A method and system for defect detection of a gear box

Abstract

The application provides a kind of gear box defect detection method and system, it is related to gear box test technical field.The application can more accurately detect the actual operating state of gear box by analyzing the infrared data and stress data of gear box under normal state and various defect states, improve defect detection accuracy.Construct stress analysis model based on infrared data and stress data, and then detect real-time infrared data, potential defects can be found in time during the operation of gear box, improve defect detection efficiency.And the application does not need to take sample regularly, only infrared detection is carried out on the inside of gear box, to improve the convenience of defect detection.The application realizes gear box multi-defect stress analysis and defect detection by infrared detection and stress detection on gear box, and improves the accuracy, convenience and detection efficiency of gear box defect detection.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of gear box testing, in particular to a gear box defect detection method and system. BACKGROUND

[0002] Under the background of the rapid development of intelligent tunneling technology, technological innovation and industrial upgrading are imperative. As the core transmission component of intelligent cantilever tunneling equipment in coal mines and the like, the stability of the performance of the gear box is directly related to the overall operation efficiency and safety of the tunneling equipment. However, in the actual working environment, the gear box faces the challenge of complex and variable working conditions, and the difficulty in obtaining real-time performance data, resulting in frequent gear defects such as tooth breakage, pitting, and severe wear. These problems not only hinder the efficient progress of tunneling operations, but also may cause equipment downtime and even safety accidents, posing a serious threat to production safety and progress.

[0003] Traditional gear box defect detection methods, such as vibration signal analysis and oil analysis, although can provide a basis for defect judgment to some extent, have significant limitations. Although vibration signal analysis can identify defects by capturing the vibration characteristics of the gear during operation, in actual application, factors such as noise interference and the complexity of the signal transmission path often result in a significant reduction in the accuracy of the analysis results. On the other hand, although oil analysis can assess the wear state of the gear by detecting wear particles in the lubricating oil, it has a long detection period and insufficient sensitivity to early defects, making it difficult to meet the needs of rapid response and accurate diagnosis.

[0004] In summary, the current traditional gear box defect detection technology faces core technical challenges such as difficulty in detection, insufficient accuracy, and the like in the application of intelligent tunneling equipment. SUMMARY

[0005] The present application provides a gear box defect detection method and system, which can improve the convenience and accuracy of gear box defect detection.

[0006] In a first aspect, the present application provides a gear box defect detection method, which comprises: acquiring infrared data and stress data of the gear box in a set period of time under various states; based on the infrared data in the set period of time under various states, performing finite element analysis to determine the temperature distribution data at each time in the set period of time under various states; based on the temperature distribution data at each time in the set period of time under various states, determining the temperature change of each point in the gear box under various states; based on the temperature change of each point in the gear box under various states and the stress data, constructing a stress analysis model; the stress analysis model is used to analyze the temperature and stress change under various states; based on the stress analysis model, detecting the real-time detected infrared data to determine the defect detection result of the gear box, the defect detection result including the defect type and the defect position.

[0007] In a possible implementation, the temperature distribution data at each time within the set time period in each state is determined based on infrared data at each time within the set time period in each state and finite element analysis, including: for any state, based on the infrared data at each time within the set time period in the state and a twin model of the preset gearbox, mapping matching is performed to determine the mapping relationship between the infrared data and the twin model; wherein the type of the state includes normal, broken tooth, wear, pitting and scratch; based on the infrared data at each time within the set time period in the state and the mapping relationship, the boundary condition at each time is determined; the boundary condition includes the temperature of the surface of each component in the twin model; based on the twin model and the boundary condition at each time, heat conduction and heat radiation analysis is performed to determine the temperature of each point in the gearbox at each time; and based on the temperature of each point in the gearbox at each time within the set time period in the state, the temperature distribution data at each time within the set time period in the state is determined.

[0008] In a possible implementation, the temperature change of each point in the gearbox in each state is determined based on the temperature distribution data at each time within the set time period in each state, including: based on the twin model of the gearbox, the temperature related point of each point is determined; the temperature related point includes a heat conduction point, a heat convection point and / or a heat radiation point; for any state, based on the temperature distribution data at each time within the set time period in the state, the temperature sequence of each point is extracted; based on the temperature sequence of any point, time window division is performed to calculate the temperature change feature of the point; the temperature change feature includes a temperature instantaneous value, a temperature average value and a temperature change rate; based on the temperature sequence of the temperature related point of the point, time window division is performed to calculate the temperature change feature of the temperature related point of the point; based on the temperature change feature of each point and the temperature change feature of the temperature related point of each point, the heat transfer information of the gearbox is determined, and the heat transfer information includes: a heat transfer path, a heat transfer efficiency and a heat source starting point; based on the temperature change feature of each point, the temperature change feature of the temperature related point of each point and the heat transfer information of the gearbox, the temperature change of each point in the gearbox in the state is determined.

[0009] In a possible implementation, the stress analysis model is constructed based on temperature changes of each point in the gearbox under various states and stress data, and includes: generating first input features based on temperature changes of each point in the gearbox under various states, and generating stress features based on stress data under various states; dividing the first input features and the stress features based on state types to obtain first input features and stress features corresponding to each state type; taking the first input features corresponding to each state type as input and taking the stress features corresponding to each state type as output to obtain first training samples; taking the stress features corresponding to each state type as input and taking each state type as output to obtain second training samples; performing neural network training based on the first training samples to obtain a first model; performing neural network training based on the second training samples to obtain a second model; and constructing the stress analysis model based on the first model and the second model.

