An automatic control method and system in a laser precision manufacturing process

Abstract

The application provides an automatic control method and system in a laser precision manufacturing process, and relates to the technical field of automatic control. The method comprises the following steps: acquiring multi-dimensional physical sensor data, and establishing a physical reference coordinate system in real time based on the multi-dimensional physical sensor data; acquiring an image captured by a machine vision system, and performing feature point positioning calculation on the image to obtain visual positioning data; comparing the visual positioning data and the physical reference coordinate system in real time to obtain the deviation therebetween; judging whether the deviation is a systematic deviation of the machine vision system according to the persistence and stability of the deviation; when the deviation is the systematic deviation of the machine vision system, adaptively adjusting the coordinate conversion parameters in the control system to correct the conversion relationship from the machine vision system to the motion platform coordinate system; and monitoring the change characteristics of the coordinate conversion parameters, and issuing a maintenance suggestion information when the change characteristics are abnormal. The application reduces the product rejection rate.

Description

Technical Field

[0001] This application relates to the field of automation control, and more specifically, to an automation control method and system for the precision manufacturing process of lasers. Background Technology

[0002] In automated production lines for precision laser manufacturing, a core element is the use of a sophisticated automated control method that relies on image information captured by a machine vision system. Feature point localization calculations in image recognition are responsible for accurately identifying key features on optical components. This positional information is then transmitted to a high-precision motion platform to align, bond, and subsequently perform UV curing operations on the optical components, ensuring extremely high optical coupling efficiency and product consistency during the laser module manufacturing process. However, under the context of long-term continuous operation of the production line, the protective coating of the camera lenses used in the machine vision system may undergo localized uneven degradation due to the cumulative and gradual effects of trace amounts of chemical vapors in daily cleaning or cleanroom environments. This causes a small, systematic shift in the output coordinate data, severely impacting product accuracy. Summary of the Invention

[0003] This application aims to address at least one of the technical problems existing in the prior art. To this end, this application proposes an automated control method and system for the precision manufacturing process of lasers, aiming to solve the problems of lens degradation caused by long-term operation of machine vision systems on automated production lines for precision laser manufacturing, resulting in systematic misalignment, and the problem of a sharp increase in product scrap rate due to the compensation operation based on error diagnosis in the prior art.

[0004] In a first aspect, this application discloses an automated control method for the precision manufacturing process of lasers, comprising the following steps:

[0005] Acquire multi-dimensional physical sensor data and establish a physical reference coordinate system in real time based on the multi-dimensional physical sensor data;

[0006] The system acquires images captured by the machine vision system and performs feature point localization calculations on the images to obtain visual localization data.

[0007] The visual positioning data is compared with the physical reference coordinate system in real time to obtain the deviation between the two.

[0008] Based on the persistence and stability of the deviation, determine whether the deviation is a systematic offset generated by the machine vision system;

[0009] When the deviation is a systematic offset generated by the machine vision system, the coordinate transformation parameters inside the control system are adaptively adjusted based on the physical reference coordinate system to correct the transformation relationship between the machine vision system and the motion platform coordinate system.

[0010] Monitor the changes in coordinate transformation parameters, and issue maintenance recommendations when abnormal changes are detected.

[0011] Furthermore, after acquiring multi-dimensional physical sensor data and establishing a physical reference coordinate system in real time based on the multi-dimensional physical sensor data, the method also includes:

[0012] Establish a pointing error distribution map for the rangefinder, which includes:

[0013] Drive the motion platform to carry a laser rangefinder to scan and measure multiple preset static reference points within the workspace;

[0014] Record the current position as fed back by the motion platform encoder;

[0015] The pointing deviation of the rangefinder is calculated based on the position fed back by the encoder of the motion platform, the installation position and direction of the laser rangefinder relative to the zero point of the platform, and the known coordinates of the static reference point.

[0016] Store the pointing deviation of the rangefinder at different locations to obtain the pointing error distribution map of the rangefinder;

[0017] The laser rangefinder measurement data is pre-corrected using the rangefinder pointing error distribution map in order to fine-tune the physical reference coordinate system.

[0018] Furthermore, in the step of monitoring the changing characteristics of coordinate transformation parameters and issuing maintenance recommendations when the changing characteristics are abnormal, the method includes:

[0019] Monitor the changing characteristics of coordinate transformation parameters;

[0020] When the change characteristics are abnormal, a preset physical cause is associated with the change characteristics;

[0021] Maintenance recommendations are issued based on physical reasons.

[0022] Based on the above, this application further proposes that the steps for associating a pre-defined physical cause with the characteristics of change include:

[0023] Extract the variation characteristics of coordinate transformation parameters;

[0024] The similarity between the change features and the typical coordinate transformation parameter change feature set in the preset physical cause feature fingerprint database is calculated. The physical cause feature fingerprint database stores the typical coordinate transformation parameter change feature set corresponding to known physical causes. The typical coordinate transformation parameter change feature set contains feature descriptions in multiple dimensions.

[0025] Filter out multiple physical reasons with similarity exceeding a preset threshold;

[0026] Based on the current environmental sensor data of the production line and historical maintenance records, confidence levels are assessed for multiple physical causes;

[0027] Select the physical cause associated change features with the highest confidence level.

[0028] Preferably, the steps for assessing the confidence level of multiple physical causes by combining current production line environmental sensor data and historical maintenance records include:

[0029] Acquire current environmental sensor data and historical maintenance records;

[0030] For each of the multiple physical causes, an indication intensity factor is generated based on the degree of matching between the physical cause and environmental sensor data and historical maintenance records;

[0031] Acquire equipment uptime, optical component characteristic data for the current production batch, and external operating condition sensor data;

[0032] The weight of the indication intensity factor is calculated based on the equipment's operating time, the optical component characteristic data of the current production batch, and the external operating condition sensor data.

[0033] The indicator intensity factor and its corresponding weight are weighted and fused to assess the confidence level of multiple physical causes.

[0034] In one implementation, the step of calculating the weight of the indication intensity factor based on the equipment's operating time, optical element characteristic data of the current production batch, and external operating condition sensor data includes:

[0035] The continuous operation time of the equipment is checked. When the operation time is interrupted, the most recent continuous operation time is used for smooth interpolation, and the confidence level of the interpolated operation time is marked.

[0036] The integrity of the optical component characteristic data of the current production batch is checked. When key characteristic parameters are missing in the optical component characteristic data, they are filled in according to the average characteristic value of the optical components in the same batch, and the confidence level is marked on the filled optical component characteristic data.

[0037] Real-time monitoring of external operating condition sensor data is performed. When the update delay of external operating condition sensor data exceeds a preset threshold, short-term predicted values ​​of historical operating condition data are used as replacements, and the confidence level of the replaced external operating condition sensor data is marked.

[0038] Based on the confidence markers of running time, optical element characteristic data, and external operating condition sensor data, the contribution ratio of each to the weight of the indication intensity factor is dynamically adjusted.

[0039] The weights of the indicator intensity factor are calculated based on the adjusted contribution ratio.

[0040] Based on the above, this application further proposes a step of dynamically adjusting the contribution ratio of each to the weight of the indication intensity factor according to the confidence marker of the running time, the confidence marker of the optical element characteristic data, and the confidence marker of the external operating condition sensor data, including:

[0041] Obtain the equipment's cumulative operating time equivalent and equipment lifecycle segmentation rules;

[0042] The stage of the equipment's life cycle is determined based on the equipment's cumulative operating time equivalent and the equipment life cycle classification rules.

[0043] The basic contribution ratio adjustment rules are invoked based on the stage of the device's lifecycle.

[0044] When a specific failure mode is identified, the basic contribution ratio adjustment rule is locally modified according to the characteristics of the failure mode.

[0045] Under the revised contribution ratio adjustment rule, the contribution ratio of each to the weight of the indication intensity factor is dynamically adjusted based on the confidence level of the running time, the confidence level of the optical element characteristic data, and the confidence level of the external operating condition sensor data.

