An intelligent control system for shot blasting machines based on multi-source data acquisition

Abstract

The present invention discloses an intelligent control system for a shot blasting machine based on multi-source data acquisition, comprising: a data acquisition module group for multi-dimensionally acquiring data related to the operation of the shot blasting machine, including: a shot blasting machine body information acquisition module for acquiring equipment model and component identification; an environmental information acquisition module for real-time acquisition of ambient temperature, humidity, and dust concentration; an image information acquisition module for synchronously acquiring images of the workpiece surface and operator behavior; an operating status acquisition module for acquiring equipment three-dimensional axial vibration waveforms, operating sound patterns, and key mechanism action images; a shot blasting quality acquisition module for real-time detection of workpiece surface cleanliness and roughness; and a data processing module for analyzing and optimizing the acquired data. The present invention achieves precise control and dynamic optimization of the shot blasting process through multi-dimensional data fusion, intelligent algorithm modeling, and human-machine collaborative mechanisms, thereby improving production efficiency and process stability.

Description

Technical Field

[0001] The present invention relates to the field of management and control systems, and in particular to an intelligent control system for a shot blasting machine based on multi-source data acquisition. Background Art

[0002] Shot blasting machines are industrial equipment that blasts shot (such as steel and cast iron shots) at high speed onto workpiece surfaces to clean, remove rust, strengthen, or roughen surfaces. They are widely used in machinery manufacturing, shipbuilding, automotive, and other fields. Their core principle is to utilize the high-speed rotation of the shot blaster's impeller to project shot onto the workpiece surface. The impact force removes impurities or alters the surface topography to meet subsequent processing or usage requirements.

[0003] The shot blasting machine control system is the core module that ensures the precise execution of the shot blasting process. Its main functions include real-time monitoring of equipment operating status (such as vibration, sound patterns, and component movement), adjusting process parameters (such as shot blasting pressure, shot flow rate, and shot time), ensuring workpiece processing quality (such as cleanliness and roughness), and managing equipment maintenance. With the development of intelligent industry, traditional control systems are increasingly unable to meet the high-precision and high-efficiency production needs due to their limited data collection dimensions and insufficient algorithm optimization capabilities.

[0004] The existing shot blasting machine control system has the following main technical problems: single data acquisition dimension, insufficient intelligence, low fault diagnosis efficiency, poor environmental adaptability and weak human-machine collaboration ability. Therefore, an intelligent control system for shot blasting machine based on multi-source data acquisition is proposed. Summary of the Invention

[0005] The technical problem to be solved by the present invention is: how to solve the problems in the existing technology such as single data acquisition dimension, insufficient intelligence, low fault diagnosis efficiency, poor environmental adaptability and weak human-computer collaboration ability, and provide an intelligent control system for shot blasting machines based on multi-source data acquisition.

[0006] The present invention solves the above-mentioned technical problems through the following technical solutions, which include:

[0007] The data acquisition module group is used to collect shot blasting machine operation-related data in multiple dimensions, including:

[0008] Shot blasting machine body information collection module, used to collect equipment model and component identification;

[0009] Environmental information collection module, used to collect environmental temperature, humidity and dust concentration in real time;

[0010] Image information acquisition module, used to synchronously capture images of the workpiece surface and operator behavior;

[0011] Operation status acquisition module, used to collect equipment's three-dimensional axial vibration waveform, operation soundprint, and key mechanism motion images;

[0012] Shot blasting quality collection module, used to detect the cleanliness and roughness of the workpiece surface in real time;

[0013] The data processing module is used to analyze and optimize the collected data, including:

[0014] Time synchronization controller, based on The protocol aligns the data timestamps of each module and outputs synchronized data packets with a time deviation of less than 1ms;

[0015] Process optimization engine generates process parameter optimization instructions based on shot blasting quality detection values ​​and synchronization data packets;

[0016] Collaborative fault diagnosis unit locates abnormal equipment components based on vibration, voiceprint, and image features;

[0017] Human-machine collaborative optimization module integrates human intervention parameters to generate final optimization instructions;

[0018] Information sending module, used to write the final optimization instructions into controller and push fault location information and operation suggestions.

[0019] Furthermore, the operating status acquisition module includes:

[0020] A vibration sensor group arranged on the bearing seat is used to output a three-dimensional axial vibration spectrum;

[0021] A directional microphone array installed on the side wall of the shot blasting room is used to collect 200Hz-10kHz operating sound patterns;

[0022] High-speed camera unit is used to capture dynamic images of the shot blasting machine impeller and shot dividing wheel.

[0023] Furthermore, the data processing module further includes a time synchronization controller, which is used to: receive the original data stream of each acquisition module;

[0024] based on The protocol aligns all data timestamps;

[0025] Output synchronization data packets with time deviation less than 1ms to the process optimization engine and fault collaborative diagnosis unit.

