Safety management system for automobile transmission shaft production based on big data

Abstract

The application relates to the technical field of automobile transmission shaft production, in particular to a safety management system for automobile transmission shaft production based on big data, which comprises a data acquisition module, which is used for collecting production parameter data, environment parameter data, operator behavior data and quality inspection data of multiple production devices in the automobile transmission shaft production process in real time, and constructing an original safety data set; a big data processing platform, which is in communication connection with the data acquisition module, is used for cleaning, fusing and standardizing the original safety data set, and generating a standardized safety data set; a safety risk research and judgment module, which is connected with the big data processing platform, is used for storing a dynamic safety threshold model, and identifying abnormal states and safety risks in the transmission shaft production process in real time based on the standardized safety data set, and outputting quantitative research and judgment results; the safety risk research and judgment module realizes the whole domain monitoring, hierarchical early warning, millisecond-level emergency shutdown and closed-loop control of the transmission shaft production safety, and significantly improves the response speed and management efficiency.

Description

Technical Field

[0001] This invention relates to the field of automotive driveshaft manufacturing technology, specifically a safety management system for automotive driveshaft manufacturing based on big data. Background Technology

[0002] As a core component of the automotive power transmission system, the production process of automotive driveshafts involves multiple steps such as forging, heat treatment, and machining, which places high demands on equipment operating status, process parameters, operating procedures, and on-site environment. With the in-depth application of big data technology in the industrial manufacturing field, data-driven safety management methods are gradually becoming an important means to improve the level of production safety. According to CN116125852A, a big data-based intelligent monitoring system and method for automotive parts production lines is disclosed. This technology includes an inventory monitoring module, a manufacturing monitoring module, a quality monitoring module, and a big data analysis module. The inventory monitoring module selects finished products and raw materials; the manufacturing module monitors the production process; the quality monitoring module performs quality checks; and the big data analysis module conducts comprehensive analysis of the production process. When a company produces parts, the inventory monitoring module manages finished products and raw materials to achieve green production; the manufacturing monitoring module monitors the production line to ensure safe production; the quality monitoring module guarantees product quality; and the big data analysis module comprehensively analyzes data to improve production line efficiency. In the existing safety management of automotive driveshaft production, data acquisition mostly relies on single-type sensors or manual inspections. There is a lack of a unified timestamp association and fusion mechanism between vibration, temperature, load, image, environmental and quality inspection data. Each data source is independent, forming information silos. This makes it difficult to comprehensively identify the coupling risks between equipment operating status, personnel operation behavior, process parameter deviations and workshop environment changes. Safety judgments are based solely on threshold comparisons of isolated parameters, and safety benchmarks cannot be adjusted according to the dynamic characteristics of different driveshaft models, materials and processing stages. At the same time, the processing delay of the central platform creates a time lag between data acquisition and instruction issuance. When sudden parameter exceedances occur, it is impossible to achieve millisecond-level independent emergency shutdown at the equipment side. Furthermore, equipment control, personnel supervision and emergency dispatch after the warning are disconnected. There is a lack of differentiated design for the response methods of low-level and high-level warnings. Low-risk events and high-risk events occupy the same emergency resources. The entire safety management process lacks a complete closed-loop chain from risk identification to equipment linkage, personnel handling and rectification verification. Summary of the Invention

[0003] To address the shortcomings of existing technologies, this invention provides a big data-based safety management system for automotive driveshaft production, which enables comprehensive monitoring, tiered early warning, millisecond-level emergency shutdown, and closed-loop control of driveshaft production safety, significantly improving response speed and management efficiency.

[0004] To achieve the above objectives, the present invention provides the following technical solution: a safety management system for automotive driveshaft production based on big data, comprising: The data acquisition module is used to collect production parameter data, environmental parameter data, operator behavior data and quality inspection data of multiple production equipment during the production process of automotive drive shafts in real time, and to build the original safety dataset. The big data processing platform communicates with the data acquisition module and is used to clean, merge and standardize the raw security dataset to generate a standardized security dataset. The safety risk assessment module is connected to the big data processing platform to store dynamic safety threshold models and identify abnormal states and safety risks in the transmission shaft production process in real time based on standardized safety datasets, and output quantitative assessment results. The graded early warning execution module is connected to the security risk assessment module. It is used to execute graded differentiated early warning push and corresponding security early warning actions based on the risk level of the quantitative assessment results. The production equipment safety management module is connected to the graded early warning execution module and the safety risk assessment module, respectively. It is used to perform automated closed-loop safety management of abnormal risks of the transmission shaft production equipment, including sending speed reduction or shutdown commands to the equipment programmable logic controller. The personnel operation safety supervision module is connected to the graded early warning execution module to monitor the on-site personnel's qualifications and operating procedures in real time, and to identify and record violations of operating procedures. The emergency response and dispatch module is connected to the production equipment safety management and control module and the personnel operation safety supervision module, respectively. It is used to receive level 2 and level 3 high-risk warnings, automatically match emergency plans, and complete intelligent dispatch and risk area isolation and control. Edge computing nodes are set up on the production equipment side of the drive shaft and are connected to the data acquisition module and the big data processing platform respectively. They are used to perform on-site real-time analysis of the collected data. When any data exceeds the preset emergency stop threshold, an emergency stop command is directly sent to the PLC of the production equipment to complete the shutdown. At the same time, the hierarchical early warning execution module is triggered to report the early warning, and the on-site analysis results are uploaded to the big data processing platform.