[0010] In a possible implementation, the stress data includes stresses of each key point at each time within a set period under various states; the stress features are generated based on stress data under various states, and include: determining stress-related point positions of each key point based on a twin model of the gearbox; the stress-related point positions include direct contact nodes and indirect contact nodes; extracting stress data of each key point and stress data of the stress-related point positions of each key point based on stress data under various states; for any state, constructing stress time series of each key point and stress time series of the stress-related point positions of each key point based on stress data of each key point at each time and stress data of the stress-related point positions of each key point under the state; and performing spatial sorting based on the stress time series of each key point and the stress time series of the stress-related point positions of each key point to obtain stress features under the state.

[0011] In a possible implementation, the real-time detected infrared data is detected based on the stress analysis model to determine a defect detection result of the gearbox, and includes: acquiring infrared data of a current period in a running process of the gearbox; performing finite element analysis based on the infrared data of the current period to obtain temperature distribution data of the current period; generating first input features based on the temperature distribution data of the current period; and determining a defect detection result of the gearbox in the current period based on the first input features and the stress analysis model.

[0012] In a possible implementation, the method further includes: acquiring physical parameters of each component of the gearbox, connection relationships between the components, and mechanical performance parameters and thermal performance parameters between the components; constructing a twin of each component based on the physical parameters of each component of the gearbox; constructing a twin of the gearbox based on the twin of each component and the connection relationships between the components; and constructing a twin model of the gearbox based on the twin of the gearbox and the mechanical performance parameters and the thermal performance parameters between the components.

[0013] In a possible implementation, based on the stress analysis model, the real-time detected infrared data is detected to determine the defect detection result of the gearbox, and then further comprising: if the defect detection result is a slight defect, a defect monitoring instruction is generated, the defect monitoring instruction is used to instruct to monitor the gearbox with a defect; the slight defect includes wear, pitting and scratch; if the defect detection result is a serious defect, a shutdown instruction is generated, the shutdown instruction is used to instruct the gearbox to shut down, and the serious defect includes tooth breakage.

[0014] In a possible implementation, if the defect detection result is a slight defect, a defect monitoring instruction is generated, and then further comprising: recording the infrared data after the defect of the gearbox; the infrared data after the defect of the gearbox is divided into a plurality of sliding time window infrared data by sliding time window division; based on the plurality of sliding time window infrared data and the stress analysis model, a plurality of sliding time window analysis results are determined, the analysis results include defect type, defect probability of each defect type and defect position; based on the analysis results of the plurality of sliding time window, the defect at the defect position in the defect detection result is analyzed to determine the defect trend; based on the defect detection result, the analysis results of the plurality of sliding time window and the defect trend, it is determined whether the gearbox is shut down.

[0015] In a second aspect, the embodiments of the present application provide a defect detection device of a gearbox, the device comprising a communication module and a processing module, the communication module is used to acquire infrared data and stress data of the gearbox in a set period of time under various states; the processing module is used to perform finite element analysis based on the infrared data in the set period of time under various states, to determine temperature distribution data at each time in the set period of time under various states; based on the temperature distribution data at each time in the set period of time under various states, the temperature change of each point in the gearbox under various states is determined; based on the temperature change of each point in the gearbox under various states and the stress data, a stress analysis model is constructed; the stress analysis model is used to analyze the temperature and stress change under various states; based on the stress analysis model, the real-time detected infrared data is detected to determine the defect detection result of the gearbox, and the defect detection result includes defect type and defect position.

[0016] In a third aspect, the embodiments of the present application provide an electronic device, which comprises a memory and a processor, the memory stores a computer program, and the processor is used to call and run the computer program stored in the memory to execute the steps of the method in the first aspect and any possible implementation manner of the first aspect.

[0017] In a fourth aspect, an embodiment of the present application provides a computer readable storage medium, the computer readable storage medium storing a computer program, and the computer program is executed by a processor to implement the steps of the method in the first aspect and any possible implementation manner of the first aspect.

[0018] The present application provides a gear box defect detection method and system, the present application can more accurately detect the actual running state of the gear box by analyzing the infrared data and stress data of the gear box in the normal state and various defect states, and improves the defect detection accuracy. Based on the infrared data and stress data, a stress analysis model is constructed, and then the real-time infrared data is detected, so that potential defects can be found in time during the operation of the gear box, and the defect detection efficiency is improved. Moreover, the present application does not need to be sampled regularly, and can be realized only by infrared detection of the inside of the gear box, thereby improving the convenience of defect detection. The present application realizes the stress analysis and defect detection of the gear box by infrared detection and stress detection of the gear box, and improves the accuracy, convenience and detection efficiency of the gear box defect detection. BRIEF DESCRIPTION OF DRAWINGS

[0019] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0020] Figure 1 is a flowchart of a gear box defect detection method provided by an embodiment of the present application;

[0021] Figure 2 is a structural schematic diagram of a gear box defect detection device provided by an embodiment of the present application;

[0022] Figure 3 is a structural schematic diagram of an electronic device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0023] In the following description, specific details are set forth in order to provide a thorough understanding of embodiments of the present application. However, persons of ordinary skill in the art will readily recognize that embodiments of the present application can be practiced without these specific details. In other instances, well-known structures, devices, circuits, and methods have not been described in detail in order to avoid obscuring the present application.

