Safe and efficient control method and system for squeeze riveter based on sensor

Abstract

The invention discloses a safe and efficient control method and system for a squeeze riveter based on a sensor, and relates to the technical field of safety control. A laser displacement sensor, a distributed pressure sensor array, an industrial camera and a vibration accelerometer are arranged at key positions of a workbench of a squeeze riveter to obtain real-time squeeze riveting data; the method comprises the following steps: initializing a control system comprising a digital twin model and a final neural network quality prediction model, processing real-time pressure riveting data, adjusting a servo motor by using a fuzzy PID control algorithm, and making a compensation decision according to a quality prediction result; meanwhile, a safety threshold matrix is set, abnormal conditions are detected in real time, and corresponding safety measures are automatically taken; according to the automatic riveting machine, the stability and the consistency of the riveting quality are effectively improved, the fluctuation of the product quality is reduced, the automation level and the production efficiency of the riveting machine are improved, meanwhile, the safety performance is enhanced, and the riveting process is more accurate and efficient.

Description

Technical Field

[0001] The present invention belongs to the technical field of safety control. Specifically, it particularly relates to a safe and efficient control method and system for a riveting machine based on sensors. Background Art

[0002] With the rapid development of modern industry, as an important device in metal processing, the requirements for the automation and intelligence level of riveting machines are increasing day by day; the traditional control system of riveting machines mainly relies on manual experience and preset parameters, lacking the ability of real-time monitoring and intelligent adjustment of the riveting process, resulting in unstable riveting quality, low production efficiency and potential safety hazards. Summary of the Invention

[0003] (1) Technical Problems to be Solved

[0004] Aiming at the problems in the related technologies, the present invention provides a safe and efficient control method for a riveting machine based on sensors. Through the combination of sensor technology and deep learning technology, the present invention realizes the real-time monitoring and intelligent adjustment ability of the riveting process, makes the riveting quality more stable, improves the production efficiency and reduces the potential safety hazards.

[0005] (2) Technical Solutions

[0006] To solve the above technical problems, the present invention is realized through the following technical solutions:

[0007] S1. Deploy sensors and initialize the control system;

[0008] S2. Collect real-time riveting data through sensors, and perform Kalman filtering, data spatio-temporal alignment and feature extraction processing on the real-time riveting data to obtain real-time riveting feature data;

[0009] S3. Combine the real-time riveting feature data, call the fuzzy PID control algorithm to adjust the servo motor; through the final neural network quality prediction model in the control system, obtain the predicted quality, judge whether the predicted quality meets the standard and make a compensation decision;

[0010] S4. Set a safety threshold matrix, extract real-time safety feature data from the real-time riveting feature data, compare the real-time safety feature data with the safety threshold matrix to obtain a comparison result, and make a safety decision according to the comparison result.

[0011] Preferably, the S1 includes the following steps:

[0012] S11. Deploy a laser displacement sensor, a distributed pressure sensor array, an industrial camera and a vibration accelerometer at key position points on the working table of the riveting machine;

[0013] S12. Start the control system, load the preset process parameter library, which includes the material-pressure-displacement matching matrix; establish the digital twin model interface for data intercommunication with the CAD / CAM system; initialize the final quality prediction model;

[0014] The above steps deploy a variety of sensors and industrial cameras at key positions on the riveting machine workbench, and start the control system to load the preset process parameter library, establish the digital twin model interface, and initialize the final quality prediction model, realizing comprehensive monitoring, data intercommunication, and intelligent prediction of the riveting process; effectively improving the accuracy, stability, and automation level of the riveting process.

[0015] Preferably, the final quality prediction model in S12 includes the following steps:

[0016] S121. Construct a neural network quality prediction model and set parameters;

[0017] S122. Collect historical riveting parameter data, and obtain historical product quality data by collecting product quality data from the historical riveting parameter data; divide the historical riveting parameter data into historical riveting parameter training data and historical riveting parameter test data; divide the historical product quality data into historical product quality training data and historical product quality test data;

[0018] S123. Set the training error threshold and the maximum number of training iterations, and use the historical riveting parameter training data and the historical product quality training data to repeatedly train the neural network quality prediction model and adjust the parameters. When the training error ≤ the training error threshold or reaches the maximum number of training iterations, obtain the trained neural network quality prediction model;

[0019] S124. Use the historical riveting parameter test data and the historical product quality test data to test the trained neural network quality prediction model. After the test is completed, obtain the final neural network quality prediction model;

[0020] The above steps construct a neural network quality prediction model, collect and divide historical riveting parameters and product quality data, set training conditions, and conduct model training and testing, and finally obtain an accurate neural network quality prediction model; realizing the intelligent prediction of the quality of riveted products, effectively improving the accuracy and efficiency of product quality control.