[0005] Preferably, the data acquisition module includes: The equipment sensing and acquisition unit is installed on the spindle and tool post of the transmission shaft machining tool to collect real-time spindle vibration spectrum data, tool vibration amplitude data, forging temperature, heat treatment furnace temperature and axial load data. The image acquisition unit includes at least one industrial camera for acquiring data on the operator's safety equipment wearing status, intrusion behavior data in hazardous areas of the equipment, and process operation behavior data; The environmental sensing unit includes a dust concentration sensor and a noise sensor, used to collect dust concentration data and noise decibel data in the production workshop; The quality inspection data docking unit is used to dock with quality inspection equipment and synchronously collect quality inspection and safety data of drive shaft products.

[0006] Preferably, the big data processing platform includes: The data cleaning unit is used to remove invalid data, outliers, and interference noise from the original secure dataset. The missing data completion unit is used to intelligently complete missing data in production conditions; The multi-source data fusion unit is used to associate and align the cleaned and completed multi-source data according to timestamps to form a time-series fusion feature set and build a secure data chain for the entire product lifecycle. The standardization mapping unit is used to map data of different dimensions in the time-series fusion feature set to a unified standardization interval through normalization processing, thereby generating a standardized security dataset.

[0007] Preferably, the security risk assessment module includes: The dynamic safety threshold database is used to store multi-level safety thresholds that are dynamically adjusted according to different drive shaft models, materials and processing stages. The multi-level safety thresholds include at least a primary threshold and a secondary threshold. The value of the secondary threshold corresponding to the same monitoring parameter is greater than that of the primary threshold. The real-time comparison unit is used to compare the current value in the standardized security dataset with the corresponding multi-level security thresholds in real time and calculate the deviation between the current value and each threshold. The multi-dimensional risk assessment unit is used to identify various safety risks throughout the entire production process based on a dedicated big data model for drive shaft production, from four dimensions: equipment failure risk, process compliance risk, personnel operation risk, and environmental safety risk. The anomaly identification unit, connected to the real-time comparison unit and the multi-dimensional risk assessment unit, is used to generate a primary anomaly status signal when the deviation exceeds the first-level threshold or any dimension of risk is identified; and to generate an advanced anomaly status signal when the deviation exceeds the second-level threshold or a high-risk risk is identified.

[0008] Preferably, the tiered early warning execution module includes: The risk level determination unit is used to determine the current warning level as Level 1, Level 2, or Level 3 based on the type of abnormal state signal received, the duration of the signal, and the results of multi-dimensional risk assessment. The multi-terminal hierarchical push unit is used to accurately push warning content and handling guidelines according to the current warning level and job position. The audible and visual alarm unit is used to issue an audible and visual alarm prompt at the corresponding workstation when the current warning level is Level 1. The early warning full data retention unit is used to completely retain the data of the entire early warning process and form a traceable early warning ledger.

[0009] Preferably, the production equipment safety management module includes: The equipment status full-domain monitoring unit is used to monitor and display the equipment's operating status and parameter trends in real time; The parameter anomaly locking unit is used to lock equipment access when parameters exceed limits, thereby preventing unauthorized production. The risk-linked start-stop control unit is used to send a speed reduction command or a shutdown command to the programmable logic controller of the corresponding drive shaft production equipment when the current warning level is level two or level three, and to perform automatic safety start-stop protection for high-risk equipment.

[0010] Preferably, the personnel operation safety supervision module includes: The on-the-job qualification verification unit is used to verify the compliance qualifications of operators. The process operation behavior recognition unit is used to intelligently identify various violations in process operations; The personal protective equipment (PPE) compliance testing unit is used to test whether workers are wearing PPE in compliance with regulations. The violation tracing assessment unit is used for evidence storage, tracing of violations, and linkage with job performance assessments.

[0011] Preferably, the emergency response dispatch module includes: The intelligent emergency response plan retrieval unit is used to automatically match specific emergency response plans based on risk type and level. A tiered personnel dispatch unit is used to allocate personnel and tasks as needed. The risk area isolation and control unit is used to achieve intelligent isolation and control of risk areas; The closed-loop feedback unit is used to track the entire handling process, linking the production equipment safety management module and the personnel operation safety supervision module to achieve a closed-loop management of risk warning, handling, rectification, and verification.

[0012] This invention provides a safety management system for automotive driveshaft production based on big data. Compared with existing technologies, it has the following advantages: 1. Through a step-by-step pipeline processing of data acquisition module, big data processing platform, safety risk assessment module, and graded early warning execution module, combined with a dual verification mechanism of dynamic safety threshold and dedicated big data model for drive shaft production, the system achieves automatic identification and quantitative assessment of four types of risks in drive shaft production: equipment failure, process compliance, personnel operation, and environmental safety. It can output graded abnormal status signals based on deviation and risk dimensions, avoiding missed and false alarms caused by a single data source or fixed threshold, and significantly improving the accuracy, comprehensiveness, and timeliness of safety risk identification.

[0013] 2. Edge computing nodes are deployed on the production equipment side to perform real-time analysis of the collected data. When any data exceeds the emergency stop threshold, an emergency stop command is sent directly to the equipment's programmable logic controller without going through the central processing flow. At the same time, the hierarchical early warning execution module is triggered to report the early warning and upload the analysis results to the big data processing platform. An independent emergency stop channel is built on the equipment side, which shortens the response time in extreme cases to the millisecond level, while not losing the recordability and traceability of the event, providing the highest priority safety protection for drive shaft production.

