A Real-Time Online Measurement System for Bearing Coating Wear Based on Ultrasonic Longitudinal and Transverse Wave Fusion

Abstract

This invention proposes a real-time online measurement system for bearing coating wear based on ultrasonic longitudinal and transverse wave fusion. It includes a signal control and acquisition module and a PC integrating a wear detection platform and a wear detection classification platform, connected sequentially. The signal control and acquisition module controls and acquires temperature signals and longitudinal / transverse wave ultrasonic signals. The PC integrating the wear detection platform and the wear detection classification platform calculates the real-time wear of the bearing coating and classifies the wear condition. The signal control and acquisition module includes multiple ultrasonic sensor groups, each including at least one ultrasonic longitudinal wave sensor and at least one ultrasonic transverse wave sensor. The ultrasonic longitudinal wave sensor and the ultrasonic transverse wave sensor in each ultrasonic sensor group are positioned at the same detection point on the bearing to be tested. This invention improves detection accuracy while enabling early fault prediction.

Description

Technical Field

[0001] This utility model relates to the technical field of bearing safety monitoring, and in particular to a real-time online measurement system and method for bearing coating wear. Background Technology

[0002] Sliding bearings, as one of the most commonly used bearing structures, are an important component of rotating mechanisms, and their structural integrity directly affects the operational safety and service life of equipment. The bearing bush, as the part of the sliding bearing structure that directly contacts the rotating shaft, often needs to bear a large load. To reduce friction between the bearing bush and the shaft and to protect the bearing structure, the inner surface of the bearing bush is often covered with a layer of anti-friction material, also known as a bearing bush coating. In practical applications, although the coating provides some lubrication as it is the part that directly contacts the shaft journal, friction and wear during operation are unavoidable. Therefore, in some cases, the coating structure is designed to be sacrificed to protect the more critical rotating shaft, such as soft metal coatings like Babbitt metal, copper alloys, and aluminum alloys. Furthermore, in environments requiring high wear resistance and low friction, non-metallic coatings such as ceramics and diamond-like carbon (DLC) are also commonly used on the bearing bush surface.

[0003] Bearing shells are structurally classified into two types: integral and split. Integral bearing shells are also known as bushings. Due to their advantages such as low cost, simple structure, convenient installation, and easy replacement, bearing shells are widely used in many fields, including industry and transportation, and are particularly suitable for high-load scenarios, such as the main shaft of internal combustion engines and the stern tube bearing system of ships. Since lubrication is usually used between the bearing shell and the shaft, under sufficient lubrication, the lubricant can separate the surfaces of the shaft and the bearing shell. Therefore, ideally, no mechanical wear should occur between the shaft and the bearing shell. However, in actual operation, due to external factors such as friction, overload, impact, and mismatch, or lubrication factors such as insufficient, failed, or contaminated lubricant, contact and friction between the shaft and the bearing shell, as well as wear on the bearing shell surface, are unavoidable. Once the bearing shell begins to wear, the gap between the shaft and the bearing shell will increase, the oil cannot flow along its normal path, the lubrication system pressure will drop, directly leading to accelerated component wear. When the wear of the bearing shell exceeds the allowable limit, dry friction may occur between the bearing shell and the journal, causing the alloy layer on the surface of the bearing shell to melt at high temperatures, further aggravating the wear, and even causing the bearing shell to burn and melt onto the journal, resulting in serious failure. Taking an engine as an example, severe bearing shell wear can cause the connecting rod to collide with the cylinder liner, causing the connecting rod to bend or break, resulting in serious consequences such as connecting rod breakage and cylinder block penetration.

[0004] Currently, acoustic emission and vibration monitoring technologies are commonly used for monitoring the operational safety of sliding bearings. These technologies aim to monitor fault and vibration signals during system operation and identify faults or abnormal vibration states based on the frequency characteristics of the signals. However, acoustic emission and vibration monitoring have significant limitations. First, they are passive measurements; sensors are often installed externally, receiving weak signals generated by faults, making them highly susceptible to external noise and prone to misjudgments and missed detections. Second, acoustic emission and vibration sensors are often designed as broadband sensors, meaning they can capture signals over a wide frequency range. However, this characteristic also brings drawbacks. Because the acquired signals are relatively complex and have small amplitudes, subsequent signal amplification is often required to meet analytical needs. This signal amplification not only increases the complexity of the equipment but also occupies a large amount of space. These factors combined make it difficult to use these sensors to accurately quantitatively analyze the operating status of bearing systems, hindering the precise assessment and judgment of the actual working condition of the bearing.

