Processing defect analysis method and system for multifunctional probe

Abstract

The invention discloses a machining defect analysis method and system for a multifunctional probe, and belongs to the technical field of high-end intelligent manufacturing. In hot bending machining, multiple technological data such as temperature, pressure and deformation are collected in real time through multiple sensors, and in combination with a quality detection result, the data are cleaned, standardized and marked; a deep learning and optimization algorithm is adopted to construct a model, process quality prediction and optimization are realized through training and optimization, accurate detection and uniform control of the temperature of the whole surface of the hot bending mold are realized, expansion change rules of materials at different temperatures are captured in real time, a heating or cooling system is intelligently adjusted according to a detection result, and the process quality is predicted and optimized. According to the method, the material is uniformly stressed and stably formed, meanwhile, the forming state is evaluated in real time through a high-precision means, deviation is corrected in time, accurate data support is provided for the process, visualization and real-time online detection of the whole production process are supported, electrical testing, aging and optical parameter detection can be carried out, and the product performance can be comprehensively evaluated.

Description

Technical Field

[0001] This invention belongs to the field of high-end intelligent manufacturing technology, and in particular relates to a method and system for analyzing processing defects in multifunctional probes. Background Technology

[0002] Hot bending is a process that involves heating a material to soften it, then shaping it in a mold, and finally cooling it to set the shape. This process is complex and involves the coupling of multiple parameters such as heat, force, and time, making it prone to defects. Defect analysis aims to identify defect types, trace their root causes, and ultimately optimize the process to improve yield.

[0003] Existing multifunctional probe-based processing defect analysis technology has the following limitations: it cannot adaptively simulate and analyze the thermal expansion curvature of materials, it is difficult to achieve full-area detection and uniform control of the temperature of hot bending dies, and it cannot comprehensively monitor the completion status of the material's thermoforming process. At the same time, this technology lacks the ability to visualize and monitor quality in real time, does not have the functions of electrical performance testing, aging testing, and optical parameter detection, and has not introduced AI large models for intelligent analysis and optimization.

[0004] Based on this, the present invention designs a method and system for analyzing processing defects in a multifunctional probe to solve the above problems. Summary of the Invention

[0005] The purpose of this invention is to address the limitations of existing multifunctional probe-based processing defect analysis technologies, which include: the inability to adaptively simulate and analyze the thermal expansion curvature of materials, the difficulty in achieving full-area detection and uniform control of the temperature of hot bending dies, the inability to comprehensively monitor the completion status of the material's thermoforming process, the lack of visualization and real-time online detection capabilities for quality, the absence of electrical performance testing, aging tests, and optical parameter detection functions, and the lack of the introduction of AI large-scale models for intelligent analysis and optimization. Therefore, this invention proposes a multifunctional probe-based processing defect analysis method and system.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: A method for analyzing manufacturing defects in a multifunctional probe, comprising: S1: Install multi-functional probes at key locations in the hot bending equipment to collect processing parameters of the heating area, mold surface, and material contact points in real time, obtain data such as temperature, pressure, and deformation, record the time-temperature curve and material expansion during the heating stage, monitor the deformation of the mold under different pressures, and obtain the internal and external temperature distribution of the material by combining infrared thermal imaging technology. At the same time, record quality indicators such as finished product appearance defects and dimensional accuracy, analyze the relationship between process parameters and quality, and collect equipment operating parameters. S2: Preprocess the data, including noise reduction, outlier handling and missing value imputation, then standardize the data units using standardization methods, and finally label the data according to quality grade and classify them into qualified products and defective products. S3: A data model is built using a deep learning framework, and hyperparameters are tuned using the whale optimization algorithm. The data is divided into training, validation and test sets. Overfitting is prevented during training using the validation set. Finally, the model performance is evaluated based on metrics such as accuracy, recall, F1 score and MSE. If the performance does not meet the criteria, the model is returned for adjustment and optimization.

[0007] As a further description of the above technical solution: The multifunctional probe includes a high-speed and optical camera, a surface temperature scanning probe, a topography scanning probe, a stress sensing probe, and a defect detection probe; The high-speed and optical camera is used for recording and identifying the process, achieving a visual presentation of the process. The high-speed and optical camera also has an optical testing function to accurately collect parameter data corpus A. The high-speed and optical camera integrates a variety of key components, including but not limited to convex lenses, concave lenses, high-resolution photosensors, and high-performance data transmission coupling chips. These components work together to meet diverse detection needs.