[0014] 3. The production equipment safety management module sends speed reduction or shutdown commands to the equipment during level 2 or 3 warnings and locks the execution authority for parameters exceeding limits. The personnel operation safety supervision module verifies the qualifications of employees, compliance of operations, and status of wearing personal protective equipment in real time, and records evidence of violations for traceability and assessment. The emergency response and dispatch module automatically matches emergency plans, assigns disposal tasks, and links with the access control and fence system to achieve risk area isolation and closed-loop management during high-risk warnings. The three modules work together to form a multi-level, all-encompassing safety protection system from automatic equipment control and personnel behavior supervision to emergency dispatch and disposal, realizing full coverage and closed-loop control of drive shaft production safety management. Attached Figure Description

[0015] Figure 1 This is a system overall block diagram of the present invention; Figure 2 This is a block diagram of the data acquisition module in this invention; Figure 3 This is a block diagram of the big data processing platform and security risk assessment module in this invention; Figure 4 This is a block diagram of the graded early warning execution module and the production equipment safety management module in this invention; Figure 5 This is a block diagram of the personnel operation safety supervision module and the emergency response dispatch module in this invention.

[0016] In the diagram: 1. Data Acquisition Module; 11. Equipment Sensor Acquisition Unit; 12. Image Acquisition Unit; 13. Environmental Sensing Unit; 14. Quality Inspection Data Integration Unit; 2. Big Data Processing Platform; 21. Data Cleaning Unit; 22. Missing Data Completion Unit; 23. Multi-Source Data Fusion Unit; 24. Standardized Mapping Unit; 3. Security Risk Assessment Module; 31. Dynamic Security Threshold Database; 32. Real-time Comparison Unit; 33. Multi-Dimensional Risk Assessment Unit; 34. Anomaly Identification Unit; 4. Hierarchical Early Warning Execution Module; 41. Risk Level Determination Unit; 42. Multi-Terminal Hierarchical Push Unit ; 43. Audible and visual alarm unit; 44. Early warning full data retention unit; 5. Production equipment safety management and control module; 51. Equipment status full-domain monitoring unit; 52. Parameter anomaly locking unit; 53. Risk linkage start-stop control unit; 6. Personnel operation safety supervision module; 61. On-the-job qualification verification unit; 62. Process operation behavior identification unit; 63. Labor protection wearing compliance detection unit; 64. Violation tracing assessment unit; 7. Emergency response dispatch module; 71. Classified emergency plan intelligent retrieval unit; 72. Hierarchical personnel dispatch unit; 73. Risk area isolation and control unit; 74. Closed-loop feedback unit for handling. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0018] Please see Figure 1 - Figure 5 This invention provides a technical solution: a safety management system for automotive driveshaft production based on big data, comprising: Data acquisition module 1 is used to collect production parameter data, environmental parameter data, operator behavior data and quality inspection data of multiple production equipment in the production process of automotive drive shaft in real time, and to build the original safety dataset; Big data processing platform 2 is connected to data acquisition module 1 and is used to clean, merge and standardize the original security dataset to generate a standardized security dataset. The safety risk assessment module 3 is connected to the big data processing platform 2. It is used to store dynamic safety threshold models and identify abnormal states and safety risks in the transmission shaft production process in real time based on standardized safety datasets, and output quantitative assessment results. The graded early warning execution module 4 is connected to the safety risk assessment module 3 and is used to execute graded differentiated early warning push and corresponding safety early warning actions based on the risk level of the quantitative assessment results. The production equipment safety management module 5 is connected to the graded early warning execution module 4 and the safety risk assessment module 3 respectively. It is used to perform automated closed-loop safety management of abnormal risks of the transmission shaft production equipment, including sending speed reduction or shutdown commands to the equipment programmable logic controller. The personnel operation safety supervision module 6 is connected to the graded early warning execution module 4 and is used to monitor the on-site personnel's qualifications and operation specifications in real time, and to identify and record violations of operation behavior. The emergency response and dispatch module 7 is connected to the production equipment safety management and control module 5 and the personnel operation safety supervision module 6 respectively. It is used to receive level 2 and level 3 high-risk warnings, automatically match emergency plans, and complete intelligent dispatch and risk area isolation and control. Edge computing node 8 is set on the side of the transmission shaft production equipment and is connected to data acquisition module 1 and big data processing platform 2 respectively. It is used to perform on-site real-time analysis of the collected data. When it is determined that any data exceeds the preset emergency stop threshold, it directly sends an emergency stop command to the production equipment PLC to complete the shutdown. At the same time, it triggers the hierarchical early warning execution module 4 to report the early warning and uploads the on-site analysis results to big data processing platform 2.

[0019] In this implementation plan, the data acquisition module 1 converts multi-dimensional physical quantities such as equipment vibration, temperature, load, on-site images, environmental dust and noise, and quality inspection output into a raw safety dataset; the big data processing platform 2 sequentially performs invalid data removal, missing data completion, timestamp-based alignment, and normalization mapping on the raw safety dataset to generate a standardized safety dataset; the safety risk assessment module 3 compares the current values ​​in the standardized safety dataset with the dynamic safety threshold model and calculates the deviation, while using a dedicated big data model for drive shaft production to identify risks from four dimensions: equipment failure, process compliance, personnel operation, and environmental safety, and outputs quantitative assessment results; the graded early warning execution module 4 determines level one, level two, or level three early warnings based on the risk level of the quantitative assessment results, and executes differentiated early warning pushes and audible and visual alarms; the production equipment safety management module 5 sends a speed reduction or shutdown command to the equipment programmable logic controller when it receives a level two or level three early warning, and locks the execution permission for parameters exceeding limits; the personnel operation safety supervision module 6 uses image recognition technology to verify on-the-job qualifications, compliance of process operations, and the status of wearing protective equipment in real time, and stores evidence and traces violations. Source and linkage assessment; the emergency response and dispatch module 7 is activated only when a level 2 or level 3 high-risk warning is received. It automatically matches the emergency plan, allocates personnel and tasks as needed, and links with the access control and fence system to isolate the risk area; the edge computing node 8 is set on the transmission shaft production equipment side, and analyzes the collected data on-site in real time. When any data is judged to exceed the preset emergency shutdown threshold, it directly issues an emergency shutdown command to the equipment programmable logic controller without going through the central processing flow. At the same time, it triggers the graded warning execution module 4 to report the warning and upload the analysis results to the big data processing platform 2. Through the collaborative work of each module, a multi-level safety protection system is formed, from regular graded warning to emergency shutdown, from automatic equipment control to personnel behavior supervision, and from risk identification to emergency dispatch. The deployment of the edge computing node 8 on the equipment side builds an independent emergency shutdown channel, which shortens the response time in extreme cases to the millisecond level without losing event traceability. The emergency response and dispatch module 7 is limited to activation of high-risk warnings to avoid the ineffective occupation of emergency resources by low-risk events. Overall, it realizes full coverage, graded response and closed-loop control of transmission shaft production safety management.