[0021] Preferably, S2 includes the following steps:

[0022] S21. Combine the timestamp synchronization method to collect real-time riveting data through sensors;

[0023] S22. Perform Kalman filtering, data spatio-temporal alignment, and feature extraction processing on the real-time riveting data to obtain real-time riveting feature data;

[0024] The above steps collect real-time riveting data through a time-stamp synchronization method, and perform Kalman filtering, spatio-temporal alignment of data, and feature extraction on it, effectively improving the quality and usability of the data.

[0025] Preferably, the S21 includes the following steps:

[0026] S211. Combine the time-stamp synchronization method, collect the displacement trajectory data of the punch through a laser displacement sensor, collect the dynamic pressure waveform data through a distributed pressure sensor, compare the deformation data of the reference image workpiece in real time through a vision system, and collect the vibration spectrum data through a vibration accelerometer to obtain real-time riveting data;

[0027] Through the above steps, multi-dimensional and high-precision monitoring of the riveting process is achieved, ensuring the comprehensiveness and real-time nature of the data.

[0028] Preferably, the S22 includes the following steps:

[0029] S221. Perform Kalman filtering calculation on the real-time riveting data to obtain real-time filtered riveting data;

[0030] S222. Perform spatio-temporal alignment on the real-time filtered riveting data to obtain real-time aligned riveting data; the spatio-temporal alignment compensation model is as follows,

[0031]

[0032] where x(t) represents the sensor data after spatio-temporal synchronization compensation, Δt represents the time delay between different sensors, C represents the calibration coefficient matrix of the sensor, x i (t + Δt) represents the data of sensor i at time t plus the time delay of Δt, y j (t) represents the data of another sensor j at time t, ||y j (t) - x i (t + α)|| 2 represents the Euclidean distance between the data of sensor j and sensor i at times t and t + α;

[0033] S223. Extract the features of the real-time aligned riveting data to obtain real-time riveting feature data;

[0034] The above steps perform fine processing on the real-time riveting data through steps such as Kalman filtering calculation, spatio-temporal alignment of data, and feature extraction, and finally obtain accurate and consistent real-time riveting feature data. This process effectively removes data noise, compensates for the time delay and calibration differences between sensors, and ensures the consistency and comparability of multi-source data.

[0035] Preferably, S3 includes the following steps:

[0036] S31. Invoke the fuzzy PID control algorithm to adjust the feed speed of the servo motor, the pressure gradient of the hydraulic system, and the compensation amount of the indenter movement trajectory according to the real-time riveting feature data;

[0037] S32. Set a preset quality and a quality error threshold, input the real-time riveting feature data into the final neural network quality prediction model to obtain the predicted quality;

[0038] Calculate the error between the predicted quality and the preset quality to obtain the quality error. If the quality error ≥ the quality error threshold, trigger the compensation mechanism to obtain a compensation plan, and adjust the pressure-displacement curve of the riveting point according to the compensation plan; otherwise, maintain the current state;

[0039] The above steps adjust the riveting parameters in real time by invoking the fuzzy PID control algorithm, monitor the predicted quality using the final neural network quality prediction model, and dynamically trigger the compensation mechanism according to the quality error, realizing the intelligent closed-loop control of the riveting process; ensuring the stability of the riveting process and the precise control of product quality, effectively reducing quality fluctuations, and improving the product qualification rate.

