Bearing machining process fault detection method and system based on Internet of Things

Abstract

The invention discloses a bearing machining process fault detection method and system based on the Internet of Things, and relates to the technical field of fault detection. A bearing machining process is divided into different machining links, a simulation model of each machining link is constructed, and first-class vibration parameters of each simulation model in an isolated state and second-class vibration parameters of simulation models of adjacent machining links in a collaborative state are obtained; constructing a vibration evaluation model according to second-class vibration parameters of each simulation model under different material parameters and first-class vibration parameters, acquiring historical processing data, constructing a product evaluation model, acquiring actual processing data, and judging whether each processing link has a fault or not by combining the vibration evaluation model and the product evaluation model; the pertinence of fault analysis can be effectively improved; and the comprehensiveness of fault detection can be improved.

Description

Technical Field

[0001] This invention relates to the field of fault detection technology, specifically to a method and system for fault detection in bearing manufacturing processes based on the Internet of Things. Background Technology

[0002] During bearing processing, failures can lead to problems such as decreased product quality, reduced production efficiency, and equipment damage. Traditional fault detection methods often rely on manual experience and periodic inspections, making it difficult to detect faults in real time and accurately. Utilizing advanced Internet of Things (IoT) technology to achieve comprehensive and real-time monitoring and fault early warning of the bearing processing process is of great significance for improving bearing processing quality and production efficiency. For processing equipment, its vibration is often the most direct reflection of the equipment's operating status and is often the main factor affecting the final quality of the processed products. In existing technologies, most of the time, the amplitude of the vibration of the processing equipment during its processing is monitored to determine whether there are any abnormalities in the processing process. However, existing technologies lack analysis of the amplitude impact during material transfer at different processing stages. Since the material transfer process also affects the final product quality, the results obtained by monitoring only the processing process lack comprehensiveness. Therefore, this invention provides a bearing processing fault detection method and system based on the Internet of Things. Summary of the Invention

[0003] The purpose of this invention is to provide a method and system for detecting faults in the bearing manufacturing process based on the Internet of Things.

[0004] The objective of this invention can be achieved through the following technical solution: a bearing machining process fault detection system based on the Internet of Things, comprising the following modules: The data acquisition module is used to divide the bearing processing process into different processing stages and acquire the equipment parameters and process parameters of each processing stage; The processing simulation module is used to set up different monitoring units in each processing stage and obtain the corresponding operating parameters. It combines equipment parameters and process parameters to build a simulation model for each processing stage. The data analysis module is used to obtain the first type of vibration parameters of each simulation model under different operating parameters in an isolated state, and to obtain the second type of vibration parameters of the simulation models of adjacent processing links under different material parameters in a collaborative state. The processing evaluation module is used to construct a vibration evaluation model based on the second type of vibration parameters under different material parameters and first type of vibration parameters of each simulation model, obtain historical processing data of qualified products, and construct a product evaluation model based on various parameters in different historical processing data. The fault diagnosis module is used to acquire actual processing data and, in conjunction with the vibration assessment model and product assessment model, determine whether there are faults in each processing stage, and generate corresponding fault signals for feedback.

[0005] Furthermore, the process of dividing the bearing manufacturing process into different processing stages and obtaining the equipment and process parameters for each stage includes: The bearing manufacturing process is divided into the following different processing stages, including bar stock preparation, forging, turning, heat treatment, grinding, and component assembly. Different processing stages correspond to different processing equipment, including band saws, forging presses, CNC lathes, heat treatment furnaces, CNC grinding machines, and assembly lines. The equipment parameters refer to the inherent performance indicators of processing equipment in different processing stages, while the process parameters refer to the control variables and operating conditions set by processing equipment in different processing stages during their processing to achieve specific quality, efficiency, and cost targets.

