days); Defect type identification accuracy: wear defects 88%, crack defects 85%, bearing damage 90%.

[0151] In this embodiment, the vibration sensing unit is integrated inside the housing of the existing wheel defect detection device 104, without changing the external structural dimensions and installation method, and is fully compatible with the original hardware.

[0152] A vibration signal processing thread was added to the ESP32 microcontroller 107, which runs in parallel with the original microswitch signal processing thread and shares the same system resources and communication interface.

[0153] Vibration data, along with the original trigger signal and crossarm number data, are encapsulated into a data packet and uploaded to the system platform layer through the same PoE fiber optic switch 108, with backward compatibility in data format.

[0154] A vibration early warning function has been added to the real-time monitoring and alarm module at the system platform layer, and a vibration spectrum data storage and analysis function has been added to the data statistics and analysis module, while maintaining a consistent user interface style.

[0155] The vibration sensing unit obtains power from the existing PoE power supply system, and the power consumption increases within the range of PoE power supply capability.

[0156] This embodiment integrates a triaxial MEMS accelerometer into a wheel defect detection system based on the physical principles of vibration spectrum analysis to form a vibration defect early warning system. When early damage occurs in the wheel material, its structural dynamic characteristics undergo fundamental changes, manifested as specific characteristic frequency components in the vibration spectrum. By analyzing the spectral characteristic parameters in the 0.5kHz to 5kHz frequency band, including fundamental frequency amplitude, harmonic components, spectral entropy, and characteristic band energy ratio, early material damage that cannot be identified by traditional geometric detection methods can be detected.

[0157] This embodiment overcomes the inherent limitations of single geometric dimension detection. Traditional time interval detection can only identify defects after the wheel has undergone significant geometric deformation, while vibration spectrum analysis can detect anomalies in the early stages of microscopic damage to the material, advancing the defect identification time from "after geometric deformation" to "the early stage of material damage." Wear defects cause a drop in fundamental frequency amplitude of more than 30% and a significant increase in harmonic components; crack defects produce characteristic resonance peaks in the 2.3kHz to 2.7kHz frequency band; bearing damage causes spectral sideband modulation.

[0158] This embodiment establishes a comprehensive defect scoring system by weightedly fusing vibration spectrum analysis results with time interval detection results, where vibration analysis accounts for 70% and geometric detection accounts for 30%. This fusion mechanism enables accurate identification of different defect stages: early warning is triggered when vibration is abnormal but geometry is normal, and an emergency alarm is triggered when the comprehensive score exceeds the standard.

[0159] Verified through 30 days of continuous testing on 50 wheels (including 10 wheels with known defects), this vibration early warning solution reduces the defect identification time by an average of 21.5 days compared to the traditional single geometric detection solution, and improves the defect identification accuracy to 92.3%, truly achieving a technological leap from post-maintenance to predictive maintenance.

[0160] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0161] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. It should be noted that any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A defect detection system for turntable wheels, characterized in that: It includes the field perception layer, network transmission layer, and system platform layer; The field sensing layer includes a wheel defect detection device, a crossarm detection device, and a periodic reset device. The wheel defect detection device is used to detect the defect status of the wheel and includes a device housing, microswitches, and a connecting mechanism. The microswitches include two sets, which are symmetrically installed inside the device housing. When the wheel passes by, the microswitches are triggered by the surface of the wheel to generate a trigger signal. The trigger signal includes a trigger timestamp and a trigger status. The field sensing layer also includes a speed detection device. The speed detection device uses an incremental rotary encoder installed on the turntable drive shaft to measure the turntable running speed in real time, or calculates the turntable running speed by the time interval between adjacent crossarms passing through the crossarm detection device. The crossarm detection device is used to identify the crossarm number and includes a photoelectric switch and an optical signal reflector. The optical signal reflector is installed on each crossarm and has a unique code. The photoelectric switch emits an optical signal and receives the reflected signal from the optical signal reflector to determine the crossarm position and number. The cycle reset device is used to determine the turntable's operating cycle. It includes a photoelectric switch and an optical signal reflector. The optical signal reflector is installed on the first crossarm of the turntable. When the photoelectric switch detects the first crossarm, it triggers the cycle reset and packages the data. The network transmission layer, based on PoE technology, includes a PoE fiber optic switch and network cables, providing DC power and data transmission to the field sensing layer. The system platform layer receives data and includes a real-time monitoring and alarm module and a data statistics and analysis module. The real-time monitoring and alarm module determines whether an alarm threshold has been reached based on the trigger signal from the wheel defect detection device and issues an alarm accordingly. The data statistics and analysis module is used for historical data recording, visualization, and maintenance reminders; the microswitch of the walking wheel defect detection device is a high-precision mechanical microswitch with a triggering force range of 0.1N to 1.0N; the field sensing layer also includes a speed detection device, which uses an incremental rotary encoder mounted on the turntable drive shaft to measure the turntable's running speed in real time, or calculates the turntable's running speed by the time interval between adjacent crossarms passing through the crossarm detection device; the field sensing layer is based on an ESP32 microcontroller, which is configured to process the input signals of the microswitch and photoelectric switch, calculate the triggering time interval, and perform data preprocessing; the walking wheel defect detection device, crossarm detection device, and periodic reset device are fixed near the turntable track by bolts; the real-time monitoring and alarm module of the system platform layer also includes a heartbeat detection mechanism to monitor the online status of the field sensing layer devices, and if the device is offline for more than a set time, an offline alarm signal is issued.