[0008] As a further description of the above technical solution: The surface temperature scanning probe detects the temperature of each layer inside the heating cavity and simultaneously collects parameter data corpus B. The surface temperature scanning probe integrates a variety of key components, including but not limited to convex lenses, concave lenses, high-resolution photosensors, and high-performance data transmission coupling chips. It can flexibly adjust the light to meet diverse optical needs. The high-resolution infrared thermal sensor can capture weak infrared radiation and convert it into accurate temperature data. The high-performance data transmission coupling chip ensures high-speed and stable data transmission.

[0009] As a further description of the above technical solution: The morphology scanning probe records the surface size and position changes of the forming material during the process and collects process parameter data corpus C. The probe incorporates, but is not limited to, convex lenses, concave lenses, fiber optic high-resolution laser ranging sensor matrix, and high-performance data transmission coupling chip.

[0010] As a further description of the above technical solution: The topography scanning probe records the changes in size and position of the surface of the formed material during the process, and comprehensively collects process parameter data corpus C, providing detailed and reliable data support for process quality assessment, process optimization, and subsequent analysis. The topography scanning probe integrates a variety of key components, including but not limited to convex lenses, concave lenses, high-resolution photosensors, and high-performance data transmission coupling chips. The convex and concave lenses are designed and matched to achieve flexible focusing and scattering control of light to meet the precise requirements of light propagation path and intensity in different scanning scenarios. The fiber optic high-resolution laser ranging sensor matrix is ​​used to capture the distance information of various points on the surface of the formed material, and then calculate the changes in surface size and position. The high-performance data transmission coupling chip ensures that the large amount of data collected by the sensor matrix can be transmitted to the AI ​​control system in a high-speed and stable manner.

[0011] As a further description of the above technical solution: The stress sensing probe records material stress data during the process and collects and analyzes process parameter data corpus D. The stress sensing probe integrates a variety of key components, including but not limited to convex lenses, concave lenses, high-resolution photosensitive sensors, and high-performance data transmission coupling chips.

[0012] As a further description of the above technical solution: The defect detection probe monitors the quality defects of products or processes throughout the entire process, capturing and recording dynamic changes in the form and degree of product quality defects from the start to the end of the process. This provides detailed evidence for in-depth analysis of the generation and development of defects. Simultaneously, it collects process parameter data corpus E. Through the integration and analysis of multi-dimensional data, it comprehensively evaluates the impact of the process on product quality. The defect detection probe integrates a variety of key components, including but not limited to convex lenses, concave lenses, high-resolution photosensitive sensors, and high-performance data transmission coupling chips.

[0013] As a further description of the above technical solution: The AI ​​control system includes a large model chip and a system control chip; The AI ​​control system performs in-depth analysis on parameter data corpora A, B, C, D, and E acquired by the multi-functional probe. It uses a large data model to comprehensively process massive amounts of data covering multiple dimensions such as time, temperature, pressure, and quality level. During the analysis, the AI ​​control system uses complex algorithm models to mine the correlation patterns of parameters and customize efficient process parameter solutions.

[0014] As a further description of the above technical solution: The whale optimization algorithm mainly includes three mathematical operations: To surround its prey, the whale moves towards the current optimal solution using the following equation: At+1=Bbestt-C∙X In the formula, Bbestt is the current optimal solution, C is the control parameter, which decreases linearly with the number of iterations, and X is the distance between the current individual and the optimal solution; In a bubble net formation, whales spiral upwards and release bubbles to create a net-like structure that traps their prey. The algorithm incorporates a spiral trajectory to update the position. At+1=X,∙ebl∙cos2πl+Bbestt In the formula, b is the spiral shape constant, l is a random number, and X is the distance between the individual and the optimal solution; In random search, when C≥1, the whale swims randomly to explore a wider area and avoids getting trapped in local optima. At+1=Arand-C∙Xrand In the formula, Arand represents the randomly selected individual location; Neural networks are a type of negative feedback neural network algorithm that primarily modifies network parameters inversely based on the output error of the output layer. The thermal error modeling steps based on the whale optimization algorithm and neural networks are as follows: Define the input and output of the hybrid model, and select parameter data corpus A, parameter data corpus B, parameter data corpus C, parameter data corpus D and parameter data corpus E as the input data of the network, and use the thermal error data of the Z-axis as the output data; The neural network structure and whale optimization algorithm initialization are determined. The neural network structure to be optimized is determined, and after obtaining the parameters to be optimized, the upper and lower limits of the parameters to be optimized and the population size are further determined. Q=i=1nyi'-yi2n-1 In the formula, yi' represents the i-th predicted value, and yi represents the i-th measured value.