[0020] Specifically, data acquisition module 1 includes: The equipment sensing and acquisition unit 11 is installed on the spindle and tool post of the transmission shaft processing machine tool and is used to collect spindle vibration spectrum data, tool vibration amplitude data, forging temperature, heat treatment furnace temperature and axial load data in real time. Image acquisition unit 12 includes at least one industrial camera for acquiring data on the operator's safety equipment wearing status, intrusion behavior data in dangerous areas of equipment, and process operation behavior data; The environmental sensing unit 13 includes a dust concentration sensor and a noise sensor, used to collect dust concentration data and noise decibel data in the production workshop; The quality inspection data docking unit 14 is used to dock with quality inspection equipment and synchronously collect quality inspection and safety data of drive shaft products.

[0021] In this embodiment, the equipment sensing and acquisition unit 11 uses vibration sensors, temperature sensors, and load sensors installed on the spindle and tool post of the transmission shaft machining tool. It converts mechanical vibration into electrical signals using the vibration signal conversion principle to obtain spindle vibration spectrum data and tool vibration amplitude data; it converts temperature into electrical signals using the thermoelectric effect or thermal radiation principle to obtain forging temperature and heat treatment furnace temperature data; and it converts stress deformation into electrical signals using strain gauges or piezoelectric effect principles to obtain axial load data. The image acquisition unit 12 acquires on-site images using at least one industrial camera and uses image recognition algorithms to extract and compare features such as safety equipment wearing characteristics, dangerous area intrusion boundaries, and process operation postures in the images. The environmental sensing unit 13 uses a dust concentration sensor... The concentration of particulate matter in the air is converted into an electrical signal using light scattering or charge induction methods to obtain dust concentration data. A noise sensor converts sound pressure level into an electrical signal using the principle of acoustic-to-electrical conversion to obtain noise decibel data. The quality inspection data interface unit 14 establishes a data communication interface with the quality inspection equipment and uses data protocol parsing technology to synchronously parse the output data of the quality inspection equipment into quality inspection safety data for the drive shaft product. The data collected by the above units jointly construct the original safety dataset. Through the collaborative collection of multiple types of sensing units, comprehensive data coverage of equipment status, personnel behavior, workshop environment, and product quality during the drive shaft production process is achieved. This provides the main big data processing platform 2 with a complete, multi-source, and real-time original data foundation, avoiding monitoring blind spots caused by a single data source.

[0022] Specifically, the big data processing platform 2 includes: Data cleaning unit 21 is used to remove invalid data, outliers and interference noise from the original secure dataset; Missing data completion unit 22 is used to intelligently complete missing data in production conditions; The multi-source data fusion unit 23 is used to associate and align the cleaned and completed multi-source data according to the timestamp to form a time-series fusion feature set and build a product lifecycle security data chain. The standardization mapping unit 24 is used to map data of different dimensions in the time series fusion feature set to a unified standardization interval through normalization processing, thereby generating a standardized security dataset.

[0023] In this embodiment, the data cleaning unit 21 uses statistical analysis principles to calculate the distribution characteristics of each data field, identifies data points exceeding the preset distribution range to remove invalid and outlier data from the original safety dataset, and uses filtering algorithms to eliminate high-frequency interference noise in the data; the missing data completion unit 22 uses time series interpolation or a regression model trained on historical data to intelligently estimate and complete missing data in the production process based on the data change trends before and after the missing data and the correlation between normal data under the same working condition; the multi-source data fusion unit 23 adds a timestamp to each set of collected data, aligns and correlates the cleaned and completed equipment sensor data, image recognition results, environmental parameters, and quality inspection data according to the same timestamp, forming a time-series fusion feature set unfolding along the time axis, thereby... A product lifecycle safety data chain is constructed, from raw material input to finished product output. The standardization mapping unit 24 uses the maximum-minimum value normalization algorithm to linearly map the original data of different dimensions in the time-series fusion feature set to a unified standardized interval according to their respective value intervals, eliminating the impact of dimension differences on data analysis and generating a standardized safety dataset. The sequential processing of the above units constitutes a pipeline for the transformation of data from the original state to the standardized state. Data quality and integrity are improved through cleaning and completion operations, spatiotemporal unification of multi-source heterogeneous data is achieved through timestamp alignment, and normalization processing enables data of different physical quantities to be compared and calculated on the same scale. This provides a high-quality, standardized data input foundation for the safety risk assessment module 3, ensuring the accuracy and reliability of subsequent risk identification.