[0024] In the description of this invention, unless otherwise stated, " / " means "or". For example, A / B can mean A or B. The term "and / or" in this document is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, and B alone. Furthermore, "at least one" and "more than one" refer to two or more. The terms "first," "second," etc., do not limit the quantity or order of execution, and "first," "second," etc., do not necessarily imply differences.

[0025] In the embodiments of this application, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design that is described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design. Specifically, the use of terms such as "exemplary" or "for example" is intended to present the relevant concepts in a specific manner to facilitate understanding.

[0026] Furthermore, the terms "comprising" and "having," and any variations thereof, used in the description of this application are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or modules is not limited to the steps or modules listed, but may optionally include other steps or modules not listed, or may optionally include other steps or modules inherent to such process, method, product, or device.

[0027] To make the objectives, technical solutions, and advantages of the present invention clearer, the following description will be provided in conjunction with the accompanying drawings and specific embodiments.

[0028] like Figure 1 As shown, this embodiment of the invention provides a method for detecting defects in a gearbox. The method includes steps S101-S104.

[0029] S101. Acquire infrared data and stress data of the gearbox under various conditions for a set period of time.

[0030] In some embodiments, stress data includes stress at each key point at each time point within a set time period under various conditions.

[0031] For example, embodiments of the present invention may use a high-precision infrared thermal imager and a stress sensor. The infrared thermal imager is used to capture the temperature distribution on the gear surface, while the stress sensor is positioned at critical locations on the gear to measure stress changes. An appropriate data sampling rate is set based on the gearbox's operating speed and defect characteristics to ensure that sufficient defect information is captured.

[0032] S102. Based on the infrared data of the set time period under various conditions, perform finite element analysis to determine the temperature distribution data at each moment within the set time period under various conditions.

[0033] It should be noted that the embodiments of the present invention can use professional finite element analysis software to construct a three-dimensional finite element model based on the gearbox's geometric dimensions and material properties. Based on infrared data, thermal boundary conditions for the gearbox are set, including thermal radiation, thermal convection, and thermal conduction. Based on the gearbox's thermal inertia and temperature change rate, an appropriate time step is set to ensure the accuracy of the finite element analysis. Through finite element analysis, the temperature distribution data of the gearbox at various times within a set time period under different conditions are calculated.

[0034] As one possible implementation, step S102 can be specifically implemented as steps S1021-S1024.

[0035] S1021. For any state, based on the infrared data at each moment within a set time period under that state, and the twin model of the preset gearbox, perform mapping matching to determine the mapping relationship between the infrared data and the twin model.

[0036] The types of conditions include normal, broken teeth, worn, pitted, and scratched.

[0037] For example, embodiments of the present invention can preprocess the acquired infrared data, including denoising, calibration, and enhancement, to ensure the accuracy and integrity of the data. A digital twin model of the gearbox is constructed using advanced modeling techniques, such as computer-aided design (CAD) or 3D scanning technology. This model should highly reproduce the actual structure and material properties of the gearbox. The preprocessed infrared data is then mapped and matched with the twin model. This typically involves spatially aligning and temporally synchronizing the temperature data in the infrared image with the corresponding components in the twin model. This process can be completed automatically or semi-automatically using advanced image processing algorithms and registration techniques.

[0038] For any given condition (such as normal, broken tooth, wear, pitting, and scratches), the mapping relationship is fine-tuned based on its unique defect characteristics. For example, a broken tooth condition may cause abnormal local temperature increases, requiring special attention to these areas during the mapping process.

[0039] S1022. Based on the infrared data at each time point within the set time period under this state, and the mapping relationship, determine the boundary conditions at each time point.

[0040] In some embodiments, boundary conditions include the temperature of the surfaces of the components in the twin model.

[0041] For example, embodiments of the present invention can extract the surface temperatures of various gearbox components from infrared data as boundary conditions based on mapping relationships. These include key components such as gears, bearings, and the gearbox housing. It is ensured that the extracted boundary conditions match the time step in the twin model to enable accurate heat conduction and thermal radiation analysis.

[0042] S1023. Based on the twin model and the boundary conditions at each time step, perform heat conduction and heat radiation analysis to determine the temperature at each point inside the gearbox at each time step.

[0043] For example, embodiments of the present invention can perform finite element mesh generation on the twin model, decomposing the model into a series of interconnected elements. The size and shape of these elements should be reasonably selected based on the analysis accuracy and computational resources. Heat conduction equations (such as Fourier's law of heat conduction) are applied to simulate the heat transfer process within the gearbox. This involves solving a system of partial differential equations that incorporate time, space, and thermophysical property parameters. For the portions of the gearbox exposed to the external environment, the effect of thermal radiation must also be considered. This is typically achieved by introducing radiative boundary conditions or using a radiative heat transfer coefficient.

[0044] S1024. Based on the temperature of each point in the gearbox at each moment within the set time period under this state, determine the temperature distribution data at each moment within the set time period under this state.

[0045] For example, embodiments of the present invention can extract temperature data at various points within the gearbox at different times from finite element analysis. This includes temperature information at key locations such as gear tooth surfaces, tooth roots, and bearing contact areas. Visualization tools (such as Unity3D) are used to display the temperature distribution data in the form of cloud maps, contour maps, etc., to intuitively analyze the thermal behavior of the gearbox. Based on the changing trends and abnormal characteristics of the temperature distribution data, the types and locations of potential defects in the gearbox can be identified. This helps to take timely and appropriate maintenance measures to prevent further deterioration of defects.

[0046] S103. Based on the temperature distribution data of each time period within a set time period under various conditions, determine the temperature change of each point in the gearbox under various conditions.