[0040] Preferably, triggering the compensation mechanism in S32 to obtain a compensation plan includes the following steps:

[0041] S321. Synchronize the real-time riveting feature data and the quality error to the digital twin model;

[0042] S322. Simulate the current riveting process in the digital twin environment and analyze the cause of the prediction deviation; generate a compensation plan based on the simulation analysis results;

[0043] The above steps synchronize the real-time riveting feature data and the quality error to the digital twin model, simulate and analyze the current riveting process in the digital twin environment, and generate a targeted compensation plan, realizing the precise optimization of the riveting process.

[0044] Preferably, S4 includes the following steps:

[0045] S41. Preset the safety threshold matrix of pressure-displacement correlation, response time, and energy consumption according to the process parameters and historical data;

[0046] S42. Extract the feature data of pressure-displacement correlation, response time, and energy consumption from the real-time riveting feature data; obtain the real-time safety data of the real-time P-S curve, real-time single-cycle time, and energy consumption per unit stroke;

[0047] Compare the real-time safety feature data with the safety threshold matrix to detect whether the curve shape, slope, and peak are abnormal, determine whether there is a timeout, and whether there is a sudden change in energy consumption; obtain the safety comparison data;

[0048] S43. Set the minor abnormality threshold, serious abnormality threshold, and extreme abnormality threshold respectively; if there is minor abnormal data in the safety comparison data, such as the P-S curve deviating slightly and the response time being slightly longer, activate the audible and visual alarm device to remind the operator, and the control system sends an instruction to reduce the speed of the servo motor to 50%, and continue to monitor the abnormal situation;

[0049] If there is serious abnormal data in the safety comparison data, such as the P-S curve deviating seriously and the response time being significantly overtime, immediately stop the movement of the punch head, activate the self-locking device, fix the position of the punch head, strengthen the audible and visual alarm, and display the abnormal information on the operation interface;

[0050] If there is extreme abnormal data in the safety comparison data, immediately cut off all power sources, including electricity and hydraulic pressure, and activate the emergency braking system; form and upload the fault code to the MES system, and notify the maintenance personnel to arrive at the scene for handling through the MES system;

[0051] The above steps can detect key indicators such as the pressure-displacement correlation, response time, and energy consumption in the riveting process in real time by presetting the safety threshold matrix, extracting the riveting feature data in real time and performing safety comparison, and classifying and responding according to the degree of abnormality, timely discovering and distinguishing minor, serious, and extreme abnormal situations, so as to take corresponding alarm, speed reduction, shutdown, or emergency braking measures, significantly improving the safety and stability of the riveting process, effectively preventing potential safety accidents; realizing real-time monitoring and intelligent early warning of the production process, and improving production efficiency and quality.

[0052] The construction hoist ride safety supervision system based on big data is used to implement the above-mentioned safety and efficient control method of a riveting machine based on sensors, including a sensor deployment and control system initialization module, a real-time data collection and processing module, a quality control and compensation decision module, and a safety monitoring and decision module;

[0053] The sensor deployment and control system initialization module is used to deploy laser displacement sensors, distributed pressure sensor arrays, industrial cameras, and vibration accelerometers; start the control system, load the preset process parameter library, establish a digital twin model interface, and initialize the final quality prediction model;

[0054] The real-time data collection and processing module is used to collect real-time riveting data through sensors; perform Kalman filtering, data spatio-temporal alignment, and feature extraction processing on the real-time riveting data to obtain real-time riveting feature data;

[0055] The quality control and compensation decision-making module is used to combine real-time riveting feature data, call the fuzzy PID control algorithm to adjust the servo motor, obtain the predicted quality through the final neural network quality prediction model, judge whether the predicted quality meets the standard, and make a compensation decision.

[0056] The safety monitoring and decision-making module is used to set a safety threshold matrix, extract real-time safety feature data from the real-time riveting feature data, compare the real-time safety feature data with the safety threshold matrix, and make a safety decision according to the comparison result.

[0057] (III) Beneficial effects

[0058] The present invention has the following beneficial effects:

[0059] By deploying a laser displacement sensor, a distributed pressure sensor array, an industrial camera, and a vibration accelerometer, and combining deep learning technology, the present invention realizes real-time monitoring and intelligent adjustment of the riveting process, effectively overcomes the limitations of traditional control systems that rely on manual experience and preset parameters, significantly improves the stability and consistency of riveting quality, and reduces product quality fluctuations.