[0005] Wear on the inner surface coating of bearing bushes can be categorized into three typical surface morphologies: plastic deformation, surface scratches, and large-area material removal. Similar surface morphologies may appear at different stages of wear, and different surface morphologies may alternate within the same stage. Compared to other non-destructive testing methods, ultrasonic sensors have the advantages of small sensor size, low hazard to humans, and robustness against electromagnetic interference. Early ultrasonic wear testing primarily utilized longitudinal wave detection. However, with the increasing demand for detecting complex microstructures and defects within materials, when longitudinal waves encounter small wear areas (such as plastic deformation or surface scratches), lateral reflection occurs due to the difference in elastic modulus and density between the wear area and the surrounding normal material, affecting wear detection. Utility Model Content

[0006] To address the technical problem that existing technologies cannot accurately reflect the wear state of bearing coatings during online operation in real time, this invention proposes a real-time online measurement system for bearing coating wear based on ultrasonic longitudinal and transverse wave fusion. By exciting ultrasonic longitudinal / transverse wave sensors to send ultrasonic signals, which are reflected at the wear point, the system receives the reflected ultrasonic longitudinal / transverse wave signals and calculates the wear degree in real time based on these signals. Furthermore, it uses machine learning methods to distinguish surface morphology, improving detection accuracy and enabling early fault prediction.

[0007] To achieve the above objectives, the technical solution of this utility model is implemented as follows:

[0008] A real-time online measurement system for bearing coating wear based on ultrasonic longitudinal and transverse wave fusion includes a signal control and acquisition module and a PC integrating a wear detection platform and a wear detection classification platform, connected in sequence. The signal control and acquisition module is used for controlling and acquiring temperature signals and longitudinal / transverse wave ultrasonic signals. The PC integrating the wear detection platform and the wear detection classification platform is used for calculating the real-time wear of the bearing coating and classifying the wear condition. The signal control and acquisition module includes multiple ultrasonic sensor groups, each ultrasonic sensor group including at least one ultrasonic longitudinal wave sensor and at least one ultrasonic transverse wave sensor. The ultrasonic longitudinal wave sensor and the ultrasonic transverse wave sensor in each ultrasonic sensor group are set at the same detection point on the bearing to be tested.

[0009] Specifically, the signal control and acquisition module further includes an ultrasonic data acquisition card. The signal receiving end of the ultrasonic data acquisition card is connected to the ultrasonic longitudinal wave sensor or ultrasonic transverse wave sensor in the multiple ultrasonic sensor groups, and the ultrasonic data acquisition card is connected to the PC.

[0010] Specifically, the signal control and acquisition module also includes a temperature data acquisition card and a temperature sensor connected to the communication network. The temperature sensor is located on one side of the ultrasonic longitudinal wave sensor or ultrasonic transverse wave sensor in each ultrasonic sensor group, and the temperature data acquisition card is connected to the PC.

[0011] Specifically, the signal control and acquisition module further includes an ultrasonic multiplexer and a microcontroller. The input terminal of the microcontroller is connected to the PC for communication, the output terminal of the microcontroller is connected to the control terminal of the ultrasonic multiplexer, the input terminal of the ultrasonic multiplexer is connected to the trigger terminal of the ultrasonic data acquisition card, and the output terminal of the ultrasonic multiplexer is connected to the ultrasonic longitudinal wave sensor or ultrasonic transverse wave sensor in each ultrasonic sensor group.

[0012] Specifically, the ultrasonic longitudinal wave sensor or ultrasonic transverse wave sensor is tightly attached to the outer surface of the bearing to be tested through a conductive adhesive layer. The surface of the ultrasonic transverse wave / longitudinal wave sensor is covered with an epoxy resin-tungsten powder mixture for electromagnetic shielding and mechanical protection. The signal transmission end of the ultrasonic longitudinal wave sensor or ultrasonic transverse wave sensor is connected to the output end of the ultrasonic data acquisition card and the ultrasonic multiplexer respectively through a coaxial cable.

