Paster bearing, positioning and calibrating method and system, storage medium and robot

Abstract

The invention relates to the field of electronic manufacturing and intelligent control, and discloses a patch bearing positioning calibration method and system, a storage medium and a robot, environmental parameters are monitored in real time through a distributed sensor network, and data preprocessing and integration are performed by using edge computing equipment to generate comprehensive environmental state data; a nanoscale micro-motor driving system is utilized to realize high-precision motion control and real-time positioning calibration of the suction nozzle; path optimization is carried out through the quantum computing technology, the path is dynamically adjusted based on real-time positioning calibration data and environment state data, the response speed and precision of the chip mounting process are improved, AI continuous learning and anomaly detection are further included, and self-adaptive optimization and long-term stability of the system are achieved through real-time data collection and model updating and optimization; according to the invention, the performance of the existing patch calibration technology is obviously improved, efficient and accurate patch bearing, positioning and calibration are realized through advanced technical means and optimized data processing flow, and the production quality and efficiency are improved.

Description

Technical Field

[0001] The present invention relates to the fields of electronic manufacturing and intelligent control, and in particular to a patch bearing positioning calibration method and system, a storage medium, and a robot. Background Art

[0002] With the increasing miniaturization of electronic components and the continuous improvement of patch speed, patch machines have put forward higher requirements for the accuracy and speed of patch load positioning calibration. The object of patch is the strain gauge of the sensor. When the strain gauge is attached to the sensor, the patch effect will affect the parameters such as the accuracy of sensor weighing. The application field of the weighing device composed of this sensor is industrial weighing such as molten iron load, coking coal weighing, and weighing in high-speed rail, subway and urban rail transit related fields. Therefore, accurate patch load positioning calibration is crucial to improving patch efficiency and quality. The robot used for patch load positioning calibration needs to work in complex environmental conditions and must have the characteristics of high precision, high speed and high stability.

[0003] The current patch calibration technology (Chinese invention patent, publication number: CN113382555B) mainly relies on CCD cameras for Mark point information collection and mechanical zeroing process, but these methods have many problems. First, CCD cameras have high requirements for light and environment. Factors such as light changes and dust will affect the accuracy of Mark point information. Secondly, the installation deviation from the CCD camera to the nozzle will cause cumulative errors and affect the calculation accuracy of the product center position. In addition, the response speed and calculation processing time of the CCD camera will affect the overall placement efficiency and may become a bottleneck in high-speed production lines. The existing technology also lacks dynamic adjustment and adaptive learning capabilities, and can only perform static compensation. It cannot cope with complex and changing production environments. Mechanical wear and environmental changes during long-term use will also lead to a decrease in accuracy. Summary of the invention

[0004] In view of the many problems existing in the above-mentioned prior art, the present invention provides a patch-carrying positioning and calibration method and system, a storage medium, and a robot. The present invention deploys a high-density distributed sensor network to monitor environmental parameters in real time, uses edge computing equipment to preprocess and integrate data, and generates comprehensive environmental status data; then, a nano-scale micro-motor drive system is used to achieve precise positioning and calibration of the nozzle, and combines the real-time collected environmental data and AI model data to perform feedback control and dynamic path adjustment; quantum computing technology is used to optimize path planning, improve calculation speed and path optimization efficiency, and ensure the efficiency and accuracy of the patch process; through AI continuous learning and anomaly detection, the system's adaptive optimization and long-term stability are achieved.

[0005] A patch bearing positioning calibration method comprises the following steps: By collecting historical patch operation records and environmental data, data preprocessing is performed to train and optimize AI models, and the generated initial AI model data is verified and deployed; Deploy high-density distributed sensors in the work area to collect environmental data in real time, including temperature, humidity, vibration, air pressure, air quality, and light intensity. Use edge computing devices to pre-process and integrate data to generate comprehensive environmental status data. By initializing the nano-scale micro-motor drive system, the nozzle is positioned and calibrated in real time, and feedback adjustment is performed based on the comprehensive environmental status data and initial AI model data to generate real-time positioning calibration data; Utilize quantum computing technology for path planning and optimization, dynamically adjust the path based on real-time positioning calibration data and comprehensive environmental status data, and generate dynamically adjusted path data through real-time monitoring and feedback re-optimization; By collecting operational data in real time, performing data preprocessing, updating and training AI models, performing anomaly detection and early warning, verifying and optimizing models, and redeploying optimized AI model data. Preferably, the data preprocessing includes data cleaning, denoising, normalization and feature extraction of the collected historical patch operation records and environmental data to generate a training data set for eliminating noise and abnormal data.

[0006] Preferably, the AI ​​model training includes using a deep learning algorithm to train the preprocessed training data set, and adjusting the model parameters through cross-validation and hyperparameter optimization to generate optimized AI model data.

[0007] Preferably, the data preprocessing performed by the edge computing device includes smoothing, abnormal data filtering and data fusion of the real-time collected environmental data to generate comprehensive environmental status data, wherein the smoothing is calculated using the following formula:

[0008] in, represents the smoothed data; Represents raw environmental data; Indicates the smoothing window size; Indicates the current data index; Indicates the offset index within the window.

[0009] Preferably, the real-time positioning and calibration of the nanoscale micro-motor drive system comprises the following steps: Determine the initial position and angle of the nozzle by integrating environmental status data and initial AI model data; Adjust the position and angle of the nozzle based on the feedback control algorithm to generate real-time positioning calibration data; The feedback control amount is calculated by the following formula to generate real-time positioning calibration data:

[0010] in, represents the feedback control amount; Indicates the set position; Indicates the measurement location; represents proportional gain; represents the integral gain; represents the differential gain; Represents a time variable.

[0011] Preferably, the path planning and optimization comprises the following steps: Use quantum computing technology to optimize the initial path planning data and generate optimized path data; Use quantum algorithms to accelerate the calculation of paths and determine the optimal patch path; Use the following formula for path optimization calculation:

[0012] in, represents the optimized path cost; Indicates The cost function of each path; Indicates The distance of the path; Indicates the number of paths.

[0013] Preferably, the dynamic adjustment and feedback re-optimization comprises the following steps: Dynamically adjust the path based on real-time positioning calibration data and comprehensive environmental status data; Fine-tune and correct the path based on real-time monitoring data and feedback data; The path is fine-tuned using the following formula:

[0014] in, Indicates the adjustment amount of the path; Indicates real-time positioning calibration data; Represents comprehensive environmental status data; Represents the adjustment function.

[0015] A patch bearing positioning and calibration system, comprising: Historical data collection module, used to collect historical patch operation records and environmental data; The data preprocessing module is used to clean, denoise, normalize and extract features of the collected historical patch operation records and environmental data to generate a training data set for eliminating noise and abnormal data; The AI ​​model training module is used to train the preprocessed training data set using a deep learning algorithm, and adjust the model parameters through cross-validation and hyperparameter optimization to generate optimized AI model data; A sensor network module is used to deploy high-density distributed sensors in the working area to collect environmental data in real time, including temperature, humidity, vibration, air pressure, air quality and light intensity; Edge computing devices are used to smooth, filter out abnormal data, and fuse data collected in real time to generate comprehensive environmental status data; Nano-scale micro-motor drive system, which is used to determine the initial position and angle of the nozzle by integrating environmental status data and initial AI model data, adjust the position and angle of the nozzle based on the feedback control algorithm, and generate real-time positioning calibration data; The quantum computing module is used to optimize the initial path planning data using quantum computing technology, generate optimized path data, accelerate the path calculation through quantum algorithms, and determine the optimal patch path; The real-time monitoring and feedback module is used to collect operation data in real time, perform data preprocessing, update and train AI models, perform anomaly detection and early warning, verify and optimize models, and redeploy optimized AI model data.

[0016] A storage medium stores a computer program, which, when executed by a processor, implements the steps of the patch carrying positioning calibration method.