[0006] Furthermore, the process of setting up different monitoring units in each processing stage, acquiring corresponding operating parameters, and constructing simulation models for each processing stage by combining equipment parameters and process parameters includes: Different monitoring units are installed in the processing equipment at each processing stage. The operating parameters of different processing stages, including pressure, load rate, temperature, current and voltage, are obtained through different monitoring units. Using 3D modeling tools, physical models of corresponding processing steps are constructed based on the equipment parameters of each processing device. Simulation software is then used to simulate the processing process based on the physical models of each processing step, combined with its process parameters and operating parameters, to obtain the corresponding simulation model.

[0007] Furthermore, the process of obtaining Class I vibration parameters for each simulation model under different operating parameters in an isolated state, and obtaining Class II vibration parameters for simulation models of adjacent processing stages under different material parameters in a collaborative state, includes: The simulation models of each processing stage are integrated in sequence to obtain a comprehensive simulation model of the bearing processing process. The process of transferring materials from the processing equipment of the previous processing stage to the processing equipment of the current processing stage is regarded as the material transfer process. The isolated state refers to the state in which each simulation model in the integrated simulation model only simulates its processing process and does not simulate its material transfer process. The collaborative state refers to the state in which each simulation model in the integrated simulation model simulates both its processing process and its material transfer process. In an isolated state, the process parameters of a single simulation model are kept constant, and its operating parameters and first material parameters are continuously adjusted to obtain a type of vibration parameter of the corresponding processing equipment at different times under different operating parameters and first material parameters. The first material parameter refers to the total weight of the material processed by the simulation model. In a collaborative state, the vibration parameters of the single simulation model and the simulation models of its adjacent processing stages are kept constant. The second material parameter is continuously adjusted to obtain the second type of vibration parameters of the corresponding processing equipment of the simulation model at different times under different second material parameters. The second material parameter refers to the total weight of the material transferred between the simulation model and the simulation model of the previous processing stage.

[0008] Furthermore, the process of constructing a vibration assessment model based on the second type of vibration parameters under different material parameters and first type of vibration parameters for each simulation model includes: Based on the different operating parameters, first material parameters and first type of vibration parameters obtained by each simulation model in isolated state, the first vibration set is generated by combining the equipment parameters and process parameters of the corresponding simulation model, and it is divided into the first training set and the first test set. Construct a first convolutional neural network by using different operating parameters, material parameters, equipment parameters, and process parameters from the first training set as input data and a corresponding type of vibration parameter from the first training set as output data. Train the first convolutional neural network to obtain an initial first convolutional neural network. The initial first convolutional neural network is validated using the first test set, and the initial first convolutional neural network whose output is less than or equal to the preset first test error threshold is used as the first evaluation model. Based on the different second material parameters and their two types of vibration parameters obtained by each simulation model under the cooperative state, a second vibration set is generated by combining the equipment parameters and process parameters of the corresponding simulation model, and it is divided into a second training set and a second test set. The first evaluation model is trained by using different second material parameters, equipment parameters, and process parameters from the second training set as input data and the corresponding two types of vibration parameters from the second training set as output data. The initial first evaluation model is validated using the second test set, and the initial first evaluation model whose output is less than or equal to the preset second test error threshold is used as the second evaluation model. The vibration evaluation model includes the first evaluation model and the second evaluation model.

[0009] Furthermore, the process of obtaining historical processing data for qualified products and constructing a product evaluation model based on various parameters from different historical processing data includes: The qualified products refer to a batch of bearings that have completed all processing steps and are free of defects. The process parameters, operating parameters, first material parameters, second material parameters, first type vibration parameters, and second type vibration parameters corresponding to different processing steps of the same batch of qualified products are used as the historical processing data of their corresponding processing steps. Based on the historical processing data of qualified products from different batches at different processing stages, a corresponding third evaluation set is generated, which is then divided into a third training set and a third test set. Construct a second convolutional neural network, using different processing stages and their historical processing data from the third training set as input data for the second convolutional neural network, and using whether the product is qualified at the corresponding processing stage as output data for the second convolutional neural network. The second convolutional neural network is trained to obtain an initial second convolutional neural network. The initial second convolutional neural network is validated using a third test set. The initial second convolutional neural network whose output is less than or equal to the preset third test error threshold is used as the product evaluation model.

