Automobile part machining control method and system

Abstract

The invention relates to the field of automobile machining, and discloses an automobile part machining control system which comprises a data acquisition module used for acquiring related data to facilitate subsequent prediction work; the virtual model module is used for designing related data of the physical equipment through related tools; the data integration module is used for transmitting the acquired data so as to realize synchronous updating; the algorithm analysis module is used for improving the prediction capability and the self-adaptability of the model through some algorithms; and the verification and calibration module is used for ensuring that the simulation result of the virtual model is consistent with the behavior of the actual equipment. According to the invention, a self-adaptive learning unit and an algorithm unit in the algorithm analysis module are adopted to carry out deep learning and analysis on historical data, and a prediction model is continuously optimized through cross validation, gradient descent and other algorithms, so that the self-adaptability of the system is improved; by means of the method, the system can automatically adjust the machining strategy according to changes in the actual machining process, and human intervention is reduced.

Description

Technical Field

[0001] The present invention relates to the field of automobile processing, and in particular to an automobile parts processing control method and system. Background Art

[0002] In the modern automobile manufacturing industry, the accuracy and efficiency of parts processing are directly related to the performance and quality of the whole vehicle. With the continuous development of automobile technology, the requirements for the accuracy of parts processing in the manufacturing process are getting higher and higher. However, due to the complex and changeable processing environment, traditional processing control methods are often difficult to adapt to the real-time changes of various working conditions, resulting in unstable processing quality. In order to improve the accuracy and efficiency of the processing process, sensor monitoring, data analysis and virtual simulation technologies have been introduced into the existing technology. However, these technologies still have problems such as inaccurate data collection, asynchrony between models and actual equipment, and insufficient prediction capabilities in actual applications.

[0003] The processing control systems in the existing technology are usually unable to achieve effective integration and processing of multi-sensor data, which makes it difficult to accurately reflect the processing status in real time. In addition, traditional processing control methods mostly rely on fixed algorithm models, lack adaptive capabilities, and are difficult to cope with the complex and changeable situations that occur in the actual processing process.

[0004] Therefore, a processing control system is needed that can comprehensively apply multi-sensor data, virtual simulation, algorithm analysis, and calibration verification to achieve high-precision and high-efficiency control of the automotive parts processing process. Summary of the invention

[0005] In order to make up for the above deficiencies, the present invention provides an automobile parts processing control method and system, aiming to improve the problem of "the traditional automobile parts processing control system has low working efficiency" mentioned in the prior art.

[0006] In order to achieve the above-mentioned purpose, the present invention adopts the following technical scheme: an automobile parts processing control system, comprising:

[0007] Data collection module, used to collect relevant data to facilitate subsequent prediction work;

[0008] The virtual model module uses relevant tools to design the relevant data of physical equipment;

[0009] Data integration module, which can transmit the acquired data to achieve synchronous update;

[0010] Algorithm analysis module, which uses some algorithms to improve the prediction ability and adaptability of the model;

[0011] A validation and calibration module to ensure that the simulation results of the virtual model are consistent with the behavior of the actual device;

[0012] The data acquisition module, virtual model module, data integration module, algorithm analysis module and verification and calibration module are connected through network communication. The data acquisition module includes multiple sensor units, a cleaning unit and a cloud database unit. The sensor units may include but are not limited to temperature sensors, pressure sensors, speed sensors, displacement sensors and vibration sensors, and they are all installed on actual physical equipment to facilitate the acquisition of relevant data in various aspects. The cleaning unit is used to perform relevant processing on the acquired data to ensure the accuracy of subsequent prediction work data. The cloud database unit is used to store the data acquired by the multiple sensor units after cleaning by the cleaning unit to facilitate subsequent use.