[0046] In one implementation, the step of determining the stage of the equipment's lifecycle based on the equipment's cumulative operating time and the equipment lifecycle segmentation rules includes:

[0047] Continuously acquire equipment operation mode data and identify high-intensity operation events within the equipment operation mode data;

[0048] Identify abnormal operating conditions in environmental sensor data;

[0049] The comprehensive degradation index of the equipment is calculated based on the frequency and duration of abnormal operating events, as well as the cumulative duration and intensity of high-intensity operating events.

[0050] Obtain the cumulative operating time of the equipment;

[0051] Calculate the equipment's cumulative operating time equivalent based on the equipment's comprehensive degradation index and cumulative operating time.

[0052] The stage of the equipment's life cycle is determined based on the equipment's cumulative operating time equivalent and the equipment life cycle classification rules.

[0053] Based on the above, this application further proposes that the steps for calculating the comprehensive equipment degradation index, based on the frequency and duration of abnormal operating events and the cumulative duration and intensity of high-intensity operating events, include:

[0054] Obtain the frequency and duration of abnormal operating conditions;

[0055] The equipment anomaly index is calculated based on the frequency and duration of abnormal operating events.

[0056] Obtain the cumulative duration and intensity of high-intensity operation events;

[0057] The equipment strength index is calculated based on the cumulative duration and intensity of high-intensity operational events;

[0058] The comprehensive equipment degradation index is obtained by weighting and fusing the equipment anomaly index and the equipment strength index.

[0059] Secondly, this application also discloses an automated control system for the precision manufacturing process of lasers, the system comprising:

[0060] The physical reference coordinate system establishment module is used to acquire multi-dimensional physical sensor data and establish a physical reference coordinate system in real time based on the multi-dimensional physical sensor data.

[0061] The visual positioning data acquisition module is used to acquire images captured by the machine vision system and perform feature point localization calculations on the images to obtain visual positioning data.

[0062] The deviation comparison module is used to compare the visual positioning data with the physical reference coordinate system in real time to obtain the deviation between the two.

[0063] The systematic offset judgment module is used to determine whether the deviation is a systematic offset generated by the machine vision system based on the persistence and stability of the deviation.

[0064] The coordinate transformation parameter adjustment module is used to adaptively adjust the coordinate transformation parameters inside the control system based on the physical reference coordinate system when the deviation is a systematic offset generated by the machine vision system, thereby correcting the transformation relationship between the machine vision system and the motion platform coordinate system.

[0065] The information sending module is used to monitor the change characteristics of coordinate transformation parameters. When the change characteristics are abnormal, maintenance suggestion information is sent.

[0066] The technical solution according to the embodiments of this application has at least the following beneficial effects:

[0067] This application effectively solves the problems in existing machine vision systems where lens degradation leads to systematic offsets due to long-term operation, and where traditional compensation methods mask these issues, resulting in increased product scrap rates. This application establishes an independent physical reference coordinate system, providing the vision system with a calibration benchmark unaffected by its own degradation. This allows for accurate identification and correction of systematic offsets in the vision system, avoiding spurious convergence or local optima. This adaptive adjustment mechanism enables the control system to dynamically adapt to changes in the vision system, ensuring the accuracy and stability of optical component alignment, significantly improving the overall optical coupling efficiency of the laser module, and reducing product scrap rates. Simultaneously, by monitoring changes in coordinate transformation parameters and issuing maintenance recommendations, early warning and intervention for potential equipment failures are achieved, further ensuring the stable and efficient operation of the production line.

[0068] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0069] The accompanying drawings are used to provide a further understanding of the technical solutions of this application and constitute a part of the specification. They are used together with the embodiments of this application to explain the technical solutions of this application and do not constitute a limitation on the technical solutions of this application.

[0070] Figure 1 A flowchart illustrating an automated control method in the precision manufacturing process of a laser, provided as an embodiment of this application;

[0071] Figure 2 This is a schematic diagram of an automated control system in the precision manufacturing process of a laser, provided as an embodiment of this application. Detailed Implementation

[0072] To make the objectives, technical methods, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0073] In automated production lines for precision laser manufacturing, traditional automated control methods can lead to localized, uneven degradation of the protective coating on camera lenses used in machine vision systems over long-term operation due to the cumulative and gradual effects of trace amounts of chemical vapors encountered during routine cleaning or in cleanroom environments. This degradation results in non-uniform optical transmittance and scattering characteristics of the camera lenses, causing slight, non-uniform blurring and distortion in the images captured by the vision system, severely impacting production.

[0074] Based on the above, this application proposes an automated control method and system for the precision manufacturing process of lasers, aiming to improve the accuracy of automated control in the precision manufacturing process of lasers.

[0075] See Figure 1 , Figure 1 This is a flowchart illustrating an automated control method for the precision manufacturing process of a laser, provided in one embodiment of this application. The automated control method for the precision manufacturing process of a laser provided in this embodiment includes, but is not limited to, steps S110 to S160, which will be described in detail below.

[0076] S110. Acquire multi-dimensional physical sensor data and establish a physical reference coordinate system in real time based on the multi-dimensional physical sensor data;

[0077] S120. Acquire the image captured by the machine vision system, and perform feature point localization calculation on the image to obtain visual localization data;

[0078] S130. Real-time comparison of visual positioning data with physical reference coordinate system to obtain the deviation between the two;

[0079] S140. Based on the persistence and stability of the deviation, determine whether the deviation is a systematic offset generated by the machine vision system.

[0080] S150. When the deviation is a systematic offset generated by the machine vision system, the coordinate transformation parameters inside the control system are adaptively adjusted based on the physical reference coordinate system to correct the transformation relationship between the machine vision system and the motion platform coordinate system.

[0081] S160. Monitor the change characteristics of coordinate transformation parameters, and issue maintenance suggestion information when the change characteristics are abnormal.

[0082] To make the technical solution of this application easier and clearer to understand, some key terms involved will be explained first.

[0083] Multidimensional physical sensor data refers to measurement information from various physical sensors, such as laser rangefinders, high-precision encoders, temperature sensors, and humidity sensors. This data is used together to accurately describe the physical state and equipment location within the workspace.

[0084] The physical reference coordinate system is a high-precision and stable spatial coordinate system established based on these multi-dimensional physical sensor data. It serves as the benchmark for the entire control system, is independent of the machine vision system, and can provide a real-world physical reference.

[0085] Machine vision systems typically consist of industrial cameras, lenses, light sources, and image processing units. They are used to capture images of the work area and identify and locate target features through image processing algorithms.

[0086] Feature point localization calculation refers to the process by which a machine vision system analyzes captured images, identifies preset key geometric feature points, calculates the pixel coordinates of these feature points in the camera image coordinate system, and finally converts them into visual localization data.

[0087] Visual positioning data is the data output by the feature point positioning calculation program that describes the position of the target feature point in the coordinate system of the machine vision system.

[0088] Deviation refers to the spatial positional difference between visual positioning data and the physical reference coordinate system.

[0089] Systematic offset refers to a positioning error with a certain degree of persistence and stability caused by reasons inherent to the machine vision system itself (such as lens optical performance degradation, camera installation looseness, etc.), rather than random noise.

[0090] Coordinate transformation parameters are mathematical parameters used within the control system to transform data from the coordinate system of the machine vision system to the coordinate system of the motion platform. They typically include rotation matrices and translation vectors.

[0091] The motion platform coordinate system is the coordinate system used by the motion control system to precisely control the movement of the motion platform in three-dimensional space.

[0092] Maintenance recommendation information refers to the information sent by the system to operators or maintenance systems when the system detects abnormal changes in coordinate transformation parameters, indicating possible causes of the fault and suggesting maintenance measures.

[0093] The core of the automated control method proposed in this application lies in achieving accurate identification and adaptive correction of systematic deviations in machine vision systems through multi-source data fusion and intelligent judgment.