[0015] A multifunctional probe-based machining defect analysis system includes: The parameter acquisition module acquires parameter data corpus A, parameter data corpus B, parameter data corpus C, parameter data corpus D, and parameter data corpus E. It records the time-temperature change curves during the heating stage of the hot bending process and the material expansion at the corresponding moment. It collects the deformation data of the mold under different pressures, including pressure value, holding time, and deformation. It uses infrared thermal imaging technology combined with probe measurement to obtain the temperature distribution and differences on the material surface and inside. It records the appearance defects and dimensional accuracy of the formed products and performs correlation analysis with the process parameters. At the same time, it collects equipment operating parameters to evaluate their impact on the hot bending quality. The data preprocessing module preprocesses the data, including removing noise, outliers, and missing values. For noisy data, filtering algorithms are used for smoothing. Outliers are identified and corrected based on data distribution characteristics and business rules. Missing values ​​are filled using interpolation or prediction methods based on historical data. Subsequently, data of different dimensions and ranges are standardized based on Z-score standardization and Min-Max standardization to enhance data comparability. Finally, the parameter data is labeled according to the product quality level, dividing it into qualified and defective product data to provide labels for subsequent model training. AI control system analysis and modeling employs a deep learning framework to construct a large data model and combines it with the whale optimization algorithm to establish a hybrid model. The preprocessed parameter data is divided into training, validation, and test sets according to a certain ratio. The training set is used to train the model, and hyperparameters are adjusted to optimize model performance. The validation set is used for evaluation and tuning during training to prevent overfitting. Finally, the test set is used to evaluate model performance, with evaluation metrics including accuracy, recall, F1 score, and mean squared error (MSE). If the model performance does not meet expectations, it returns to the training phase for further adjustment and optimization.

[0016] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are: 1. In this invention, the expansion and change patterns of materials under different temperature conditions are captured in real time, providing accurate data support for subsequent processes. It enables full-surface temperature detection of the hot bending mold, employs advanced sensing technology to comprehensively and meticulously acquire temperature distribution information of various areas on the mold surface, ensuring accurate temperature control. It also has the ability to uniformly control hot bending forming across the entire surface. Based on the mold temperature detection results, the heating or cooling system is intelligently adjusted to ensure uniform stress on the material during hot bending, resulting in stable and reliable forming quality. It can detect the completion status of material hot forming across the entire surface, and through high-precision detection methods, it can evaluate in real time whether the material has reached the expected forming state, promptly identifying and correcting deviations in the forming process.

[0017] 2. In terms of quality inspection, this invention supports visualization and real-time online inspection, which can intuitively present the entire process of material heat treatment and hot bending forming, allowing operators to keep abreast of the production status. At the same time, it can also perform electrical testing, aging and optical parameter testing to comprehensively evaluate the electrical performance, stability and optical properties of the material, ensuring that the product quality meets high standards. In addition, the equipment is equipped with advanced AI large model analysis function, which can deeply mine and analyze the large amount of data collected, providing a scientific basis for process optimization and quality control. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of a processing defect analysis method and system for a multifunctional probe proposed in this invention; Figure 2 This is a flowchart illustrating a processing defect analysis method and system for a multifunctional probe proposed in this invention. Figure 3 This is a block diagram of a processing defect analysis method and system for a multifunctional probe proposed in this invention.