[0013] As a further description of the above technical solution:

[0014] The virtual model module includes a data acquisition unit, a simulation unit and a modeling unit. The data acquisition unit is used to acquire the shape, size and structure of the physical device, wherein the model can be created by using, but not limited to, CAD tools to acquire data or by actually measuring the physical device. The simulation unit is used to model the physical characteristics of the device, wherein the physical characteristics include, but are not limited to, kinematics, dynamics, and thermodynamics, and can be used by, but not limited to, CAE tools or other simulation software. The modeling unit describes the behavior of the device by modeling based on physical laws, so that its operating state can be simulated more accurately. Fluid dynamics simulation can be used, and the formula used is as follows:

[0015]

[0016] Where: u is the velocity field;

[0017] For a compressible fluid, the continuity equation is:

[0018]

[0019] Where: ρ is the density of the fluid;

[0020] The Navier-Stokes equation can also be used, expressed as follows:

[0021]

[0022] Where: ρ is the density of the fluid;

[0023] P is pressure;

[0024] f is the external body force;

[0025] The term on the left describes the acceleration of the fluid (i.e., the change in fluid momentum), and the term on the right describes the It shows the force generated by the pressure gradient. is the viscous force, and f is the external force

[0026] As a further description of the above technical solution:

[0027] The data integration module includes a data synchronization unit and an adjustment unit. The data synchronization unit is used to connect the sensor unit on the physical device with the virtual model to maintain synchronous updates between the virtual and the actual. The adjustment unit is used to make timely adjustments based on the data transmitted from the sensor unit to the data synchronization unit, so as to more accurately reflect the current status of the physical device.

[0028] As a further description of the above technical solution:

[0029] The algorithm analysis module includes an algorithm unit and an analysis unit. The algorithm unit is used to analyze historical data to improve the prediction ability and adaptability of the entire system. The analysis unit is used to discover some problems that may exist in the system. The algorithm unit can use a cross-validation method for verification, and the formula used is as follows:

[0030] For a data set {D1,D2,…,D K , divide it into K subsets, use one K-1 subset to train the model each time, and verify it on the remaining subsets. The average error is:

[0031]

[0032] Where: E validation (D k ) is the validation error of the k-th fold.

[0033] As a further description of the above technical solution:

[0034] The verification and calibration module includes a model verification unit and a continuous calibration unit. The model verification unit is used to ensure that the simulation result of the virtual model is consistent with the behavior of the actual device, and the continuous calibration unit is used to ensure that the model verification unit can always maintain accuracy.

[0035] As a further description of the above technical solution:

[0036] The model verification unit adopts error analysis method for verification, and the formula used is as follows:

[0037] Absolute error:

[0038] ΔA=|A measured -A true |

[0039] Among them: A measured is a measured or calculated value;

[0040] Relative error:

[0041]

[0042] Where: ΔA is the absolute error.

[0043] As a further description of the above technical solution:

[0044] The continuous calibration unit adopts the calibration error method, and the reference formula is as follows:

[0045] E=X measured -X true

[0046] Where: E is the calibration error;

[0047] X true It is a known standard value or reference value.

[0048] As a further description of the above technical solution:

[0049] The data that the data acquisition module needs to collect includes but is not limited to the temperature, pressure, speed, displacement and vibration data of the actual equipment. The cleaning unit includes but is not limited to processing missing values, abnormal value detection, and noise filtering in the data, thereby improving the data quality and improving the accuracy of subsequent simulation and prediction work. The data integration module also includes a data conversion unit for converting different types of sensor data into a unified format and standardizing the data according to actual needs in order to better adapt to the application in the virtual model module and the algorithm analysis module. The data conversion unit can normalize the data through the following formula:

[0050]

[0051] Where: X normalized is the normalized data;

[0052] X max and X min are the minimum and maximum values ​​of this type of sensor data respectively;

[0053]

[0054] Where: T aligned (t) is the timestamp aligned on t.

[0055] As a further description of the above technical solution:

[0056] The algorithm analysis module also includes an adaptive learning unit, which continuously optimizes the prediction model through deep learning analysis of historical data to enable it to have stronger self-adjustment capabilities. The adaptive learning can adopt an optimization algorithm based on gradient descent, and its formula is as follows:

[0057]

[0058] Where: θ t are the parameters of the current model;

[0059] is the gradient of the loss function Jθ with respect to θ.