[0094] Firstly, regarding the step of acquiring multi-dimensional physical sensor data and establishing a physical reference coordinate system in real time based on this data, one approach is to use multiple high-precision laser rangefinders to perform triangulation on multiple fixed reference points within the workspace. Combined with the position information fed back by the motion platform's encoder, the physical reference coordinate system is calculated and established in real time using the least squares method or other optimization algorithms. For example, three or more static reference points with known coordinates are preset in the workspace. The laser rangefinders measure the distances to these reference points, while the encoders on the motion platform provide the current position of the rangefinders. Using this data, the physical reference coordinate system of the workspace can be accurately calculated. Another approach is to employ a physical measurement network consisting of multiple high-precision absolute encoders and tilt sensors. These sensors are installed on the motion platform and key structural components to continuously monitor their attitude and position. Through the fusion processing of this sensor data, a high-precision physical reference coordinate system can be constructed in real time. For example, multiple encoders are installed at different positions on the motion platform to acquire the platform's linear and angular positions in real time, while the tilt sensors provide the platform's attitude information. After these data are processed by the fusion algorithm, a stable physical reference coordinate system can be established.

[0095] Secondly, regarding the step of acquiring images captured by a machine vision system and performing feature point localization calculations on those images to obtain visual localization data, one approach is to employ a deep learning-based image recognition model. This model, trained on a large amount of labeled image data, can directly identify and locate target feature points from the original image and output their precise position information in the vision system's coordinate system. For example, a convolutional neural network (CNN) model can be trained to recognize multiple alignment features on laser components and directly output the sub-pixel coordinates of these feature points.

[0096] Next, regarding the step of comparing the visual positioning data with the physical reference coordinate system in real time to obtain the deviation between the two, one implementation method is to transform the visual positioning data into the physical reference coordinate system using the current coordinate transformation parameters, and then directly calculate the coordinate difference between the corresponding points in the transformed visual positioning data and the physical reference coordinate system to obtain the deviation. For example, if the vision system recognizes a feature point whose position in its own coordinate system is (x_v, y_v, z_v), while the actual position of the feature point in the physical reference coordinate system is (x_p, y_p, z_p), then (x_v, y_v, z_v) is transformed into (x'_p, y'_p, z'_p) using the current transformation matrix T, and then (x'_p - x_p, y'_p - y_p, z'_p - z_p) is calculated as the deviation.

[0097] Then, regarding the step of determining whether the deviation is a systematic offset generated by the machine vision system based on its persistence and stability, one implementation method is to perform time series analysis on the deviation obtained from real-time comparisons. If the deviation persists for a period of time (e.g., dozens of consecutive measurements) and its numerical variation ranges within a preset small threshold, it is determined to be a systematic offset. For example, a time window is set, and the mean and standard deviation of the deviation within the window are calculated. If the mean is significantly non-zero and the standard deviation is less than a certain preset value, a systematic offset is considered to exist.

[0098] Next, regarding the step of adaptively adjusting the coordinate transformation parameters within the control system to correct the transformation relationship between the machine vision system and the motion platform coordinate system when the deviation is a systematic offset generated by the machine vision system, using the physical reference coordinate system as a reference, one implementation method is that when a systematic offset is detected, the control system will initiate an optimization algorithm (e.g., based on gradient descent or the Levenberg-Marquardt algorithm). Using the physical reference coordinate system as the truth, it maps the visual positioning data to the physical reference coordinate system through the adjusted coordinate transformation parameters, minimizing the error between the mapped visual positioning data and the physical reference coordinate system. For example, if the systematic offset manifests as the vision system continuously deviating from the physical reference coordinate system in the X direction, the optimization algorithm will adjust the X-direction translation component in the coordinate transformation parameters to gradually reduce this offset until the visual positioning data is highly consistent with the physical reference coordinate system.

[0099] Finally, regarding the monitoring of the changes in the coordinate transformation parameters, one approach to issuing maintenance recommendations when these changes are abnormal is to continuously record the historical values ​​of the coordinate transformation parameters and perform trend analysis. If the rate, magnitude, or pattern of change of a parameter exceeds a preset normal range, it is considered abnormal, and maintenance recommendations are issued. For example, if a rotation parameter experiences a large jump within a short period, or if a translation parameter continuously drifts in a non-linear manner, the system will trigger an alarm and recommend checking the corresponding physical components.

[0100] The automated control method proposed in this application establishes a physical reference coordinate system by introducing multi-dimensional physical sensor data and comparing it in real time with the visual positioning data captured by the machine vision system. This allows for accurate identification and quantification of the deviation between the two. More importantly, the method further intelligently determines whether the deviation is a systematic offset generated by the machine vision system itself, rather than simple random noise, based on the persistence and stability of the deviation. Once a systematic offset is confirmed, the system can adaptively adjust the coordinate transformation parameters within the control system, using the high-precision physical reference coordinate system as a benchmark, thereby correcting the transformation relationship from the machine vision system to the motion platform coordinate system. This series of steps forms a closed-loop adaptive control mechanism, effectively solving the problem of accumulated positioning errors caused by lens optical performance degradation and loose installation due to long-term operation of the machine vision system in traditional methods, and avoiding the negative impact of blind compensation.

[0101] In one embodiment of this application, the steps following the acquisition of multi-dimensional physical sensor data and the establishment of a physical reference coordinate system in real time based on the multi-dimensional physical sensor data include:

[0102] Establishing a pointing error distribution map for the rangefinder, the establishment of the pointing error distribution map for the rangefinder includes:

[0103] Drive the motion platform to carry a laser rangefinder to scan and measure multiple preset static reference points within the workspace;

[0104] Record the current position as fed back by the motion platform encoder;

[0105] The pointing deviation of the rangefinder is calculated based on the position fed back by the encoder of the motion platform, the installation position and direction of the laser rangefinder relative to the zero point of the platform, and the known coordinates of the static reference point.

[0106] The pointing deviation of the rangefinder at different locations is stored to obtain the pointing error distribution map of the rangefinder;

[0107] The laser rangefinder measurement data is pre-corrected using the rangefinder pointing error distribution map to fine-tune the physical reference coordinate system.

[0108] Specifically, establishing a rangefinder pointing error distribution map aims to systematically quantify and record the measurement deviation of the laser rangefinder throughout the entire workspace. Driving the motion platform to carry the laser rangefinder to scan and measure multiple pre-set static reference points within the workspace involves precisely controlling the motion platform to move to multiple known positions within the workspace and using the laser rangefinder at each position to measure the pre-placed static reference points with known coordinates. These static reference points are typically high-precision standard parts with precisely calibrated coordinates. Recording the current position fed back by the motion platform encoder means synchronously acquiring the precise position information of the motion platform itself during each measurement. This position information is provided by the encoder inside the motion platform, ensuring the accuracy of the measurement position. Calculating the rangefinder pointing error based on the position fed back by the motion platform encoder, the installation position and direction of the laser rangefinder relative to the platform's zero point, and the known coordinates of the static reference points involves comparing the laser rangefinder's measurement results with theoretical values ​​calculated based on the motion platform's position and the known coordinates of the static reference points to determine the measurement error at a specific position and direction. The installation position and direction of the laser rangefinder relative to the platform's zero point are pre-calibrated fixed parameters. Storing the pointing deviations of the rangefinder at different locations to obtain the pointing error distribution map means associating these calculated deviation data with the corresponding measurement positions to form a comprehensive error database or model, which is the pointing error distribution map. This distribution map can be a three-dimensional lookup table, a fitting function, or a machine learning model, used to describe the pointing error of the rangefinder at different spatial locations. Using the pointing error distribution map to pre-correct the laser rangefinder measurement data to fine-tune the physical reference coordinate system means that in subsequent actual measurements, based on the current measurement position of the laser rangefinder, the corresponding pointing deviation is found or calculated from the error distribution map and applied back to the original measurement data, thereby eliminating or reducing the systematic error of the rangefinder itself, making the physical reference coordinate system established based on the corrected measurement data more accurate.