[0026] Furthermore, the process optimization engine performs:

[0027] Receive the actual cleanliness and roughness measurement values ​​of the workpiece surface, that is, the shot blasting quality test value and synchronization data packets;

[0028] Calculate quality prediction value The specific process is:

[0029] ;

[0030] Where, The workpiece surface texture feature matrix is ​​collected by the image information acquisition module. Feature extraction is obtained;

[0031] It is the three-dimensional axial vibration spectrum feature vector of the equipment, which is collected by the vibration sensor group of the operation status acquisition module. Temporal coding acquisition;

[0032] Mel cepstral coefficients for running voiceprints , through the microphone array of the running status acquisition module Feature extraction is obtained;

[0033] is the model weight, which is called from the process parameter library based on the equipment model;

[0034] Solve Minimal parameter set .

[0035] Furthermore, the process optimization engine includes an adaptive feature extraction unit, which reconstructs the voiceprint features and replaces the model input when the ambient dust concentration exceeds a threshold or the temperature and humidity deviate from the standard range, wherein:

[0036] The specific process of reconstructing voiceprint features is as follows:

[0037] ;

[0038] Where, is the compensation function of environmental parameters (temperature, humidity) on voiceprint;

[0039] The specific process of updating the model input is:

[0040] .

[0041] Furthermore, the fault collaborative diagnosis unit performs the following process:

[0042] Extract vibration spectrum features from synchronization data packets Voiceprint features ;

[0043] Build a spatiotemporal alignment model:

[0044] ;

[0045] Output Fault Part Number:

[0046] ;

[0047] In the formula is the Pearson correlation coefficient (range [-1,1]), which is used to calculate the linear correlation strength between vibration and voiceprint;

[0048] is the spatiotemporal overlap (range [0,1]), which is used to quantify the matching ratio between the vibration peak area and the image motion area;

[0049] is the weight, is the optical flow variation of the key mechanism action.

[0050] Furthermore, the update mechanism of the process optimization engine is:

[0051] When the faulty part number Severity coefficient hour;

[0052] Update weights:

[0053] ;

[0054] in, represents the update amount of the model weight in the process optimization engine, where The weights of different features in the corresponding model, It is the learning rate parameter for weight update, which is used to control the step size of weight update and determine the amplitude of each weight adjustment. Loss is the loss function in the process optimization engine, which is used to quantify the deviation between the shot blasting quality prediction value and the actual detection value. To express the loss function for the Weight The partial derivative of reflects the rate and direction of change of the loss function with the weight, which is used to indicate the direction and basic amplitude of the weight adjustment; is the default value. Trigger the weight update mechanism.

[0055] Furthermore, the human-machine collaborative optimization module performs:

[0056] When it is detected that the operator manually adjusts the parameters, the behavior operation action feature vector of the image information acquisition module is synchronously extracted. pass Network prediction adjustment intent confidence ;

[0057] Record adjustment amount , generate the final optimization instructions:

[0058] ;

[0059] in, The calculation is based on the correlation between the gesture trajectory in the behavior image and the parameter adjustment history. It is the weight coefficient of the manual intervention parameter, which is used to adjust the influence of the operator's manual adjustment on the final instruction.

[0060] Furthermore, the human-machine collaborative optimization module responds to the output of the fault collaborative diagnosis unit:

[0061] According to the fault part number Generate 3D fault markers and overlay abnormal action hotspots in operator behavior images;

[0062] Spatial coordinates are identified by components and 3D mapping functions Determine the abnormal action hotspots by analyzing the pixel motion intensity distribution in the behavior image using the optical flow method generate;

[0063] The distribution of pixel motion intensity in the image and the operation action feature vector The spatial and temporal distribution of

[0064] The information sending module performs:

[0065] Will Write Control registers;

[0066] Generate component coordinates fault reports and push action suggestions.

[0067] Furthermore, the operator behavior images collected by the image information collection module are processed by a computer vision algorithm, and the specific processing process includes: using The model recognizes human postures in behavioral images and extracts action feature vectors ;

[0068] The operation action feature vector and preset compliance operation template library Dynamic Time Warping Matching, generating operational compliance scores ;

[0069] when When the warning signal is sent to the human-machine collaborative optimization module, is the preset threshold.