[0060] A method for controlling automobile parts processing comprises the following steps:

[0061] S1: First, sensors are installed on the actual equipment related to automobile parts processing, including temperature sensors, pressure sensors, speed sensors, displacement sensors and vibration sensors, and then the data are transmitted to the data conversion unit through these sensors and the data transmitted by the sensors are converted into a unified format through the data conversion unit, and then the data is purified by the cleaning unit of the data acquisition module, and the purified data is transmitted to the cloud database unit for storage, so as to facilitate subsequent use;

[0062] S2: The data acquisition unit acquires data related to the actual device, and the simulation unit models the device based on its physical characteristics, and then the modeling unit describes the device's behavior and simulates its operating state;

[0063] S3: Connect the sensor unit with the cloud database unit through the data synchronization unit to keep the data updated synchronously, and adjust the data in time during the update process through the adjustment unit to ensure that the current state of the physical device is accurately reflected;

[0064] S4: After the SS step, the data can be analyzed through the algorithm unit and analysis unit in the algorithm analysis module to improve the prediction ability and adaptability of the entire system and timely discover the problems that may exist in the system, so that the system has the ability of autonomous learning and reduces the failure rate;

[0065] S5: The accuracy of the simulation work is ensured through the algorithms of the model verification unit and the continuous calibration unit. Finally, the prediction model can be continuously optimized through the adaptive learning unit in the algorithm analysis module to make it have stronger self-adjustment capabilities and improve the intelligence of the system.

[0066] The present invention has the following beneficial effects:

[0067] 1. In the present invention, the adaptive learning unit and the algorithm unit in the algorithm analysis module are used to perform deep learning and analysis on the historical data, and the prediction model is continuously optimized through algorithms such as cross-validation and gradient descent to improve the adaptability of the system. This method enables the system to automatically adjust the processing strategy according to the changes in the actual processing process, reduces human intervention, improves the intelligence of the system, and reduces the failure rate in the processing process.

[0068] 2. In the present invention, the key parameters of the physical equipment, such as temperature, pressure, speed, displacement and vibration, are monitored in real time through the data acquisition module, and the physical equipment and the virtual model are synchronously updated through the data synchronization unit and the adjustment unit in the data integration module, thereby ensuring the real-time and accuracy of the data during the processing, enabling the system to adjust the processing parameters in time, ensuring that the processing state is always in the optimal state, and improving the processing accuracy of parts.

[0069] 3. In the present invention, the simulation results of the virtual model are continuously verified and calibrated by the model verification unit and the continuous calibration unit in the verification and calibration module to ensure that the virtual model behaves in accordance with the actual equipment. Through the application of error analysis and calibration error method, the deviations in the virtual model can be discovered and corrected in time to ensure a high degree of consistency between the virtual simulation and the actual processing environment, thereby improving the reliability of the processing control system and the processing quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] Figure 1 A schematic diagram of the system framework of an automobile parts processing control method and system in the present invention;

[0071] Figure 2 A detailed framework diagram of a data acquisition module of an automobile parts processing control method and system in the present invention;

[0072] Figure 3 A detailed framework diagram of a virtual model module of an automobile parts processing control method and system in the present invention;

[0073] Figure 4 A detailed framework diagram of a data integration module of an automobile parts processing control method and system in the present invention;

[0074] Figure 5 A detailed framework diagram of an algorithm analysis module of an automobile parts processing control method and system in the present invention;

[0075] Figure 6 It is a detailed framework diagram of a verification and calibration module of an automobile parts processing control method and system in the present invention.