[0047] It should be noted that the embodiments of the present invention can use the post-processing function of finite element analysis software to extract the temperature changes at various points inside the gearbox. The temperature changes at each point are plotted as curves to visually analyze the temperature distribution and trend of the gearbox.

[0048] As one possible implementation, step S103 can be specifically implemented as steps S1031-S1036.

[0049] S1031. Based on the twin model of the gearbox, determine the temperature-related points for each location.

[0050] In some embodiments, temperature-related points include heat conduction points, heat convection points, and / or heat radiation points.

[0051] For example, embodiments of the present invention can utilize a pre-constructed digital twin model of a gearbox, which should accurately reflect the gearbox's geometry, material properties, and thermophysical characteristics. In the twin model, points closely related to temperature changes are identified and marked. These points include, but are not limited to, heat conduction points (such as gear contact surfaces, gear shafts, etc.), heat convection points (such as the gearbox's inner wall, cooling system interfaces, etc.), and heat radiation points (such as the gearbox's external exposed surfaces).

[0052] S1032. For any state, based on the temperature distribution data of each time point within a set time period under that state, extract the temperature sequence of each point.

[0053] S1033. Based on the temperature sequence at any point, divide the time window and calculate the temperature change characteristics at that point.

[0054] In some embodiments, temperature change characteristics include instantaneous temperature values, average temperature values, and rate of temperature change.

[0055] For example, embodiments of the present invention can divide the temperature sequence into multiple time windows, each containing a certain number of continuous temperature data points. For the temperature data within each time window, the instantaneous temperature value (i.e., the average or median temperature within the window), the average temperature value (i.e., the average temperature over the entire set time period), and the rate of temperature change (i.e., the ratio of the difference in temperature values ​​between adjacent time windows to the time interval) are calculated.

[0056] S1034. Based on the temperature sequence of temperature-related points at this point, divide the time window and calculate the temperature change characteristics of the temperature-related points at this point.

[0057] For example, embodiments of the present invention can also perform time window division and temperature change characteristic calculation for temperature points related to that point (such as heat conduction points, heat convection points, etc.). This helps to analyze the temperature interaction and transmission between different points.

[0058] S1035. Based on the temperature change characteristics of each point and the temperature change characteristics of temperature-related points at each point, determine the heat transfer information of the gearbox.

[0059] In some embodiments, heat transfer information includes: heat transfer path, heat transfer efficiency, and heat source origin.

[0060] For example, embodiments of the present invention can analyze the heat transfer path inside the gearbox by combining the temperature change characteristics of each point and the temperature change characteristics of temperature-related points. Heat transfer efficiency is evaluated by comparing the rate of temperature change and temperature differences between different points. The heat source origin is identified, i.e., the point or region with the most significant temperature change and the highest temperature value.

[0061] S1036. Based on the temperature change characteristics of each point, the temperature change characteristics of temperature-related points of each point, and the heat transfer information of the gearbox, determine the temperature change of each point in the gearbox under this state.

[0062] For example, embodiments of the present invention can synthesize the above analysis to determine the temperature distribution of the gearbox under specific conditions. Abnormal temperature points or regions can be identified, which may correspond to potential defects or performance problems.

[0063] S104. Based on the temperature changes at various points inside the gearbox under various conditions, as well as stress data, a stress analysis model is constructed.

[0064] In this embodiment, the stress analysis model is used to analyze temperature and stress changes under various conditions.

[0065] As one possible implementation, step S104 can be specifically implemented as steps S1041-S1048.

[0066] S1041. Based on the temperature changes at various points inside the gearbox under various conditions, generate the first input feature.

[0067] For example, embodiments of the present invention can organize the temperature changes (including temperature values, temperature change rates, etc.) at various points within the gearbox under various conditions to form time series data. The time series data is preprocessed, such as through noise reduction and normalization, to improve data quality. As needed, statistical features (such as mean, variance, maximum, minimum, etc.) or time-frequency features (such as Fourier transform results, etc.) of the time series data are extracted as the first input features.

[0068] S1042. Generate stress features based on stress data under various conditions.

[0069] In some embodiments, the present invention can organize stress data under various conditions to form a stress time series or stress distribution map. The stress data is preprocessed, such as removing outliers and smoothing. Features of the stress data, such as peak stress, mean stress, and stress distribution range, are extracted as stress characteristics.

[0070] For example, step S1042 can be specifically implemented as steps A1-A4.

[0071] A1. Based on the twin model of the gearbox, determine the stress-related points of each key point.

[0072] In some embodiments, stress-related points include direct contact nodes and indirect contact nodes.

[0073] For example, embodiments of the present invention can utilize a digital twin model of a gearbox, which should accurately reflect the gearbox's geometry, material properties, and stress conditions. Points closely related to stress changes are identified and marked in the model. These points include direct contact nodes (such as gear tooth contact points, bearing contact points, etc.) and indirect contact nodes (such as points that transmit stress through connectors, drive shafts, etc.). Direct contact nodes are points in the gearbox that directly bear loads and contact stresses, while indirect contact nodes are points that transmit stress through other structures.

[0074] A2. Based on stress data under various conditions, extract stress data for each key point, as well as stress data for stress-related points of each key point.

[0075] For example, embodiments of the present invention can utilize stress measurement equipment (such as strain gauges, stress sensors, etc.) or finite element analysis software to continuously collect stress data at key points and stress-related points of the gearbox within a set time period. For each key point and its stress-related points, the stress value for the corresponding time period is extracted from the stress data to form time series data.