[0070] This invention offers the following advantages over existing technologies: This intelligent control system for shot blasting machines, based on multi-source data acquisition, collects and intelligently processes operational data from multiple dimensions, enabling precise optimization of the shot blasting process and efficient diagnosis of equipment faults. The system utilizes a multi-source data acquisition module to comprehensively capture information such as equipment operating status, environmental parameters, and workpiece quality. Time synchronization technology is used to ensure data consistency. Intelligent algorithms dynamically optimize process parameters, enabling real-time adjustment based on actual shot blasting quality to enhance process performance. Furthermore, a collaborative fault diagnosis mechanism based on vibration, voiceprint, and image characteristics accurately locates faulty equipment components, improving troubleshooting efficiency. The human-machine collaborative optimization module integrates manual intervention with intelligent prediction, leveraging the optimization capabilities of artificial intelligence and leveraging operator experience to enhance system adaptability and ease of use. Furthermore, the system features environmental adaptive processing capabilities, automatically adjusting model inputs to changes in environmental parameters such as dust concentration, temperature, and humidity, ensuring stable system operation. Computer vision algorithms monitor operator behavior to ensure operational compliance and reduce the risk of human error. Overall, the system realizes intelligent and precise control of the shot blasting process, improves production efficiency, processing quality and equipment maintenance level, making the system more worthy of promotion and use. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] Figure 1 It is the overall structural diagram of the present invention. DETAILED DESCRIPTION

[0072] The following is a detailed description of an embodiment of the present invention. This embodiment is implemented based on the technical solution of the present invention, and provides a detailed implementation method and specific operation process. However, the protection scope of the present invention is not limited to the following embodiment.

[0073] like Figure 1 As shown, this embodiment provides a technical solution: an intelligent control system for a shot blasting machine based on multi-source data acquisition, comprising:

[0074] The data acquisition module group is used to collect shot blasting machine operation-related data in multiple dimensions, including:

[0075] Shot blasting machine body information collection module, used to collect equipment model and component identification;

[0076] Environmental information collection module, used to collect environmental temperature, humidity and dust concentration in real time;

[0077] Image information acquisition module, used to synchronously capture images of the workpiece surface and operator behavior;

[0078] Operation status acquisition module, used to collect equipment's three-dimensional axial vibration waveform, operation soundprint, and key mechanism motion images;

[0079] Shot blasting quality collection module, used to detect the cleanliness and roughness of the workpiece surface in real time;

[0080] The data processing module is used to analyze and optimize the collected data, including:

[0081] Time synchronization controller, based on The protocol aligns the data timestamps of each module and outputs synchronized data packets with a time deviation of less than 1ms;

[0082] Process optimization engine generates process parameter optimization instructions based on shot blasting quality detection values ​​and synchronization data packets;

[0083] Collaborative fault diagnosis unit locates abnormal equipment components based on vibration, voiceprint, and image features;

[0084] Human-machine collaborative optimization module integrates human intervention parameters to generate final optimization instructions;

[0085] Information sending module, used to write the final optimization instructions into controller and push fault location information and operation suggestions.

[0086] The operating status acquisition module includes:

[0087] A vibration sensor group arranged on the bearing seat is used to output a three-dimensional axial vibration spectrum;

[0088] A directional microphone array installed on the side wall of the shot blasting room is used to collect 200Hz-10kHz operating sound patterns;

[0089] High-speed camera unit, used to capture dynamic images of the shot blasting machine impeller and shot dividing wheel;

[0090] By installing a vibration sensor group, a directional microphone array, and a high-speed camera unit in the operating status acquisition module, key operating data of the shot blasting machine can be collected in real time from multiple dimensions. The vibration sensor group can capture the three-dimensional axial vibration spectrum of the bearing seat to monitor the operating stability of the equipment's mechanical components; the directional microphone array can capture the operating soundprint of the shot blasting room to assist in identifying abnormal mechanical sounds; and the high-speed camera unit captures dynamic images of key mechanisms such as the shot blasting machine's impeller and shot-dispensing wheel to intuitively reflect the operating status of the components. The coordinated arrangement of multiple types of sensors enables all-round monitoring of equipment vibration, sound, and mechanical movement, providing multi-source data support for subsequent fault location and operating status analysis, improving the system's perception accuracy and real-time performance of equipment anomalies, and ensuring the comprehensiveness and accuracy of shot blasting machine operating status monitoring.

[0091] The data processing module further comprises a time synchronization controller, which is used to: receive the original data streams from each acquisition module;

[0092] based on The protocol aligns all data timestamps;

[0093] Output synchronization data packets with time deviation less than 1ms to the process optimization engine and fault collaborative diagnosis unit;

[0094] By eliminating the time deviation of the data from each acquisition module (the time deviation of the output synchronization data packet is less than 1ms), the consistency of multi-dimensional data such as vibration spectrum, voiceprint, and image in the time dimension is ensured, providing a precise time reference for parameter calculation in the subsequent process optimization engine and anomaly location in the fault collaborative diagnosis unit. This mechanism avoids analysis errors caused by data asynchrony, enabling the system to perform more accurate process parameter optimization and equipment fault feature extraction based on time-space aligned data streams, improving the accuracy and reliability of the intelligent control system's response to dynamic changes in the shot blasting process.