[0017] A patch carrying positioning and calibration robot, comprising: A control unit, used to control the overall operation of the robot; Historical data collection unit, used to collect historical patch operation records and environmental data; The data preprocessing unit is used to perform data cleaning, denoising, normalization and feature extraction on the collected historical patch operation records and environmental data to generate a training data set for eliminating noise and abnormal data; An AI model training unit is used to train the preprocessed training data set using a deep learning algorithm, and to adjust model parameters through cross-validation and hyperparameter optimization to generate optimized AI model data; A sensor network unit is used to deploy high-density distributed sensors in the working area to collect environmental data in real time, including temperature, humidity, vibration, air pressure, air quality and light intensity; Edge computing unit, used to smooth, filter abnormal data and fuse data collected in real time to generate comprehensive environmental status data; Nano-scale micro-motor drive unit, used to determine the initial position and angle of the nozzle by integrating environmental status data and initial AI model data, adjust the position and angle of the nozzle based on the feedback control algorithm, and generate real-time positioning calibration data; A quantum computing unit is used to optimize the initial path planning data using quantum computing technology, generate optimized path data, accelerate the path calculation using quantum algorithms, and determine the optimal patch path; The real-time monitoring and feedback unit is used to collect operation data in real time, perform data preprocessing, update and train AI models, perform anomaly detection and early warning, verify and optimize models, and redeploy optimized AI model data.

[0018] Compared with the prior art, the advantages and beneficial effects of the present invention are: The present invention realizes real-time monitoring and processing of environmental parameters through distributed sensor networks and edge computing technology, ensuring the real-time and accuracy of data; The present invention realizes high-precision motion control and real-time positioning calibration of the nozzle through a nano-scale micro-motor drive system, thereby reducing error accumulation; The present invention uses quantum computing technology to optimize the path, realizes efficient path planning and dynamic adjustment, and improves the response speed and efficiency of the patch process; The present invention realizes adaptive optimization and long-term stability of the system through AI continuous learning and anomaly detection, ensuring the efficiency and maintainability of the patch process. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 It is a schematic diagram of the process of the present invention; Figure 2 A schematic diagram of a nano-scale micro-motor driving system of the present invention; Figure 3 The quantum computing path planning flow chart of the present invention; Figure 4 It is a structural schematic block diagram of the system of the present invention; Figure 5 It is a schematic structural block diagram of the robot of the present invention. DETAILED DESCRIPTION

[0020] Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of the present disclosure. In the following detailed description, for ease of explanation, many specific details are set forth to provide a comprehensive understanding of the embodiments of the present disclosure. However, it is obvious that one or more embodiments may also be implemented without these specific details. In addition, in the following description, descriptions of known structures and technologies are omitted to avoid unnecessary confusion of the concepts of the present disclosure.

[0021] The terms used herein are only for describing specific embodiments and are not intended to limit the present disclosure. The terms "comprise", "include", etc. used herein indicate the existence of the features, steps, operations and / or components, but do not exclude the existence or addition of one or more other features, steps, operations or components.

[0022] All terms (including technical and scientific terms) used herein have the meanings commonly understood by those skilled in the art unless otherwise defined. It should be noted that the terms used herein should be interpreted as having a meaning consistent with the context of this specification and should not be interpreted in an idealized or overly rigid manner.

[0023] In the case of using expressions such as "at least one of A, B, and C, etc.", it should generally be interpreted in accordance with the meaning of the expression generally understood by those skilled in the art (for example, "a system having at least one of A, B, and C" should include but is not limited to a system having A alone, B alone, C alone, A and B, A and C, B and C, and / or A, B, C, etc.). In the case of using expressions such as "at least one of A, B, or C, etc.", it should generally be interpreted in accordance with the meaning of the expression generally understood by those skilled in the art (for example, "a system having at least one of A, B, or C" should include but is not limited to a system having A alone, B alone, C alone, A and B, A and C, B and C, and / or A, B, C, etc.).

[0024] Some block diagrams and / or flow charts are shown in the accompanying drawings. It should be understood that some boxes or combinations thereof in the block diagrams and / or flow charts can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that these instructions can create a device for implementing the functions / operations described in these block diagrams and / or flow charts when executed by the processor. The technology of the present disclosure can be implemented in the form of hardware and / or software (including firmware, microcode, etc.). In addition, the technology of the present disclosure can take the form of a computer program product on a computer-readable storage medium storing instructions, which can be used by an instruction execution system or in combination with an instruction execution system.

[0025] like Figure 1 As shown, a patch bearing positioning calibration method includes the following steps: By collecting historical patch operation records and environmental data, data preprocessing is performed to train and optimize AI models, and the generated initial AI model data is verified and deployed; The patch-bearing positioning calibration method realizes efficient data preprocessing, accurate training and optimization of AI models, and effective verification and deployment of models, thereby providing high-precision decision support in actual operations and significantly improving the quality and efficiency of patch operations.

[0026] Preferably, the data preprocessing includes data cleaning, denoising, normalization and feature extraction of the collected historical patch operation records and environmental data to generate a training data set for eliminating noise and abnormal data.

[0027] Data cleaning is the process of improving data quality by processing missing values, outliers, and duplicate data. In patch operation records and environmental data, sensor failure or human error may cause missing or anomalies in data. Data cleaning ensures the integrity and reliability of the data set by deleting or correcting these erroneous data.

[0028] Missing value processing includes: using interpolation, mean filling or deleting missing records to handle missing data.

[0029] Outlier processing includes: using statistical methods (such as box plots, Z-Score) to identify and delete outliers.

[0030] The present invention improves the integrity of the data set and reduces the model deviation caused by erroneous data.

[0031] In one embodiment, in a patch operation record, missing temperature data and abnormal vibration data due to sensor failure are found. The missing temperature data is supplemented by interpolation, and the abnormal vibration data is eliminated by using a box plot method.

[0032] Data denoising is to remove random noise from data through filtering and other methods to retain useful signals. Environmental data such as temperature and vibration may be affected by random noise, and denoising can extract more stable features.

[0033] Moving average filtering includes: smoothing data and eliminating short-term fluctuations.

[0034] Kalman filtering includes: filtering time series data based on a dynamic model to reduce noise interference.

[0035] The present invention eliminates random noise, improves the signal-to-noise ratio of data, and extracts more reliable features.

[0036] In one embodiment, moving average filtering is performed on the ambient temperature data to smooth the temperature change trend and eliminate short-term random fluctuations.

[0037] Data normalization is to convert data of different dimensions to the same scale range for model processing. The parameters such as temperature, humidity and vibration in environmental data have different dimensions and need to be normalized to eliminate the dimension differences.

[0038] Min-max scaling involves scaling the data to the range [0,1].

[0039] Z-Score standardization involves converting the data into a standard normal distribution with a mean of 0 and a standard deviation of 1.

[0040] The present invention eliminates dimensional differences and enables different data to be compared and processed on the same scale, which is convenient for model training.

[0041] In one embodiment, Z-Score standardization is performed on the vibration data to convert the vibration signal into a standard normal distribution with a mean of 0 and a standard deviation of 1, thereby unifying the data scale.

[0042] Feature extraction is to extract key features useful for model training from preprocessed data. The patch operation records and environmental data contain a large amount of raw data. Through feature extraction, the key factors affecting the quality of the patch can be found.

[0043] Time series feature extraction includes: extracting time features such as time intervals and operation sequences from operation records.

[0044] Statistical feature extraction includes: extracting statistical features such as mean value, variance, extreme value, etc. from environmental data.

[0045] Frequency domain feature extraction includes: performing Fourier transform on vibration data and extracting frequency domain features.

[0046] The present invention extracts features useful for model training, reduces data dimensions, and improves model training efficiency and prediction performance.

[0047] In one embodiment, the time interval feature of each operation step is extracted from the patch operation record; the vibration data is subjected to Fourier transform to extract the main frequency component as the frequency domain feature.

[0048] In summary, the present invention eliminates noise and abnormal data and improves the quality of training data through data cleaning, denoising and normalization. The feature extraction process helps to identify and retain key features related to the patch operation effect, and improves the model's ability to understand and interpret data. High-quality training data sets can significantly improve the training effect and prediction performance of AI models, ensuring efficiency and accuracy in practical applications.

[0049] In one embodiment, in a certain patch operation, the ambient temperature, humidity and equipment vibration have a significant impact on the patch quality. These influencing factors can be discovered through data preprocessing and feature extraction.

[0050] Implementation steps: Remove erroneous data due to sensor failure and handle missing values.

[0051] Apply Kalman filtering to the environmental data to remove random noise.

[0052] Perform min-max scaling on the environmental data and normalize it to the range [0,1].