[0109] Through the above technical solution, this application can significantly improve the accuracy of establishing the physical reference coordinate system during the precision manufacturing process of lasers. Because the pointing error of the laser rangefinder is systematically quantified and pre-corrected, the physical reference coordinate system established based on physical sensor data is more accurate and reliable, effectively avoiding coordinate system inaccuracies caused by sensor errors. This not only improves the accuracy of comparing visual positioning data with the physical reference coordinate system, making the judgment of systematic offsets more precise, but also provides a more robust benchmark for the subsequent adaptive adjustment of coordinate transformation parameters within the control system. This ensures high precision and high stability of the entire automated control system during the precision manufacturing process of lasers, reducing the manufacturing defect rate caused by measurement errors.

[0110] In one embodiment of this application, the step of issuing maintenance recommendation information when the change characteristics of the coordinate transformation parameters are abnormal can be further refined.

[0111] The steps described above for monitoring the change characteristics of the coordinate transformation parameters and issuing maintenance recommendation information when the change characteristics are abnormal include:

[0112] Monitor the variation characteristics of the coordinate transformation parameters;

[0113] When the change characteristics are abnormal, a preset physical cause is associated with the change characteristics;

[0114] Maintenance recommendations are issued based on the physical reasons stated.

[0115] Monitoring the changes in the coordinate transformation parameters refers to continuously tracking and recording the changes in these parameters within the control system over time. These parameters may include elements of transformation matrices such as translation, rotation, and scaling, and their changes can exhibit trends, periodicity, abrupt changes, or fluctuations. Monitoring can be performed through real-time data acquisition, historical data analysis, or statistical methods.

[0116] When the aforementioned change characteristics are abnormal, the system associates these characteristics with a pre-defined physical cause. This means that when the monitored changes in coordinate transformation parameters deviate from the normal range or pattern, the system attempts to identify the root physical cause of this anomaly. This typically involves comparing the current abnormal characteristics with a pre-established knowledge base or fingerprint database, which stores various known physical fault modes and their corresponding coordinate transformation parameter change characteristics. For example, wear on a certain axis may cause a continuous drift in specific translation parameters, while contamination of optical components may cause periodic fluctuations in rotational parameters. Through this association, meaningful fault information can be extracted from complex parameter changes.

[0117] Issuing maintenance recommendations based on the physical causes means that after identifying a specific physical cause, the system generates and sends corresponding maintenance suggestions. These suggestions can be specific maintenance operation instructions, such as checking and cleaning the machine vision system lens, calibrating the X-axis of the motion platform, or replacing the laser collimation module, or higher-level warning messages that alert operators to specific components or areas. Maintenance recommendations can be issued in various forms, such as through human-machine interfaces, email or SMS notifications, or integrated into the factory's maintenance management system.

[0118] The above technical solutions enable early warning and accurate diagnosis of equipment failures during the precision manufacturing process of lasers. By correlating abnormal changes in coordinate transformation parameters with specific physical causes, the system can provide highly targeted maintenance suggestions, thereby effectively shortening troubleshooting time, reducing maintenance costs, and minimizing production downtime caused by equipment failures.

[0119] In one embodiment of this application, regarding the monitoring of the change characteristics of the coordinate transformation parameters, when the change characteristics are abnormal, the step of associating a preset physical cause based on the change characteristics includes:

[0120] Extract the variation characteristics of the coordinate transformation parameters;

[0121] The similarity between the change features and the set of typical coordinate transformation parameter change features in the preset physical cause feature fingerprint database is calculated. The physical cause feature fingerprint database stores the set of typical coordinate transformation parameter change features corresponding to known physical causes. The set of typical coordinate transformation parameter change features contains feature descriptions in multiple dimensions.

[0122] Filter out multiple physical reasons with similarity exceeding a preset threshold;

[0123] Based on the current environmental sensor data of the production line and historical maintenance records, a confidence level assessment is performed on the multiple physical causes.

[0124] The physical cause with the highest confidence level was selected to associate the change feature.

[0125] Specifically, extracting the variation features of the coordinate transformation parameters refers to identifying and quantifying specific patterns, trends, or values ​​that characterize their abnormal states from the monitored coordinate transformation parameter data. For example, the fluctuation amplitude, rate of change, periodicity, and degree of deviation from the normal range of the parameters can be extracted. These features can be extracted using statistical methods, signal processing techniques, or machine learning algorithms. The aim is to transform the raw parameter variation data into a structured and comparable feature representation.

[0126] The pre-defined physical cause feature fingerprint database can be understood as a knowledge base or database, storing sets of typical coordinate transformation parameter change features corresponding to various known physical fault causes (e.g., sensor aging, mechanical wear, ambient temperature changes, power fluctuations, etc.). Each typical coordinate transformation parameter change feature set contains multi-dimensional feature descriptions; for example, one physical cause may cause the parameter to exhibit periodic fluctuations, while another physical cause may cause the parameter to drift slowly. The extracted change features are compared with the typical features in this fingerprint database to preliminarily identify the potential physical cause that best matches the current anomaly. Similarity calculation can employ various mathematical methods such as Euclidean distance, cosine similarity, and correlation coefficient.

[0127] In practical applications, filtering out multiple physical causes with similarity exceeding a preset threshold means that after the initial similarity calculation, instead of selecting only the most similar cause, all potential causes with a certain level of similarity are retained. This is because in complex systems, a single anomalous feature may correspond to multiple physical causes, or multiple physical causes may collectively lead to the current anomaly. The preset threshold needs to be adjusted based on practical experience and system sensitivity, with the aim of avoiding overlooking potential root causes of failures.

[0128] Furthermore, by combining current production line environmental sensor data and historical maintenance records, a confidence assessment is conducted on the multiple physical causes. This involves introducing more contextual information for a comprehensive judgment after initially identifying several potential physical causes. Environmental sensor data (such as temperature, humidity, vibration, and air pressure) provides information on the external conditions of current equipment operation, while historical maintenance records contain valuable information such as the time, type, maintenance measures, and effects of past failures. By analyzing the correlation between these auxiliary data and each potential physical cause, a confidence score can be assigned to each cause, aiming to improve the accuracy of the final diagnosis.

[0129] Therefore, selecting the physical cause with the highest confidence level to associate with the aforementioned change characteristics means determining the most likely physical cause of the current coordinate transformation parameter anomaly after a comprehensive evaluation of multi-dimensional information. This process ensures the reliability of the diagnostic results and provides a solid foundation for subsequent maintenance recommendations.

[0130] Through the above technical solution, this application can significantly improve the accuracy and reliability of fault diagnosis in the automated control system during the precision manufacturing process of lasers. Compared with traditional methods that rely solely on a single changing feature for correlation, this solution introduces multi-dimensional feature extraction, fingerprint-based similarity matching, and a confidence assessment mechanism that combines environmental data and historical records. This enables a more comprehensive and in-depth analysis of abnormal phenomena, effectively distinguishing similar abnormal features caused by different physical reasons in complex and ever-changing production environments, thus avoiding misjudgments. Consequently, the maintenance recommendations issued by the system will more accurately pinpoint the root cause of the fault, significantly shortening troubleshooting time, reducing maintenance costs, and minimizing production losses caused by equipment downtime, thereby improving the stability and efficiency of the entire manufacturing process.

[0131] In one embodiment of this application, the step of assessing the confidence level of the multiple physical causes by combining current production line environmental sensor data and historical maintenance records includes:

[0132] Acquire current environmental sensor data and historical maintenance records;

[0133] For each of the plurality of physical causes, an indication intensity factor is generated based on the degree of matching between the physical cause and the environmental sensor data and the historical maintenance records;

[0134] Acquire equipment uptime, optical component characteristic data for the current production batch, and external operating condition sensor data;

[0135] The weight of the indication intensity factor is calculated based on the equipment's operating time, the optical component characteristic data of the current production batch, and the external operating condition sensor data.

[0136] The confidence level of the multiple physical causes is assessed by weighting and fusing the indicator intensity factor with its corresponding weight.