[0095] The process optimization engine performs:

[0096] Receive the actual cleanliness and roughness measurement values ​​of the workpiece surface, that is, the shot blasting quality test value and synchronization data packets;

[0097] Calculate quality prediction value The specific process is:

[0098] ;

[0099] Where, The workpiece surface texture feature matrix is ​​collected by the image information acquisition module. Feature extraction is obtained;

[0100] It is the three-dimensional axial vibration spectrum feature vector of the equipment, which is collected by the vibration sensor group of the operation status acquisition module. Temporal coding acquisition;

[0101] To generate the Mel-frequency cepstral coefficient (MFCC) of the running voiceprint, the microphone array of the running status acquisition module is used to collect the Feature extraction is obtained;

[0102] is the model weight, which is called from the process parameter library based on the equipment model;

[0103] Solve Minimal parameter set ;

[0104] Combined with the workpiece surface image ( Extract texture features), equipment vibration spectrum ( Processing time series features), running voiceprint ( Extract acoustic features), build a comprehensive quality prediction model, avoid analysis deviation of a single data dimension, and make process parameter optimization more in line with actual production needs.

[0105] Intelligent prediction and precise parameter adjustment: Through weighted fusion of multiple model outputs, , calculate the deviation between the quality prediction value and the actual value in real time, and automatically solve the optimal parameter set , realizing closed-loop control from detection, analysis to optimization, and reducing the cost of manual trial and error.

[0106] For example, if the surface cleanliness of a workpiece does not meet the standard after shot blasting, such as residual oxide scale.

[0107] The image information acquisition module captures the uneven surface texture of the workpiece, the vibration sensor detects abnormal axial vibration of the shot blasting wheel, and the microphone array collects abnormal sound patterns, such as high-frequency noise.

[0108] Extracting texture feature matrix of oxide scale residue from images ;

[0109] Analyzing the vibration spectrum, we found the abnormal vector of the timing vibration when the impeller of the shot blasting machine rotates. ;

[0110] Extracting voiceprints feature , identifying acoustic patterns associated with component wear.

[0111] Process optimization engine through The model calculates the quality prediction deviation under the current parameters (such as shot blasting pressure and shot flow), automatically adjusts the parameters, and generates the optimal parameter set. , such as increasing the shot blasting pressure by 10% and extending the shot blasting time by 5s to make the cleanliness meet the standard.

[0112] This process does not require human intervention. Through the collaboration of multi-source data and intelligent algorithms, it achieves adaptive optimization of the shot blasting process and improves production efficiency and quality stability.

[0113] The process optimization engine includes an adaptive feature extraction unit. When the ambient dust concentration exceeds a threshold or the temperature and humidity deviate from the standard range, it reconstructs the voiceprint features and replaces the model input, wherein:

[0114] The specific process of reconstructing voiceprint features is as follows:

[0115] ;

[0116] Where, is the compensation function of environmental parameters (temperature, humidity) on voiceprint;

[0117] The specific process of updating the model input is:

[0118] ;

[0119] Through the adaptive feature extraction unit, the system can automatically reconstruct the voiceprint features and adjust the model input when the dust concentration changes suddenly or the temperature and humidity deviate from the standard range, avoiding the optimization deviation of process parameters caused by environmental interference and ensuring the stability of shot blasting quality under different working conditions.

[0120] Without human intervention, the system can dynamically correct voiceprint characteristics according to real-time environmental parameters, such as filtering dust noise and compensating for the impact of temperature and humidity on sound waves, so that the quality prediction model can continue to maintain high accuracy and improve the reliability of intelligent control.

[0121] If the dust removal system in the shot blasting workshop fails, the dust concentration will suddenly increase from 50mg / m 3 Increased to 200 mg / m 3 , exceeding the threshold of 100mg / m 3 , affecting the accuracy of voiceprint collection.

[0122] The environmental information collection module detects dust concentration Exceeding the threshold triggers the adaptive feature extraction unit.

[0123] Voiceprint feature reconstruction: The system reconstructs the original voiceprint Applying the Wiener filter , filter out dust noise interference and get the corrected voiceprint ;

[0124] At the same time, call the environment compensation function ,According to the current temperature, such as 35℃, and humidity, such as 70%, parameter compensation is performed on the impact of voiceprint propagation to further optimize the features.

[0125] The process optimization engine switches the model input and no longer uses the original voiceprint , but based on the revised and vibration spectrum Recalculate quality predictions Ensure that parameter optimization instructions are not disturbed by dust and maintain stable shot blasting quality.

[0126] This mechanism enables the system to have self-purification capabilities in complex industrial environments, avoid fluctuations in shot blasting effects caused by environmental fluctuations, and reduce production anomalies and quality defects caused by environmental factors.