[0060] Through the control system and the digital twin model interface, the present invention realizes data intercommunication with the CAD / CAM system, as well as the initialization and optimization of the final quality prediction model. This integrated control method greatly improves the automation level of the riveting machine, reduces manual intervention, thus significantly improving production efficiency and shortening the production cycle.

[0061] By setting a safety threshold matrix and extracting real-time riveting feature data, the present invention can quickly detect and judge abnormal situations during the riveting process. According to the severity of the abnormality, the system will automatically take corresponding safety measures, effectively avoiding equipment damage and personal injury, and greatly improving the safety performance of the riveting machine.

[0062] The present invention makes real-time adjustments to the servo motor through the fuzzy PID control algorithm, predicts and compensates the riveting quality using the neural network quality prediction model. When there is a large error between the predicted quality and the preset quality, the system will trigger a compensation mechanism, generate an optimized compensation plan through simulation and analysis in the digital twin environment, and further adjust the riveting parameters to ensure that the product quality always meets the requirements, making the riveting process more accurate and efficient.

[0063] Of course, it is not necessary for any product implementing the present invention to simultaneously achieve all the above-mentioned advantages. Description of the drawings

[0064] To more clearly illustrate the technical solutions of the embodiments of the invention, the following will briefly introduce the accompanying drawings required for the description of the embodiments. Obviously, the accompanying drawings in the following description are only some embodiments of the invention. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can also be obtained based on these drawings.

[0065] Figure 1 It is a schematic flow chart of a safety and efficient control method for a riveting machine based on sensors according to the present invention. Detailed implementation manners

[0066] The following will clearly and completely describe the technical solutions in the embodiments of the invention with reference to the accompanying drawings in the embodiments of the invention. Obviously, the described embodiments are only some embodiments of the invention, rather than all embodiments. Based on the embodiments of the invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the invention.

[0067] In the description of the present invention, it should be understood that the terms "openings", "upper", "lower", "top", "middle", "inner", etc. indicating orientation or positional relationships are only for the convenience of describing the invention and simplifying the description, rather than indicating or implying that the components or elements referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus should not be construed as a limitation of the invention.

[0068] Embodiment 1:

[0069] Please refer to Figure 1 , the present invention discloses a safety and efficient control method for a riveting machine based on sensors, including the following steps:

[0070] S1. Deploy sensors and initialize the control system;

[0071] The S1 includes the following steps:

[0072] S11. Deploy laser displacement sensors, distributed pressure sensor arrays, industrial cameras, and vibration accelerometers at key position points on the working table of the riveting machine;

[0073] S12. Start the control system, load the preset process parameter library, including the material-pressure-displacement matching matrix; establish a digital twin model interface for data intercommunication with the CAD / CAM system; initialize the final quality prediction model;

[0074] The final quality prediction model in the S12 includes the following steps:

[0075] S121. Construct a neural network quality prediction model and set parameters;

[0076] S122. Collect historical riveting parameter data, and obtain historical product quality data by collecting product quality data from the historical riveting parameter data; divide the historical riveting parameter data into historical riveting parameter training data and historical riveting parameter test data; divide the historical product quality data into historical product quality training data and historical product quality test data;

[0077] S123. Set the training error threshold and the maximum number of training iterations, repeatedly train the neural network quality prediction model using the historical riveting parameter training data and the historical product quality training data, and adjust the parameters. When the training error ≤ the training error threshold or the maximum number of training iterations is reached, obtain the trained neural network quality prediction model;

[0078] S124. Use the historical riveting parameter test data and the historical product quality test data to test the trained neural network quality prediction model. After the test is completed, obtain the final neural network quality prediction model;

[0079] S2. Collect real-time riveting data through sensors, and perform Kalman filtering, data spatio-temporal alignment, and feature extraction on the real-time riveting data to obtain real-time riveting feature data;

[0080] S2 includes the following steps:

[0081] S21. Combine the timestamp synchronization method to collect real-time riveting data through sensors;

[0082] S21 includes the following steps:

[0083] S211. Combine the timestamp synchronization method to collect the indenter displacement trajectory data through a laser displacement sensor, collect the dynamic pressure waveform data through a distributed pressure sensor, compare the workpiece deformation data of the reference image in real time through a vision system, and collect the vibration spectrum data through a vibration accelerometer to obtain real-time riveting data;