[0076] Legend:

[0077] 1. Data acquisition module; 2. Virtual model module; 3. Data integration module; 4. Algorithm analysis module; 5. Verification and calibration module; 101. Sensor unit; 102. Cleaning unit; 103. Cloud database unit; 201. Data acquisition unit; 202. Simulation unit; 203. Modeling unit; 301. Data synchronization unit; 302. Adjustment unit; 303. Data conversion unit; 401. Algorithm unit; 402. Analysis unit; 403. Adaptive learning unit; 501. Model verification unit; 502. Continuous calibration unit. DETAILED DESCRIPTION

[0078] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0079] Example

[0080] Reference 1- Figure 6 The present invention provides an automobile parts processing control system, including a data acquisition module 1, a virtual model module 2, a data integration module 3, and a verification and calibration module 5. The data acquisition module 1, the virtual model module 2, the data integration module 3, the algorithm analysis module 4 and the verification and calibration module 5 are connected through network communication, wherein the data that the data acquisition module 1 needs to collect includes but is not limited to the temperature, pressure, speed, displacement and vibration data of the actual equipment.

[0081] Furthermore, the data acquisition module 1 includes multiple sensor units 101, a cleaning unit 102 and a cloud database unit 103. The sensor unit 101 may include but is not limited to temperature sensors, pressure sensors, speed sensors, displacement sensors and vibration sensors, and they are all installed on actual physical devices to facilitate the acquisition of relevant data in various aspects. The sensor is a prior art and can be implemented by technicians in this field, so it will not be described in detail in this case. The cleaning unit 102 is used to perform relevant processing on the acquired data to ensure the accuracy of subsequent prediction work data. Specifically, the main functions of the cleaning unit 102 are not limited to in-depth processing of the data to ensure the accuracy and reliability of the data. Specifically, it involves multiple steps, including but not limited to processing missing values ​​in the data to ensure the integrity of the data set; performing outlier detection to identify and process those that do not meet the expected model The cloud database unit 103 is used to store the data acquired by the multiple sensor units 101 after being cleaned by the cleaning unit 102, so as to facilitate subsequent use. The cloud database unit 103 provides an efficient and reliable data storage solution. By centrally storing the data collected by the sensors in the cloud, efficient management and rapid access to large-scale data can be achieved. Compared with traditional local storage methods, the cloud database has stronger scalability and flexibility, and allows multiple modules and systems to access and share data in real time through the network, so that the data acquisition module 1 and other modules, such as the virtual model module 2 and the algorithm analysis module 4, can obtain the latest sensor data in a timely manner, thereby improving the response speed and processing efficiency of the system.

[0082] Furthermore, the virtual model module 2 includes a data acquisition unit 201, a simulation unit 202 and a modeling unit 203. The data acquisition unit 201 is used to acquire the shape, size and structure of the physical device, wherein the model may be created using, but not limited to, CAD tools to acquire data or by actually measuring the physical device. The simulation unit 202 is used to model the physical characteristics of the device, which include, but are not limited to, kinematics, dynamics, and thermodynamics, and may use, but are not limited to, CAE tools or other simulation software. The modeling unit 203 describes the behavior of the device by modeling based on physical laws, so that its operating state can be simulated more accurately, and fluid dynamics simulation may be used, and the formula used is as follows:

[0083]

[0084] Where: u is the velocity field;

[0085] For a compressible fluid, the continuity equation is:

[0086]

[0087] Where: ρ is the density of the fluid;

[0088] The Navier-Stokes equation can also be used, expressed as follows:

[0089]

[0090] Where: ρ is the density of the fluid;

[0091] P is pressure;

[0092] f is the external body force;

[0093] The term on the left describes the acceleration of the fluid (i.e., the change in fluid momentum), and the term on the right describes the It shows the force generated by the pressure gradient. is the viscous force, and f is the external force

[0094] The virtual model module 2 accurately models the physical device through the simulation unit 202 and the modeling unit 203, and can simulate various situations that may be encountered during the processing, making the processing more controllable, thereby improving the processing accuracy.