[0053] Time series features are extracted from operation records, and statistical features and frequency domain features are extracted from environmental data.

[0054] Generate a high-quality training data set, including time series features, statistical features, and frequency domain features, to provide a reliable data foundation for the training of AI models.

[0055] The data preprocessing process not only eliminates noise and abnormal data, but also extracts key features, laying a solid foundation for subsequent AI model training, thereby improving the accuracy and efficiency of patch-bearing positioning calibration.

[0056] Preferably, the AI ​​model training includes using a deep learning algorithm to train the preprocessed training data set, and adjusting the model parameters through cross-validation and hyperparameter optimization to generate optimized AI model data.

[0057] The purpose of AI model training is to use deep learning algorithms to train the preprocessed training data set, adjust the model parameters through cross-validation and hyperparameter optimization, and generate optimized AI model data for patch-bearing positioning calibration.

[0058] Select an appropriate deep learning algorithm for model training. Common deep learning models include convolutional neural networks (CNN), recurrent neural networks (RNN), etc. These models can automatically extract high-level features of data and are suitable for processing complex time series and multi-dimensional data. Select a suitable model architecture based on data characteristics and task requirements. For example, convolutional neural networks are suitable for processing images and two-dimensional data, and recurrent neural networks are suitable for processing time series data. Convolutional neural networks (CNN) are selected to process pre-processed patch operation records and environmental data because these data can be regarded as a combination of multi-dimensional features, and CNN performs well in extracting spatial features.

[0059] The model is trained through forward propagation and back propagation algorithms. Forward propagation calculates the output of the model, and back propagation updates the model parameters by calculating the gradient of the loss function to minimize the prediction error.

[0060] The steps include: Input the preprocessed training data set and calculate the output of the model; Calculate the error between the predicted value and the true value using an appropriate loss function (e.g., mean squared error, cross entropy); Calculate the gradient of the loss function relative to the model parameters and update the model parameters using the gradient descent method; The forward propagation and backpropagation process is repeated until the model converges or the predetermined number of training rounds is reached.

[0061] For example: using mean square error as the loss function, inputting the preprocessed training data set, after multiple iterations of training, the model parameters are gradually optimized and the error continues to decrease.

[0062] Cross-validation is a method for evaluating model performance. By dividing the data set into a training set and a validation set, training and validation are repeated to ensure the generalization ability of the model. Common methods include K-fold cross-validation and leave-one-out method.

[0063] The steps include: Divide the data set into K subsets; In each iteration, a subset is selected as the validation set and the remaining subsets are used as the training set. The model is trained and its performance on the validation set is evaluated. Calculate the average performance index of K training and validation to evaluate the generalization ability of the model.

[0064] For example, use K-fold cross validation (K=5), divide the data set into 5 subsets, perform 5 training and validation cycles, and calculate the average accuracy and loss of the model on the validation set.

[0065] Hyperparameter optimization refers to adjusting the model's hyperparameters (such as learning rate, batch size, number of layers, number of neurons, etc.) to improve the model's performance. Common optimization methods include grid search, random search, and Bayesian optimization.

[0066] The steps include: Determine the hyperparameters that need to be optimized and their possible value ranges; Choose an appropriate hyperparameter optimization method (such as grid search); Search within a predefined range of hyperparameters, train the model and evaluate its performance to find the best hyperparameter combination.

[0067] For example: Use grid search to optimize the learning rate and batch size, define the learning rate range as 0.001 to 0.01, the batch size range as 16 to 64, perform grid search, and find the hyperparameter combination that makes the model perform best on the validation set.

[0068] Through the training and optimization of the deep learning algorithm, the model of the present invention can accurately predict the key parameters in the patch operation and reduce patch errors; through cross-validation and hyperparameter optimization, it is ensured that the model can perform well on unseen data and adapt to different operating environments and conditions; the optimized AI model data has higher stability and reliability, and can provide efficient and accurate decision support in practical applications.

[0069] Deploy high-density distributed sensors in the work area to collect environmental data in real time, including temperature, humidity, vibration, air pressure, air quality, and light intensity. Use edge computing devices to pre-process and integrate data to generate comprehensive environmental status data. Deploy high-density distributed sensors and collect environmental data in real time, pre-process and integrate the data through edge computing devices, and generate comprehensive environmental status data for patch-bearing positioning calibration.

[0070] High-density distributed sensors are deployed in the working area to monitor environmental parameters in real time. These sensors cover the entire working area of ​​the placement machine to ensure the comprehensiveness and accuracy of the data. High-density deployment can capture tiny environmental changes and provide fine-grained environmental data.

[0071] Select appropriate sensor types, including temperature sensors, humidity sensors, accelerometers (vibration monitoring), air pressure sensors, air quality sensors, and light intensity sensors. Arrange sensors appropriately according to the layout of the work area to ensure comprehensive coverage and redundancy of data collection. Calibrate sensors to ensure accurate and consistent readings from each sensor.

[0072] In one embodiment, temperature sensors and humidity sensors are arranged at different locations of the placement machine, with one sensor placed at a certain interval to ensure that the temperature and humidity changes in the entire working area are fully captured.

[0073] Environmental data is collected in real time through the deployed sensor network, including temperature, humidity, vibration, air pressure, air quality and light intensity. Real-time data collection can provide dynamic change information of the working environment, which helps to adjust the patch operation in time. A reasonable data collection frequency is set according to the characteristics of environmental changes to ensure the timeliness and accuracy of the data. Sensor data is transmitted to the edge computing device for processing through wireless or wired networks.

[0074] In one embodiment, the temperature and humidity sensors are set to collect data at a frequency of once per second, and the data is transmitted to the edge computing device in real time via Wi-Fi.

[0075] Preprocess the collected environmental data on the edge computing device, including data cleaning, denoising, and normalization. Data preprocessing can eliminate noise and outliers in the data to ensure data quality and consistency. Identify and remove erroneous and abnormal data points, and handle missing values ​​caused by sensor failures. Use filtering methods (such as moving average filtering and Kalman filtering) to remove random noise. Convert environmental data of different dimensions to the same scale range for subsequent processing.

[0076] In one embodiment, Kalman filtering is performed on the collected temperature data to remove random fluctuations; and minimum-maximum scaling is performed on the light intensity data to normalize it to the range of [0, 1].

[0077] Integrate various pre-processed environmental data to generate comprehensive environmental status data. Comprehensive environmental status data can fully reflect the environmental conditions of the current working area and provide a reliable reference for patch operations. Time synchronization is performed on data from different sensors to ensure data consistency. The synchronized data is fused to generate a unified environmental status data set.

[0078] In one embodiment, the denoised and normalized temperature, humidity, vibration and other data are synchronized by timestamp and then fused into a multi-dimensional comprehensive environmental status data set.

[0079] High-density distributed sensors are deployed in the working area of ​​the placement machine to collect environmental data in real time. The data is pre-processed and integrated through edge computing devices to generate comprehensive environmental status data to guide real-time adjustments to the placement operation.

[0080] The specific implementation steps include: Arrange multiple temperature, humidity, and vibration sensors in the work area to ensure full coverage; Set the sensor to collect environmental data once per second and transmit it to the edge computing device via Wi-Fi; Clean, denoise and normalize the collected data to eliminate anomalies and noise; The preprocessed data are synchronized by timestamp and fused to generate comprehensive environmental status data.

[0081] The generated comprehensive environmental status data can accurately reflect the environmental changes in the working area, provide reliable environmental information support for the precise positioning and calibration of patch operations, and improve patch quality and production efficiency.

[0082] Through the above detailed description, the environmental data collection and preprocessing process not only deploys high-density distributed sensors to realize real-time collection of environmental data, but also performs data preprocessing and integration through edge computing devices to generate high-quality comprehensive environmental status data, providing a solid foundation for patch-bearing positioning calibration.

[0083] Preferably, the data preprocessing performed by the edge computing device includes smoothing, abnormal data filtering and data fusion of the real-time collected environmental data to generate comprehensive environmental status data, wherein the smoothing is calculated using the following formula:

[0084] in, represents the smoothed data; Represents raw environmental data; Indicates the smoothing window size; Indicates the current data index; Indicates the offset index within the window.