[0019] Legend: 1. Multifunctional probe; 11. High-speed and optical camera; 12. Surface temperature scanning probe; 13. Topography scanning probe; 14. Stress sensing probe; 15. Defect detection probe; 2. AI control system; 21. Large model chip; 22. System control chip. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] Please see the appendix Figure 1 - Appendix Figure 3 This invention provides a technical solution: a method for analyzing manufacturing defects in a multifunctional probe, comprising: S1: Install multi-functional probes at key locations in the hot bending equipment to collect processing parameters of the heating area, mold surface, and material contact points in real time, obtain data such as temperature, pressure, and deformation, record the time-temperature curve and material expansion during the heating stage, monitor the deformation of the mold under different pressures, and obtain the internal and external temperature distribution of the material by combining infrared thermal imaging technology. At the same time, record quality indicators such as finished product appearance defects and dimensional accuracy, analyze the relationship between process parameters and quality, and collect equipment operating parameters. S2: Preprocess the data, including noise reduction, outlier handling and missing value imputation, then standardize the data units using standardization methods, and finally label the data according to quality grade and classify them into qualified products and defective products. S3: A data model is built using a deep learning framework, and hyperparameters are tuned using the whale optimization algorithm. The data is divided into training, validation and test sets. Overfitting is prevented during training using the validation set. Finally, the model performance is evaluated based on metrics such as accuracy, recall, F1 score and MSE. If the performance does not meet the criteria, the model is returned for adjustment and optimization.

[0022] Specifically, the multifunctional probe includes a high-speed and optical camera, a surface temperature scanning probe, a topography scanning probe, a stress sensing probe, and a defect detection probe; The high-speed and optical camera is used for recording and identifying the process, achieving a visual presentation of the process. The high-speed and optical camera also has an optical testing function to accurately collect parameter data corpus A. The high-speed and optical camera integrates a variety of key components, including but not limited to convex lenses, concave lenses, high-resolution photosensors, and high-performance data transmission coupling chips. These components work together to meet diverse detection needs.

[0023] Specifically, the surface temperature scanning probe detects the temperature of each layer inside the heating cavity and simultaneously collects parameter data corpus B. The surface temperature scanning probe integrates a variety of key components, including but not limited to convex lenses, concave lenses, high-resolution photosensors, and high-performance data transmission coupling chips. It can flexibly adjust the light to meet diverse optical needs. The high-resolution infrared thermal sensor can capture weak infrared radiation and convert it into accurate temperature data. The high-performance data transmission coupling chip ensures high-speed and stable data transmission.

[0024] Specifically, the morphology scanning probe records the surface size and position changes of the forming material during the process and collects process parameter data corpus C. The probe incorporates, but is not limited to, convex lenses, concave lenses, fiber optic high-resolution laser ranging sensor matrix, and high-performance data transmission coupling chip.

[0025] Specifically, the topography scanning probe records the changes in size and position of the surface of the molding material during the process, and comprehensively collects process parameter data corpus C, providing detailed and reliable data support for process quality assessment, process optimization, and subsequent analysis. The topography scanning probe integrates a variety of key components, including but not limited to convex lenses, concave lenses, high-resolution photosensors, and high-performance data transmission coupling chips. The convex and concave lenses are designed and matched to achieve flexible focusing and scattering control of light, so as to meet the precise requirements of light propagation path and intensity under different scanning scenarios. The fiber optic high-resolution laser ranging sensor matrix is ​​used to capture the distance information of various points on the surface of the molding material, and then calculate the changes in surface size and position. The high-performance data transmission coupling chip ensures that the large amount of data collected by the sensor matrix can be transmitted to the AI ​​control system in a high-speed and stable manner.

[0026] Specifically, the stress sensing probe records material stress data during the process and collects and analyzes process parameter data corpus D. The stress sensing probe integrates a variety of key components, including but not limited to convex lenses, concave lenses, high-resolution photosensitive sensors, and high-performance data transmission coupling chips.

[0027] Specifically, the defect detection probe monitors the quality defects of a product or process object throughout the entire process, capturing and recording dynamic changes in the form and degree of product quality defects from the start to the end of the process. This provides detailed evidence for in-depth analysis of the generation and development of defects. Simultaneously, it collects process parameter data corpus E. Through the integration and analysis of multi-dimensional data, it comprehensively evaluates the impact of the process on product quality. The defect detection probe integrates a variety of key components, including but not limited to convex lenses, concave lenses, high-resolution photosensitive sensors, and high-performance data transmission coupling chips.