[0010] Furthermore, the process of acquiring actual processing data, combining vibration assessment models and product assessment models to determine whether faults exist in each processing stage, and generating corresponding fault signals for feedback includes: The actual processing data refers to the operating parameters, first material parameters, process parameters, and second material parameters of the same batch of products in the current processing stage in a real application scenario. Input the current operating parameters, the first material parameters, the process parameters, and the equipment parameters of the current processing stage into the first evaluation model to obtain the corresponding type of vibration parameters; The current second material parameters and process parameters, combined with the equipment parameters of the current processing stage, are input into the second evaluation model to obtain the corresponding second type of vibration parameters; The current actual processing data, along with the acquired Class I and Class II vibration parameters, are input into the product evaluation model to determine whether the product is qualified at the current processing stage. If the result is satisfactory, the subsequent processing steps will continue; if the result is unsatisfactory, a fault signal for the corresponding processing step will be generated and fed back to the relevant personnel.

[0011] A method for fault detection in bearing manufacturing process based on the Internet of Things includes the following steps: Step S1: Divide the bearing processing process into different processing stages and obtain the equipment parameters and process parameters for each processing stage; Step S2: Set up different monitoring units in each processing stage and obtain the corresponding operating parameters. Combine the equipment parameters and process parameters to build a simulation model for each processing stage. Step S3: Obtain the first type of vibration parameters of each simulation model under different operating parameters in the isolated state, and obtain the second type of vibration parameters of the simulation models of adjacent processing links under different material parameters in the cooperative state; Step S4: Construct a vibration evaluation model based on the second type of vibration parameters under different material parameters and first type of vibration parameters of each simulation model, obtain historical processing data of qualified products, and construct a product evaluation model based on various parameters in different historical processing data; Step S5: Obtain actual processing data, and combine the vibration assessment model and product assessment model to determine whether there are faults in each processing stage, and generate corresponding fault signals for feedback.

[0012] Compared with the prior art, the beneficial effects of the present invention are: This invention divides the bearing processing process into different processing stages and constructs simulation models for each processing stage. It can obtain the first type of vibration parameters during the processing and the second type of vibration parameters during the material transfer process. This is beneficial for analyzing the vibration of the processing equipment under different working conditions. In particular, it provides a method for analyzing the vibration of the processing equipment during the material transfer process, which can effectively improve the pertinence of fault analysis. By constructing a vibration assessment model, we can directly obtain the Class I and Class II vibration parameters corresponding to various parameters in the current processing stage. By using historical processing data of qualified products to construct a product assessment model, we can effectively determine whether the product in the current processing stage is within the qualified range based on real-time relevant data. This can also indirectly determine whether there is a fault in the current processing process and provide feedback, which is conducive to improving the comprehensiveness of fault detection. Attached Figure Description

[0013] Figure 1 This is a schematic diagram of the modules of the present invention. Detailed Implementation

[0014] like Figure 1 As shown, an Internet of Things (IoT) based bearing manufacturing process fault detection system includes the following modules: The data acquisition module is used to divide the bearing processing process into different processing stages and acquire the equipment parameters and process parameters of each processing stage; The processing simulation module is used to set up different monitoring units in each processing stage and obtain the corresponding operating parameters. It combines equipment parameters and process parameters to build a simulation model for each processing stage. The data analysis module is used to obtain the first type of vibration parameters of each simulation model under different operating parameters in an isolated state, and to obtain the second type of vibration parameters of the simulation models of adjacent processing links under different material parameters in a collaborative state. The processing evaluation module is used to construct a vibration evaluation model based on the second type of vibration parameters under different material parameters and first type of vibration parameters of each simulation model, obtain historical processing data of qualified products, and construct a product evaluation model based on various parameters in different historical processing data. The fault diagnosis module is used to acquire actual processing data and, in conjunction with the vibration assessment model and product assessment model, determine whether there are faults in each processing stage, and generate corresponding fault signals for feedback.