[0076] A3. For any state, based on the stress data of each key point at each time point in that state, and the stress data of the stress-related points of each key point, construct the stress time series sequence of each key point and the stress time series sequence of the stress-related points of each key point.

[0077] For example, embodiments of the present invention can organize the extracted stress time series data for each key point and its stress-related points to form a stress time series. The stress time series should include the stress value of the point at each time within a set time period, as well as other possible relevant information (such as timestamps, measurement conditions, etc.).

[0078] A4. Based on the stress time series of each key point and the stress time series of stress-related points of each key point, spatial sorting is performed to obtain the stress characteristics under this state.

[0079] For example, embodiments of the present invention can spatially sort the stress time series for each state. That is, based on the structure and stress conditions of the gearbox, the stress time series of each key point and stress-related point is sorted according to their positional relationship within the system. The sorted stress time series can more intuitively reflect the distribution and changes of stress inside the gearbox. Based on the sorted stress time series, stress features are extracted. These features may include statistical features such as stress peak value, stress mean, stress fluctuation range, and stress change trend, as well as spatial features such as stress distribution map and stress gradient. The stress features should be able to comprehensively and accurately reflect the stress situation of the gearbox under specific conditions, providing an important basis for subsequent multi-defect stress analysis.

[0080] S1043. Based on the state type, the first input feature and stress feature are divided to obtain the first input feature and stress feature corresponding to each state type.

[0081] For example, embodiments of the present invention can divide the first input feature and stress feature according to the state type of the gearbox (such as normal, broken tooth, worn, pitted, scratched, etc.) to obtain the first input feature and stress feature corresponding to each state type.

[0082] S1044. Using the first input feature corresponding to each state type as input and the stress feature corresponding to each state type as output, the first training sample is obtained.

[0083] For example, embodiments of the present invention can use the first input feature corresponding to each state type as input and the stress feature corresponding to each state type as output to form input-output pairs. These input-output pairs will be used to train a neural network model to predict the stress features under a given temperature change.

[0084] S1045. Using the stress characteristics corresponding to each state type as input and each state type as output, the second training sample is obtained.

[0085] For example, embodiments of the present invention can use the stress characteristics corresponding to each state type as input and each state type as output to form input-output pairs. These input-output pairs will be used to train another neural network model to identify the state type of the gearbox based on the stress characteristics.

[0086] S1046. Based on the first training sample, perform neural network training to obtain the first model.

[0087] For example, embodiments of the present invention may select a suitable neural network architecture (such as a multilayer perceptron, convolutional neural network, etc.) and set corresponding hyperparameters (such as learning rate, number of iterations, etc.). The neural network is trained using a first training sample, and the network weights are adjusted through a backpropagation algorithm so that the network can accurately predict stress characteristics under a given temperature change. After training, a first model is obtained, which can receive temperature changes as input and output the predicted stress characteristics.

[0088] S1047. Based on the second training samples, perform neural network training to obtain the second model.

[0089] For example, embodiments of the present invention may select appropriate neural network architectures and hyperparameters. The neural network is trained using second training samples, enabling it to identify the gearbox's state type based on stress characteristics. After training, a second model is obtained, which can receive stress characteristics as input and output the identified state type.

[0090] S1048. Based on the first and second models, construct a stress analysis model.

[0091] For example, embodiments of the present invention can integrate the first model and the second model to construct a complete stress analysis model. This model can receive the temperature changes of the gearbox as input, first predict stress characteristics using the first model, and then identify the state type using the second model. The model can also be further optimized and adjusted as needed to improve its accuracy and robustness.

[0092] S105. Based on the stress analysis model, the real-time infrared data is detected to determine the defect detection results of the gearbox.

[0093] In this embodiment of the application, the defect detection result includes the defect type and the defect location.

[0094] As one possible implementation, step S105 can be specifically implemented as steps S1051-S1054.

[0095] S1051. Obtain infrared data for the current time period during gearbox operation.

[0096] For example, embodiments of the present invention can utilize an infrared thermal imager or other infrared detection equipment to perform real-time infrared detection on a running gearbox. The infrared thermal imager should be set to an appropriate resolution and sampling frequency to ensure that detailed temperature changes on the gearbox surface can be captured. The infrared image or video data acquired by the infrared thermal imager is converted into digital format and saved as infrared data for the current time period.

[0097] S1052. Based on the infrared data of the current time period, perform finite element analysis to obtain the temperature distribution data of the current time period.

[0098] For example, in this embodiment of the invention, infrared data for the current time period can be imported into finite element analysis software. A finite element model is established based on the gearbox's geometry, material properties, and boundary conditions. The model is then solved using the finite element analysis software to obtain the temperature distribution data inside the gearbox for the current time period. The temperature distribution data should include the temperature value at each point and information such as the temperature gradient.

[0099] S1053. Generate the first input feature based on the temperature distribution data of the current time period.

[0100] For example, embodiments of the present invention can preprocess the temperature distribution data for the current time period, such as denoising and smoothing, to improve data quality. Statistical features (such as mean, variance, maximum, minimum, etc.) or time-frequency features (such as Fourier transform results, etc.) of the temperature distribution data can be extracted as needed. The extracted features are used as the first input features for subsequent defect detection.

[0101] S1054. Based on the first input features and the stress analysis model, determine the defect detection results of the gearbox in the current time period.