[0024] Specifically, security risk assessment module 3 includes: The dynamic safety threshold database 31 is used to store multi-level safety thresholds that are dynamically adjusted according to different drive shaft models, materials and processing stages. The multi-level safety thresholds include at least a first-level threshold and a second-level threshold. The value of the second-level threshold corresponding to the same monitoring parameter is greater than that of the first-level threshold. The real-time comparison unit 32 is used to compare the current value in the standardized security dataset with the corresponding multi-level security thresholds in real time and calculate the deviation between the current value and each threshold. The multi-dimensional risk assessment unit 33 is used to identify various safety risks throughout the production process based on the dedicated big data model for drive shaft production, from four dimensions: equipment failure risk, process compliance risk, personnel operation risk, and environmental safety risk. The anomaly identification unit 34 is connected to the real-time comparison unit 32 and the multi-dimensional risk assessment unit 33. It is used to generate a primary anomaly status signal when the deviation exceeds the first-level threshold or any dimension risk is identified; and to generate an advanced anomaly status signal when the deviation exceeds the second-level threshold or a high-risk risk is identified.

[0025] In this embodiment, the dynamic safety threshold database 31 stores multi-level safety thresholds pre-calibrated according to different drive shaft models, material grades, and different processing stages such as forging, heat treatment, and machining for each monitoring parameter. The value of the second-level threshold corresponding to the same monitoring parameter is greater than the first-level threshold to reflect the progressive degree of risk. The real-time comparison unit 32 compares the value of the current moment in the standardized safety dataset with the corresponding multi-level safety thresholds in the dynamic safety threshold database 31 for the current drive shaft model, material, and processing stage. The deviation is obtained by calculating the difference or ratio between the current value and each threshold. This deviation quantifies the distance of the current production state from the safety boundary. The multi-dimensional risk assessment unit 33 uses a dedicated big data model trained based on the historical production data of the drive shaft to perform trend analysis on parameters such as vibration, temperature, and load from the perspective of equipment failure risk to identify equipment degradation or abnormal modes, and compares the actual process parameters from the perspective of process compliance risk. The system uses data and standard process specifications to identify deviations or out-of-tolerances. It combines the identification results from the image acquisition unit 12 with the personnel operation risk dimension to determine whether there are any violations. It monitors dust concentration and noise decibels from the environmental safety risk dimension to predict environmental safety hazards. The anomaly identification unit 34 receives the deviation degree output by the real-time comparison unit 32 and the risk identification results of each dimension output by the multi-dimensional risk assessment unit 33. When the deviation degree exceeds the first-level threshold or any dimension risk is identified, a primary anomaly status signal is generated. When the deviation degree exceeds the second-level threshold or is judged as a high-risk risk, a high-level anomaly status signal is generated. Through the dual verification mechanism of dynamic thresholds and dedicated big data models, it not only uses threshold comparison to achieve rapid response to known anomaly boundaries, but also uses multi-dimensional models to identify complex coupled risks and unknown pattern risks. At the same time, the progressive design of the first-level and second-level thresholds makes the anomaly identification have clear severity levels, providing accurate and hierarchical anomaly judgment basis for the graded early warning execution module 4.

[0026] Specifically, the tiered early warning execution module 4 includes: Risk level determination unit 41 is used to determine the current warning level as Level 1, Level 2 or Level 3 based on the type of abnormal state signal received, the duration of the signal and the results of multi-dimensional risk assessment. The multi-terminal hierarchical push unit 42 is used to accurately push warning content and handling guidelines according to the current warning level and job position. The audible and visual alarm unit 43 is used to issue an audible and visual alarm prompt at the corresponding workstation when the current warning level is Level 1. The early warning full data retention unit 44 is used to fully retain the data of the entire early warning process and form a traceable early warning ledger.

[0027] In this embodiment, the risk level determination unit 41 receives the primary or advanced abnormal state signal output by the safety risk assessment module 3, determines the initial risk level based on the signal type, and simultaneously calculates the duration of the abnormal state signal. Combining this with the risk dimension type and confidence level output by the multi-dimensional risk assessment unit 33, it comprehensively determines the current warning level as a Level 1, Level 2, or Level 3 warning according to a preset level mapping rule. Level 1 warning corresponds to the lowest risk level, and Level 3 warning corresponds to the highest risk level. The multi-terminal hierarchical push unit 42, based on the Level 1, Level 2, or Level 3 warning result determined by the risk level determination unit 41, calls a preset job posting strategy table to push the warning content and corresponding handling instructions to the on-site operator terminal, team leader terminal, safety management personnel terminal, and workshop management terminal, respectively, achieving hierarchical and precise push of information at different levels to different job positions. When the current warning level determined by the risk level determination unit 41 is a Level 1 warning, the audible and visual alarm unit 43 activates the alarm by placing an audible and visual alarm at the corresponding workstation. The light alarm emits a specific frequency buzzer and flashing light, distinct from other warning levels, to alert on-site operators to the current abnormal state without requiring immediate shutdown. The full-data retention unit 44 automatically records the time of each warning, warning level, abnormal signal type, signal duration, associated device identifier, multi-dimensional risk assessment results, and push status and time of each terminal, storing this data in chronological order in the warning database. This forms a warning ledger that can be retrieved and traced by time, device, risk type, and other dimensions. By comprehensively considering the abnormal signal type, duration, and multi-dimensional risk assessment, false alarms caused by instantaneous fluctuations are avoided, improving the accuracy of warnings. Simultaneously, multi-terminal hierarchical push ensures that information at different risk levels accurately reaches the corresponding responsible positions. The audible and visual alarm unit's work restrictions during a Level 1 warning provide a gentle reminder of low-risk events, while the full-data retention unit provides complete data evidence for subsequent safety performance evaluations, risk trend analysis, and accountability.