[0085] Smoothing is the process of reducing short-term fluctuations in data through methods such as moving average, making the data more stable and easier to analyze. Smoothing can eliminate random noise in sensor data and highlight the main trends. Using a moving average algorithm to smooth the data reduces fluctuations by averaging the data over a period of time.

[0086] The specific steps are: Choose an appropriate smoothing window size based on the characteristics of data fluctuations ; For each data point, calculate its smoothed value ; Smooth the value Applied to the original data, it produces a smoothed dataset.

[0087] Abnormal data filtering is to identify and remove abnormal values ​​in the data to ensure the accuracy and reliability of the data. Abnormal data may be caused by sensor failure or unexpected events. Abnormal values ​​are identified through statistical methods and then removed or corrected.

[0088] The specific steps include: Use statistical methods (e.g., Z-Score, box plots) to detect outliers in the data.

[0089] For detected outliers, select appropriate methods to handle them, such as removal, correction, or interpolation.

[0090] For example: perform outlier detection on the collected vibration data. Use the box plot method to identify abnormal points in the vibration data and remove these abnormal values.

[0091] Data fusion is the process of processing data from multiple sensors to generate a unified environmental status data set. Data fusion can eliminate the redundancy between different sensor data and improve the overall consistency and reliability of the data.

[0092] The specific steps include: Time synchronization is performed on data from different sensors to ensure data consistency.

[0093] The synchronized data are fused to generate a multi-dimensional comprehensive environmental status dataset.

[0094] For example: the temperature, humidity, vibration and other data collected at different time points are synchronized by timestamp and then fused into a multi-dimensional comprehensive environmental status data set.

[0095] The effect of the present invention of performing data preprocessing through edge computing devices is to generate a high-quality comprehensive environmental status data set, providing reliable environmental information support for patch-bearing positioning calibration. Random noise in the data is eliminated through smoothing, and erroneous data is eliminated through abnormal data filtering to improve the accuracy and consistency of the data. Through a high-density distributed sensor network and edge computing devices, environmental data is collected and processed in real time to dynamically reflect environmental changes in the working area. Through data fusion and integration of multiple environmental parameters, the generated comprehensive environmental status data set comprehensively reflects the environmental conditions of the current working area, providing a full range of environmental references for patch operations.

[0096] In one embodiment, high-density distributed sensors are deployed in the working area of ​​the placement machine to collect environmental data in real time, and smoothing, abnormal data filtering and data fusion are performed through edge computing devices to generate comprehensive environmental status data for guiding real-time adjustment of placement operations.

[0097] Arrange multiple temperature, humidity, and vibration sensors in the work area to ensure full coverage; set the sensors to collect environmental data once per second and transmit it to the edge computing device through the wireless network; apply moving average filtering to the temperature data, and the window size , calculate the smoothed temperature value . The outliers in the vibration data are detected and removed using the box plot method. The data after smoothing and outlier filtering are synchronized by timestamp and fused into a multi-dimensional comprehensive environmental status data set.

[0098] The generated comprehensive environmental status data can accurately reflect the environmental changes in the working area, provide reliable environmental information support for the precise positioning and calibration of patch operations, and improve patch quality and production efficiency.

[0099] Through the present invention, the data preprocessing process performed by the edge computing device not only includes smoothing, abnormal data filtering and data fusion, but also ensures the accuracy and consistency of the data through specific formulas and methods, generates high-quality comprehensive environmental status data, and provides a solid foundation for patch-bearing positioning calibration.

[0100] like Figure 2 As shown, by initializing the nano-scale micro-motor drive system, the nozzle is positioned and calibrated in real time, and feedback adjustment is performed based on the comprehensive environmental state data and the initial AI model data to generate real-time positioning calibration data; Preferably, the real-time positioning and calibration of the nanoscale micro-motor drive system comprises the following steps: Determine the initial position and angle of the nozzle by integrating environmental status data and initial AI model data; Adjust the position and angle of the nozzle based on the feedback control algorithm to generate real-time positioning calibration data; The feedback control amount is calculated by the following formula to generate real-time positioning calibration data:

[0101] in, represents the feedback control amount; Indicates the set position; Indicates the measurement location; represents proportional gain; represents the integral gain; represents the differential gain; Represents a time variable.

[0102] The purpose of real-time positioning and calibration of the nanoscale micromotor drive system is to determine the initial position and angle of the nozzle by integrating environmental status data and initial AI model data, adjust the position and angle of the nozzle based on the feedback control algorithm, and generate real-time positioning calibration data to achieve high-precision placement operations.

[0103] The initial position and angle of the nozzle are determined using comprehensive environmental status data and initial AI model data. The comprehensive environmental status data provides accurate information about the current environment, and the initial AI model data is based on historical data and trained models to predict the best initial position and angle.

[0104] The steps include: obtaining comprehensive environmental status data from the edge computing device, obtaining initial AI model data from the AI ​​system, and determining the initial position and angle of the nozzle using the optimal position and angle predicted by the AI ​​model and the current environmental status.

[0105] Before the patch operation begins, the current ambient temperature, humidity, vibration and other data are obtained from the edge computing device, and the initial model trained based on historical data is obtained from the AI ​​system to determine the optimal initial position and angle of the nozzle.

[0106] The position and angle of the nozzle are adjusted in real time based on the feedback control algorithm, which uses the deviation between real-time measurement data and the set position to achieve high-precision positioning and calibration through proportional, integral and differential (PID) control.

[0107] The steps include: Set target position: Set the target position and angle of the nozzle according to the initial AI model data .

[0108] Real-time position measurement: The current position and angle of the nozzle are measured in real time through the sensor .

[0109] Calculate the feedback control amount: Use the PID control formula to calculate the feedback control amount , The calculation formula is the above formula.

[0110] Adjust position and angle: control amount based on feedback The position and angle of the nozzle can be adjusted in real time.

[0111] During the patch operation, the position and angle of the nozzle are measured in real time, the position deviation is calculated using the PID control formula, the feedback control quantity is generated, and the nozzle position is adjusted to ensure that it is accurately aligned with the patch position.

[0112] The PID control algorithm calculates the feedback control quantity through three parts: proportional, integral and differential. The proportional control part is adjusted according to the current error, the integral control part is adjusted according to the accumulated error, and the differential control part is adjusted according to the error change rate. This comprehensive adjustment method can quickly respond to position deviations and make corrections.

[0113] The specific steps include: Calculate the current error: and multiplied by the proportional gain .

[0114] Integral control: Calculate the integral of the accumulated error and multiply it by the integral gain .

[0115] Derivative control: Calculate the error rate of change and multiply by the derivative gain .

[0116] Feedback control amount calculation: Add the above three parts to get the final feedback control amount .

[0117] During the feedback control process, the error between the current position and the set position is calculated in real time, and the error is input into the PID controller. By calculating the proportional, integral and differential control quantities, the feedback control quantity for adjusting the nozzle position is generated to achieve high-precision calibration.

[0118] The present invention realizes high-precision positioning and calibration of the nozzle during the patch process through real-time positioning and calibration of the nano-scale micro-motor drive system, thereby improving the patch quality and production efficiency. The optimal initial position and angle of the nozzle are determined by using comprehensive environmental status data and initial AI model data to achieve precise positioning; based on the feedback control algorithm, the position and angle of the nozzle are adjusted in real time to ensure that it is always in the best state and reduce patch errors; through high-precision real-time positioning and calibration, rework caused by patch errors is reduced, and the overall efficiency of the production line is improved.

[0119] In one embodiment, during the patch operation, the initial position and angle of the nozzle are determined by integrating environmental status data and initial AI model data, and the position and angle of the nozzle are adjusted in real time using a PID control algorithm to generate real-time positioning calibration data.

[0120] Implementation steps include: (1) Determine the initial position and angle: Obtain current environment data from the edge computing device and initial model data from the AI ​​system to determine the optimal initial position and angle of the nozzle.

[0121] (2) Real-time positioning and calibration: (2.1) Set the target position: Set the target position of the nozzle according to the AI ​​model data .

[0122] (2.2) Real-time position measurement: The current position of the nozzle is measured by the sensor .

[0123] (2.3) Calculate the feedback control quantity: Use the PID control formula to calculate the feedback control quantity .

[0124] (2.4) Adjust position and angle: Adjust the nozzle position according to the feedback control amount to ensure that it is accurately aligned with the patch position.