[0084] S22. Perform Kalman filtering, data spatio-temporal alignment, and feature extraction on the real-time riveting data to obtain real-time riveting feature data;

[0085] S22 includes the following steps:

[0086] S221. Perform Kalman filtering calculation on the real-time riveting data to obtain real-time filtered riveting data;

[0087] S222. Perform data spatio-temporal alignment on the real-time filtered riveting data to obtain real-time aligned riveting data; The spatio-temporal alignment compensation model is as follows,

[0088]

[0089] Among them, x(t) represents the sensor data after spatio-temporal synchronization compensation, Δt represents the time delay between different sensors, C represents the calibration coefficient matrix of the sensor, and x i (t + Δt) represents the data of sensor i at time t plus the time delay of Δt, and y j (t) represents the data of another sensor j at time t, and ||y j (t) - x i (t + α)|| 2 represents the Euclidean distance between the data of sensor j and sensor i at times t and t + α;

[0090] S223. Extract the real-time alignment riveting data features to obtain the real-time riveting feature data;

[0091] S3. Combine the real-time riveting feature data, call the fuzzy PID control algorithm to adjust the servo motor; through the final neural network quality prediction model in the control system, obtain the predicted quality, judge whether the predicted quality meets the standard and make a compensation decision;

[0092] The said S3 includes the following steps:

[0093] S31. Call the fuzzy PID control algorithm to adjust the feed speed of the servo motor, the pressure gradient of the hydraulic system, and the compensation amount of the indenter movement trajectory according to the real-time riveting feature data;

[0094] S32. Set a preset quality and a quality error threshold, input the real-time riveting feature data into the final neural network quality prediction model to obtain the predicted quality;

[0095] Calculate the error between the predicted quality and the preset quality to obtain the quality error. If the quality error ≥ the quality error threshold, trigger the compensation mechanism to obtain a compensation plan, and adjust the pressure-displacement curve of the riveting point according to the compensation plan; otherwise, maintain the current state;

[0096] In the said S32, triggering the compensation mechanism to obtain a compensation plan includes the following steps:

[0097] S321. Synchronize the real-time riveting feature data and the quality error to the digital twin model;

[0098] S322. Simulate the current riveting process in the digital twin environment to analyze the cause of the prediction deviation; based on the simulation analysis results, generate a compensation plan;

[0099] S4. Set a safety threshold matrix, extract the real-time safety feature data from the real-time riveting feature data, compare the real-time safety feature data with the safety threshold matrix to obtain a comparison result, and make a safety decision according to the comparison result;

[0100] S4 includes the following steps:

[0101] S41. Preset a safety threshold matrix for pressure-displacement correlation, response time, and energy consumption according to process parameters and historical data;

[0102] S42. Extract feature data of pressure-displacement correlation, response time, and energy consumption from real-time riveting feature data; obtain real-time safety data of real-time P-S curve, real-time single-cycle time, and energy consumption per unit stroke;

[0103] Compare the real-time safety feature data with the safety threshold matrix, detect whether the curve shape, slope, and peak are abnormal, judge whether it times out, and whether the energy consumption mutates; obtain safety comparison data;

[0104] S43. Set a slight abnormality threshold, a serious abnormality threshold, and an extreme abnormality threshold respectively; if there is slight abnormal data in the safety comparison data, such as the P-S curve deviating slightly and the response time being slightly longer, activate the audible and visual alarm device to remind the operator, and the control system sends an instruction to reduce the speed of the servo motor to 50%, and continue to monitor the abnormal situation;

[0105] If there is serious abnormal data in the safety comparison data, such as the P-S curve deviating seriously and the response time significantly timing out, immediately stop the movement of the punch head, activate the self-locking device, fix the position of the punch head, strengthen the audible and visual alarm, and display the abnormal information on the operation interface;

[0106] If there is extreme abnormal data in the safety comparison data, immediately cut off all power sources, including electricity and hydraulics, and activate the emergency braking system; form and upload the fault code to the MES system, and notify the maintenance personnel to arrive at the scene for handling through the MES system.