[0028] Specifically, production equipment safety management module 5 includes: The equipment status full-domain monitoring unit 51 is used to monitor and display the equipment operating status and parameter trends in real time; The parameter anomaly locking unit 52 is used to lock equipment access when parameters exceed limits, thereby preventing unauthorized production. The risk-linked start-stop control unit 53 is used to send a speed reduction command or a shutdown command to the programmable logic controller of the corresponding drive shaft production equipment when the current warning level is a level 2 or level 3 warning, and to perform automatic safety start-stop protection for high-risk equipment.

[0029] In this embodiment, the equipment status monitoring unit 51 is connected to the programmable logic controllers and sensor networks of each drive shaft production equipment via a communication interface. It reads the equipment operating parameters and status indicators in real time and displays the real-time operating status, current values ​​of key parameters, and historical trend curves of each equipment in a graphical interface. When the parameter anomaly locking unit 52 detects that any production parameter exceeds the second-level threshold of the multi-level safety threshold corresponding to that parameter, it sends an access lock command to the equipment programmable logic controller to prohibit operators from modifying process parameters or starting the next production cycle through the human-machine interface until authorized personnel unlock the lock, thereby preventing the continued production in violation of regulations under abnormal parameter conditions. The risk linkage start-stop control unit 53 receives the current warning level determined by the graded warning execution module 4. When the current warning level is a second-level warning, it sends a speed reduction command to the programmable logic controller of the corresponding drive shaft production equipment to reduce the equipment operating frequency to a preset safety level. Within the speed range, when the current warning level is Level 3, a shutdown command is sent to cause the equipment to perform an emergency stop or orderly shutdown. At the same time, for scenarios that are judged as high-risk equipment by the multi-dimensional risk assessment unit 33 and whose risk level reaches the preset linkage threshold, automatic safety start-stop protection actions are triggered directly without waiting for manual confirmation. The equipment status full-domain monitoring unit 51 provides a visual and transparent display of the equipment's operating status, making it easy for operators and managers to grasp the equipment's health status in real time. The parameter anomaly locking unit 52 prevents unauthorized continued production behavior after parameters exceed the limits from the permission level, making up for the lack of control based solely on alarm prompts. The risk linkage start-stop control unit 53 realizes the direct association between the warning level and the equipment control command. When a Level 2 warning occurs, speed reduction can avoid production interruption losses caused by immediate shutdown and reduce risks. When a Level 3 warning occurs, shutdown ensures absolute safety under extreme risks, thus forming a complete closed-loop safety control chain on the equipment side from status monitoring, permission blocking to linkage start-stop.

[0030] Specifically, personnel operation safety supervision module 6 includes: Unit 61, used to verify the compliance qualifications of operators for their positions; The process operation behavior recognition unit 62 is used to intelligently identify various violations in process operations; The personal protective equipment (PPE) compliance testing unit 63 is used to test the compliance of workers' wearing of PPE. Unit 64, which is used for evidence storage, source tracing, and linkage with job performance evaluation, is used to trace and verify violations.

[0031] In this embodiment, the qualification verification unit 61 reads the operator's identity information and compares it with a pre-stored job qualification database to verify whether the operator has the required training records, job certificates, and authorization status for the current process. Only after the verification is passed is the operator allowed to start the corresponding production equipment. The process operation behavior recognition unit 62 receives the process operation image sequence collected by the image acquisition unit 12, uses a deep learning-based behavior recognition algorithm to extract the key points of the operator's human skeleton and hand movement trajectory, and compares the identified operation actions with the standard action sequence specified in the standard operating procedure to identify violations such as missing sequence, wrong sequence, and unauthorized cross-regional operation. The safety protection compliance detection unit 63 receives the personnel images collected by the image acquisition unit 12, uses a target detection algorithm to perform feature detection and positioning of safety helmets, protective glasses, protective gloves, work clothes, and safety shoes in the images, and determines whether the operator wears the corresponding safety protection equipment completely as required by the job. When the process operation behavior identification unit 62 or the labor protection wearing compliance detection unit 63 detects a violation, the violation tracing and assessment unit 64 automatically captures the image sequence before and after the violation as evidence, extracts the identity information of the violator, the time of the violation, the type of violation, and the process to which it belongs, writes the above information into the violation database to form a traceable violation record, and automatically deducts the operator's safety assessment score or generates an assessment processing form according to the preset assessment rules. The qualification verification unit 61 eliminates the safety hazards of unqualified personnel operating special equipment or key processes from the source. The process operation behavior identification unit 62 and the labor protection wearing compliance detection unit 63 realize all-weather automated supervision of personnel's work behavior from the two dimensions of operation standardization and protection integrity, respectively, overcoming the intermittent and subjective defects of manual inspection. The violation tracing and assessment unit 64 forms a management closed loop from violation discovery to responsibility implementation through the evidence storage, tracing and assessment linkage mechanism, which enhances the deterrent power and traceability of personnel safety management.

[0032] Specifically, the emergency response dispatch module 7 includes: The intelligent emergency response unit 71 is used to automatically match specific emergency response plans based on risk type and level. The hierarchical personnel dispatching unit 72 is used to allocate personnel and tasks as needed. Risk area isolation and control unit 73 is used to realize intelligent isolation and control of risk areas; The closed-loop feedback unit 74 is used to follow up on the entire handling process, linking the production equipment safety management module 5 and the personnel operation safety supervision module 6 to achieve full closed-loop management of risk warning, handling, rectification and verification.