[0028] Specifically, the AI ​​control system includes a large model chip and a system control chip; The AI ​​control system performs in-depth analysis on parameter data corpora A, B, C, D, and E acquired by the multi-functional probe. It uses a large data model to comprehensively process massive amounts of data covering multiple dimensions such as time, temperature, pressure, and quality level. During the analysis, the AI ​​control system uses complex algorithm models to mine the correlation patterns of parameters and customize efficient process parameter solutions.

[0029] Specifically, the whale optimization algorithm mainly includes three mathematical operations: To surround its prey, the whale moves towards the current optimal solution using the following equation: At+1=Bbestt-C∙X In the formula, Bbestt is the current optimal solution, C is the control parameter, which decreases linearly with the number of iterations, and X is the distance between the current individual and the optimal solution; In a bubble net formation, whales spiral upwards and release bubbles to create a net-like structure that traps their prey. The algorithm incorporates a spiral trajectory to update the position. At+1=X,∙ebl∙cos2πl+Bbestt In the formula, b is the spiral shape constant, l is a random number, and X is the distance between the individual and the optimal solution; In random search, when C≥1, the whale swims randomly to explore a wider area and avoids getting trapped in local optima. At+1=Arand-C∙Xrand In the formula, Arand represents the randomly selected individual location; Neural networks are a type of negative feedback neural network algorithm that primarily modifies network parameters inversely based on the output error of the output layer. The thermal error modeling steps based on the whale optimization algorithm and neural networks are as follows: Define the input and output of the hybrid model, and select parameter data corpus A, parameter data corpus B, parameter data corpus C, parameter data corpus D and parameter data corpus E as the input data of the network, and use the thermal error data of the Z-axis as the output data; The neural network structure and whale optimization algorithm initialization are determined. The neural network structure to be optimized is determined, and after obtaining the parameters to be optimized, the upper and lower limits of the parameters to be optimized and the population size are further determined. Q=i=1nyi'-yi2n-1 In the formula, yi' represents the i-th predicted value, and yi represents the i-th measured value.

[0030] A multifunctional probe-based machining defect analysis system includes: The parameter acquisition module acquires parameter data corpus A, parameter data corpus B, parameter data corpus C, parameter data corpus D, and parameter data corpus E. It records the time-temperature change curves during the heating stage of the hot bending process and the material expansion at the corresponding moment. It collects the deformation data of the mold under different pressures, including pressure value, holding time, and deformation. It uses infrared thermal imaging technology combined with probe measurement to obtain the temperature distribution and differences on the material surface and inside. It records the appearance defects and dimensional accuracy of the formed products and performs correlation analysis with the process parameters. At the same time, it collects equipment operating parameters to evaluate their impact on the hot bending quality. The data preprocessing module preprocesses the data, including removing noise, outliers, and missing values. For noisy data, filtering algorithms are used for smoothing. Outliers are identified and corrected based on data distribution characteristics and business rules. Missing values ​​are filled using interpolation or prediction methods based on historical data. Subsequently, data of different dimensions and ranges are standardized based on Z-score standardization and Min-Max standardization to enhance data comparability. Finally, the parameter data is labeled according to the product quality level, dividing it into qualified and defective product data to provide labels for subsequent model training. AI control system analysis and modeling employs a deep learning framework to construct a large data model and combines it with the whale optimization algorithm to establish a hybrid model. The preprocessed parameter data is divided into training, validation, and test sets according to a certain ratio. The training set is used to train the model, and hyperparameters are adjusted to optimize model performance. The validation set is used for evaluation and tuning during training to prevent overfitting. Finally, the test set is used to evaluate model performance, with evaluation metrics including accuracy, recall, F1 score, and mean squared error (MSE). If the model performance does not meet expectations, it returns to the training phase for further adjustment and optimization.