[0013] Specifically, the ultrasonic longitudinal wave sensor or ultrasonic transverse wave sensor in each ultrasonic sensor group is connected to the same selection channel of the ultrasonic multiplexer.

[0014] Specifically, the center frequency of the ultrasonic longitudinal wave sensor is 10MHz-30MHz, and the center frequency of the ultrasonic transverse wave sensor is 5MHz-10MHz.

[0015] Specifically, the conductive adhesive layer is a thin conductive adhesive layer with a size of 0.1-0.3 mm.

[0016] Specifically, the temperature sensor is thermally connected to the outer surface of the bearing bush via conductive epoxy adhesive.

[0017] Specifically, the temperature sensor is a K-type thermocouple sensor.

[0018] The beneficial effects of this utility model are as follows:

[0019] This invention combines the advantages of ultrasonic longitudinal wave and ultrasonic transverse wave detection. The signal control and acquisition module and the LabVIEW virtual instrument wear testing platform work together to perform in-depth analysis of the signal through machine learning algorithms, thereby achieving real-time and accurate monitoring of the wear on the inner surface of the bearing bush. The sensor has strong adaptability, and multi-channel acquisition and efficient control improve measurement efficiency. The measurement results are accurate and reliable, and faults can be predicted in advance, reducing equipment maintenance costs and improving operating efficiency, resulting in significant economic and social benefits. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of this utility model or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this utility model. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 An ultrasonic sensor integrated into the outer surface of a bearing bush is designed for this utility model.

[0022] Figure 2 This is a schematic diagram of the ultrasonic signal control and acquisition module of this utility model.

[0023] Figure 3 This is a flowchart illustrating the data control and acquisition process of the LabVIEW virtual instrument testing platform of this utility model. Detailed Implementation

[0024] The technical solutions of the present utility model will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of the present utility model, and not all embodiments. Based on the embodiments of the present utility model, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the protection scope of the present utility model.

[0025] A real-time online measurement system for bearing coating wear based on ultrasonic longitudinal and transverse wave fusion, such as Figure 2As shown, it includes a signal control and acquisition module and a PC that integrates a wear detection platform and a wear detection classification platform, connected in sequence.

[0026] The system's signal control and acquisition module is used for controlling and acquiring temperature signals and different ultrasonic signals, and sends the returned temperature signals and returned longitudinal / transverse ultrasonic signals to the wear detection platform and the wear detection classification platform.

[0027] The wear detection platform (based on LabVIEW virtual instrument testing platform) is used to perform real-time analysis and processing of the returned temperature signal and the returned longitudinal / transverse ultrasonic signal to obtain time-domain and frequency-domain characteristics, and calculate the wear degree based on the time-domain characteristics and display and record it in real time.

[0028] The wear detection and classification platform described above extracts features and classifies wear types based on the time-domain and frequency-domain features obtained by the wear detection platform.

[0029] Specifically, the signal control and acquisition module of the system includes multiple ultrasonic sensor groups, an ultrasonic data acquisition card (PicoScope), an ultrasonic multiplexer (TMUX1109), a microcontroller (Arduino Uno), a temperature data acquisition card (TC-08), a temperature sensor (K-type thermocouple sensor), and a DC power supply.

[0030] The PC integrating the wear detection platform and wear detection classification platform is connected to the ultrasonic data acquisition card, microcontroller, and temperature data acquisition card via USB interfaces. The ultrasonic data acquisition card has an internal trigger terminal and multiple signal receiving terminals (eight in this embodiment), designated as ports 1 / 5, 2 / 6, 3 / 7, and 4 / 8 for the ultrasonic transverse / longitudinal wave sensors. Each signal receiving terminal is connected to a corresponding ultrasonic sensor group, which is mounted on the bearing to be tested. The internal trigger terminal is connected to the input terminal of an ultrasonic multiplexer. The output terminal of the ultrasonic multiplexer is connected to the multiple ultrasonic sensor groups, and the control terminal of the ultrasonic multiplexer is connected to the microcontroller. The temperature data acquisition card is connected to a temperature sensor mounted on the bearing to be tested. The microcontroller and temperature data acquisition card are powered by the PC. A DC power supply is connected to both the ultrasonic data acquisition card and the ultrasonic multiplexer to meet their power requirements.