[0125] Through real-time positioning and calibration, the nozzle is ensured to maintain a high-precision position during the entire patch operation, improving patch quality, reducing rework and improving production efficiency.

[0126] In summary, in the present invention, the real-time positioning and calibration process of the nanoscale micromotor drive system not only includes determining the initial position and angle, but also realizes real-time calibration through the PID control algorithm, ensuring that the suction nozzle remains in the best state during the entire operation process, thereby significantly improving the accuracy and efficiency of the patch carrying positioning.

[0127] like Figure 3 As shown, quantum computing technology is used for path planning and optimization, the path is dynamically adjusted based on real-time positioning calibration data and comprehensive environmental status data, and dynamically adjusted path data is generated through real-time monitoring and feedback re-optimization; Quantum computing technology is used for path planning and optimization, and the path is dynamically adjusted based on real-time positioning calibration data and comprehensive environmental status data. Through real-time monitoring and feedback optimization, dynamically adjusted path data is generated to achieve efficient and accurate patch-bearing positioning calibration.

[0128] Path planning and optimization are performed using quantum computing technology. Quantum computing can efficiently solve complex optimization problems and quickly find the global optimal path through quantum superposition and quantum parallel computing. The goal of path planning is to determine the best path for the nozzle from the initial position to the target position, taking into account various environmental factors and real-time positioning data.

[0129] The specific steps include: Build a quantum computing model to transform the path planning problem into a quantum optimization problem; for example, you can use a quantum annealing algorithm or a quantum variational algorithm to solve the path optimization problem; Input real-time positioning calibration data and comprehensive environmental status data into the quantum computing model; Run the quantum computing model and calculate the optimal path through quantum algorithms; Obtain quantum computing results and generate optimized path data.

[0130] In one embodiment, before the patch operation begins, a path optimization model is constructed using a quantum annealing algorithm, and the current nozzle position, target position, and environmental status data are input. The optimal path is obtained through quantum calculation to ensure that the nozzle reaches the target position accurately in the shortest time.

[0131] Based on real-time positioning calibration data and comprehensive environmental status data, the movement path of the nozzle is dynamically adjusted. Real-time positioning calibration data provides the current position and status of the nozzle, and comprehensive environmental status data provides real-time change information of the current environment. Dynamic adjustment can respond to environmental changes and positioning errors in a timely manner and optimize the path.

[0132] The specific steps include: Through sensors and edge computing devices, the nozzle positioning data and environmental status data are collected in real time; Fine-tune the current path and correct path deviation based on real-time data; The movement status of the nozzle is continuously monitored and the path is dynamically adjusted to ensure that it reaches the target position accurately.

[0133] In one embodiment, during the patch operation, the position and environmental data of the nozzle are monitored in real time. If a path deviation is found, the nozzle movement path is immediately adjusted through a path fine-tuning algorithm to ensure that it operates according to the optimized path.

[0134] By real-time monitoring of the nozzle's motion state and environmental changes, the feedback mechanism is used to re-optimize the path. Real-time monitoring can promptly detect path deviations and environmental changes, and re-optimize through feedback control to ensure continuous optimization and calibration of the path.

[0135] The specific steps include: Through sensors and edge computing devices, the movement status and environmental changes of the nozzle are continuously monitored; According to the real-time monitoring data, the deviation of the current path is calculated and a feedback control signal is generated; Utilizing feedback control signals, the current path is re-optimized through a quantum computing model to generate dynamically adjusted path data.

[0136] In one embodiment, during the patch operation, the position and environmental status of the nozzle are monitored in real time. If it is detected that temperature fluctuations affect the patch accuracy, the path is immediately adjusted through feedback control signals, and the quantum computing model is run again to optimize the path and generate a new dynamic adjustment path to ensure operation accuracy.

[0137] The present invention uses quantum computing technology to plan and optimize the path, and through dynamic path adjustment and real-time monitoring feedback re-optimization, it can achieve efficient and accurate patch bearing positioning calibration, and improve the quality and production efficiency of the patch. Quantum computing technology can quickly find the global optimal path, reduce calculation time, and improve the efficiency of path planning; fine-tune and dynamically adjust the path based on real-time data, respond to environmental changes and positioning errors in a timely manner, and ensure the precise movement of the nozzle; through real-time monitoring and feedback mechanisms, the path is continuously optimized to ensure that the nozzle always operates according to the optimal path, thereby improving the quality of the patch.

[0138] In one embodiment, quantum computing technology is used for path planning and dynamic adjustment. During the patch operation, quantum computing technology is used for path planning and optimization, and the path is dynamically adjusted based on real-time positioning calibration data and comprehensive environmental status data. Dynamically adjusted path data is generated through real-time monitoring and feedback re-optimization.

[0139] Implementation steps include: Construct a quantum annealing algorithm model, input the nozzle's initial position, target position and environmental data, and calculate the optimal path; monitor the nozzle's position and environmental status in real time, fine-tune the path based on real-time data, and correct path deviations; continuously monitor the nozzle's motion status and environmental changes, generate control signals through a feedback mechanism, run the quantum computing model again to optimize the path, and generate a new dynamically adjusted path.

[0140] Through quantum computing technology and dynamic path adjustment, the nozzle is ensured to always move along the optimal path during the patch operation, thereby improving the patch accuracy and production efficiency.

[0141] Through the present invention, quantum computing technology is used for path planning and optimization, which can not only quickly find the optimal path, but also ensure continuous optimization and calibration of the path through dynamic adjustment and real-time monitoring feedback mechanism, thereby significantly improving the accuracy and efficiency of patch bearing positioning.

[0142] Preferably, the dynamic adjustment and feedback re-optimization comprises the following steps: Dynamically adjust the path based on real-time positioning calibration data and comprehensive environmental status data; Fine-tune and correct the path based on real-time monitoring data and feedback data; The path is fine-tuned using the following formula:

[0143] in, Indicates the adjustment amount of the path; Indicates real-time positioning calibration data; Represents comprehensive environmental status data; Represents the adjustment function.

[0144] The purpose of dynamic adjustment and feedback re-optimization is to dynamically adjust the path based on real-time positioning calibration data and comprehensive environmental status data, and to fine-tune and correct the path through real-time monitoring data and feedback data to ensure high accuracy and efficiency in the patch process.

[0145] Dynamic adjustment uses real-time positioning calibration data and comprehensive environmental status data to update the nozzle path in real time. Real-time positioning calibration data provides the current position and status of the nozzle, and comprehensive environmental status data reflects the real-time changes in the current environment. Through dynamic adjustment, the nozzle path is ensured to be consistent with the actual situation.

[0146] The specific steps include: Collect the nozzle positioning calibration data and environmental status data in real time; Based on the current positioning calibration data and environmental status data , calculate the adjustment amount of the path ; The calculated adjustment Applied to the current path, update the motion trajectory of the nozzle.

[0147] In one embodiment, during the patch operation, the position and environmental status of the nozzle are monitored in real time. If it is detected that the ambient temperature rise affects the positioning accuracy of the nozzle, the path adjustment amount is immediately calculated. , adjust the path of the nozzle to ensure that it runs along the predetermined trajectory.

[0148] Fine-tuning and correction is to make subtle adjustments and corrections to the path by real-time monitoring of the nozzle's motion status and feedback data. Real-time monitoring data provides the nozzle's real-time position, and feedback data reflects the path deviation. Through fine-tuning and correction, the path can be further optimized.

[0149] The specific steps include: The movement status of the nozzle is monitored in real time through sensors and monitoring systems.

[0150] Analyze feedback data and calculate path deviation.

[0151] Calculate the fine-tuning amount based on the path deviation , making minor adjustments and corrections to the path.

[0152] The path fine-tuning formula is used to calculate the adjustment amount of the path , based on real-time positioning calibration data and comprehensive environmental status data, dynamically adjust the motion trajectory of the nozzle.

[0153] During the patch operation, the positioning data collected in real time Displays the nozzle position deviating from the predetermined trajectory, and the environmental data Display temperature rise, by adjusting the function , calculate the path adjustment amount , apply it to the path adjustment of the nozzle to ensure that the nozzle runs along the optimized path.