[0015] It should be further explained that, in the specific implementation process, the process of dividing the bearing processing into different processing stages and obtaining the equipment parameters and process parameters for each processing stage includes: The bearing processing process refers to the entire process of processing a bearing from raw materials to finished products in a factory. During this process, different technological processes are involved, and each technological process corresponds to different processing equipment. A complete bearing processing process is divided into processes such as bar stock preparation, forging, turning, heat treatment, grinding, and component assembly. The main processing equipment corresponding to different process flows are band saws, forging presses, CNC lathes, heat treatment furnaces, CNC grinding machines, assembly lines, etc. The processing process of the processing equipment corresponding to different process flows is taken as a processing link. By adopting this method, the complete bearing processing process can be divided into several different processing links. The equipment parameters and process parameters for different processing stages are obtained separately. The equipment parameters refer to the inherent performance indicators of the processing equipment in different processing stages, which are used to reflect the specifications, capacity, accuracy range and working limits of the corresponding processing equipment. The process parameters refer to the control variables and operating conditions set by the processing equipment in different processing stages during their processing to achieve specific quality, efficiency and cost targets.

[0016] It should be further explained that, in the specific implementation process, different monitoring units are set up in each processing stage to obtain the corresponding operating parameters. The process of constructing a simulation model for each processing stage by combining equipment parameters and process parameters includes: Different monitoring units are set up in the processing equipment corresponding to each processing stage, including pressure monitoring unit, load monitoring unit, temperature monitoring unit, current monitoring unit and voltage monitoring unit. The operating parameters of different processing stages are obtained through each monitoring unit, including pressure, load rate, temperature, current and voltage. The operating parameters of different processing stages are not exactly the same. Using 3D modeling tools, physical models of corresponding processing steps are constructed based on the equipment parameters of each processing device. Simulation software is then used to simulate the processing process of the processing equipment based on the physical models of each processing step, combined with its process parameters and operating parameters, to obtain corresponding simulation models. Different processing steps have corresponding simulation models.

[0017] It should be further explained that, in the specific implementation process, the process of obtaining the first type of vibration parameters of each simulation model under different operating parameters in an isolated state, and obtaining the second type of vibration parameters of the simulation models of adjacent processing stages under different material parameters in a collaborative state, includes: The simulation models of each processing stage are integrated sequentially to obtain a comprehensive simulation model of the bearing processing process. The comprehensive simulation model includes simulation models of different processing stages, which can simulate the processing of each processing stage separately, and can also simulate the material transfer process between adjacent processing stages and the resulting parameter changes. The material transfer process refers to the process of transferring materials from the processing equipment of the previous processing stage to the processing equipment of the current processing stage. The isolated state refers to the state in which each simulation model in the comprehensive simulation model only simulates its own processing process and does not simulate the material transfer process between itself and the previous processing stage. The collaborative state refers to the state in which each simulation model in the comprehensive simulation model simulates both its own processing process and the material transfer process between itself and the previous processing stage. Taking a single simulation model as an example, in an isolated state, its process parameters are kept fixed, and its operating parameters and first material parameters are continuously adjusted to obtain the vibration amplitude of the corresponding processing equipment at different times under different operating parameters and first material parameters. This is recorded as a type of vibration parameter. At this time, the first material parameter refers to the total weight of the material processed by the corresponding processing equipment of the simulation model. In the collaborative state, the first type of vibration parameters of the simulation model and the simulation model of the adjacent processing stage are kept constant. The second material parameter is continuously adjusted to obtain the vibration amplitude of the corresponding processing equipment of the simulation model at different times under different second material parameters. This is recorded as the second type of vibration parameter. At this time, the second material parameter refers to the total weight of the material transferred between the simulation model and the corresponding processing equipment of the simulation model of the previous processing stage.