[0102] For example, in this embodiment of the invention, the first input feature can be input into the first model (i.e., the model used to predict stress features) in the previously constructed stress analysis model. The first model predicts the stress characteristics of the gearbox in the current time period based on the input temperature distribution data (i.e., the first input feature). The predicted stress features are then input into the second model (i.e., the model used to identify the state type) to obtain the defect detection results of the gearbox in the current time period. The defect detection results may include the state type of the gearbox (e.g., normal, broken tooth, wear, pitting, etc.), as well as information such as the possible defect location and severity.

[0103] This invention provides a method for defect detection in gearboxes. By analyzing infrared and stress data of the gearbox under normal and various defect conditions, the actual operating state of the gearbox can be detected more accurately, improving the accuracy of defect detection. A stress analysis model is constructed based on infrared and stress data, and then real-time infrared data is detected, allowing for timely detection of potential defects during gearbox operation, thus improving defect detection efficiency. Furthermore, this invention eliminates the need for periodic sampling; infrared detection only needs to be performed inside the gearbox, improving the convenience of defect detection. This invention, through infrared and stress detection of the gearbox and the construction of a stress analysis model, achieves multi-defect stress analysis and defect detection in gearboxes, improving the accuracy, convenience, and efficiency of gearbox defect detection.

[0104] Optionally, the gearbox defect detection method provided in this embodiment of the invention further includes steps S201-S204.

[0105] S201. Obtain the physical parameters of each component of the gearbox, the connection relationship between each component, and the mechanical and thermal performance parameters of each component.

[0106] For example, geometric parameters include the dimensions, shape, and positional relationships of each component. These parameters can be obtained from design drawings, CAD models, or actual measurements. Material parameters include the material type, density, elastic modulus, thermal conductivity, and coefficient of thermal expansion of each component. These parameters can usually be obtained from material handbooks or data provided by suppliers.

[0107] For example, mechanical performance parameters include the yield strength, tensile strength, hardness, and toughness of each component. These parameters can be obtained through material testing or by consulting relevant standards. Thermal performance parameters include the thermal conductivity, heat capacity, and coefficient of thermal expansion of each component. These parameters are crucial for analyzing the stress distribution of the gearbox under temperature changes.

[0108] S202. Based on the physical parameters of each component of the gearbox, construct a twin of each component.

[0109] For example, using CAD software or 3D modeling tools, a 3D model of each component is constructed based on its geometric and material parameters. Material properties, such as density and elastic modulus, are embedded in the model to simulate the mechanical properties of each component.

[0110] S203. Based on the twins of each component and the connection relationships between the components, construct the twin of the gearbox.

[0111] For example, the twins of each component are assembled according to their actual connection relationships to form a complete twin of the gearbox. This ensures the accuracy and stability of the connection points to reflect the actual structure of the gearbox.

[0112] S204. Based on the twin of the gearbox and the mechanical and thermal performance parameters between each component, a twin model of the gearbox is constructed.

[0113] For example, the mechanical and thermal performance parameters of each component are embedded in a twin of the gearbox. Using finite element analysis software or other simulation tools, the twin is meshed and boundary conditions are set to construct a gearbox twin model suitable for simulation analysis. It is ensured that the twin model accurately reflects the stress and temperature changes of the gearbox during actual operation.

[0114] In this way, embodiments of the present invention can construct a twin model of the gearbox before performing multi-defect stress analysis, thereby improving the accuracy, convenience, and efficiency of gearbox defect detection.

[0115] Optionally, the gearbox defect detection method provided in this embodiment of the invention further includes steps S301-S302 after step S105.

[0116] S301. If the defect detection result is a minor defect, a defect monitoring instruction is generated.

[0117] In some embodiments, the defect monitoring instruction is used to instruct the monitoring of gearbox operation with defects.

[0118] In some embodiments, minor defects include wear, pitting, and scratches.

[0119] For example, if the defects are wear, pitting, or scratches, the gearbox will not fail immediately, but their long-term presence may affect its performance and lifespan. The gearbox can be monitored and operated with these defects present.

[0120] For example, steps B1-B5 are included after step S301.

[0121] B1. Record infrared data after gearbox defects.

[0122] For example, after a defect is found in a gearbox, its presence is first confirmed through visual inspection, vibration analysis, or other diagnostic methods. Infrared images or video data of the defective gearbox are recorded using an infrared thermal imager, ensuring coverage of the entire defect period. The infrared data is then converted to a digital format for subsequent processing and analysis.

[0123] B2. Divide the infrared data after the gearbox defect into sliding time windows to obtain infrared data for multiple sliding time windows.

[0124] For example, embodiments of the present invention can set the size of the sliding time window (i.e., the length of time contained in each window) according to the time range of defect occurrence and the sampling frequency of infrared data. The sliding time window should be small enough to capture the dynamic changes of the defect, and large enough to contain sufficient defect information. The sliding step size is determined, i.e., the time overlap between two adjacent windows. The recorded infrared data is divided according to the set sliding time window to obtain multiple continuous and non-overlapping (or partially overlapping) time window data. The infrared data within each time window should contain sufficient temperature distribution information for subsequent stress analysis and defect detection.

[0125] B3. Based on infrared data from multiple sliding time windows and a stress analysis model, determine the analysis results for multiple sliding time windows.

[0126] In some embodiments, the analysis results include defect type, defect probability for each defect type, and defect location.

[0127] For example, the infrared data for each sliding time window is processed using a pre-built stress analysis model. The model should be able to predict the stress distribution and possible defect types in the gearbox based on the temperature distribution information in the infrared data. Information such as defect type, defect probability for each defect type, and defect location is extracted from the model's output. Defect types may include wear, pitting, tooth breakage, etc.; defect probability represents the likelihood of each defect type occurring within the current time window; and defect location indicates the specific location of the defect within the gearbox.