[0154] The present invention realizes high precision and high efficiency of the nozzle in the patch process through dynamic adjustment and feedback re-optimization, ensures continuous optimization and calibration of the path, and improves the patch quality and production efficiency. The path is dynamically adjusted based on real-time data to ensure that the movement trajectory of the nozzle is always consistent with the current environment and positioning state; through real-time monitoring and feedback data, the path is fine-tuned and corrected to further optimize the movement trajectory of the nozzle and reduce path deviation; the path fine-tuning formula is used to dynamically adjust the nozzle path to achieve continuous optimization and high-precision calibration, ensuring the efficiency and accuracy of the patch operation.

[0155] In one embodiment, dynamic adjustment and feedback re-optimization are used to achieve high-precision patching. During the patching operation, by collecting positioning calibration data and environmental status data in real time, dynamic adjustment and feedback re-optimization technology are used to ensure that the nozzle runs along the optimal path.

[0156] The specific steps include: Real-time collection of nozzle positioning data and environmental data , calculate the path adjustment , adjust the nozzle path.

[0157] By monitoring the movement of the nozzle in real time, analyzing the path deviation and calculating the fine-tuning amount , and make path corrections.

[0158] Using the Adjustment Function Calculate path adjustment , ensuring continuous optimization of the nozzle path.

[0159] Through dynamic adjustment and feedback re-optimization technology, it is ensured that the nozzle always runs along the optimal path during the patch operation, thereby improving the patch accuracy and production efficiency.

[0160] The present invention not only includes path adjustment based on real-time data through dynamic adjustment and feedback re-optimization, but also performs path fine-tuning and correction through real-time monitoring and feedback mechanism, ensuring that the suction nozzle remains in the best state during the entire operation process, thereby significantly improving the accuracy and efficiency of patch carrying positioning.

[0161] By collecting operation data in real time, preprocessing data, updating and training AI models, detecting anomalies and early warning, verifying and optimizing models, and redeploying optimized AI model data, the accuracy and efficiency of patch-bearing positioning calibration can be improved.

[0162] Real-time data collection refers to the acquisition of various data in the patch operation in real time through sensors and monitoring equipment. These data include the position of the nozzle, operation steps, environmental parameters, etc. Real-time data collection can dynamically reflect various changes in the operation process and provide the latest input for model update and optimization.

[0163] Data preprocessing refers to cleaning, denoising, normalizing and feature extraction of real-time collected data to ensure data quality and consistency. Preprocessed data can improve the training effect and prediction accuracy of AI models.

[0164] By updating and training the AI ​​model, it can adapt to the latest operational data and environmental changes. During the training process, the preprocessed data set is used to adjust the model parameters and improve the model's prediction accuracy and generalization ability.

[0165] Abnormal detection refers to identifying abnormal situations during operation by analyzing real-time data and issuing early warning signals. Abnormal detection can detect problems in a timely manner, prevent incorrect operations, and improve production quality. During the patch process, the position data of the nozzle is analyzed in real time. If it is detected that the nozzle position deviates from the predetermined trajectory and exceeds the threshold, an early warning signal is immediately issued to remind the operator to check the equipment.

[0166] Through the verification and optimization process, the performance of the AI ​​model is evaluated, its deficiencies are identified and improved. Model verification includes evaluating the performance of the model on the verification data set, and model optimization further improves the model performance by adjusting hyperparameters and improving the model structure. The cross-validation method is used to evaluate the performance of the model on different verification sets. It is found that the model performs poorly in high temperature environments. After adjusting the model structure and hyperparameters, the model is retrained to improve its robustness in various environments.

[0167] Deploy the optimized AI model to the production environment, replace the old model, and ensure that the new model can run in actual operations. The model deployment process includes converting the model into a format suitable for the production environment and testing and verifying it. The specific steps include: Convert the trained model into a production-friendly format (such as TensorFlow Lite, ONNX, etc.). Test in a production environment to ensure the model works properly; The model is formally deployed to the placement machine control system to replace the old model.

[0168] For example: convert the optimized CNN model into TensorFlow Lite format, deploy it to the control system of the placement machine, and test it to ensure that the model can predict and adjust the nozzle position in real time in actual production.

[0169] Through real-time data collection, AI model update and optimization, we ensure high accuracy and efficiency of patch bearing positioning calibration, and significantly improve production quality and efficiency. Provide the latest operation data and environmental information to provide reliable data support for model update and optimization; improve data quality and consistency, and enhance model training effect; enable the model to adapt to the latest data and environmental changes, improve prediction accuracy and generalization ability; timely discover and handle abnormal situations during operation, prevent erroneous operations, and improve production quality; through verification and optimization, further improve model performance and ensure its robustness in various environments; apply the optimized model to the production environment, realize real-time prediction and adjustment, and improve patch accuracy and production efficiency.

[0170] The real-time data collection, AI model update and optimization process of the present invention not only includes the collection and preprocessing of operation data, but also realizes anomaly detection and early warning through the training, verification and optimization of AI models, and finally deploys the optimized model into the production environment to ensure the high precision and efficiency of patch bearing positioning calibration.

[0171] like Figure 4 As shown, a patch carrying positioning and calibration system includes: The historical data collection module is used to collect historical patch operation records and environmental data; this module provides key data of past operation processes through historical patch operation records and environmental data, including patch location, operation time, environmental conditions, etc.

[0172] Extract historical data from placement machines and environmental monitoring systems. Store the collected data in a database for subsequent analysis and processing. Provide detailed historical operation records and environmental data to provide basic data support for AI model training and optimization.

[0173] In one embodiment, patch operation data for the past month is collected, including the specific location, time, and ambient temperature and humidity data of each operation.

[0174] The data preprocessing module is used to clean, denoise, normalize and extract features of the collected historical patch operation records and environmental data to generate a training data set to eliminate noise and abnormal data; the data preprocessing module cleans, denoises, normalizes and extracts features of historical data to ensure data quality and consistency. Further eliminate noise and abnormal data to improve the quality of the training data set.

[0175] Eliminate errors and invalid data and handle missing values. Apply filtering techniques (such as moving average and Kalman filtering) to remove noise from the data. Scale the data to the same scale for subsequent processing. Extract features useful for patch operations from the data to generate a training data set.

[0176] Generate high-quality training data sets to ensure that the AI ​​model can accurately reflect the actual situation of patch operations.

[0177] In one embodiment, moving average filtering is performed on the collected ambient temperature data to remove short-term fluctuations, the temperature data is normalized to a range of [0, 1], and time series features of temperature changes are extracted.

[0178] The AI ​​model training module is used to train the preprocessed training data set using a deep learning algorithm, and adjust the model parameters through cross-validation and hyperparameter optimization to generate optimized AI model data; Use deep learning algorithms to train the preprocessed training data set, adjust model parameters through cross-validation and hyperparameter optimization, and generate optimized AI model data.

[0179] Select a suitable deep learning model (such as a convolutional neural network); use a training dataset to train the model; use cross-validation methods to evaluate model performance; and adjust the model's hyperparameters to improve model performance.

[0180] Generate optimized AI model data, which can accurately predict key parameters in patch operations and improve patch accuracy.

[0181] In one embodiment, a convolutional neural network is used to train the training data, and the learning rate and batch size are optimized through grid search to generate an optimized AI model for predicting patch locations.

[0182] A sensor network module is used to deploy high-density distributed sensors in the working area to collect environmental data in real time, including temperature, humidity, vibration, air pressure, air quality and light intensity; Deploy high-density distributed sensors in the working area to collect environmental data in real time to ensure that the environmental data is comprehensive and accurate.

[0183] Temperature, humidity, vibration, air pressure, air quality and light intensity sensors are placed around the placement machine; environmental data is continuously collected and transmitted to edge computing devices for processing; comprehensive environmental data is acquired in real time to provide accurate environmental information for placement operations.

[0184] In one embodiment, temperature sensors are arranged around the placement machine to monitor the ambient temperature in real time, and transmit the data to the edge computing device via a wireless network.

[0185] Edge computing devices are used to smooth, filter out abnormal data and fuse data collected in real time to generate comprehensive environmental status data; edge computing devices are used to smooth, filter out abnormal data and fuse data collected in real time to generate comprehensive environmental status data.