[0031] Each ultrasonic sensor group consists of one longitudinal wave sensor and one transverse wave sensor, positioned at the same detection point on the bearing to be tested. The longitudinal wave sensor (Panametrics V101-RM) with a center frequency of 10MHz-30MHz or the transverse wave sensor (Olympus V312-SU) with a center frequency of 5MHz-10MHz is selected. Taking the arrangement of one ultrasonic sensor as an example, as shown in the figure, a thin conductive adhesive layer (0.1-0.3mm, Epo-Tek) is applied to one side of the ultrasonic transverse / longitudinal wave sensor. The H2OE (hydrocarbon epoxy) is tightly bonded to the outer surface of the bearing to be tested, ensuring electrical conductivity between the ultrasonic shear / longitudinal wave sensor and the outer surface of the bearing. The surface of the ultrasonic shear / longitudinal wave sensor is covered with an epoxy resin-tungsten powder mixture (tungsten powder volume percentage 60%-70%) for electromagnetic shielding and mechanical protection. After the signal transmission end of the ultrasonic shear / longitudinal wave sensor is soldered to the coaxial cable, it is connected to the corresponding signal receiving end of the ultrasonic data acquisition card and the output end of the ultrasonic multiplexer via the coaxial cable. Temperature sensors are set on one side of the ultrasonic longitudinal wave sensor or ultrasonic shear wave sensor in each ultrasonic sensor group. The temperature sensors are thermally connected to the outer surface of the bearing through conductive epoxy adhesive. In this embodiment, there are 8 temperature sensors corresponding to the ultrasonic shear / longitudinal wave sensors. Figure 2 It is not shown in the middle.

[0032] A real-time online measurement method for bearing coating wear based on ultrasonic longitudinal and transverse wave fusion includes the following steps:

[0033] S1: Install ultrasonic transverse wave sensors, ultrasonic longitudinal wave sensors, and temperature sensors at different locations on the bearing bush, and configure the acquisition test matrix, channel parameters, and sampling parameters through the wear detection platform.

[0034] Four measuring points are evenly arranged circumferentially at 90° intervals on the outer surface of the bearing bush. Each measuring point is simultaneously equipped with an ultrasonic longitudinal wave sensor and an ultrasonic transverse wave sensor. The ultrasonic longitudinal wave sensor is a Panametrics V101-RM with a center frequency of 20MHz, and the ultrasonic transverse wave sensor is an Olympus V312-SU with a center frequency of 5MHz. The sensors are tightly bonded to the bearing bush surface through a 0.2mm thick Epo-Tek H20E conductive adhesive layer, and the surface is covered with a shielding layer of epoxy resin-tungsten powder mixture with a 70% tungsten powder volume ratio. A temperature sensor (K-type thermocouple sensor) with an accuracy of ±0.5℃ is attached next to each measuring point and is thermally connected to the outer surface of the bearing bush through conductive epoxy adhesive.

[0035] The acquisition test matrix is ​​used to determine the transmission and reception modes of ultrasonic signals; four sets of "transmit-receive" modes are defined, corresponding to four measurement points, with each set having a working cycle of 100μs, including: transmission phase (10μs): longitudinal wave / transverse wave sensors sequentially transmit ultrasonic signals (longitudinal wave pulse width 0.5μs, transverse wave pulse width 1μs); reception phase (90μs): synchronous acquisition of primary / secondary echo signals.

[0036] Channel parameters are used to select the working channels of the ultrasonic data acquisition card and the ultrasonic multiplexer, and to define the signal transmission path. On the ultrasonic data acquisition card (PicoScope6404D), channels 1-4 correspond to the longitudinal wave sensor, and channels 5-8 correspond to the transverse wave sensor. The ultrasonic multiplexer (TMUX1109) is configured with 4 input channels, mapped to measurement points 1-4 via I2C address 0x70. For ease of representation... Figure 1 The longitudinal wave sensor channel 1 and the transverse wave sensor channel 5 are referred to as channel 1 / 5.

[0037] The sampling parameters determine the frequency and accuracy of the data acquired by the ultrasonic data acquisition card and the temperature data acquisition card. The ultrasonic signal trigger signal frequency is 10kHz, and the sampling rate is set to 100MHz to meet the Nyquist criterion and cover a bandwidth of 30MHz longitudinal wave and 10MHz transverse wave. The temperature signal sampling rate is set to 10Hz, and synchronization with the ultrasonic signal acquisition cycle is achieved through the DAQmx function library of the LabVIEW virtual instrument testing platform.