[0127] The fault collaborative diagnosis unit performs the following process:

[0128] Extract vibration spectrum features from synchronization data packets Voiceprint features ;

[0129] Build a spatiotemporal alignment model:

[0130] ;

[0131] Output Fault Part Number:

[0132] ;

[0133] In the formula is the Pearson correlation coefficient (range [-1,1]), which is used to calculate the linear correlation strength between vibration and voiceprint;

[0134] is the spatiotemporal overlap (range [0,1]), which is used to quantify the matching ratio between the vibration peak area and the image motion area;

[0135] is the weight, is the optical flow variation of the key mechanism action;

[0136] Collaborative fault location using multi-source data: By integrating vibration spectrum, voiceprint features, and image optical flow changes, a spatiotemporal aligned fault diagnosis model is constructed, breaking through the detection limitations of a single sensor, enabling accurate location of abnormal equipment components and reducing misjudgment rates.

[0137] Dynamic correlation features improve diagnostic accuracy: using Pearson correlation coefficient Quantify the linear correlation between vibration and voiceprint, combined with the temporal and spatial overlap Matching vibration peaks with image motion areas enables fault location to have both synchronization in the temporal dimension and correlation in the spatial dimension, improving diagnostic efficiency.

[0138] Such as shot blasting machine shot blasting wheel bearing wear leads to abnormal operation.

[0139] The vibration sensor group detects the three-dimensional axial vibration spectrum of the bearing seat High-frequency abnormal fluctuations occur;

[0140] The directional microphone array collects periodic abnormal sound characteristics in the 200Hz-10kHz soundprint. ;

[0141] The high-speed camera unit captures the slight deviation of the shot blasting wheel during rotation, and the optical flow method is used to analyze the Abnormal exercise intensity.

[0142] The system calculates the optimal alignment time through the formula , make vibration and voiceprint The difference and optical flow change are minimized to ensure that multi-source data are synchronized in the time dimension;

[0143] Calculate the Pearson correlation coefficient between vibration and voiceprint , and found that the two have significant linear correlation, such as At the same time, the overlap between the vibration peak area and the motion area of ​​the bearing part in the image Reaching 0.78.

[0144] Fault collaborative diagnosis unit according to weight Integration and value, calculate the failure probability of each component, and finally output the bearing component number , locate the source of the fault.

[0145] This mechanism avoids misjudgment of a single signal through cross-validation of multi-dimensional data, such as missed diagnosis based on vibration or voiceprint alone, and implements fault location using a triple chain of evidence: vibration, sound, and image. It is particularly suitable for detecting anomalies in equipment such as shot blasting machines in high-vibration and high-noise environments, improving maintenance efficiency and reducing downtime.

[0146] The update mechanism of the process optimization engine is:

[0147] When the faulty part number Severity coefficient hour;

[0148] Update weights:

[0149] ;

[0150] in, represents the update amount of the model weight in the process optimization engine, where The weights of different features in the corresponding model, The learning rate parameter for weight update is used to control the step size of weight update and determine the magnitude of each weight adjustment. It is the loss function in the process optimization engine, which is used to quantify the deviation between the shot blasting quality prediction value and the actual detection value. To express the loss function for the Weight The partial derivative of reflects the rate and direction of change of the loss function with the weight, which is used to indicate the direction and basic amplitude of the weight adjustment; is the default value. Trigger the weight update mechanism;

[0151] By binding the severity coefficient of the faulty component to the model weight update mechanism, the system can dynamically adjust the parameter sensitivity of the process optimization model according to the severity of the equipment abnormality, avoid process parameter optimization deviations caused by component failure, and improve the robustness of the system under non-ideal working conditions.

[0152] Model self-evolution capability: When equipment fails, the system automatically increases the weight of data dimensions related to the failure, such as vibration and soundprint characteristics, so that the process optimization model pays more attention to the status changes of abnormal components. This forms a closed-loop adaptive mechanism from fault detection to weight adjustment and parameter optimization, reducing manual intervention costs.

[0153] If the impeller of the shot blasting machine is worn due to long-term use, the serious coefficient , exceeding the threshold , resulting in unstable shot blasting force.

[0154] The fault collaborative diagnosis unit locates the impeller part number , calculate its value based on the abnormal degree of vibration spectrum, voiceprint feature offset, etc. .

[0155] The process optimization engine triggers an update mechanism that increases the model weight related to the impeller state based on the gradient descent formula.

[0156] For example: the weight corresponding to the vibration feature (Original weight 0.3) Due to the wear of the impeller, the vibration is abnormal, and its gradient Positive, combined , calculated , after update ;

[0157] Voiceprint feature weight (Original weight 0.2) Due to the significant abnormal sound related features, the gradient is positive, , after update .

[0158] After the model is updated, the process optimization engine calculates the quality prediction value When more dependent on the vibration spectrum Harmony Voiceprint The feature is to automatically adjust the shot blasting parameters (such as increasing the shot material supply by 15%), compensate for the insufficient shot blasting force caused by impeller wear, and maintain the surface cleanliness of the workpiece to meet the standard.