[0107] Embodiment 2:

[0108] A construction hoist ride safety supervision system based on big data, used to implement the above-mentioned safety and efficient control method of a riveting machine based on sensors, includes a sensor deployment and control system initialization module, a real-time data collection and processing module, a quality control and compensation decision module, and a safety monitoring and decision module;

[0109] The sensor deployment and control system initialization module is used to deploy a laser displacement sensor, a distributed pressure sensor array, an industrial camera, and a vibration accelerometer; start the control system, load the preset process parameter library, establish a digital twin model interface, and initialize the final quality prediction model;

[0110] The real-time data collection and processing module is used to collect real-time riveting data through sensors; perform Kalman filtering, data spatio-temporal alignment, and feature extraction processing on the real-time riveting data to obtain real-time riveting feature data;

[0111] The quality control and compensation decision-making module is used to combine real-time riveting feature data, call the fuzzy PID control algorithm to adjust the servo motor; obtain the predicted quality through the final neural network quality prediction model, judge whether the predicted quality meets the standard and make a compensation decision;

[0112] The safety monitoring and decision-making module is used to set a safety threshold matrix, extract real-time safety feature data from the real-time riveting feature data; compare the real-time safety feature data with the safety threshold matrix, and make a safety decision according to the comparison result.

[0113] In the description of this specification, the descriptions referring to terms such as "one embodiment", "example", "specific example", etc. mean that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in a suitable manner in any one or more embodiments or examples.

[0114] The preferred embodiments of the invention disclosed above are only used to help illustrate the invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to the specific embodiments described. Obviously, many modifications and variations can be made according to the content of this specification. These embodiments are selected and specifically described in this specification to better explain the principle and practical application of the invention, so that those skilled in the art can understand and utilize the invention well.

Claims

1. A safe and efficient control method for a riveting machine based on a sensor, characterized in that: The following steps are involved: S1, deploy sensors and initialize control systems; S2, collecting real-time riveting data through sensors, performing Kalman filtering, data time-space alignment and feature extraction on the real-time riveting data to obtain real-time riveting feature data; S3. Combined with the real-time riveting feature data, the fuzzy PID control algorithm is called to adjust the servo motor; the predicted quality is obtained through the final neural network quality prediction model in the control system, and whether the predicted quality meets the standard is judged and compensation decisions are made; S4. Set a safety threshold matrix, extract real-time safety feature data from the real-time riveting feature data, compare the real-time safety feature data with the safety threshold matrix, obtain a comparison result, and make a safety decision based on the comparison result.

2. A sensor-based safe and efficient control method for a riveting machine according to claim 1, characterized in that: The S1 comprises the following steps: S11. Deploy laser displacement sensors, distributed pressure sensor arrays, industrial cameras and vibration accelerometers at key locations on the riveting machine workbench; S12. Start the control system and load the preset process parameter library; establish a digital twin model interface to exchange data with the CAD / CAM system; and initialize the final quality prediction model.

3. A sensor-based safe and efficient control method for a riveting machine according to claim 2, characterized in that: The final quality prediction model in S12 includes the following steps: S121, constructing a neural network quality prediction model and setting parameters; S122, collecting historical press riveting parameter data, collecting product quality data obtained through the historical press riveting parameter data to obtain historical product quality data; dividing the historical press riveting parameter data into historical press riveting parameter training data and historical press riveting parameter test data; dividing the historical product quality data into historical product quality training data and historical product quality test data; S123, setting a training error threshold and a maximum number of training iterations, using historical riveting parameter training data and historical product quality training data to repeatedly train the neural network quality prediction model, and adjusting the parameters, when the training error is ≤ the training error threshold or reaches the maximum number of training iterations, the trained neural network quality prediction model is obtained; S124. Use historical riveting parameter test data and historical product quality test data to test the trained neural network quality prediction model. After the test is completed, the final neural network quality prediction model is obtained.

4. A sensor-based safe and efficient control method for a riveting machine according to claim 1, characterized in that: The S2 comprises the following steps: S21, combining the time stamp synchronization method, collecting real-time riveting data through sensors; S22, performing Kalman filtering, data time-space alignment and feature extraction processing on the real-time riveting data to obtain real-time riveting feature data.