[0018] It should be further explained that, in the specific implementation process, the process of constructing a vibration assessment model based on the second type of vibration parameters under different material parameters and first type of vibration parameters for each simulation model includes: Based on the different operating parameters and first material parameters and their corresponding type of vibration parameters obtained by each simulation model in isolated state, the first vibration set is generated by combining the equipment parameters and process parameters of the corresponding simulation model, and it is divided into the first training set and the first test set. Construct a first convolutional neural network by using different operating parameters, material parameters, equipment parameters, and process parameters from the first training set as input data and a corresponding type of vibration parameter from the first training set as output data. Train the first convolutional neural network to obtain an initial first convolutional neural network. The initial first convolutional neural network is validated using the first test set, and the initial first convolutional neural network whose output is less than or equal to the preset first test error threshold is used as the first evaluation model. Based on the different second material parameters and their corresponding second type of vibration parameters obtained by each simulation model under the cooperative state, a second vibration set is generated by combining the equipment parameters and process parameters of the corresponding simulation model, and it is divided into a second training set and a second test set. The first evaluation model is trained by using different second material parameters, equipment parameters, and process parameters from the second training set as input data and the corresponding two types of vibration parameters from the second training set as output data. The initial first evaluation model is validated using the second test set, and the initial first evaluation model whose output is less than or equal to the preset second test error threshold is used as the second evaluation model. The vibration evaluation model includes the first evaluation model and the second evaluation model.

[0019] It should be further explained that, in the specific implementation process, the process of obtaining historical processing data of qualified products and constructing a product evaluation model based on various parameters in different historical processing data includes: The qualified products refer to a batch of bearings that have completed all processing steps and are free of defects. The process parameters, operating parameters, first material parameters, second material parameters, first type vibration parameters, and second type vibration parameters corresponding to different processing steps of the same batch of qualified products are obtained and recorded as historical processing data. Based on the historical processing data of qualified products from different batches at different processing stages, a corresponding third evaluation set is generated, which is then divided into a third training set and a third test set. Construct a second convolutional neural network, using different processing stages and their historical processing data from the third training set as input data for the second convolutional neural network, and using whether the product is qualified at the corresponding processing stage as output data for the second convolutional neural network. The second convolutional neural network is trained to obtain an initial second convolutional neural network. The initial second convolutional neural network is validated using a third test set. The initial second convolutional neural network whose output is less than or equal to the preset third test error threshold is used as the product evaluation model.

[0020] It should be further explained that, in the specific implementation process, the process of acquiring actual processing data, combining it with vibration assessment models and product assessment models to determine whether there are faults in each processing stage, and generating corresponding fault signals for feedback includes: The actual processing data refers to the operating parameters, first material parameters, process parameters, and second material parameters of the same batch of products in the current processing stage in a real application scenario. The current operating parameters, first material parameters, and process parameters, combined with the equipment parameters of the current processing stage, are input into the first evaluation model to obtain the corresponding first type of vibration parameters. The current second material parameters and process parameters, combined with the equipment parameters of the current processing stage, are input into the second evaluation model to obtain the corresponding second type of vibration parameters. The current actual processing data, along with the acquired Class I and Class II vibration parameters, are input into the product evaluation model to determine whether the product in the current processing stage is qualified. If qualified, the subsequent processing stages continue. If unqualified, a fault signal for that processing stage is generated and fed back to the relevant personnel to prompt them to promptly inspect and repair that processing stage.

[0021] The embodiments of the present invention also include an Internet of Things-based method for detecting faults in the bearing manufacturing process, comprising the following steps: Step S1: Divide the bearing processing process into different processing stages and obtain the equipment parameters and process parameters for each processing stage; Step S2: Set up different monitoring units in each processing stage and obtain the corresponding operating parameters. Combine the equipment parameters and process parameters to build a simulation model for each processing stage. Step S3: Obtain the first type of vibration parameters of each simulation model under different operating parameters in the isolated state, and obtain the second type of vibration parameters of the simulation models of adjacent processing links under different material parameters in the cooperative state; Step S4: Construct a vibration evaluation model based on the second type of vibration parameters under different material parameters and first type of vibration parameters of each simulation model, obtain historical processing data of qualified products, and construct a product evaluation model based on various parameters in different historical processing data; Step S5: Obtain actual processing data, and combine the vibration assessment model and product assessment model to determine whether there are faults in each processing stage, and generate corresponding fault signals for feedback.