[0159] The human-machine collaborative optimization module performs:

[0160] When it is detected that the operator manually adjusts the parameters, the behavior operation action feature vector of the image information acquisition module is synchronously extracted. ,pass Network prediction adjustment intent confidence ;

[0161] Record adjustment amount , generate the final optimization instructions:

[0162] ;

[0163] in, The calculation is based on the correlation between the gesture trajectory in the behavior image and the parameter adjustment history. The weight coefficient of the manual intervention parameter is used to adjust the influence of the operator's manual adjustment on the final instruction;

[0164] By combining the operator's manual adjustment with the system's intelligent prediction, the dynamic complementarity of human experience and algorithm optimization is achieved, which not only gives full play to the human's intuitive judgment ability on complex working conditions, but also takes advantage of the system's precise calculation to improve the flexibility and adaptability of shot blasting process adjustment. The network analyzes the operator's behavioral characteristics (such as gesture trajectories and operating habits) and predicts the confidence level of the intention of manual adjustment, enabling the system to automatically judge the rationality of manual intervention, avoid blindly following invalid operations, and reduce the trial and error cost of parameter adjustment.

[0165] The operator found that the surface roughness of a batch of workpieces was too high after shot blasting, so he manually increased the speed of the shot blasting motor. .

[0166] The image information acquisition module captures the operator's gesture trajectory when rotating the speed adjustment knob. Model extracts action feature vector (such as gesture speed, rotation amplitude).

[0167] Network Based Calculate the confidence level of adjustment intention by comparing with historical adjustment data (such as adjustment patterns when similar workpieces had insufficient roughness in the past). (High confidence, indicating that operational intention is highly correlated with roughness optimization).

[0168] Final optimization instruction generation: system original optimization parameter set The speed is increased by 3%, the shot flow rate remains unchanged, and the manual adjustment is combined With confidence level 0.9, the final optimization instructions are generated:

[0169] , , that is, the speed is increased by 3% + 0.8 × 5% × 0.9 = 3% + 3.6% = 6.6%, and the shot flow rate is adjusted synchronously to match the speed change.

[0170] After shot blasting, the roughness of the workpiece meets the standard. By integrating manual experience and algorithm optimization, the system reduces the number of parameter iterations by two compared to relying solely on automatic optimization, thereby improving production efficiency.

[0171] The human-machine collaborative optimization module responds to the output of the fault collaborative diagnosis unit:

[0172] According to the fault part number Generate 3D fault markers and overlay abnormal action hotspots in operator behavior images;

[0173] Spatial coordinates are identified by components and 3D mapping functions Determine the abnormal action hotspots by analyzing the pixel motion intensity distribution in the behavior image using the optical flow method generate;

[0174] The distribution of pixel motion intensity in the image and the operation action feature vector The spatial and temporal distribution of

[0175] The information sending module performs:

[0176] Will Write to PLC control register;

[0177] Generate component coordinates fault reports and push action suggestions;

[0178] By converting the faulty component number into three-dimensional spatial coordinates and superimposing abnormal action hotspots, an intuitive mapping of the fault location to the operating behavior is achieved, helping operators quickly locate the physical fault point and the area associated with manual operation, reducing maintenance and troubleshooting time.

[0179] By combining the optical flow method to analyze operator behavior images, equipment failures are dynamically associated with abnormal operating actions, and fault reports and operation suggestions with spatial coordinates are generated, providing dual support of data-driven and visual guidance for maintenance, lowering the technical threshold for maintenance.

[0180] If the shot blasting machine bearing is worn (fault part number ), and operators frequently accidentally touch nearby sensors during debugging.

[0181] Fault diagnosis and 3D mark generation: The fault collaborative diagnosis unit outputs the bearing part number , the system uses a three-dimensional mapping function Calculate its spatial coordinates (x=2.5m, y=1.8m, z=0.5m) and superimpose a red three-dimensional marker (such as a cube highlight at the bearing position) on the operator's behavior image.

[0182] Abnormal action hotspot analysis: Optical flow analysis of behavioral images reveals the distribution of pixel motion intensity in the corresponding area of ​​the bearing by the operator. Abnormalities, such as high brightness in areas where hands frequently approach the sensor, generate yellow hot zone marks that overlap with the 3D fault marks.

[0183] The system will operate the action feature vector By correlating with the spatiotemporal distribution of hot spots, the abnormal operation pattern of the mis-touch sensor is identified.

[0184] Information push and maintenance guidance: The information sending module will eventually optimize the instructions Write , and generate a fault report, including: 3D coordinates The corresponding device structure diagram is highlighted;

[0185] Operation video clips of abnormal action hotspots (e.g., a 10-second video of a hand accidentally touching a sensor);

[0186] Operation suggestion: Check the bearing lubrication status and avoid touching sensor A (coordinate x=2.6m, y=1.7m) during debugging.