[0137] Specifically, acquiring current environmental sensor data and historical maintenance records refers to the system collecting environmental parameters such as temperature, humidity, vibration, and dust on the production line in real time, and retrieving historical maintenance records related to the current equipment or production batch from the database, such as the time of failure, maintenance content, and replaced parts. This data provides basic information for the preliminary assessment of physical causes. For each of the multiple physical causes, generating an indication strength factor based on the degree of matching between the physical cause and the environmental sensor data and historical maintenance records can be understood as the system analyzing the correlation between each selected potential physical cause and the current environmental data and historical maintenance records. For example, if a physical cause (such as optical component contamination) highly matches the high dust concentration data in the current environment and the frequent optical component cleaning records in historical maintenance records, a high indication strength factor will be generated for that physical cause. This factor quantifies the degree of support of existing evidence for a specific physical cause. In practical applications, acquiring equipment operating time, optical component characteristic data of the current production batch, and external operating condition sensor data refers to the system further collecting deeper levels of operating status and condition information. Equipment uptime reflects wear and aging; optical component characteristic data from the current production batch reveals batch-to-batch differences or potential defects; external sensor data provides broader information on external influencing factors, such as power supply stability and cooling system status. This data is used to more comprehensively assess the reliability of the indication intensity factor. Furthermore, calculating the weight of the indication intensity factor based on equipment uptime, optical component characteristic data from the current production batch, and external sensor data means that the system assigns different weights to previously generated indication intensity factors based on this supplementary data. For example, if the equipment uptime is too long, indicating that the equipment may be in the later stages of wear, the indication intensity factor for wear-related physical causes may be assigned a higher weight; if there are anomalies in the optical component characteristic data, the weight of the indication intensity factor for optical component-related physical causes will increase accordingly. The weight calculation can be based on preset rules, machine learning models, or expert experience. Therefore, weighted fusion of the indication intensity factor and its corresponding weight to assess the confidence of multiple physical causes means multiplying the indication intensity factor of each physical cause by its calculated weight and summing the results to obtain the final confidence score for that physical cause. This weighted fusion approach can more accurately reflect the contribution of different sources of evidence to the determination of physical causes, thus obtaining a more reliable confidence assessment result.

[0138] This application's solution addresses the issues of uneven data source weighting and insufficient consideration of reliability differences in traditional confidence assessments by introducing an indicator strength factor and its weight calculation. This hierarchical, weighted approach makes the final confidence assessment results more comprehensive, objective, and accurate.

[0139] In one embodiment of this application, the method for calculating the weight of the indicator intensity factor includes the following steps:

[0140] The running time of the device is checked for continuity. When the running time is interrupted, the most recent continuous running time is used for smooth interpolation, and the interpolated running time is marked with a confidence level.

[0141] The integrity of the optical component characteristic data of the current production batch is checked. When key characteristic parameters in the optical component characteristic data are missing, they are filled in according to the average characteristic value of the optical components in the same batch, and the confidence level of the filled optical component characteristic data is marked.

[0142] Real-time monitoring of external operating condition sensor data is performed. When the update delay of the external operating condition sensor data exceeds a preset threshold, the short-term predicted value of historical operating condition data is used as a substitute, and the confidence level of the replaced external operating condition sensor data is marked.

[0143] Based on the confidence markers of the running time, the optical element characteristic data, and the external operating condition sensor data, the contribution ratio of each to the weight of the indication intensity factor is dynamically adjusted.

[0144] The weight of the indicator intensity factor is calculated based on the adjusted contribution ratio.

[0145] Specifically, continuous monitoring of equipment runtime is performed to ensure the integrity and reliability of runtime data. When an interruption in runtime data is detected, to avoid data loss affecting subsequent calculations, smooth interpolation methods, such as linear interpolation, spline interpolation, or moving average interpolation, can be used to estimate and fill the gaps based on the most recent continuous runtime data before the interruption. The interpolated runtime data is assigned a confidence level, which reflects the reliability of the interpolated data. For example, the longer the interpolation period or the higher the uncertainty of the interpolation method, the lower the confidence level.

[0146] The integrity verification of optical component characteristic data for the current production batch is to ensure that the optical component characteristic parameters used for weighting calculations are comprehensive and accurate. When a key characteristic parameter is found to be missing, it can be filled by the average characteristic value of other optical components in the same production batch to compensate for the data gap. The filled optical component characteristic data will also be assigned a confidence level label. For example, if there are many missing key parameters or if the filled value may deviate significantly from the actual value, the confidence level label will be lowered accordingly.

[0147] In practical applications, real-time monitoring of external operating condition sensor data is crucial to ensuring the timeliness of this data. When the sensor data update delay exceeds a preset threshold, it indicates that the current data may no longer reflect the real-time operating conditions. In this case, short-term predictions from historical operating condition data can be used as a substitute, such as predictions based on time series analysis (e.g., ARIMA models) or machine learning models. The replaced external operating condition sensor data will also be assigned a confidence level; for example, the longer the prediction time window or the greater the uncertainty of the prediction model, the lower the confidence level.

[0148] Furthermore, based on the confidence markers for the running time, optical component characteristic data, and external operating condition sensor data, the system dynamically adjusts the contribution ratio of each to the weight calculation of the indication intensity factor. For example, if a data source has a low confidence marker, its contribution ratio in the weight calculation will be appropriately reduced to minimize its potential negative impact; conversely, high-confidence data sources will receive a higher contribution ratio. Thus, based on the adjusted contribution ratios, the final weight of the indication intensity factor is calculated.

[0149] This application's solution effectively addresses the issue of potential interruptions, missing data, or delays in the original data by rigorously managing data on equipment uptime, optical component characteristics, and external sensor data. This includes continuity checks, integrity verification, and real-time monitoring, and by assigning confidence levels to the processed data. This preprocessing and confidence assessment of data quality allows for dynamic adjustment of the contribution ratio of each data source based on data reliability when calculating the weights of the indicator strength factor. This mechanism ensures that even with some data of poor quality, the system can still calculate weights based on more reliable information, avoiding weight distortion caused by data defects. This provides a more solid and accurate data foundation for subsequent physical cause confidence assessments.

[0150] In one embodiment of this application, the step of dynamically adjusting the contribution ratio of each of the confidence markers based on the running time, the optical element characteristic data, and the external operating condition sensor data to the weight of the indication intensity factor includes:

[0151] Obtain the equipment's cumulative operating time equivalent and equipment lifecycle segmentation rules;

[0152] The life cycle stage of the equipment is determined based on the cumulative operating time equivalent of the equipment and the equipment life cycle division rules.

[0153] The basic contribution ratio adjustment rules are invoked based on the stage of the device's lifecycle.

[0154] When a specific fault mode is identified, the basic contribution ratio adjustment rule is locally modified according to the characteristics of the fault mode.

[0155] Under the revised contribution ratio adjustment rule, the contribution ratio of each of the following to the weight of the indication intensity factor is dynamically adjusted according to the confidence marker of the running time, the confidence marker of the optical element characteristic data, and the confidence marker of the external operating condition sensor data.

[0156] Specifically, the cumulative operating time equivalent refers to the equivalent operating time obtained by weighting or converting the actual operating time of the equipment after considering factors such as the actual operating intensity and environmental conditions. Its purpose is to more accurately reflect the actual wear and aging of the equipment. Equipment lifecycle segmentation rules can be understood as pre-defined standards or models that divide the entire lifecycle of equipment from commissioning to scrapping into several stages (e.g., break-in period, stabilization period, decline period, etc.). Their purpose is to provide guidance for the operating characteristics of equipment at different stages. Determining the equipment's current lifecycle stage means judging the specific lifecycle stage of the equipment based on the calculated cumulative operating time equivalent and the pre-defined equipment lifecycle segmentation rules. Its purpose is to provide macro-level contextual information for subsequent weight adjustments.