[0031] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for analyzing manufacturing defects in a multifunctional probe, characterized in that, include: S1: Install multi-functional probes at key locations in the hot bending equipment to collect processing parameters of the heating area, mold surface, and material contact points in real time, obtain data such as temperature, pressure, and deformation, record the time-temperature curve and material expansion during the heating stage, monitor the deformation of the mold under different pressures, and obtain the internal and external temperature distribution of the material by combining infrared thermal imaging technology. At the same time, record quality indicators such as finished product appearance defects and dimensional accuracy, analyze the relationship between process parameters and quality, and collect equipment operating parameters. S2: Preprocess the data, including noise reduction, outlier handling and missing value imputation, then standardize the data units using standardization methods, and finally label the data according to quality grade and classify them into qualified products and defective products. S3: A data model is built using a deep learning framework, and hyperparameters are tuned using the whale optimization algorithm. The data is divided into training, validation and test sets. Overfitting is prevented during training using the validation set. Finally, the model performance is evaluated based on metrics such as accuracy, recall, F1 score and MSE. If the performance does not meet the criteria, the model is returned for adjustment and optimization.

2. The method for analyzing manufacturing defects in a multifunctional probe according to claim 1, characterized in that, The multifunctional probe includes a high-speed and optical camera, a surface temperature scanning probe, a topography scanning probe, a stress sensing probe, and a defect detection probe; The high-speed and optical camera is used for recording and identifying the process, achieving a visual presentation of the process. The high-speed and optical camera also has an optical testing function to accurately collect parameter data corpus A. The high-speed and optical camera integrates a variety of key components, including but not limited to convex lenses, concave lenses, high-resolution photosensors, and high-performance data transmission coupling chips. These components work together to meet diverse detection needs.

3. The method for analyzing manufacturing defects in a multifunctional probe according to claim 2, characterized in that, The surface temperature scanning probe detects the temperature of each layer inside the heating cavity and simultaneously collects parameter data corpus B. The surface temperature scanning probe integrates a variety of key components, including but not limited to convex lenses, concave lenses, high-resolution photosensors, and high-performance data transmission coupling chips. It can flexibly adjust the light to meet diverse optical needs. The high-resolution infrared thermal sensor can capture weak infrared radiation and convert it into accurate temperature data. The high-performance data transmission coupling chip ensures high-speed and stable data transmission.

4. The method for analyzing manufacturing defects in a multifunctional probe according to claim 3, characterized in that, The morphology scanning probe records the surface size and position changes of the forming material during the process and collects process parameter data corpus C. The probe incorporates, but is not limited to, convex lenses, concave lenses, fiber optic high-resolution laser ranging sensor matrix, and high-performance data transmission coupling chip.

5. The method for analyzing manufacturing defects in a multifunctional probe according to claim 4, characterized in that, The topography scanning probe records the changes in size and position of the surface of the formed material during the process, and comprehensively collects process parameter data corpus C, providing detailed and reliable data support for process quality assessment, process optimization, and subsequent analysis. The topography scanning probe integrates a variety of key components, including but not limited to convex lenses, concave lenses, high-resolution photosensors, and high-performance data transmission coupling chips. The convex and concave lenses are designed and matched to achieve flexible focusing and scattering control of light to meet the precise requirements of light propagation path and intensity in different scanning scenarios. The fiber optic high-resolution laser ranging sensor matrix is ​​used to capture the distance information of various points on the surface of the formed material, and then calculate the changes in surface size and position. The high-performance data transmission coupling chip ensures that the large amount of data collected by the sensor matrix can be transmitted to the AI ​​control system in a high-speed and stable manner.

6. The method for analyzing manufacturing defects in a multifunctional probe according to claim 5, characterized in that, The stress sensing probe records material stress data during the process and collects and analyzes process parameter data corpus D. The stress sensing probe integrates a variety of key components, including but not limited to convex lenses, concave lenses, high-resolution photosensitive sensors, and high-performance data transmission coupling chips.

7. The method for analyzing manufacturing defects in a multifunctional probe according to claim 6, characterized in that, The defect detection probe monitors the quality defects of products or processes throughout the entire process, capturing and recording dynamic changes in the form and degree of product quality defects from the start to the end of the process. This provides detailed evidence for in-depth analysis of the generation and development of defects. Simultaneously, it collects process parameter data corpus E. Through the integration and analysis of multi-dimensional data, it comprehensively evaluates the impact of the process on product quality. The defect detection probe integrates a variety of key components, including but not limited to convex lenses, concave lenses, high-resolution photosensitive sensors, and high-performance data transmission coupling chips.