[0038] S2: The wear detection platform sends instructions to the ultrasonic data acquisition card and controller. The ultrasonic data acquisition card opens the corresponding receiving channel according to the channel parameters, and the microcontroller selects the channel of the ultrasonic multiplexer according to the channel parameters.

[0039] The LabVIEW virtual instrument testing platform sends initialization commands to the ultrasonic data acquisition card (PicoScope6404D) via USB interface. The ultrasonic data acquisition card opens the receiving channel of the corresponding measurement point, such as channel 1 / 5 corresponding to measurement point 1. The trigger signal parameters are set to rise edge trigger, trigger level 1V. The microcontroller (Arduino Uno R3) receives the LabVIEW command and sends a 4-bit digital level signal to the ultrasonic multiplexer (TMUX1109) via I2C bus to select the current working measurement point. After each set of measurement point acquisition is completed, wait 5ns to switch to the next measurement point to ensure that the signals of the four measurement points are connected to the acquisition card in a time-division manner.

[0040] S3: The ultrasonic data acquisition card sends a trigger signal from the selected channel according to the sampling parameters. At the same time, the ultrasonic longitudinal wave sensor and ultrasonic transverse wave sensor transmit ultrasonic signals and receive longitudinal / transverse wave ultrasonic signals returned from the bearing according to the trigger signal. Meanwhile, the temperature sensor collects the bearing temperature signal in real time.

[0041] The ultrasonic data acquisition card generates a 10kHz repetition frequency trigger signal according to the sampling parameters, driving the ultrasonic longitudinal wave sensor and ultrasonic transverse wave sensor to emit ultrasonic signals at their respective center frequencies of 20MHz and 5MHz. The signals propagate through the bearing substrate and generate primary / secondary reflection echoes upon encountering the substrate-coating interface (double-layer structure) or the inner surface (single-layer structure). A temperature sensor (K-type thermocouple sensor) collects the bearing surface temperature in real time and synchronously uploads it to the PC at a frequency of 10Hz via the TC-08 temperature data acquisition card. The ultrasonic longitudinal wave sensor and ultrasonic transverse wave sensor receive the echo signals and transmit them to the ultrasonic data acquisition card via a 50Ω coaxial cable.

[0042] S4: The ultrasonic longitudinal / transverse wave sensor sends the longitudinal / transverse wave ultrasonic signals returned by the bearing to the ultrasonic data acquisition card. The ultrasonic data acquisition card processes the received longitudinal / transverse wave ultrasonic signals and sends them to the wear detection platform. The temperature signal is processed by the temperature data acquisition card and then sent to the wear detection platform.

[0043] The ultrasonic data acquisition card performs bandpass filtering on the echo signal (longitudinal wave: 10-30MHz, transverse wave: 5-10MHz) to remove noise and DC drift; the temperature signal is converted to digital by the TC-08 temperature data acquisition card and aligned with the ultrasonic signal timestamp.

[0044] S5: The wear detection platform calculates the coating wear thickness based on the ultrasonic pulse echo method according to the longitudinal / transverse ultrasonic signals and temperature signals returned by the bearing, and performs real-time analysis on the returned longitudinal / transverse ultrasonic signals and temperature signals to obtain the time-domain and frequency-domain characteristics of the returned signals.

[0045] Each ultrasonic sensor functions as both a transmitter and receiver. The emitted ultrasonic signal is reflected upon encountering the bearing boundary; the reflection pattern varies depending on the bearing structure. When the bearing is a single-layer structure, the ultrasonic signal incident from the outer surface is reflected once by the inner surface of the bearing, then reflected back to the outer surface and received; this is denoted as the "first ultrasonic reflection signal." A portion of the ultrasonic signal is then reflected again back to the inner surface of the bearing, undergoing a second reflection, and then reflected back to the outer surface and received; this is denoted as the "second ultrasonic reflection signal." When the bearing employs a double-layer structure of substrate + coating, the ultrasonic signal incident from the outer surface of the bearing undergoes an initial reflection at the interface between the substrate and coating. This reflected signal then returns to the outer surface of the bearing and is received; this signal is denoted as the "first ultrasonic reflection signal." Simultaneously, some ultrasonic signals penetrate the substrate-coating interface and enter the coating, resulting in a second reflection at the boundary between the coating and the outside. This second reflected signal then undergoes reflections at the substrate-coating interface and the outer surface of the bearing before finally being received; this signal is denoted as the "second ultrasonic reflection signal." The coating thickness can be obtained by comparing the time-of-flight (ToF) of the two reflected signals. The expression for the coating thickness d is as follows:

[0046] d=c(T)·Δt / 2

[0047] In the formula, the sound velocity c is a function of temperature T, obtained by calibrating a standard bearing with the same substrate material as the bearing under test. During calibration, the longitudinal / transverse wave sound velocities are collected every 10°C, and the function c(T) is obtained by fitting using the least squares method. Δt represents the time of flight (ToF) between two reflected signals. The reflection position is determined by the peak position of the reflected pulse. The wear state of the bearing can be analyzed by calculating the difference between the original thickness and the obtained coating thickness.

[0048] In this embodiment, sound velocity calibration was performed using a standard specimen (base steel + 0.5mm Babbitt alloy coating) made of the same material as the bearing to be tested. Sound velocity was collected every 10°C within the range of 20-120°C and obtained by least squares fitting.

[0049] Longitudinal wave velocity formula:

[0050] c p (T)=0.0001T 2 -0.005T+5900

[0051] Formula for transverse wave sound velocity:

[0052] c s (T)=-0.03T 2 +3200

[0053] The final wear thickness is calculated by weighting the coating thickness results of longitudinal and transverse waves, Δd = αΔd. p +βΔd s α and β are weighting coefficients, which are obtained by fitting a large amount of experimental data using the least squares method. In actual calculations, they can be implemented using Python's NumPy library.

[0054] S6: Based on the time-domain and frequency-domain features of the returned longitudinal / transverse ultrasonic signals and temperature signals, a hybrid classification network of 1DCNN-LSTM-attention mechanism is used for classification to obtain the bearing coating wear classification results.

[0055] The classification network consists of an input layer, a first 1DCNN module, a second 1DCNN module, a first LSTM layer, a second LSTM layer, an additive attention mechanism module, and a classification layer, which are connected in sequence. The first 1DCNN module is used for basic feature extraction. It uses 64 Conv1D convolutions with 5-core filters, a stride of 1, and ReLU activation function to capture feature combinations from 5 consecutive time points. Batch normalization (BatchNorm) is used to standardize the feature distribution, accelerating training and improving stability. MaxPooling1D (size 2) is used for dimensionality reduction to retain the main features. The second 1DCNN module is used for complex feature extraction. Its structure is similar to the first 1DCNN module. The Conv1D convolution uses 128 3-core filters to extract more refined local features, and LeakyReLU activation function is used. The first LSTM layer uses 256 memory units to process complex temporal relationships and provides a complete temporal feature sequence for the subsequent attention mechanism. The second LSTM uses 128 memory units to retain key temporal information while compressing dimensionality. The additive temporal attention mechanism module is used to focus on key signal features at the moment of defect occurrence (such as the time point of transverse wave distortion or longitudinal wave attenuation) to obtain the global context vector. The classification layer maps the global context vector to 3 class probabilities (surface scratch, plastic deformation, large area material removal) through a fully connected layer.

[0056] In this embodiment, 36,000 bearing samples with known wear states were collected. Each sample included three features (transverse wave signal, longitudinal wave signal, and temperature signal). The samples were labeled into three categories: surface scratch, plastic deformation, and large-area material removal. These samples were divided into training, validation, and test sets in an 8:1:1 ratio. The training set was subjected to noise enhancement, translation, and scaling to increase data diversity. A cosine annealing learning rate adjustment strategy was used to train the classification network using the training and validation sets, and the test set was used for testing. Finally, the trained and tested classification network was deployed in a computer for coating wear classification. Training parameters: batch size 32, epochs = 50, cosine annealing learning rate (initial 0.001, minimum 0.0001), data augmentation with Gaussian noise σ = 0.1, time shift ±50 points, and amplitude scaling ±10%.

[0057] The above description is only a preferred embodiment of the present utility model and is not intended to limit the present utility model. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present utility model should be included within the protection scope of the present utility model.