[0157] The basic contribution ratio adjustment rule can be understood as a default adjustment strategy for the contribution ratio of the indicator intensity factor to the equipment's operating time, optical component characteristic data, and external operating condition sensor data at different stages of the equipment's life cycle. Its purpose is to provide an initial weight allocation scheme based on the general patterns of the equipment. In practical applications, when a specific fault mode is identified, the basic contribution ratio adjustment rule is locally modified according to the characteristics of the fault mode. This means that when the system detects a specific fault mode in the equipment through other monitoring methods (e.g., abnormal operating conditions, high-intensity operation events, etc.), the basic contribution ratio adjustment rule corresponding to the current life cycle stage is fine-tuned or overridden based on the specific manifestation and impact of the fault mode. The purpose is to enable the weight adjustment to respond more accurately to specific abnormal situations and avoid misjudgments caused by general rules.

[0158] Through the above technical solution, this application overcomes the limitations of relying solely on confidence level markers for weight adjustment, significantly improving the accuracy and adaptability of the indicator intensity factor weight calculation. By incorporating considerations of equipment lifecycle stages and specific failure modes, the adjustment of contribution ratios becomes more refined and intelligent. For example, when equipment is in an aging phase, the system can automatically increase its focus on runtime data and external operating condition data; when a specific vibration failure mode is detected, the rules can be locally modified to respond more sensitively to relevant sensor data. Therefore, maintenance recommendations will more accurately reflect the actual health status and potential risks of the equipment, avoiding false alarms or missed alarms caused by improper weight allocation, thereby effectively extending equipment life, reducing maintenance costs, and ensuring the stability and efficiency of the laser precision manufacturing process.

[0159] In one embodiment of this application, the step of determining the lifecycle stage of the device based on the equivalent of the device's cumulative operating time and the device lifecycle division rules includes:

[0160] Continuously acquire device operation mode data and identify high-intensity operation events in the device operation mode data;

[0161] Identify abnormal operating conditions in the environmental sensor data;

[0162] The comprehensive degradation index of the equipment is calculated based on the frequency and duration of the abnormal operating conditions, as well as the cumulative duration and intensity of the high-intensity operation events.

[0163] Obtain the cumulative operating time of the equipment;

[0164] Calculate the equipment's cumulative operating time equivalent based on the equipment's comprehensive degradation index and cumulative operating time.

[0165] The life cycle stage of the equipment is determined based on the cumulative operating time equivalent of the equipment and the equipment life cycle division rules.

[0166] Specifically, equipment operation mode data refers to data recording various operating parameters and states of the equipment during operation, such as load, speed, work cycle, and processing task type. Analyzing this data can identify high-intensity operating events, such as prolonged high-load operation, frequent start-stop cycles, high-speed movement, or operation under extreme conditions. These events typically cause equipment components to experience greater stress, accelerating wear and aging. Identifying these events can be achieved through setting thresholds, pattern matching, or machine learning algorithms.

[0167] Environmental sensor data refers to various physical quantities acquired from the operating environment of equipment, such as temperature, humidity, vibration, and dust concentration. By monitoring this data, abnormal operating conditions can be identified, such as excessively high or low ambient temperatures, abnormal humidity, severe vibration, or the presence of corrosive gases. These abnormal conditions can negatively impact equipment performance and lifespan. Identification of abnormal operating conditions can be achieved by comparing data with preset safety ranges or through anomaly detection algorithms.

[0168] The Equipment Comprehensive Degradation Index (ECDI) is a quantifiable indicator of the overall health and wear level of equipment. Its calculation is based on the frequency and duration of abnormal operating events, as well as the cumulative duration and intensity of high-intensity operational events. For example, the more frequent and longer the abnormal operating events, the higher the degree of equipment degradation is likely to be; the longer the cumulative duration and the greater the intensity of high-intensity operational events, the more severe the wear and tear on the equipment is likely to be. This index aims to comprehensively reflect the cumulative damage to equipment under various adverse conditions.

[0169] The cumulative operating time of equipment refers to the total operating time of the equipment since it was put into use.

[0170] The equipment cumulative operating time equivalent is a modified operating time index that goes beyond simple physical operating time. It considers the actual wear and tear effects of equipment under different operating conditions and intensities. By combining the overall equipment degradation index with the equipment cumulative operating time, the actual equivalent wear and tear experienced by the equipment over a specific period can be assessed more accurately. For example, operating for one hour under harsh conditions or high-intensity operation may result in wear and tear equivalent to operating for several hours under normal conditions.

[0171] Equipment lifecycle segmentation rules refer to pre-defined criteria used to divide the entire lifecycle of equipment into different stages (e.g., initial stage, stable stage, decline stage, and failure stage). These rules are typically based on the equipment's cumulative operating time equivalent, performance indicators, historical failure data, etc. By comparing the calculated cumulative operating time equivalent of the equipment with these rules, the current lifecycle stage of the equipment can be determined.

[0172] Through the above technical solution, this application overcomes the limitations of traditional methods in assessing equipment degradation, significantly improving the accuracy of equipment lifecycle stage determination. By fully considering the cumulative impact of high-intensity operational events and abnormal operating conditions on equipment wear, the calculated cumulative operating time equivalent more accurately reflects the actual health status of the equipment. This makes maintenance recommendations more precise and timely, helping to take preventative measures before the equipment enters a critical decline phase, thereby effectively extending equipment lifespan, reducing the risk of unplanned downtime, optimizing maintenance resource allocation, and ultimately improving the overall stability and production efficiency of the laser precision manufacturing process.

[0173] In one embodiment of this application, the step of calculating the comprehensive degradation index of equipment based on the frequency and duration of abnormal operating condition events, and the cumulative duration and intensity of high-intensity operating events includes:

[0174] Obtain the frequency and duration of abnormal operating conditions;

[0175] The equipment anomaly index is calculated based on the frequency and duration of the abnormal operating events.

[0176] Obtain the cumulative duration and intensity of high-intensity operation events;

[0177] The equipment strength index is calculated based on the cumulative duration and intensity of the high-intensity operation events.

[0178] The equipment anomaly index and the equipment strength index are weighted and fused to obtain the comprehensive equipment degradation index.

[0179] Specifically, obtaining the frequency and duration of abnormal operating events refers to extracting event information related to abnormal equipment operation from environmental sensor data and equipment operation logs. For example, it is possible to count the number of times (frequency) the equipment exceeds normal temperature, humidity, and vibration thresholds within a specific time period, as well as the duration of each exceedance. This data forms the basis for assessing abnormal operating conditions of the equipment.

[0180] The equipment anomaly index, calculated based on the frequency and duration of the abnormal operating events, can be understood as reflecting the abnormal state of the equipment by quantifying the severity and scope of impact of these events. For example, a weighted average, exponential decay, or rule-based scoring system can be used for calculation. Higher frequency and longer duration result in a higher equipment anomaly index, indicating greater abnormal pressure on the equipment.

[0181] In practical applications, obtaining the cumulative duration and intensity of high-intensity operational events refers to identifying and quantifying operations that significantly impact equipment wear or fatigue from equipment operation mode data. For example, this involves recording the total time (cumulative duration) the equipment operates in high-speed, high-load, or high-precision modes, as well as the average or peak intensity of these operating modes. This data is a key factor in assessing whether equipment is operating normally but experiencing accelerated degradation.

[0182] Furthermore, an equipment strength index is calculated based on the cumulative duration and intensity of the high-intensity operating events. The purpose of this calculation is to quantify the degree of wear and tear on the equipment under normal operating conditions but high load. For example, this can be achieved by multiplying the cumulative duration by the operating intensity, or by defining weights for different intensity levels. The greater the cumulative duration and intensity of the high-intensity operating events, the higher the equipment strength index, indicating greater accumulated wear and tear on the equipment during normal use.

[0183] Finally, the equipment anomaly index and the equipment strength index are weighted and fused to obtain the comprehensive equipment degradation index. Weighted fusion is a common method for combining multiple indicators into a single indicator. For example, the weights of the anomaly index and the strength index can be determined based on practical experience or through machine learning models to reflect their relative importance to the overall equipment degradation. In this way, a comprehensive index reflecting the current degradation state of the equipment can be obtained.