8. The method for analyzing manufacturing defects in a multifunctional probe according to claim 7, characterized in that, The AI ​​control system includes a large-scale model chip and a system control chip; The AI ​​control system performs in-depth analysis on parameter data corpora A, B, C, D, and E acquired by the multi-functional probe. It uses a large data model to comprehensively process massive amounts of data covering multiple dimensions such as time, temperature, pressure, and quality level. During the analysis, the AI ​​control system uses complex algorithm models to mine the correlation patterns of parameters and customize efficient process parameter solutions.

9. The method for analyzing manufacturing defects in a multifunctional probe according to claim 8, characterized in that, The whale optimization algorithm mainly includes three mathematical operations: To surround its prey, the whale moves towards the current optimal solution using the following equation: At+1=Bbestt-C∙X In the formula, Bbestt is the current optimal solution, C is the control parameter, which decreases linearly with the number of iterations, and X is the distance between the current individual and the optimal solution; In a bubble net formation, whales spiral upwards and release bubbles to create a net-like structure that traps their prey. The algorithm incorporates a spiral trajectory to update the position. At+1=X,∙ebl∙cos2πl+Bbestt In the formula, b is the spiral shape constant, l is a random number, and X is the distance between the individual and the optimal solution; In random search, when C≥1, the whale swims randomly to explore a wider area and avoids getting trapped in local optima. At+1=Arand-C∙Xrand In the formula, Arand represents the randomly selected individual location; Neural networks are a type of negative feedback neural network algorithm that primarily modifies network parameters inversely based on the output error of the output layer. The thermal error modeling steps based on the whale optimization algorithm and neural networks are as follows: Define the input and output of the hybrid model, and select parameter data corpus A, parameter data corpus B, parameter data corpus C, parameter data corpus D and parameter data corpus E as the input data of the network, and use the thermal error data of the Z-axis as the output data; The neural network structure and whale optimization algorithm initialization are determined. The neural network structure to be optimized is determined, and after obtaining the parameters to be optimized, the upper and lower limits of the parameters to be optimized and the population size are further determined. Q=i=1nyi'-yi2n-1 In the formula, yi' represents the i-th predicted value, and yi represents the i-th measured value.

10. A multifunctional probe machining defect analysis system, comprising the multifunctional probe machining defect analysis method according to claim 9, characterized in that, include: The parameter acquisition module acquires parameter data corpus A, parameter data corpus B, parameter data corpus C, parameter data corpus D, and parameter data corpus E. It records the time-temperature change curves during the heating stage of the hot bending process and the material expansion at the corresponding moment. It collects the deformation data of the mold under different pressures, including pressure value, holding time, and deformation. It uses infrared thermal imaging technology combined with probe measurement to obtain the temperature distribution and differences on the material surface and inside. It records the appearance defects and dimensional accuracy of the formed products and performs correlation analysis with the process parameters. At the same time, it collects equipment operating parameters to evaluate their impact on the hot bending quality. The data preprocessing module preprocesses the data, including removing noise, outliers, and missing values. For noisy data, filtering algorithms are used for smoothing. Outliers are identified and corrected based on data distribution characteristics and business rules. Missing values ​​are filled using interpolation or prediction methods based on historical data. Subsequently, data of different dimensions and ranges are standardized based on Z-score standardization and Min-Max standardization to enhance data comparability. Finally, the parameter data is labeled according to the product quality level, dividing it into qualified and defective product data to provide labels for subsequent model training. AI control system analysis and modeling employs a deep learning framework to construct a large data model and combines it with the whale optimization algorithm to establish a hybrid model. The preprocessed parameter data is divided into training, validation, and test sets according to a certain ratio. The training set is used to train the model, and hyperparameters are adjusted to optimize model performance. The validation set is used for evaluation and tuning during training to prevent overfitting. Finally, the test set is used to evaluate model performance, with evaluation metrics including accuracy, recall, F1 score, and mean squared error (MSE). If the model performance does not meet expectations, it returns to the training phase for further adjustment and optimization.

Images