[0128] B4. Based on the analysis results of multiple sliding time windows, analyze the defects at the defect locations in the defect detection results to determine the defect trend.

[0129] For example, based on the analysis results of multiple sliding time windows, the changing trends of defect type, defect probability, and defect location over time can be observed. Methods such as statistical analysis, machine learning, or data mining can be used to identify defect development trends, such as whether defects have worsened or spread to other components.

[0130] B5. Based on the defect detection results, the analysis results of multiple sliding time windows, and the defect trend, determine whether the gearbox should be shut down.

[0131] For example, the overall condition of the gearbox is comprehensively evaluated by combining the original defect detection results (such as the preliminary defect judgment obtained through stress analysis model), the analysis results of multiple sliding time windows, and defect trend analysis. A decision on whether to shut down the gearbox is made considering factors such as defect type, defect probability, defect location, defect development trend, as well as the importance of the gearbox and downtime costs.

[0132] If the assessment indicates that the gearbox has serious defects or the defect trend is uncontrollable, a shutdown order will be generated. The shutdown order should include information such as the reason for shutdown, shutdown time, post-shutdown safety measures, and maintenance plan. Relevant personnel should be notified to perform the shutdown operation and take necessary safety and maintenance measures as required by the shutdown order.

[0133] S302. If the defect detection result is a serious defect, a stop command is generated.

[0134] In some embodiments, a stop command is used to instruct the gearbox to stop.

[0135] In some embodiments, a serious defect includes a broken tooth.

[0136] For example, if the defect type is a broken tooth, the gearbox may fail immediately or pose a serious safety hazard, and the gearbox should be instructed to be stopped immediately for maintenance.

[0137] Thus, the present invention can classify and process defects according to their severity, reduce gearbox downtime, and improve the accuracy, convenience, and efficiency of gearbox defect detection.

[0138] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0139] The following are device embodiments of the present invention. For details not described in detail, please refer to the corresponding method embodiments described above.

[0140] Figure 2 A schematic diagram of a defect detection device for a gearbox provided in an embodiment of the present invention is shown. The defect detection device 400 includes a communication module 401 and a processing module 402.

[0141] The communication module 401 is used to acquire infrared data and stress data of the gearbox under various conditions for a set period of time.

[0142] The processing module 402 is used to perform finite element analysis based on infrared data of a set time period under various conditions to determine the temperature distribution data at each moment within the set time period under various conditions; based on the temperature distribution data at each moment within the set time period under various conditions, to determine the temperature change at each point inside the gearbox under various conditions; based on the temperature change at each point inside the gearbox under various conditions and stress data, to construct a stress analysis model; the stress analysis model is used to analyze the temperature and stress changes under various conditions; based on the stress analysis model, the real-time infrared data is detected to determine the defect detection results of the gearbox, including the defect type and defect location.

[0143] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. For example... Figure 3 As shown, the electronic device 500 of this embodiment includes: a processor 501, a memory 502, and a computer program 503 stored in the memory 502 and executable on the processor 501. When the processor 501 executes the computer program 503, it implements the steps in the above-described method embodiments, for example... Figure 1 The steps S101-S105 are shown. Alternatively, when the processor 501 executes the computer program 503, it implements the functions of each module / unit in the above-described device embodiments, for example... Figure 2The functions of the communication module 401 and the processing module 402 shown are illustrated.

[0144] For example, the computer program 503 can be divided into one or more modules / units, which are stored in the memory 502 and executed by the processor 501 to complete the present invention. The one or more modules / units can be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program 503 in the electronic device 500. For example, the computer program 503 can be divided into... Figure 2 The communication module 401 and the processing module 402 are shown.

[0145] The processor 501 may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor.

[0146] The memory 502 can be an internal storage unit of the electronic device 500, such as a hard disk or memory of the electronic device 500. The memory 502 can also be an external storage device of the electronic device 500, such as a plug-in hard disk, smart media card (SMC), secure digital card (SD), flash card, etc., equipped on the electronic device 500. Furthermore, the memory 502 can include both internal and external storage units of the electronic device 500. The memory 502 is used to store the computer program and other programs and data required by the terminal. The memory 502 can also be used to temporarily store data that has been output or will be output.

[0147] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for detecting defects in a gearbox, characterized in that, include: Acquire infrared and stress data of the gearbox under various conditions for a set period of time; The infrared data refers to the infrared data of the gear surface inside the gearbox; The stress data includes the stress at each key point at each moment within a set time period under various conditions; Based on infrared data for a set time period under various conditions, finite element analysis is performed to determine the temperature distribution data at each moment within the set time period under various conditions. Based on the temperature distribution data at each moment within a set time period under various conditions, the temperature changes at each point inside the gearbox under various conditions are determined. Based on the temperature changes at various points inside the gearbox under different conditions, as well as stress data, a stress analysis model is constructed. The stress analysis model is used to analyze temperature and stress changes under various conditions; Based on the stress analysis model, the real-time infrared data is detected to determine the defect detection results of the gearbox, including the defect type and defect location. The method for constructing a stress analysis model based on temperature changes and stress data at various points within the gearbox under different conditions includes: generating a first input feature based on temperature changes at various points within the gearbox under different conditions; generating stress features based on stress data under different conditions; dividing the first input feature and the stress features according to the state type to obtain a first input feature and a stress feature corresponding to each state type; obtaining a first training sample by using the first input feature corresponding to each state type as input and the stress feature corresponding to each state type as output; obtaining a second training sample by using the stress feature corresponding to each state type as input and each state type as output; training a neural network based on the first training sample to obtain a first model; training a neural network based on the second training sample to obtain a second model; and constructing the stress analysis model based on the first model and the second model.