Claims

1. A real-time online measurement system for bearing coating wear based on ultrasonic longitudinal and transverse wave fusion, characterized in that, The system includes a signal control and acquisition module connected in sequence, and a PC integrating a wear detection platform and a wear detection classification platform. The signal control and acquisition module is used for controlling and acquiring temperature signals and longitudinal / transverse ultrasonic signals. The PC integrating the wear detection platform and the wear detection classification platform is used for calculating the real-time wear of the bearing coating and classifying the wear condition. The signal control and acquisition module includes multiple ultrasonic sensor groups. Each ultrasonic sensor group includes at least one ultrasonic longitudinal wave sensor and at least one ultrasonic transverse wave sensor. The ultrasonic longitudinal wave sensor and the ultrasonic transverse wave sensor in each ultrasonic sensor group are set at the same detection point on the bearing to be tested.

2. The real-time online measurement system for bearing coating wear based on ultrasonic longitudinal and transverse wave fusion as described in claim 1, characterized in that, The signal control and acquisition module further includes an ultrasonic data acquisition card. The signal receiving end of the ultrasonic data acquisition card is connected to the ultrasonic longitudinal wave sensor or ultrasonic transverse wave sensor in the multiple ultrasonic sensor groups. The ultrasonic data acquisition card is also connected to the PC.

3. The real-time online measurement system for bearing coating wear based on ultrasonic longitudinal and transverse wave fusion as described in claim 2, characterized in that, The signal control and acquisition module also includes a temperature data acquisition card and a temperature sensor connected to the communication network. The temperature sensor is located on one side of the ultrasonic longitudinal wave sensor or ultrasonic transverse wave sensor in each ultrasonic sensor group. The temperature data acquisition card is connected to the PC for communication.

4. The real-time online measurement system for bearing coating wear based on ultrasonic longitudinal and transverse wave fusion as described in claim 2 or 3, characterized in that, The signal control and acquisition module further includes an ultrasonic multiplexer and a microcontroller. The input terminal of the microcontroller is connected to a PC for communication, and the output terminal of the microcontroller is connected to the control terminal of the ultrasonic multiplexer. The input terminal of the ultrasonic multiplexer is connected to the trigger terminal of the ultrasonic data acquisition card, and the output terminal of the ultrasonic multiplexer is connected to the ultrasonic longitudinal wave sensor or ultrasonic transverse wave sensor in each ultrasonic sensor group.

5. The real-time online measurement system for bearing coating wear based on ultrasonic longitudinal and transverse wave fusion as described in claim 4, characterized in that, The ultrasonic longitudinal wave sensor or ultrasonic transverse wave sensor is tightly attached to the outer surface of the bearing to be tested by a conductive adhesive layer. The surface of the ultrasonic transverse wave / longitudinal wave sensor is covered with an epoxy resin-tungsten powder mixture for electromagnetic shielding and mechanical protection. The signal transmission end of the ultrasonic longitudinal wave sensor or ultrasonic transverse wave sensor is connected to the output end of the ultrasonic data acquisition card and the ultrasonic multiplexer respectively through a coaxial cable.

6. The real-time online measurement system for bearing coating wear based on ultrasonic longitudinal and transverse wave fusion as described in claim 4, characterized in that, The ultrasonic longitudinal wave sensor or ultrasonic transverse wave sensor in each ultrasonic sensor group is connected to the same selection channel of the ultrasonic multiplexer.

7. The real-time online measurement system for bearing coating wear based on ultrasonic longitudinal and transverse wave fusion as described in any one of claims 1-3 or 5, characterized in that, The ultrasonic longitudinal wave sensor has a center frequency of 10MHz-30MHz, and the ultrasonic transverse wave sensor has a center frequency of 5MHz-10MHz.

8. The real-time online measurement system for bearing coating wear based on ultrasonic longitudinal and transverse wave fusion as described in claim 5, characterized in that, The conductive adhesive layer is a thin conductive adhesive layer with a size of 0.1-0.3 mm.

9. The real-time online measurement system for bearing coating wear based on ultrasonic longitudinal and transverse wave fusion as described in claim 3, characterized in that, The temperature sensor is thermally connected to the outer surface of the bearing bush via conductive epoxy adhesive.

10. The real-time online measurement system for bearing coating wear based on ultrasonic longitudinal and transverse wave fusion as described in claim 9, characterized in that, The temperature sensor described is a K-type thermocouple sensor.

Images