[0184] The above technical solution provides a more refined and comprehensive method for calculating the overall equipment degradation index. This method distinguishes the different impacts of abnormal operating conditions and high-intensity operational events on equipment degradation, and calculates the equipment abnormality index and equipment intensity index separately, avoiding the potential issues of index confusion or weight imbalance in traditional methods. Furthermore, by weighted fusion of these two indices, the final overall equipment degradation index more accurately reflects the actual degradation state of the equipment, thereby significantly improving the accuracy and reliability of equipment lifecycle stage assessment and providing a more solid data foundation for subsequent maintenance recommendations and production decisions.

[0185] See Figure 2 , Figure 2 This is a schematic diagram of an automated control system in a laser precision manufacturing process according to one embodiment of this application. The automated control system 1000 in the laser precision manufacturing process includes:

[0186] The physical reference coordinate system establishment module 1010 is used to acquire multi-dimensional physical sensor data and establish a physical reference coordinate system in real time based on the multi-dimensional physical sensor data.

[0187] The visual positioning data acquisition module 1020 is used to acquire images captured by the machine vision system and perform feature point localization calculations on the images to obtain visual positioning data.

[0188] The deviation comparison module 1030 is used to compare the visual positioning data with the physical reference coordinate system in real time to obtain the deviation between the two.

[0189] The systematic offset judgment module 1040 is used to determine whether the deviation is a systematic offset generated by the machine vision system based on the persistence and stability of the deviation.

[0190] The coordinate transformation parameter adjustment module 1050 is used to adaptively adjust the coordinate transformation parameters inside the control system based on the physical reference coordinate system when the deviation is a systematic offset generated by the machine vision system, thereby correcting the transformation relationship between the machine vision system and the motion platform coordinate system.

[0191] The information sending module 1060 is used to monitor the change characteristics of the coordinate transformation parameters, and when the change characteristics are abnormal, it sends maintenance suggestion information.

[0192] Specifically, the physical reference coordinate system establishment module can be configured as a standalone hardware unit, such as an embedded controller, which integrates sensor interfaces and data processing units. It is specifically responsible for acquiring data from physical sensors such as laser rangefinders and encoders, and running pre-defined algorithms to construct and maintain the physical reference coordinate system in real time. Alternatively, the module can be a software service running on the central controller's operating system. It interacts with the physical sensor array via standard communication protocols and utilizes multi-threaded or distributed computing frameworks to ensure the real-time performance and high accuracy of the physical reference coordinate system.

[0193] The visual positioning data acquisition module can consist of an industrial camera, an image acquisition card, and an image processing unit. The image processing unit can be a high-performance graphics processing unit (GPU) or a field-programmable gate array (FPGA), specifically designed to accelerate image preprocessing, feature point detection, and coordinate calculation. This module can be a cloud-based vision service; images captured by the industrial camera are uploaded to a cloud server, where powerful computing resources perform feature point localization calculations, and the visual positioning data is then transmitted back to the local control system.

[0194] The deviation comparison module can be a data fusion processor that receives physical reference coordinate system data from the physical reference coordinate system establishment module and visual positioning data from the visual positioning data acquisition module. Internally, this processor runs coordinate transformation and deviation calculation algorithms, enabling it to convert the visual positioning data to the physical reference coordinate system in real time and calculate the spatial deviation vector between the two. Alternatively, this module can employ a dedicated digital signal processor (DSP) to ensure high-speed, high-precision real-time comparison calculations.

[0195] The systematic shift judgment module can be an intelligent analysis unit with built-in statistical analysis models and time series analysis algorithms. This unit continuously receives deviation data output from the deviation comparison module and monitors and analyzes it in real time, such as calculating the moving average, standard deviation, and trend line of the deviation to determine whether the deviation is persistent and stable, thereby distinguishing systematic shift from random noise. Alternatively, this module can be a fuzzy logic-based judge that, based on preset fuzzy rules, integrates multiple fuzzy inputs such as the magnitude, duration, and rate of change of the deviation to output a confidence level for systematic shift.

[0196] The coordinate transformation parameter adjustment module can be a parameter optimization controller that receives the judgment results from the systematic offset judgment module and the deviation data from the deviation comparison module. Internally, this controller runs an adaptive control algorithm, such as an iterative optimization algorithm based on the least squares method, which continuously adjusts the coordinate transformation parameters (such as the rotation matrix and translation vector) to minimize the deviation between the visual positioning data and the physical reference coordinate system after transformation.

[0197] The information transmission module can be a diagnostic and early warning unit, which continuously receives parameter values ​​output by the coordinate transformation parameter adjustment module and performs historical and trend analysis on them. This unit has a built-in anomaly detection algorithm, such as a statistical threshold-based or machine learning-based anomaly detection model. When the rate, magnitude, or pattern of parameter change is detected to be outside the normal range, an alarm is immediately triggered and maintenance suggestion information is generated.

[0198] The automated control system of this application provides a physical reference independent of the vision system through a physical reference coordinate system establishment module, and a deviation comparison module compares the visual positioning data with this physical reference in real time. Furthermore, the systematic offset judgment module can intelligently distinguish between systematic offsets and random noise, avoiding misjudgments of normal fluctuations. Once a systematic offset is confirmed, the coordinate transformation parameter adjustment module can adaptively adjust the coordinate transformation parameters within the control system based on the physical reference coordinate system, thereby accurately correcting the transformation relationship from the machine vision system to the motion platform coordinate system. This adaptive correction mechanism based on modular design and intelligent judgment fundamentally solves the error accumulation problem caused by blind compensation in traditional methods, ensuring that the correct state within the control system always remains consistent with the correct state of the actual physical world. In addition, the information transmission module, by monitoring the changing characteristics of the coordinate transformation parameters and issuing maintenance suggestions, also provides preventative maintenance capabilities, enabling early warning of potential equipment failures, thereby significantly improving the accuracy, stability, and production efficiency of the laser precision manufacturing process and reducing product scrap rates.

Claims

1. An automated control method for the precision manufacturing process of lasers, characterized in that, Includes the following steps: Acquire multi-dimensional physical sensor data, and establish a physical reference coordinate system in real time based on the multi-dimensional physical sensor data; The machine vision system captures images and performs feature point localization calculations on the images to obtain visual localization data. The visual positioning data is compared with the physical reference coordinate system in real time to obtain the deviation between the two. Based on the persistence and stability of the deviation, determine whether the deviation is a systematic offset generated by the machine vision system; When the deviation is a systematic offset generated by the machine vision system, the coordinate transformation parameters inside the control system are adaptively adjusted based on the physical reference coordinate system to correct the transformation relationship between the machine vision system and the motion platform coordinate system. Monitor the change characteristics of the coordinate transformation parameters, and issue maintenance suggestion information when the change characteristics are abnormal; The steps following the acquisition of multi-dimensional physical sensor data and the establishment of a physical reference coordinate system in real time based on the multi-dimensional physical sensor data include: Establishing a pointing error distribution map for the rangefinder, the establishment of the pointing error distribution map for the rangefinder includes: Drive the motion platform to carry a laser rangefinder to scan and measure multiple preset static reference points within the workspace; Record the current position as fed back by the motion platform encoder; The pointing deviation of the rangefinder is calculated based on the position fed back by the encoder of the motion platform, the installation position and direction of the laser rangefinder relative to the zero point of the platform, and the known coordinates of the static reference point. The pointing deviation of the rangefinder at different locations is stored to obtain the pointing error distribution map of the rangefinder; The laser rangefinder measurement data is pre-corrected using the rangefinder pointing error distribution map to fine-tune the physical reference coordinate system; The step of monitoring the change characteristics of the coordinate transformation parameters and issuing maintenance suggestion information when the change characteristics are abnormal includes: Monitor the variation characteristics of the coordinate transformation parameters; When the change characteristics are abnormal, a preset physical cause is associated with the change characteristics; Based on the aforementioned physical reasons, maintenance recommendations are issued. The step of associating a preset physical cause with the change characteristics includes: Extract the variation characteristics of the coordinate transformation parameters; The similarity between the change features and the set of typical coordinate transformation parameter change features in the preset physical cause feature fingerprint database is calculated. The physical cause feature fingerprint database stores the set of typical coordinate transformation parameter change features corresponding to known physical causes. The set of typical coordinate transformation parameter change features contains feature descriptions in multiple dimensions. Filter out multiple physical reasons with similarity exceeding a preset threshold; Based on the current environmental sensor data of the production line and historical maintenance records, a confidence level assessment is performed on the multiple physical causes. Select the physical cause with the highest confidence level to associate with the change feature; The step of combining current production line environmental sensor data and historical maintenance records to assess the confidence level of the multiple physical causes includes: Acquire current environmental sensor data and historical maintenance records; For each of the plurality of physical causes, an indication intensity factor is generated based on the degree of matching between the physical cause and the environmental sensor data and the historical maintenance records; Acquire equipment uptime, optical component characteristic data for the current production batch, and external operating condition sensor data; The weight of the indication intensity factor is calculated based on the equipment's operating time, the optical component characteristic data of the current production batch, and the external operating condition sensor data. The confidence level of the multiple physical causes is assessed by weighting and fusing the indicator intensity factor with its corresponding weight.