[0033] In this embodiment, the intelligent retrieval unit 71 receives the warning level output by the hierarchical warning execution module 4 and the risk type identifier output by the safety risk assessment module 3. Based on the combination of risk type and level, it retrieves a matching specific emergency plan from the emergency plan database, extracting the handling process, required resources, and responsible position information from the plan. The hierarchical personnel dispatch unit 72, based on the responsible position information output by the intelligent retrieval unit 71 and the current on-duty status and location information of personnel in each position, allocates handling personnel and corresponding tasks according to preset dispatch priority rules, and pushes the task instructions to the relevant departments. The system includes mobile terminals carried by personnel; when the risk area isolation and control unit 73 receives a level 2 or 3 warning, it sends a locking command to the access control system around the risk area to prevent unauthorized personnel from entering, sends an area isolation prompt command to the audible and visual alarm device in the risk area, and links with the on-site electronic fence or barrier gate system to physically block the access channels to the risk area, thereby achieving intelligent isolation and control of the risk area; the closed-loop feedback unit 74 provides full-process tracking from the intelligent retrieval unit 71 of the classified plan to the hierarchical personnel dispatch unit 72 completing personnel allocation, and the risk area isolation and control unit 73 completing isolation. The entire process, from the departure of personnel to on-site handling, involves feedback on the completion of the handling on mobile terminals, and the verification and confirmation of the rectification results by management personnel. Timestamps and operator information are recorded at each stage. The system also links with the production equipment safety management module 5 to obtain equipment status changes and with the personnel operation safety supervision module 6 to obtain personnel behavior records during the handling process. Once all preset handling steps have been completed and verified, the emergency response event is automatically closed, forming a closed-loop record encompassing early warning triggering, contingency plan retrieval, personnel dispatch, area isolation, on-site handling, and rectification verification. The categorized contingency plan intelligent retrieval unit 71 eliminates... The system eliminates the time delay caused by manual search for contingency plans, ensuring that emergency plans can be quickly and accurately matched to specific risk scenarios. The hierarchical personnel dispatch unit 72 enables the reasonable allocation of emergency human resources as needed. The risk area isolation and control unit 73 actively blocks the risk spread path through physical means such as access control and fences, making up for the shortcomings of passive control relying solely on warning signs. The closed-loop feedback unit 74 ensures that every risk event undergoes a complete handling and verification process through full-process node tracking and multi-module linkage, preventing omissions or half-hearted handling, thereby achieving closed-loop management of the entire life cycle from the occurrence of risk to its complete elimination.

[0034] The working principle and usage process of this invention are as follows: First, after the system is started, the equipment sensing acquisition unit 11, image acquisition unit 12, environmental sensing unit 13 and quality inspection data docking unit 14 in the data acquisition module 1 collect equipment vibration, temperature, load, on-site images, environmental dust and noise and quality inspection data in real time throughout the entire transmission shaft production process, construct the original safety dataset and send it to the big data processing platform 2 and edge computing node 8 at the same time. The big data processing platform 2 sequentially performs cleaning, missing data completion, timestamp alignment, and normalization mapping on the raw data to generate a standardized safety dataset, which is then sent to the safety risk assessment module 3. This module compares the current value with the dynamic safety threshold and calculates the deviation. On the other hand, it uses a dedicated big data model for the drive shaft to identify risks from four dimensions: equipment, process, personnel, and environment, and outputs quantitative assessment results to the graded early warning execution module 4. The graded early warning execution module 4 determines the first, second, or third level of early warning based on the assessment results, executes differentiated push notifications and audible and visual alarms, and retains all data from the entire early warning process. When the production equipment safety management module 5 receives a level 2 or 3 warning, it sends a speed reduction or shutdown command to the equipment PLC and locks the execution authority for parameters exceeding limits; the personnel operation safety supervision module 6 verifies the qualifications of employees, compliance of operations, and status of wearing personal protective equipment in real time through image recognition, and records, traces, and links the assessment of violations; the emergency response and dispatch module 7 is only activated when a level 2 or 3 high-risk warning is issued, automatically matches the emergency plan, assigns personnel and tasks, and links the access control and fence system to achieve risk area isolation and closed-loop management. Edge computing node 8 is always in an independent monitoring state. When it is determined that any collected data exceeds the preset emergency shutdown threshold, it directly issues an emergency shutdown command to the equipment PLC without going through the above central processing process. At the same time, it triggers the hierarchical early warning execution module 4 to report the early warning and upload the analysis results to the big data processing platform 2, forming a millisecond-level emergency safety protection channel.

[0035] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus.

[0036] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A safety management system for automotive driveshaft production based on big data, characterized in that, include: The data acquisition module (1) is used to collect production parameter data, environmental parameter data, operator behavior data and quality inspection data of multiple production equipment in the production process of automobile drive shaft in real time, and to build the original safety dataset. The big data processing platform (2) is connected to the data acquisition module (1) for cleaning, fusion and standardization of the original security dataset to generate a standardized security dataset. The safety risk assessment module (3) is connected to the big data processing platform (2) to store dynamic safety threshold models and identify abnormal states and safety risks in the transmission shaft production process in real time based on standardized safety datasets, and output quantitative assessment results. The graded early warning execution module (4) is connected to the safety risk assessment module (3) and is used to execute graded differentiated early warning push and corresponding safety early warning actions based on the risk level of the quantitative assessment results. The production equipment safety management module (5) is connected to the graded early warning execution module (4) and the safety risk assessment module (3) respectively. It is used to perform automated closed-loop safety management of abnormal risks of transmission shaft production equipment, including sending speed reduction or shutdown commands to the equipment programmable logic controller. The personnel operation safety supervision module (6) is connected to the graded early warning execution module (4) to monitor the on-site personnel's qualifications and operation specifications in real time, and to identify and record violations of operation behavior. The emergency response and dispatch module (7) is connected to the production equipment safety management and control module (5) and the personnel operation safety supervision module (6) respectively. It is used to receive the level 2 and level 3 high-risk warnings, automatically match the emergency plan and complete intelligent dispatch and risk area isolation and control. Edge computing node (8) is set on the side of the transmission shaft production equipment and is connected to the data acquisition module (1) and the big data processing platform (2) respectively. It is used to perform on-site real-time analysis of the collected data. When it is determined that any data exceeds the preset emergency shutdown threshold, it directly sends an emergency shutdown command to the production equipment PLC to complete the shutdown. At the same time, it triggers the hierarchical early warning execution module (4) to report the early warning and uploads the on-site analysis results to the big data processing platform (2).