[0022] The above embodiments are only used to illustrate the technical methods of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of the present invention without departing from the spirit and scope of the technical methods of the present invention.

Claims

1. A bearing machining process fault detection system based on the Internet of Things, characterized in that, Includes the following modules: The data acquisition module is used to divide the bearing processing process into different processing stages and acquire the equipment parameters and process parameters of each processing stage; The processing simulation module is used to set up different monitoring units in each processing stage and obtain the corresponding operating parameters. It combines equipment parameters and process parameters to build a simulation model for each processing stage. The data analysis module is used to obtain the first type of vibration parameters of each simulation model under different operating parameters in an isolated state, and to obtain the second type of vibration parameters of the simulation models of adjacent processing links under different material parameters in a collaborative state. The processing evaluation module is used to construct a vibration evaluation model based on the second type of vibration parameters under different material parameters and first type of vibration parameters of each simulation model, obtain historical processing data of qualified products, and construct a product evaluation model based on various parameters in different historical processing data. The fault diagnosis module is used to acquire actual processing data and, in conjunction with the vibration assessment model and product assessment model, determine whether there are faults in each processing stage, and generate corresponding fault signals for feedback.

2. The bearing machining process fault detection system based on the Internet of Things according to claim 1, characterized in that, The process of dividing processing steps and obtaining equipment and process parameters includes: The bearing manufacturing process is divided into the following different processing stages, including bar stock preparation, forging, turning, heat treatment, grinding, and component assembly. Different processing stages correspond to different processing equipment, including band saws, forging presses, CNC lathes, heat treatment furnaces, CNC grinding machines, and assembly lines. The equipment parameters refer to the inherent performance indicators of processing equipment in different processing stages, while the process parameters refer to the control variables and operating conditions set by processing equipment in different processing stages during their processing to achieve specific quality, efficiency, and cost targets.

3. The bearing machining process fault detection system based on the Internet of Things according to claim 2, characterized in that, The process of obtaining operating parameters and constructing simulation models for each processing stage includes: Different monitoring units are installed in the processing equipment at each processing stage. The operating parameters of different processing stages, including pressure, load rate, temperature, current and voltage, are obtained through different monitoring units. Using 3D modeling tools, physical models of corresponding processing steps are constructed based on the equipment parameters of each processing device. Simulation software is then used to simulate the processing process based on the physical models of each processing step, combined with its process parameters and operating parameters, to obtain the corresponding simulation model.

4. The bearing machining process fault detection system based on the Internet of Things according to claim 3, characterized in that, The process of obtaining Class I and Class II vibration parameters includes: The simulation models of each processing stage are integrated in sequence to obtain a comprehensive simulation model of the bearing processing process. The process of transferring materials from the processing equipment of the previous processing stage to the processing equipment of the current processing stage is regarded as the material transfer process. The isolated state refers to the state in which each simulation model in the integrated simulation model only simulates its processing process and does not simulate its material transfer process. The collaborative state refers to the state in which each simulation model in the integrated simulation model simulates both its processing process and its material transfer process. In an isolated state, the process parameters of a single simulation model are kept constant, and its operating parameters and first material parameters are continuously adjusted to obtain a type of vibration parameter of the corresponding processing equipment at different times under different operating parameters and first material parameters. The first material parameter refers to the total weight of the material processed by the simulation model. In a collaborative state, the vibration parameters of the single simulation model and the simulation models of its adjacent processing stages are kept constant. The second material parameter is continuously adjusted to obtain the second type of vibration parameters of the corresponding processing equipment of the simulation model at different times under different second material parameters. The second material parameter refers to the total weight of the material transferred between the simulation model and the simulation model of the previous processing stage.