2. The automated control method for the precision manufacturing process of a laser according to claim 1, characterized in that, The step of calculating the weight of the indication intensity factor based on the equipment's operating time, the optical component characteristic data of the current production batch, and external operating condition sensor data includes: The running time of the device is checked for continuity. When the running time is interrupted, the most recent continuous running time is used for smooth interpolation, and the interpolated running time is marked with a confidence level. The integrity of the optical component characteristic data of the current production batch is checked. When key characteristic parameters in the optical component characteristic data are missing, they are filled in according to the average characteristic value of the optical components in the same batch, and the confidence level of the filled optical component characteristic data is marked. Real-time monitoring of external operating condition sensor data is performed. When the update delay of the external operating condition sensor data exceeds a preset threshold, the short-term predicted value of historical operating condition data is used as a substitute, and the confidence level of the replaced external operating condition sensor data is marked. Based on the confidence markers of the running time, the optical element characteristic data, and the external operating condition sensor data, the contribution ratio of each to the weight of the indication intensity factor is dynamically adjusted. The weight of the indicator intensity factor is calculated based on the adjusted contribution ratio.

3. The automated control method for the precision manufacturing process of a laser according to claim 2, characterized in that, The step of dynamically adjusting the contribution ratio of each of the confidence markers for the running time, the optical element characteristic data, and the external operating condition sensor data to the weight of the indication intensity factor includes: Obtain the equipment's cumulative operating time equivalent and equipment lifecycle segmentation rules; The stage of the equipment's life cycle is determined based on the equipment's cumulative operating time equivalent and the equipment's life cycle division rules. The basic contribution ratio adjustment rules are invoked based on the stage of the device's lifecycle. When a specific fault mode is identified, the basic contribution ratio adjustment rule is locally modified according to the characteristics of the fault mode. Under the revised contribution ratio adjustment rule, the contribution ratio of each of the following to the weight of the indication intensity factor is dynamically adjusted according to the confidence marker of the running time, the confidence marker of the optical element characteristic data, and the confidence marker of the external operating condition sensor data.

4. The automated control method for the precision manufacturing process of a laser according to claim 3, characterized in that, The step of determining the lifecycle stage of the equipment based on the cumulative operating time equivalent and the equipment lifecycle division rules includes: Continuously acquire device operation mode data and identify high-intensity operation events in the device operation mode data; Identify abnormal operating conditions in the environmental sensor data; The comprehensive degradation index of the equipment is calculated based on the frequency and duration of the abnormal operating conditions, as well as the cumulative duration and intensity of the high-intensity operation events. Obtain the cumulative operating time of the equipment; Calculate the equipment's cumulative operating time equivalent based on the equipment's comprehensive degradation index and cumulative operating time. The life cycle stage of the equipment is determined based on the cumulative operating time equivalent of the equipment and the equipment life cycle division rules.

5. The automated control method for the precision manufacturing process of a laser according to claim 4, characterized in that, The step of calculating the comprehensive equipment degradation index based on the frequency and duration of the abnormal operating conditions and the cumulative duration and intensity of the high-intensity operation events includes: Obtain the frequency and duration of abnormal operating conditions; The equipment anomaly index is calculated based on the frequency and duration of the abnormal operating events. Obtain the cumulative duration and intensity of high-intensity operation events; The equipment strength index is calculated based on the cumulative duration and intensity of the high-intensity operation events. The equipment anomaly index and the equipment strength index are weighted and fused to obtain the comprehensive equipment degradation index.

6. An automated control system for the precision manufacturing process of a laser, characterized in that, The system includes: The physical reference coordinate system establishment module is used to acquire multi-dimensional physical sensor data and establish a physical reference coordinate system in real time based on the multi-dimensional physical sensor data. The visual positioning data acquisition module is used to acquire images captured by the machine vision system and perform feature point localization calculations on the images to obtain visual positioning data. The deviation comparison module is used to compare the visual positioning data with the physical reference coordinate system in real time to obtain the deviation between the two. The systematic offset judgment module is used to determine whether the deviation is a systematic offset generated by the machine vision system based on the persistence and stability of the deviation. The coordinate transformation parameter adjustment module is used to adaptively adjust the coordinate transformation parameters inside the control system based on the physical reference coordinate system when the deviation is a systematic offset generated by the machine vision system, thereby correcting the transformation relationship between the machine vision system and the motion platform coordinate system. The information sending module is used to monitor the change characteristics of the coordinate transformation parameters, and when the change characteristics are abnormal, it sends maintenance suggestion information. The monitoring of the changes in the coordinate transformation parameters, and the issuance of maintenance recommendations when the changes are abnormal, include: Monitor the variation characteristics of the coordinate transformation parameters; When the change characteristics are abnormal, a preset physical cause is associated with the change characteristics; Based on the aforementioned physical reasons, maintenance recommendations are issued. The association of preset physical causes based on the aforementioned change characteristics includes: Extract the variation characteristics of the coordinate transformation parameters; The similarity between the change features and the set of typical coordinate transformation parameter change features in the preset physical cause feature fingerprint database is calculated. The physical cause feature fingerprint database stores the set of typical coordinate transformation parameter change features corresponding to known physical causes. The set of typical coordinate transformation parameter change features contains feature descriptions in multiple dimensions. Filter out multiple physical reasons with similarity exceeding a preset threshold; Based on the current environmental sensor data of the production line and historical maintenance records, a confidence level assessment is performed on the multiple physical causes. Select the physical cause with the highest confidence level to associate with the change feature; The confidence assessment of the multiple physical causes, combining current production line environmental sensor data and historical maintenance records, includes: Acquire current environmental sensor data and historical maintenance records; For each of the plurality of physical causes, an indication intensity factor is generated based on the degree of matching between the physical cause and the environmental sensor data and the historical maintenance records; Acquire equipment uptime, optical component characteristic data for the current production batch, and external operating condition sensor data; The weight of the indication intensity factor is calculated based on the equipment's operating time, the optical component characteristic data of the current production batch, and the external operating condition sensor data. The confidence level of the multiple physical causes is assessed by weighting and fusing the indicator intensity factor with its corresponding weight. After acquiring multi-dimensional physical sensor data and establishing a physical reference coordinate system in real time based on the multi-dimensional physical sensor data, the system further includes: Establishing a pointing error distribution map for the rangefinder, the establishment of the pointing error distribution map for the rangefinder includes: Drive the motion platform to carry a laser rangefinder to scan and measure multiple preset static reference points within the workspace; Record the current position as fed back by the motion platform encoder; The pointing deviation of the rangefinder is calculated based on the position fed back by the encoder of the motion platform, the installation position and direction of the laser rangefinder relative to the zero point of the platform, and the known coordinates of the static reference point. The pointing deviation of the rangefinder at different locations is stored to obtain the pointing error distribution map of the rangefinder; The laser rangefinder measurement data is pre-corrected using the rangefinder pointing error distribution map to fine-tune the physical reference coordinate system.

Images