[0187] Based on visual guidance, operators can locate bearing wear problems and correct operating habits within 10 minutes, reducing troubleshooting time compared to traditional maintenance processes and avoiding secondary failures caused by incorrect operations.

[0188] The operator behavior images collected by the image information acquisition module are processed by computer vision algorithms. The specific processing process includes: using The model recognizes human postures in behavioral images and extracts action feature vectors ;

[0189] The operation action feature vector and preset compliance operation template library Dynamic Time Warping Matching, generating operational compliance scores ;

[0190] when When the warning signal is sent to the human-machine collaborative optimization module, is the preset threshold;

[0191] Computer vision algorithms are used to analyze operator behavior in real time, automatically identify illegal operations (such as failure to debug according to the process, dangerous actions), generate compliance scores and issue warnings, reduce the risk of equipment failure or quality defects caused by human errors, dynamically match preset compliance operation templates with actual actions, provide real-time feedback to operators, and promote the standardization of operating habits. This is especially suitable for new employee training or complex process scenarios, and can reduce production anomalies caused by non-standard operations.

[0192] If the operator does not close the protective door according to the standard process when the shot blasting machine is running, and directly reaches out to adjust the shot blasting machine parameters.

[0193] Behavior image acquisition and feature extraction: The image information acquisition module captures the operator's hand movement images. The model recognizes the gesture features of the hand approaching the running shot blasting machine and extracts the action feature vector , such as hand movement trajectory and distance parameters from the device.

[0194] Compliance score calculation: The system will and preset compliance template library Dynamic time regularization using the shutdown debugging action template in Matching, calculating the deviation between the operation trajectory and the standard process. Because the operator did not close the protective door and operated while the equipment was running, the matching similarity is low, and a compliance score is generated. (Below the preset threshold ).

[0195] The system sends an early warning signal to the human-machine collaborative optimization module, and displays a red warning box on the operation interface, marking the illegal action area, such as the video clip of the hand approaching the shot blasting wheel, and pushes a text prompt: Please close the protective door before performing debugging operations.

[0196] The operator stopped the illegal action in time and followed the standard process, avoiding downtime accidents caused by mechanical damage or accidental contact of equipment. At the same time, the system recorded the violation and provided data support for subsequent training.

[0197] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of the technical features being referred to. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one such feature. In the description of the present invention, "plurality" means at least two, such as two, three, etc., unless otherwise specifically defined.

[0198] In the description of this specification, the reference terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" mean that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present 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 any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art can combine and combine different embodiments or examples described in this specification and features of different embodiments or examples without contradiction.

[0199] Although the embodiments of the present invention have been shown and described above, it will be understood that the above embodiments are illustrative and are not to be construed as limitations on the present invention. A person skilled in the art may change, modify, replace and modify the above embodiments within the scope of the present invention.

Claims

1. An intelligent control system for shot blasting machines based on multi-source data acquisition, characterized in that: include: The data acquisition module group is used to collect shot blasting machine operation-related data in multiple dimensions, including: Shot blasting machine body information collection module, used to collect equipment model and component identification; Environmental information collection module, used to collect environmental temperature, humidity and dust concentration in real time; Image information acquisition module, used to synchronously capture images of the workpiece surface and operator behavior; Operation status acquisition module, used to collect equipment's three-dimensional axial vibration waveform, operation soundprint, and key mechanism motion images; Shot blasting quality collection module, used to detect the cleanliness and roughness of the workpiece surface in real time; Data processing module, used to analyze and optimize the collected data, including: The time synchronization controller aligns the data timestamps of each module based on the preset type protocol and outputs synchronization data packets with time deviation less than the preset value; Process optimization engine generates process parameter optimization instructions based on shot blasting quality detection values ​​and synchronization data packets; Collaborative fault diagnosis unit locates abnormal equipment components based on vibration, voiceprint, and image features; Human-machine collaborative optimization module integrates human intervention parameters to generate final optimization instructions; The information sending module is used to write the final optimization instructions to the PLC controller and push fault location information and operation suggestions.

2. The intelligent control system for shot blasting machine based on multi-source data acquisition according to claim 1, characterized in that: The operating status acquisition module includes: A vibration sensor group arranged on the bearing seat is used to output a three-dimensional axial vibration spectrum; A directional microphone array installed on the side wall of the shot blasting room is used to collect operating sound patterns at preset frequencies; High-speed camera unit is used to capture dynamic images of the shot blasting machine impeller and shot dividing wheel.

3. The intelligent control system for shot blasting machine based on multi-source data acquisition according to claim 1, characterized in that: The data processing module further comprises a time synchronization controller, which is used to: receive the original data streams from each acquisition module; Align all data timestamps based on a preset type protocol; Output synchronization data packets with time deviations less than a preset value to the process optimization engine and the fault collaborative diagnosis unit.