2. The turntable wheel defect detection system according to claim 1, characterized in that: The optical signal reflector of the crossarm detection device is an infrared optical signal reflector, and the photoelectric switch is an infrared photoelectric switch with an emission frequency of 38kHz. The unique code is achieved by combining different reflectivity regions set on the optical signal reflector, and the photoelectric switch identifies the code value by detecting changes in the intensity of the reflected signal.

3. The turntable wheel defect detection system according to claim 1, characterized in that: The data packaging of the periodic reset device includes encapsulating the trigger signals, crossarm numbers and timestamps of all traveling wheels in the current cycle into data packets, and uploading them to the system platform layer through the network transmission layer.

4. A method for detecting defects in turntable wheels, applied to a turntable wheel defect detection system as described in any one of claims 1-3, characterized in that, Includes the following steps: PoE provides DC power to the wheel defect detection device, crossarm detection device, and periodic reset device; Each turntable crossarm has a traveling wheel on its upper and lower sides. The crossarms and traveling wheels are numbered and bound together to ensure precise positioning of the traveling wheels. Crossarm data is collected in real time using a crossarm detection device; Data on the walking wheels is collected in real time using a walking wheel defect detection device; The turntable's operating cycle is checked by a cycle reset device. After the turntable starts running, if the crossarm detection value matches the preset code value of the first crossarm, it is determined that the turntable has completed one cycle, and the crossarm number is reset. If they do not match, the operating cycle determination continues. After the crossarm number is reset, the collected data of the walking wheels will be uploaded to the host computer monitoring and alarm system via the lower-level microcontroller and PoE fiber optic switch. The monitoring and alarm system analyzes the data and displays it through a visual human-computer interaction interface; On the one hand, it determines whether the damage to the walking wheel has reached the set alarm threshold. If the alarm threshold is reached, the system issues a walking wheel fault alarm signal; if the alarm threshold is not reached, the walking wheel data continues to be analyzed in real time. On the other hand, the online status of the field device is monitored in real time through a heartbeat detection mechanism. If the device is determined to be offline, an offline alarm signal is issued; if the device recovers from the offline state, the alarm signal is cleared. If the device is not offline, it will continue to monitor; The monitoring and alarm system records and archives data and alarm event information, and the data can be exported for in-depth analysis or reporting. The data statistics and analysis module of the system platform layer is also used to train a trend prediction model based on historical trigger data, predict the remaining life of the walking wheels and generate a maintenance plan. The trend prediction model is a linear regression model. The inputs include the trigger time interval change rate and the crossarm load data. The trigger time interval change rate is calculated from historical Δt data. The crossarm load data is obtained in any of the following ways: (a) real-time measurement by a pressure sensor installed on the crossarm. The pressure sensor has a range of 0-100kg and an accuracy of ±1%; (b) obtained from the real-time baggage weight data table in the airport baggage system control database. The data table contains the crossarm number, timestamp and load weight fields, and the data update frequency is not less than 1Hz.

5. The method for detecting defects in a turntable wheel according to claim 4, characterized in that, The formula for calculating the trigger time interval is: ,in For the trigger timestamp of the first group of microswitches, The second set of microswitches is triggered by a time stamp with a preset normal interval. Calculations based on the diameter D of the traveling wheel and the linear velocity v of the turntable: Where L is the installation distance between the two sets of microswitches, and v is the tangential linear velocity at the contact point between the traveling wheel and the track.

Images