[0095] Furthermore, the data integration module 3 includes a data synchronization unit 301 and an adjustment unit 302. The data synchronization unit 301 is used to connect the sensor unit 101 on the physical device with the virtual model to maintain synchronous updates between the virtual and the actual. The adjustment unit 302 is used to make timely adjustments based on the data transmitted from the sensor unit 101 to the data synchronization unit 301, so as to more accurately reflect the current state of the physical device. The data integration module 3 also includes a data conversion unit 303, which is used to convert different types of sensor data into a unified format and standardize the data according to actual needs in order to better adapt to the application in the virtual model module 2 and the algorithm analysis module 4. The data conversion unit 303 can normalize the data by the following formula:

[0096]

[0097] Where: X normalized is the normalized data;

[0098] X max and X min are the minimum and maximum values ​​of this type of sensor data respectively;

[0099]

[0100] Where: T aligned(t) is the timestamp aligned on t.

[0101] The data integration module 3 improves the system's collaborative capabilities, allowing the virtual model module 2 and the data analysis module 4 to make decisions and adjustments based on the same data, thereby optimizing the overall system performance.

[0102] Furthermore, the algorithm analysis module 4 includes an algorithm unit 401 and an analysis unit 402. The algorithm unit 401 is used to analyze historical data, thereby improving the prediction ability and adaptability of the entire system. The analysis unit 402 is used to discover some problems that may exist in the system. The algorithm unit 401 can be verified by a cross-validation method, and the formula used is as follows:

[0103] For a data set {D1,D2,…,D K , divide it into K subsets, use one K-1 subset to train the model each time, and verify it on the remaining subsets. The average error is:

[0104]

[0105] Where: E validation (D k ) is the validation error of the k-th fold.

[0106] The algorithm analysis module 4 also includes an adaptive learning unit 403, which continuously optimizes the prediction model through deep learning analysis of historical data to make it have stronger self-adjustment ability. The adaptive learning can adopt an optimization algorithm based on gradient descent, and its formula is as follows:

[0107]

[0108] Where: θ t are the parameters of the current model;

[0109] is the gradient of the loss function Jθ with respect to θ.

[0110] The algorithm analysis module 4 uses comprehensive technology to conduct in-depth analysis of large amounts of data during the processing process, providing accurate fault diagnosis, optimization suggestions and decision support.

[0111] Furthermore, the verification and calibration module 5 includes a model verification unit 501 and a continuous calibration unit 502. The model verification unit 501 is used to ensure that the simulation result of the virtual model is consistent with the behavior of the actual device. Specifically, the model verification unit 501 uses an error analysis method for verification, and the formula used is as follows:

[0112] Absolute error:

[0113] ΔA=|A measured -Atrue |

[0114] Among them: A measured is a measured or calculated value;

[0115] Relative error:

[0116]

[0117] Where: ΔA is the absolute error.

[0118] The continuous calibration unit 502 adopts a calibration error method, and the reference formula is as follows:

[0119] E=X measured -X true

[0120] Where: E is the calibration error;

[0121] X true It is a known standard value or reference value.

[0122] The continuous calibration unit 502 is used to ensure that the model verification unit 501 can always maintain accuracy.

[0123] The present invention also provides an automobile parts processing control method, comprising the following steps:

[0124] S1: First, sensors are installed on the actual equipment related to automobile parts processing, including temperature sensors, pressure sensors, speed sensors, displacement sensors and vibration sensors, and then the data are transmitted to the data conversion unit 303 through these sensors and the data transmitted by the sensors are converted into a unified format through the data conversion unit 303, and then the data is purified by the cleaning unit 102 of the data acquisition module 1, and the purified data is transmitted to the cloud database unit 103 for storage, so as to facilitate subsequent use;

[0125] S2: The data acquisition unit 201 acquires data related to the actual device, and the simulation unit 202 models the device according to its physical characteristics, and then the modeling unit 203 describes the behavior of the device to simulate its operating state;

[0126] S3: The sensor unit 101 is connected to the cloud database unit 103 through the data synchronization unit 301 to maintain synchronous update of the data, and the data is adjusted in time through the adjustment unit 302 during the update process to ensure that the current state of the physical device is accurately reflected;