4. The intelligent control system for shot blasting machine based on multi-source data acquisition according to claim 1, characterized in that: The process optimization engine performs: Receive the actual cleanliness and roughness measurement values ​​of the workpiece surface, that is, the shot blasting quality test value Q actua and synchronization data packets; Calculate the quality prediction value Q predict , the specific process is: Q predict =ω1×CNN(I img )+ω2×LSTM(V vib )+ω3×SVM(S audio ); Where, I img The workpiece surface texture feature matrix is ​​obtained by performing CNN feature extraction on the workpiece surface image collected by the image information acquisition module; V vib The three-dimensional axial vibration spectrum feature vector of the equipment is obtained by performing LSTM time series coding on the data collected by the vibration sensor group of the operation status acquisition module; S audio The Mel-frequency cepstral coefficients of the running voiceprint are obtained by performing SVM feature extraction on the microphone array of the running status acquisition module; ω1, ω2, and ω3 are model weights, which are called from the process parameter library based on the equipment model; Solve ∣Q actua −Q predict | Minimum parameter set P ∗ .

5. The intelligent control system for shot blasting machine based on multi-source data acquisition according to claim 4 is characterized in that: The process optimization engine includes an adaptive feature extraction unit. When the ambient dust concentration exceeds a threshold or the temperature and humidity deviate from the standard range, it reconstructs the voiceprint features and replaces the model input, wherein: The specific process of reconstructing voiceprint features is as follows: ； Where EnvComp(T,H) is the compensation function of environmental parameters to voiceprint, that is, the voiceprint environment compensation function based on temperature and humidity. is the Wiener filter coefficient based on dust concentration, It is a collaborative correction operator for voiceprint features; The specific process of updating the model input is: Q predict =ω2×LSTM(V vib )+ω3×SVM( ).

6. The intelligent control system for shot blasting machine based on multi-source data acquisition according to claim 1, characterized in that: The fault collaborative diagnosis unit performs the following process: Extract the vibration spectrum feature V in the synchronization data packet k With voiceprint feature S m ; Build a spatiotemporal alignment model: ； Output Fault Part Number: ； Where Corr(A,B) is the Pearson correlation coefficient, which is used to calculate the linear correlation strength between vibration and voiceprint; Overlap(A,B) is the spatiotemporal overlap, which is used to quantify the matching ratio between the vibration peak area and the image motion area; and β are weights, and OptFlow(I(t)) is the optical flow variation of the key mechanism action.

7. The intelligent control system for shot blasting machine based on multi-source data acquisition according to claim 6, characterized in that: The update mechanism of the process optimization engine is: When the severity coefficient of the faulty component number k, Severity(k)>θ; Update weights: ； in is the update amount of the i-th model weight, that is, the adjustment range of a certain weight among ω1, ω2 and ω3, Update the learning rate for the weights, is the gradient of the loss function with respect to the weights.

8. The intelligent control system for shot blasting machine based on multi-source data acquisition according to claim 1, characterized in that: The human-machine collaborative optimization module performs: When it is detected that the operator manually adjusts the parameters, the behavioral operation action feature vector Fop of the image information acquisition module is synchronously extracted, and the adjustment intention confidence Confidence(Ioperator) is predicted through the LSTM network; Record the adjustment value ΔPhuman and generate the final instruction: ； The calculation of Confidence (Ioperator) is based on the correlation between the gesture trajectory in the behavior image and the parameter adjustment history.

9. The intelligent control system for shot blasting machine based on multi-source data acquisition according to claim 8, characterized in that: The human-machine collaborative optimization module responds to the output of the fault collaborative diagnosis unit: Generate a 3D fault marker based on the fault component number k and overlay the abnormal action hotspot in the operator behavior image; The spatial coordinates are determined by the component identification and the three-dimensional mapping function (x, y, z) = Map(k, Model3D). The abnormal action hotspot is analyzed by the optical flow method to analyze the pixel motion intensity distribution M in the behavior image. flow generate; The pixel motion intensity distribution M in the image flow Associated with the spatiotemporal distribution of the operation action feature vector Fop; The information sending module performs: Will Write to PLC control register; Generate a fault report containing component coordinates Map(k) and push action suggestions.

10. The intelligent control system for shot blasting machine based on multi-source data acquisition according to claim 1, characterized in that: The operator behavior images collected by the image information acquisition module are processed by a computer vision algorithm. The specific processing process includes: using the YOLO model to perform human posture recognition on the behavior image and extracting the operation action feature vector Fop; Perform dynamic time warping matching on the operation action feature vector Fop and the preset compliance operation template library Ttemplate to generate the operation compliance score Scorecomp; When Scorecomp < τ, a warning signal is sent to the human-machine collaborative optimization module, where τ is the preset threshold.

Images