2. The safety management system for automotive driveshaft production based on big data as described in claim 1, characterized in that, The data acquisition module (1) includes: The equipment sensing and acquisition unit (11) is installed on the spindle and tool post of the transmission shaft processing machine tool and is used to collect spindle vibration spectrum data, tool vibration amplitude data, forging temperature, heat treatment furnace temperature and axial load data in real time. Image acquisition unit (12) includes at least one industrial camera for acquiring data on the operator's safety equipment wearing status, intrusion behavior data in dangerous areas of equipment, and process operation behavior data; The environmental sensing unit (13) includes a dust concentration sensor and a noise sensor, which are used to collect dust concentration data and noise decibel data in the production workshop. The quality inspection data docking unit (14) is used to dock with the quality inspection equipment and synchronously collect the quality inspection and safety data of the drive shaft product.

3. The safety management system for automotive driveshaft production based on big data as described in claim 1, characterized in that, The big data processing platform (2) includes: The data cleaning unit (21) is used to remove invalid data, outliers and noise from the original secure dataset; The missing data completion unit (22) is used to intelligently complete missing data in production conditions; The multi-source data fusion unit (23) is used to associate and align the cleaned and completed multi-source data according to the timestamp to form a time-series fusion feature set and build a product lifecycle security data chain. The standardization mapping unit (24) is used to map data of different dimensions in the time series fusion feature set to a unified standardization interval through normalization processing, thereby generating a standardized security dataset.

4. The safety management system for automotive driveshaft production based on big data as described in claim 1, characterized in that, The security risk assessment module (3) includes: The dynamic safety threshold database (31) is used to store multi-level safety thresholds that are dynamically adjusted according to different drive shaft models, materials and processing stages. The multi-level safety thresholds include at least a first-level threshold and a second-level threshold. The value of the second-level threshold corresponding to the same monitoring parameter is greater than that of the first-level threshold. The real-time comparison unit (32) is used to compare the current value in the standardized security dataset with the corresponding multi-level security thresholds in real time and calculate the deviation between the current value and each threshold. The multi-dimensional risk assessment unit (33) is used to identify various safety risks in the entire production process from four dimensions: equipment failure risk, process compliance risk, personnel operation risk and environmental safety risk, based on the dedicated big data model for drive shaft production. Anomaly identification unit (34) is connected to real-time comparison unit (32) and multi-dimensional risk assessment unit (33) to generate a primary anomaly status signal when the deviation exceeds the first-level threshold or any dimension risk is identified; and to generate an advanced anomaly status signal when the deviation exceeds the second-level threshold or a high-risk risk is identified.

5. The safety management system for automotive driveshaft production based on big data as described in claim 1, characterized in that, The hierarchical early warning execution module (4) includes: The risk level determination unit (41) is used to determine the current warning level as a Level 1 warning, Level 2 warning, or Level 3 warning based on the type of abnormal state signal received, the duration of the signal, and the results of multi-dimensional risk assessment. The multi-terminal hierarchical push unit (42) is used to accurately push warning content and handling guidelines according to the current warning level and job position. The audible and visual alarm unit (43) is used to issue an audible and visual alarm prompt at the corresponding workstation when the current warning level is Level 1. The early warning full data retention unit (44) is used to retain all data of the early warning process and form a traceable early warning ledger.

6. The safety management system for automotive driveshaft production based on big data as described in claim 1, characterized in that, The production equipment safety management module (5) includes: The equipment status full-domain monitoring unit (51) is used to monitor and display the equipment operating status and parameter trends in real time; The parameter abnormality locking unit (52) is used to lock the equipment permissions when the parameters exceed the limit, so as to prevent illegal production; The risk-linked start-stop control unit (53) is used to send a speed reduction command or a shutdown command to the programmable logic controller of the corresponding transmission shaft production equipment when the current warning level is a level 2 or level 3 warning, and to perform automatic safety start-stop protection for high-risk equipment.

7. The safety management system for automotive driveshaft production based on big data as described in claim 1, characterized in that, The personnel operation safety supervision module (6) includes: The qualification verification unit (61) is used to verify the compliance qualifications of the operators. The process operation behavior recognition unit (62) is used to intelligently identify various violations in process operations; The personal protective equipment (PPE) wearing compliance testing unit (63) is used to test the compliance of workers' wearing of PPE; The violation tracing assessment unit (64) is used for evidence storage, tracing of violations, and linkage with job performance assessment.

8. The safety management system for automotive driveshaft production based on big data as described in claim 1, characterized in that, The emergency response dispatch module (7) includes: The intelligent retrieval unit for classified contingency plans (71) is used to automatically match specific emergency plans according to risk type and level; A tiered personnel dispatching unit (72) is used to allocate personnel and tasks as needed. Risk area isolation and control unit (73) is used to realize intelligent isolation and control of risk areas; The closed-loop feedback unit (74) is used to follow up on the entire handling process, link the production equipment safety management module (5) and the personnel operation safety supervision module (6) to realize the full closed-loop management of risk warning, handling, rectification and verification.

Images