[0127] S4: After the steps S1-S3, the data can be analyzed by the algorithm unit 401 and the analysis unit 402 in the algorithm analysis module 4 to improve the prediction ability and adaptive ability of the whole system and timely discover the problems that may exist in the system, so that the system has the ability of autonomous learning and reduces the failure rate;

[0128] S5: The accuracy of the simulation work is ensured through the algorithms of the model verification unit 501 and the continuous calibration unit 502, and finally the prediction model can be continuously optimized through the adaptive learning unit 403 in the algorithm analysis module 4 to make it have stronger self-adjustment ability and improve the intelligence of the system.

Claims

1. An automobile parts processing control system, characterized in that: include: A data collection module (1) is used to collect relevant data to facilitate subsequent prediction work; Virtual model module (2), designs relevant data of physical equipment through relevant tools; Data integration module (3), which transmits the acquired data to achieve synchronous update; Algorithm analysis module (4), which uses some algorithms to improve the prediction ability and adaptability of the model; A verification and calibration module (5) is used to ensure that the simulation results of the virtual model are consistent with the behavior of the actual device; The data acquisition module (1), the virtual model module (2), the data integration module (3), the algorithm analysis module (4) and the verification and calibration module (5) are connected via network communication. The data acquisition module (1) comprises a plurality of sensor units (101), a cleaning unit (102) and a cloud database unit (103). The sensor units (101) may include but are not limited to temperature sensors, pressure sensors, speed sensors, displacement sensors and vibration sensors, and are all installed on actual physical devices to facilitate the acquisition of relevant data in various aspects. The cleaning unit (102) is used to perform relevant processing on the acquired data to ensure the accuracy of subsequent prediction work data. The cloud database unit (103) is used to store the data acquired by the plurality of sensor units (101) after being cleaned by the cleaning unit (102) to facilitate subsequent use.

2. The automotive parts processing control system according to claim 1, characterized in that: The virtual model module (2) comprises a data acquisition unit (201), a simulation unit (202) and a modeling unit (203), wherein the data acquisition unit (201) is used to acquire the shape, size and structure of the physical device, wherein the model may be created using, but not limited to, CAD tools or by actually measuring the physical device, the simulation unit (202) is used to model the physical characteristics of the device, wherein the physical characteristics include, but are not limited to, kinematics, dynamics and thermodynamics, and may be modeled using, but not limited to, CAE tools or other simulation software, the modeling unit (203) describes the behavior of the device by modeling based on physical laws, thereby simulating its operating state more accurately, and may use fluid dynamics simulation, using the following formula: Where: u is the velocity field; For a compressible fluid, the continuity equation is: Where: ρ is the density of the fluid; The Navier-Stokes equation can also be used, expressed as follows: Where: ρ is the density of the fluid; P is pressure; f is the external body force; The term on the left describes the acceleration of the fluid (i.e. the change in the fluid momentum), and the term on the right describes the acceleration of the fluid (i.e. the change in the fluid momentum). represents the force due to the pressure gradient, is the viscous force and f is the external force.

3. The automobile parts processing control system according to claim 1, characterized in that: The data integration module (3) comprises a data synchronization unit (301) and an adjustment unit (302). The data synchronization unit (301) is used to connect the sensor unit (101) on the physical device with the virtual model to maintain synchronous updates between the virtual and the actual. The adjustment unit (302) is used to make timely adjustments based on the data transmitted from the sensor unit (101) to the data synchronization unit (301), thereby being able to more accurately reflect the current state of the physical device.

4. The automobile parts processing control system according to claim 1, characterized in that: The algorithm analysis module (4) comprises an algorithm unit (401) and an analysis unit (402), wherein the algorithm unit (401) is used to analyze historical data, thereby improving the prediction capability and adaptability of the entire system, and the analysis unit (402) is used to discover some problems existing in the system, wherein the algorithm unit (401) can be verified by a cross-validation method, and the formula used is as follows: For a data set {D1,D2,…,D K , divide it into K subsets, use one K-1 subset to train the model each time, and verify it on the remaining subsets. The average error is: Where: E validation (D k ) is the validation error of the k-th fold.