[0186] Smooth the environmental data and filter out abnormal data. Merge multiple environmental data into comprehensive environmental status data. Generate high-quality comprehensive environmental status data to ensure the real-time and accuracy of environmental data.

[0187] In one embodiment, Kalman filtering is performed on the temperature data collected in real time to remove random noise, and the processed temperature, humidity and vibration data are fused to generate comprehensive environmental status data.

[0188] The nanoscale micromotor drive system is used to determine the initial position and angle of the nozzle through comprehensive environmental status data and initial AI model data, adjust the position and angle of the nozzle based on the feedback control algorithm, and generate real-time positioning calibration data; the initial position and angle of the nozzle are determined through comprehensive environmental status data and initial AI model data, and the position and angle of the nozzle are adjusted based on the feedback control algorithm to generate real-time positioning calibration data.

[0189] The initial position and angle of the nozzle are determined based on the environmental status data and AI model data; the position and angle of the nozzle are adjusted in real time using the feedback control algorithm. High-precision real-time positioning and calibration of the nozzle are achieved to improve the accuracy of patch placement.

[0190] In one embodiment, comprehensive environmental data and AI models are used to predict the optimal initial position of the nozzle, and the nozzle position is monitored in real time and calibrated through feedback control to ensure that the nozzle is accurately aligned with the patch position.

[0191] The quantum computing module is used to optimize the initial path planning data using quantum computing technology, generate optimized path data, accelerate the path calculation through quantum algorithms, and determine the optimal patch path; the quantum computing technology is used to optimize the initial path planning data, generate optimized path data, accelerate the path calculation through quantum algorithms, and determine the optimal patch path.

[0192] Use traditional algorithms to generate initial path data; optimize path data through quantum computing technology to determine the optimal path.

[0193] Achieve efficient path optimization and accelerated calculation to ensure the efficiency and accuracy of the patch process.

[0194] In one embodiment, a quantum annealing algorithm is used to optimize the movement path of the nozzle to ensure that the nozzle reaches the target position by the shortest path and at the fastest speed.

[0195] Real-time monitoring and feedback module, used to collect operation data in real time, perform data preprocessing, update and train AI models, perform anomaly detection and early warning, verify and optimize models, and redeploy optimized AI model data. Real-time collection of operation data, data preprocessing, update and train AI models, perform anomaly detection and early warning, verify and optimize models, and redeploy optimized AI model data.

[0196] Collect patch operation data in real time; clean and process the operation data; update and train the AI ​​model based on the new data; detect operation anomalies in real time and issue warnings; verify and optimize the AI ​​model, and redeploy the optimized model.

[0197] Realize adaptive optimization and real-time monitoring of the system to improve the stability and reliability of the system.

[0198] In one embodiment, during the patch process, the movement state of the nozzle is monitored in real time. If an abnormal deviation from the trajectory is detected, a warning signal is immediately issued, and the new data is used to update the AI ​​model to optimize the operation path.

[0199] Through the coordinated work of the above modules, the patch carrying positioning and calibration system of the present invention can achieve high-precision and high-efficiency patch operation. The detailed description of each module shows the principle and effect of the system, and helps readers understand the practical application and advantages of each module through specific embodiments.

[0200] A storage medium stores a computer program, which, when executed by a processor, implements the steps of the patch carrying positioning calibration method.

[0201] like Figure 5 As shown, a patch carrying positioning and calibration robot comprises: The control unit is used to control the overall operation of the robot; the control unit is responsible for managing and coordinating the overall operation of the robot and ensuring effective communication and collaboration between the units. It is the core of the entire system and controls the movement and operation of the robot through preset instructions and algorithms.

[0202] Assign work to each unit according to the patch task; manage the operation process of each unit to ensure the smooth progress of the patch process; monitor the system status and take timely measures when abnormalities are found. Ensure efficient collaboration between the various units of the robot to achieve the smooth completion of the patch task.

[0203] In one embodiment, the control unit instructs the sensor network unit to start collecting environmental data according to the production plan, and simultaneously starts the nano-scale micro-motor drive unit to perform initial position calibration.

[0204] The historical data collection unit is used to collect historical patch operation records and environmental data; this unit provides key data of past operation processes by collecting historical patch operation records and environmental data, including patch location, operation time, environmental conditions, etc.

[0205] Extract historical data from placement machines and environmental monitoring systems; store the collected data in a database for subsequent analysis and processing.

[0206] Provides detailed historical operation records and environmental data, and provides basic data support for AI model training and optimization.

[0207] In one embodiment, patch operation data for the past month is collected, including the specific location, time, and ambient temperature and humidity data of each operation.

[0208] The data preprocessing unit is used to clean, denoise, normalize and extract features of the collected historical patch operation records and environmental data to generate a training data set to eliminate noise and abnormal data; the data preprocessing unit cleans, denoises, normalizes and extracts features of historical data to ensure data quality and consistency. Further eliminate noise and abnormal data to improve the quality of the training data set.

[0209] Eliminate errors and invalid data, and handle missing values; apply filtering techniques (such as moving average and Kalman filtering) to remove noise in the data; scale the data to the same scale for subsequent processing; extract features useful for patch operations from the data to generate training data sets. Generate high-quality training data sets to ensure that the AI ​​model can accurately reflect the actual situation of patch operations.

[0210] In one embodiment, moving average filtering is performed on the collected ambient temperature data to remove short-term fluctuations, the temperature data is normalized to a range of [0, 1], and time series features of temperature changes are extracted.

[0211] The AI ​​model training unit is used to train the preprocessed training data set using a deep learning algorithm, and adjust the model parameters through cross-validation and hyperparameter optimization to generate optimized AI model data; use a deep learning algorithm to train the preprocessed training data set, and adjust the model parameters through cross-validation and hyperparameter optimization to generate optimized AI model data.

[0212] Select a suitable deep learning model (such as a convolutional neural network); use a training data set to train the model; use a cross-validation method to evaluate model performance; adjust the model's hyperparameters to improve model performance. Generate optimized AI model data that can accurately predict key parameters in patch operations and improve patch accuracy.

[0213] In one embodiment, a convolutional neural network is used to train the training data, and the learning rate and batch size are optimized through grid search to generate an optimized AI model for predicting patch locations.

[0214] The sensor network unit is used to deploy high-density distributed sensors in the working area to collect environmental data in real time, wherein the environmental data includes temperature, humidity, vibration, air pressure, air quality and light intensity; high-density distributed sensors are deployed in the working area to collect environmental data in real time to ensure that the environmental data is comprehensive and accurate.

[0215] Place temperature, humidity, vibration, air pressure, air quality, and light intensity sensors around the placement machine. Continuously collect environmental data and transmit it to edge computing devices for processing. Obtain comprehensive environmental data in real time to provide accurate environmental information for placement operations.

[0216] In one embodiment, temperature sensors are arranged around the placement machine to monitor the ambient temperature in real time, and transmit the data to the edge computing device via a wireless network.

[0217] The edge computing unit is used to smooth, filter out abnormal data and fuse the environmental data collected in real time to generate comprehensive environmental status data; the edge computing unit is used to smooth, filter out abnormal data and fuse the environmental data collected in real time to generate comprehensive environmental status data.

[0218] Smooth the environmental data and filter out abnormal data; merge multiple environmental data into comprehensive environmental status data. Generate high-quality comprehensive environmental status data to ensure the real-time and accuracy of environmental data.

[0219] In one embodiment, Kalman filtering is performed on the temperature data collected in real time to remove random noise, and the processed temperature, humidity and vibration data are fused to generate comprehensive environmental status data.

[0220] The nanoscale micro-motor drive unit is used to determine the initial position and angle of the nozzle through comprehensive environmental status data and initial AI model data, adjust the position and angle of the nozzle based on the feedback control algorithm, and generate real-time positioning calibration data; the initial position and angle of the nozzle are determined through comprehensive environmental status data and initial AI model data, and the position and angle of the nozzle are adjusted based on the feedback control algorithm to generate real-time positioning calibration data.

[0221] The initial position and angle of the nozzle are determined based on the environmental status data and AI model data; the position and angle of the nozzle are adjusted in real time using the feedback control algorithm. High-precision real-time positioning and calibration of the nozzle are achieved to improve the accuracy of patch placement.