5. A sensor-based safe and efficient control method for a riveting machine according to claim 4, characterized in that: The S21 comprises the following steps: S211. Combined with the timestamp synchronization method, the laser displacement sensor is used to collect the pressure head displacement trajectory data, the distributed pressure sensor is used to collect the dynamic pressure waveform data, the visual system is used to compare the workpiece deformation data with the reference image in real time, and the vibration accelerometer is used to collect the vibration spectrum data to obtain the real-time riveting data.

6. A sensor-based safe and efficient control method for a riveting machine according to claim 4, characterized in that: The S22 comprises the following steps: S221, performing Kalman filter calculation on the real-time riveting data to obtain real-time filtered riveting data; S222, perform spatial and temporal alignment on the real-time filtered riveting data to obtain real-time aligned riveting data; the spatial and temporal alignment compensation model is as follows: Where x(t) represents the sensor data after time-space synchronization compensation, Δt represents the time delay between different sensors, C represents the calibration coefficient matrix of the sensor, and x i (t+Δt) represents the data of sensor i at time t plus Δt time delay, y j (t) represents the data of another sensor j at time t, ||y j (t)-x i (t+α)||2 represents the Euclidean distance between the data of sensor j and sensor i at time t and t+α; S223, extracting real-time alignment riveting data features to obtain real-time riveting feature data.

7. A sensor-based safe and efficient control method for a riveting machine according to claim 1, characterized in that: The S3 comprises the following steps: S31, calling the fuzzy PID control algorithm to adjust the servo motor feed speed, hydraulic system pressure gradient and pressure head motion trajectory compensation according to the real-time riveting feature data; S32, setting a preset quality, setting a quality error threshold, inputting the real-time riveting feature data into the final neural network quality prediction model, and obtaining the predicted quality; The error between the predicted quality and the preset quality is calculated to obtain the quality error. If the quality error is greater than or equal to the quality error threshold, the compensation mechanism is triggered to obtain a compensation plan. The pressure-displacement curve of the riveting point is adjusted according to the compensation plan; otherwise, the current state is maintained.

8. A sensor-based safe and efficient control method for a riveting machine according to claim 7, characterized in that: Triggering the compensation mechanism in S32 to obtain a compensation solution includes the following steps: S321, synchronizing the real-time riveting feature data and quality error into the digital twin model; S322. Simulate the current riveting process in the digital twin environment and analyze the causes of prediction deviations; generate compensation solutions based on the simulation analysis results.

9. A sensor-based safe and efficient control method for a riveting machine according to claim 1, characterized in that: The S4 comprises the following steps: S41. Preset a safety threshold matrix of pressure-displacement correlation, response time and energy consumption according to process parameters and historical data; S42, extracting characteristic data of pressure-displacement correlation, response time and energy consumption from the real-time pressure riveting characteristic data; obtaining real-time safety data including real-time PS curve, real-time single cycle time and energy consumption per unit stroke; Compare the real-time safety feature data with the safety threshold matrix to detect whether the curve shape, slope, and peak value are abnormal, determine whether there is a timeout and whether the energy consumption has a sudden change, and obtain safety comparison data; S43, respectively set a slight abnormality threshold, a severe abnormality threshold, and an extreme abnormality threshold; if there is slight abnormality data in the safety comparison data, start the sound and light alarm device to remind the operator, and the control system sends a command to reduce the speed of the servo motor to 50%; If there are serious abnormal data in the safety comparison data, the pressure head movement will be stopped immediately, the self-locking device will be activated, the pressure head position will be fixed, the sound and light alarm will be strengthened, and the abnormal information will be displayed on the operation interface; If there are extremely abnormal data in the safety comparison data, all power sources will be cut off immediately and the emergency braking system will be activated; the fault code will be uploaded to the MES system, and the maintenance personnel will be notified through the MES system to come to the site for processing.

10. A construction elevator riding safety supervision system based on big data, used to implement a sensor-based safe and efficient control method for a riveting machine as described in any one of claims 1 to 9, characterized in that: The system includes a sensor deployment and control system initialization module, a real-time data collection and processing module, a quality control and compensation decision module, and a safety monitoring and decision module.

Images