5. The automobile parts processing control system according to claim 1, characterized in that: The verification and calibration module (5) comprises a model verification unit (501) and a continuous calibration unit (502), wherein the model verification unit (501) is used to ensure that the simulation result of the virtual model is consistent with the behavior of the actual device, and the continuous calibration unit (502) is used to ensure that the model verification unit (501) can always maintain accuracy.

6. The automobile parts processing control system according to claim 5, characterized in that: The model verification unit (501) uses an error analysis method for verification, using the following formula: Absolute error: ΔA=|A measured -IN true | Among them: A measured is a measured or calculated value; Relative error: Where: ΔA is the absolute error.

7. The automobile parts processing control system according to claim 1, characterized in that: The continuous calibration unit (502) adopts a calibration error method, and the reference formula is as follows: E=X measured -X true Where: E is the calibration error; X true It is a known standard value or reference value.

8. The automobile parts processing control system according to claim 1, characterized in that: The data that the data acquisition module (1) needs to collect includes, but is not limited to, temperature, pressure, speed, displacement and vibration data of the actual equipment. The cleaning unit (102) includes, but is not limited to, processing missing values, abnormal value detection and noise filtering in the data, thereby improving data quality and improving the accuracy of subsequent simulation and prediction work. The data integration module (3) also includes a data conversion unit (303) for converting different types of sensor data into a unified format and performing standardization processing on the data according to actual needs so as to better adapt to the application in the virtual model module (2) and the algorithm analysis module (4). The data conversion unit (303) can perform normalization processing on the data by the following formula: Where: X normalized is the normalized data; X max and X min are the minimum and maximum values ​​of this type of sensor data respectively; Where: T aligned (t) is the timestamp aligned on t.

9. The automobile parts processing control method according to claim 1, characterized in that: The algorithm analysis module (4) also includes an adaptive learning unit (403), which continuously optimizes the prediction model through deep learning analysis of historical data to enable it to have stronger self-adjustment ability. The adaptive learning can adopt an optimization algorithm based on gradient descent, and its formula is as follows: Where: θ t are the parameters of the current model; is the gradient of the loss function Jθ with respect to θ.

10. An automobile parts processing control method according to any one of claims 1 to 9, characterized in that: The following steps are involved: S1: First, sensors are installed on the actual equipment related to automobile parts processing, including temperature sensors, pressure sensors, speed sensors, displacement sensors and vibration sensors, and then the sensors transmit data to the data conversion unit (303) and convert the data transmitted by the sensors into a unified format, and then the cleaning unit (102) of the data acquisition module (1) purifies the data, and transmits the purified data to the cloud database unit (103) for storage, so as to facilitate subsequent use; S2: acquiring data related to the actual device through the data acquisition unit (201), modeling the device through the physical characteristics of the device through the simulation unit (202), and then describing the behavior of the device through the modeling unit (203) to simulate its operating state; S3: connecting the sensor unit (101) and the cloud database unit (103) through the data synchronization unit (301) to maintain synchronous update of data, and timely adjusting the data through the adjustment unit (302) during the update process to ensure that the current state of the physical device is accurately reflected; S4: After the steps S1 to S3, the algorithm unit (401) and the analysis unit (402) in the algorithm analysis module (4) can analyze the data to improve the prediction ability and adaptive ability of the entire system and timely discover the problems that may exist in the system, so that the system has the ability of autonomous learning and reduces the failure rate; S5: The accuracy of the simulation work is ensured through the algorithms of the model verification unit (501) and the continuous calibration unit (502), and finally the prediction model can be continuously optimized through the adaptive learning unit (403) in the algorithm analysis module (4) to make it have stronger self-adjustment ability and improve the intelligence of the system.

Images