[0222] In one embodiment, comprehensive environmental data and AI models are used to predict the optimal initial position of the nozzle, and the nozzle position is monitored in real time and calibrated through feedback control to ensure that the nozzle is accurately aligned with the patch position.

[0223] The quantum computing unit is used to optimize the initial path planning data using quantum computing technology, generate optimized path data, accelerate the path calculation through quantum algorithms, and determine the optimal patch path; the initial path planning data is optimized using quantum computing technology, generate optimized path data, accelerate the path calculation through quantum algorithms, and determine the optimal patch path.

[0224] Use traditional algorithms to generate initial path data; use quantum computing technology to optimize path data and determine the optimal path. Achieve efficient path optimization and accelerated calculation to ensure the efficiency and accuracy of the patch process.

[0225] In one embodiment, a quantum annealing algorithm is used to optimize the movement path of the nozzle to ensure that the nozzle reaches the target position by the shortest path and at the fastest speed.

[0226] Real-time monitoring and feedback unit, used to collect operation data in real time, perform data preprocessing, update and train AI models, perform anomaly detection and early warning, verify and optimize models, and redeploy optimized AI model data. Real-time collection of operation data, data preprocessing, update and train AI models, perform anomaly detection and early warning, verify and optimize models, and redeploy optimized AI model data.

[0227] Collect patch operation data in real time; clean and process operation data; update and train AI models based on new data; detect operation anomalies in real time and issue warnings; verify and optimize AI models, and redeploy optimized models. Achieve adaptive optimization and real-time monitoring of the system, and improve system stability and reliability.

[0228] In one embodiment, during the patch process, the movement state of the nozzle is monitored in real time. If an abnormal deviation from the trajectory is detected, a warning signal is immediately issued, and the new data is used to update the AI ​​model to optimize the operation path.

[0229] Through the coordinated work of the above-mentioned units, the patch carrying positioning and calibration robot of the present invention can achieve high-precision and high-efficiency patch operation. The detailed description of each unit shows the principle and effect of the system, and helps readers understand the practical application and advantages of each unit through specific embodiments.

[0230] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0231] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems) and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1A device that provides the functions specified in a block or multiple blocks.

[0232] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0233] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0234] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0235] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0236] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0237] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0238] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. A patch bearing positioning calibration method, characterized in that: The following steps are involved: By collecting historical patch operation records and environmental data, data preprocessing is performed to train and optimize AI models, and the generated initial AI model data is verified and deployed; Deploy high-density distributed sensors in the work area to collect environmental data in real time, including temperature, humidity, vibration, air pressure, air quality, and light intensity. Use edge computing devices to pre-process and integrate data to generate comprehensive environmental status data. By initializing the nano-scale micro-motor drive system, the nozzle is positioned and calibrated in real time, and feedback adjustment is performed based on the comprehensive environmental status data and initial AI model data to generate real-time positioning calibration data; Utilize quantum computing technology for path planning and optimization, dynamically adjust the path based on real-time positioning calibration data and comprehensive environmental status data, and generate dynamically adjusted path data through real-time monitoring and feedback re-optimization; By collecting operational data in real time, performing data preprocessing, updating and training AI models, performing anomaly detection and early warning, verifying and optimizing models, and redeploying optimized AI model data.

2. The patch carrying positioning calibration method according to claim 1, characterized in that: The data preprocessing includes data cleaning, denoising, normalization and feature extraction of the collected historical patch operation records and environmental data to generate a training data set for eliminating noise and abnormal data.

3. The patch bearing positioning calibration method according to claim 2, characterized in that: The AI ​​model training includes using a deep learning algorithm to train the preprocessed training data set, and adjusting the model parameters through cross-validation and hyperparameter optimization to generate optimized AI model data.

4. The patch carrying positioning calibration method according to claim 1, characterized in that: The data preprocessing performed by the edge computing device includes smoothing, abnormal data filtering and data fusion of the real-time collected environmental data to generate comprehensive environmental status data, wherein the smoothing is calculated using the following formula: ； in, represents the smoothed data; Represents raw environmental data; Indicates the smoothing window size; Indicates the current data index; Indicates the offset index within the window.

5. The patch carrying positioning calibration method according to claim 1, characterized in that: The real-time positioning and calibration of the nanoscale micromotor drive system comprises the following steps: Determine the initial position and angle of the nozzle by integrating environmental status data and initial AI model data; Adjust the position and angle of the nozzle based on the feedback control algorithm to generate real-time positioning calibration data; The feedback control amount is calculated by the following formula to generate real-time positioning calibration data: ； in, represents the feedback control amount; Indicates the set position; Indicates the measurement location; represents proportional gain; represents the integral gain; represents the differential gain; Represents a time variable.

6. The patch carrying positioning calibration method according to claim 1, characterized in that: The path planning and optimization includes the following steps: Use quantum computing technology to optimize the initial path planning data and generate optimized path data; Use quantum algorithms to accelerate the calculation of paths and determine the optimal patch path; Use the following formula for path optimization calculation: ； in, represents the optimized path cost; Indicates The cost function of each path; Indicates The distance of the path; Indicates the number of paths.

7. The patch carrying positioning calibration method according to claim 6, characterized in that: The dynamic adjustment and feedback re-optimization comprises the following steps: Dynamically adjust the path based on real-time positioning calibration data and comprehensive environmental status data; Fine-tune and correct the path based on real-time monitoring data and feedback data; The path is fine-tuned using the following formula: ； in, Indicates the adjustment amount of the path; Indicates real-time positioning calibration data; Represents comprehensive environmental status data; Represents the adjustment function.

8. A patch carrying positioning and calibration system, characterized in that: include: Historical data collection module, used to collect historical patch operation records and environmental data; The data preprocessing module is used to clean, denoise, normalize and extract features of the collected historical patch operation records and environmental data to generate a training data set for eliminating noise and abnormal data; The AI ​​model training module is used to train the preprocessed training data set using a deep learning algorithm, and adjust the model parameters through cross-validation and hyperparameter optimization to generate optimized AI model data; A sensor network module is used to deploy high-density distributed sensors in the working area to collect environmental data in real time, including temperature, humidity, vibration, air pressure, air quality and light intensity; Edge computing devices are used to smooth, filter out abnormal data, and fuse data collected in real time to generate comprehensive environmental status data; Nano-scale micro-motor drive system, which is used to determine the initial position and angle of the nozzle by integrating environmental status data and initial AI model data, adjust the position and angle of the nozzle based on the feedback control algorithm, and generate real-time positioning calibration data; The quantum computing module is used to optimize the initial path planning data using quantum computing technology, generate optimized path data, accelerate the path calculation through quantum algorithms, and determine the optimal patch path; The real-time monitoring and feedback module is used to collect operation data in real time, perform data preprocessing, update and train AI models, perform anomaly detection and early warning, verify and optimize models, and redeploy optimized AI model data.

9. A storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the patch carrying positioning calibration method described in any one of claims 1 to 7 are implemented.

10. A patch-carrying positioning and calibration robot, characterized in that: include: A control unit, used to control the overall operation of the robot; Historical data collection unit, used to collect historical patch operation records and environmental data; The data preprocessing unit is used to perform data cleaning, denoising, normalization and feature extraction on the collected historical patch operation records and environmental data to generate a training data set for eliminating noise and abnormal data; An AI model training unit is used to train the preprocessed training data set using a deep learning algorithm, and to adjust model parameters through cross-validation and hyperparameter optimization to generate optimized AI model data; A sensor network unit is used to deploy high-density distributed sensors in the working area to collect environmental data in real time, including temperature, humidity, vibration, air pressure, air quality and light intensity; Edge computing unit, used to smooth, filter abnormal data and fuse data collected in real time to generate comprehensive environmental status data; Nano-scale micro-motor drive unit, used to determine the initial position and angle of the nozzle by integrating environmental status data and initial AI model data, adjust the position and angle of the nozzle based on the feedback control algorithm, and generate real-time positioning calibration data; A quantum computing unit is used to optimize the initial path planning data using quantum computing technology, generate optimized path data, accelerate the path calculation using quantum algorithms, and determine the optimal patch path; The real-time monitoring and feedback unit is used to collect operation data in real time, perform data preprocessing, update and train AI models, perform anomaly detection and early warning, verify and optimize models, and redeploy optimized AI model data.

Images