5. The bearing machining process fault detection system based on the Internet of Things according to claim 4, characterized in that, The process of constructing a vibration assessment model includes: Based on the different operating parameters, first material parameters and first type of vibration parameters obtained by each simulation model in isolated state, the first vibration set is generated by combining the equipment parameters and process parameters of the corresponding simulation model, and it is divided into the first training set and the first test set. Construct a first convolutional neural network by using different operating parameters, material parameters, equipment parameters, and process parameters from the first training set as input data and a corresponding type of vibration parameter from the first training set as output data. Train the first convolutional neural network to obtain an initial first convolutional neural network. The initial first convolutional neural network is validated using the first test set, and the initial first convolutional neural network whose output is less than or equal to the preset first test error threshold is used as the first evaluation model. Based on the different second material parameters and their two types of vibration parameters obtained by each simulation model under the cooperative state, a second vibration set is generated by combining the equipment parameters and process parameters of the corresponding simulation model, and it is divided into a second training set and a second test set. The first evaluation model is trained by using different second material parameters, equipment parameters, and process parameters from the second training set as input data and the corresponding two types of vibration parameters from the second training set as output data. The initial first evaluation model is validated using the second test set, and the initial first evaluation model whose output is less than or equal to the preset second test error threshold is used as the second evaluation model. The vibration evaluation model includes the first evaluation model and the second evaluation model.

6. The bearing machining process fault detection system based on the Internet of Things according to claim 5, characterized in that, The process of acquiring historical processing data and building a product evaluation model includes: The qualified products refer to a batch of bearings that have completed all processing steps and are free of defects. The process parameters, operating parameters, first material parameters, second material parameters, first type vibration parameters, and second type vibration parameters corresponding to different processing steps of the same batch of qualified products are used as the historical processing data of their corresponding processing steps. Based on the historical processing data of qualified products from different batches at different processing stages, a corresponding third evaluation set is generated, which is then divided into a third training set and a third test set. Construct a second convolutional neural network, using different processing stages and their historical processing data from the third training set as input data for the second convolutional neural network, and using whether the product is qualified at the corresponding processing stage as output data for the second convolutional neural network. The second convolutional neural network is trained to obtain an initial second convolutional neural network. The initial second convolutional neural network is validated using a third test set. The initial second convolutional neural network whose output is less than or equal to the preset third test error threshold is used as the product evaluation model.

7. The bearing machining process fault detection system based on the Internet of Things according to claim 6, characterized in that, The process of obtaining actual processing data and determining whether there are faults in each processing stage includes: The actual processing data refers to the operating parameters, first material parameters, process parameters, and second material parameters of the same batch of products in the current processing stage in a real application scenario. Input the current operating parameters, the first material parameters, the process parameters, and the equipment parameters of the current processing stage into the first evaluation model to obtain the corresponding type of vibration parameters; The current second material parameters and process parameters, combined with the equipment parameters of the current processing stage, are input into the second evaluation model to obtain the corresponding second type of vibration parameters; The current actual processing data, along with the acquired Class I and Class II vibration parameters, are input into the product evaluation model to determine whether the product is qualified at the current processing stage. If the result is satisfactory, the subsequent processing steps will continue; if the result is unsatisfactory, a fault signal for the corresponding processing step will be generated and fed back to the relevant personnel.

8. A bearing machining process fault detection method based on the Internet of Things, implemented based on the bearing machining process fault detection system according to any one of claims 1-7, characterized in that, The method includes: Step S1: Divide the bearing processing process into different processing stages and obtain the equipment parameters and process parameters for each processing stage; Step S2: Set up different monitoring units in each processing stage and obtain the corresponding operating parameters. Combine the equipment parameters and process parameters to build a simulation model for each processing stage. Step S3: Obtain the first type of vibration parameters of each simulation model under different operating parameters in the isolated state, and obtain the second type of vibration parameters of the simulation models of adjacent processing links under different material parameters in the cooperative state; Step S4: Construct a vibration evaluation model based on the second type of vibration parameters under different material parameters and first type of vibration parameters of each simulation model, obtain historical processing data of qualified products, and construct a product evaluation model based on various parameters in different historical processing data; Step S5: Obtain actual processing data, and combine the vibration assessment model and product assessment model to determine whether there are faults in each processing stage, and generate corresponding fault signals for feedback.

Images