2. The defect detection method for a gearbox according to claim 1, characterized in that, The process involves performing finite element analysis on infrared data from various states over a set time period to determine the temperature distribution data at each moment within that time period under different states, including: For any given state, based on the infrared data at each moment within a set time period under that state, and the twin model of the preset gearbox, a mapping match is performed to determine the mapping relationship between the infrared data and the twin model; among which, the state types include normal, broken tooth, wear, pitting, and scratch; Based on the infrared data at each moment within the set time period under this state, and the mapping relationship, the boundary conditions at each moment are determined; the boundary conditions include the temperature of the surface of each component in the twin model; Based on the twin model and the boundary conditions at each time point, heat conduction and heat radiation analyses are performed to determine the temperature at each point inside the gearbox at each time point. Based on the temperature of each point in the gearbox at each time within a set time period under this condition, the temperature distribution data at each time within a set time period under this condition is determined.

3. The defect detection method for a gearbox according to claim 1, characterized in that, The determination of temperature changes at various points within the gearbox under different conditions, based on temperature distribution data at different times within a set time period under various conditions, includes: Based on the twin model of the gearbox, temperature-related points are determined for each point; the temperature-related points include heat conduction points, heat convection points, and / or heat radiation points. For any given state, based on the temperature distribution data of each time point within a set time period under that state, extract the temperature sequence of each point; Based on the temperature sequence at any point, a time window is divided, and the temperature change characteristics at that point are calculated; the temperature change characteristics include instantaneous temperature value, average temperature value, and temperature change rate. Based on the temperature series of temperature-related points at this point, a time window is divided, and the temperature change characteristics of the temperature-related points at this point are calculated. Based on the temperature change characteristics at each point and the temperature change characteristics at temperature-related points at each point, the heat transfer information of the gearbox is determined. The heat transfer information includes: heat transfer path, heat transfer efficiency, and heat source starting point. Based on the temperature change characteristics of each point, the temperature change characteristics of temperature-related points, and the heat transfer information of the gearbox, the temperature change of each point in the gearbox under this state is determined.

4. The defect detection method for a gearbox according to claim 1, characterized in that, The stress features generated based on stress data under various conditions include: Based on the twin model of the gearbox, stress-related points of each key point are determined; the stress-related points include direct contact nodes and indirect contact nodes. Based on the stress data under the various states, the stress data of each key point and the stress data of the stress-related points of each key point are extracted. For any given state, based on the stress data of each key point at each time point in that state, and the stress data of stress-related points of each key point, a stress time series sequence of each key point and a stress time series sequence of stress-related points of each key point are constructed. Based on the stress time series of each key point and the stress time series of stress-related points of each key point, spatial sorting is performed to obtain the stress characteristics under this state.

5. The defect detection method for a gearbox according to claim 1, characterized in that, The step of detecting defects in the gearbox based on the stress analysis model and real-time infrared data to determine the defect detection results includes: Acquire infrared data for the current time period during the operation of the gearbox; Finite element analysis is performed based on the infrared data of the current time period to obtain the temperature distribution data of the current time period; Based on the temperature distribution data of the current time period, a first input feature is generated; Based on the first input feature and the stress analysis model, the defect detection result of the gearbox in the current time period is determined.

6. The defect detection method for a gearbox according to claim 1, characterized in that, The method further includes: Obtain the physical parameters of each component of the gearbox, the connection relationships between the components, and the mechanical and thermal performance parameters of each component; Based on the physical parameters of each component of the gearbox, construct twins of each component; Based on the twins of each component and the connection relationships between the components, a twin of the gearbox is constructed; Based on the twin of the gearbox and the mechanical and thermal performance parameters of each component, a twin model of the gearbox is constructed.

7. The defect detection method for a gearbox according to claim 1, characterized in that, Based on the stress analysis model, the real-time infrared data is detected to determine the defect detection results of the gearbox, and then the process further includes: If the defect detection result is a minor defect, a defect monitoring instruction is generated. The defect monitoring instruction is used to instruct the gearbox to operate with defects. The minor defects include wear, pitting, and scratches. If the defect detection result is a serious defect, a stop command is generated. The stop command is used to instruct the gearbox to stop. The serious defect includes broken teeth.

8. The defect detection method for a gearbox according to claim 7, characterized in that, If the defect detection result is a minor defect, a defect monitoring instruction is generated, which then includes: Record infrared data after gearbox defects; The infrared data after the gearbox defect is divided into sliding time windows to obtain infrared data for multiple sliding time windows; Based on infrared data from multiple sliding time windows and the stress analysis model, the analysis results for multiple sliding time windows are determined. The analysis results include defect type, defect probability for each defect type, and defect location. Based on the analysis results of the multiple sliding time windows, the defects at the defect locations in the defect detection results are analyzed to determine the defect trend; Based on the defect detection results, the analysis results of the multiple sliding time windows, and the defect trend, it is determined whether the gearbox should be shut down.

9. A defect detection system for a gearbox, characterized in that, The defect detection system includes an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor is used to execute the steps of the method as described in any one of claims 1 to 8 when it calls and runs the computer program stored in the memory.

Images