A door and window processing process data analysis method and system

Abstract

The application provides a door and window processing process data analysis method and system, relates to the field of data analysis, obtains equipment operation parameters of an upstream process, carries out extraction processing to obtain microscopic information representing the microscopic state of the process, processes the microscopic information by using preset deviation information, and obtains an upstream process deviation value. Based on the upstream process deviation value, the system can generate process compensation parameters of a downstream process, and obtain process parameters of the downstream process accordingly. In the actual operation process of the downstream process, the system monitors the operation parameters, adjusts the process parameters of the downstream process according to the deviation value of the actual operation parameters and expected state parameters, and finally obtains the adjusted process parameters. The application can find potential technical defects of risks in advance, effectively avoid quality problems, save production costs and reduce economic losses for the factory, and significantly improve the quality control level and production efficiency of door and window manufacturing.

Description

Technical Field

[0001] This application relates to the field of data analysis technology, and more specifically, to a method and system for data analysis of door and window manufacturing processes. Background Technology

[0002] In modern door and window manufacturing, production lines are increasingly automated, with equipment operating information and process settings meticulously recorded. However, current technologies, primarily relying on monitoring individual devices or single parameters, struggle to meet the higher demands for product quality and production efficiency. Especially in multi-process, discrete production environments like door and window manufacturing, existing data analysis methods mainly identify anomalies by setting fixed normal ranges for each processing unit. This method is insufficient for effectively identifying and providing early warnings of specific quality risks.

[0003] Specifically, when upstream processes (such as milling) experience minor process deviations due to slight wear or other reasons, resulting in deviations that do not exceed their own alarm range (e.g., a slight increase in the surface roughness of the profile end face), and these deviations cause small, high-frequency fluctuations in the operating data of downstream processes (such as welding) (e.g., the instantaneous power curve of the heating plate), but these fluctuations also do not exceed their independently set abnormal ranges, existing systems cannot effectively link these seemingly normal, subtle data changes across processes. This prevents existing systems from automatically understanding and identifying situations where hidden deviations in upstream processes trigger specific changes in downstream process data.

[0004] Therefore, existing technologies cannot predict and avoid potential quality problems (such as microcracks in welds) that may appear after long-term exposure to the external environment due to factors such as minute stress concentrations inside the weld before the product leaves the factory. This results in quality problems being exposed late and causing economic losses to the factory. Summary of the Invention

[0005] This application discloses a data analysis method and system for door and window processing, which aims to solve the problem that it is difficult to effectively identify and warn of potential quality defects in downstream processes caused by minor deviations in upstream processes in existing door and window processing.

[0006] The technical solution of this application is as follows: In a first aspect, this application discloses a data analysis method for door and window manufacturing processes, comprising the following steps: Obtain the equipment operating parameters of upstream processes; Equipment operating parameters are extracted and processed to obtain microscopic information; microscopic information represents the microscopic state of the process. By processing the microscopic information using preset deviation information, the deviation value of the upstream process is obtained; Based on the deviation value of the upstream process, process compensation parameters for the downstream process are generated, and the process parameters of the downstream process are obtained through the process compensation parameters. Monitor the actual operating parameters of downstream processes; Based on the deviation between the actual operating parameters and the desired state parameters, the process parameters of the downstream processes are adjusted to obtain the adjusted process parameters of the downstream processes.

[0007] Secondly, this application also discloses a data analysis system for door and window manufacturing processes, the system comprising: The parameter acquisition module is used to acquire the operating parameters of the equipment in the upstream process. The information extraction module is used to extract and process equipment operating parameters to obtain microscopic information; The deviation identification module is used to process microscopic information through preset deviation information to obtain the deviation value of the upstream process; The compensation parameter module is used to generate process compensation parameters for downstream processes based on the deviation values ​​of upstream processes, and to obtain the process parameters of downstream processes through the process compensation parameters. The operation monitoring module is used to monitor the actual operating parameters of downstream processes; The parameter adjustment module is used to adjust the process parameters of downstream processes based on the deviation between the actual operating parameters and the desired state parameters, so as to obtain the adjusted process parameters of the downstream processes.

[0008] Beneficial Effects: The door and window processing data analysis method disclosed in this application obtains the equipment operating parameters of upstream processes and extracts and processes them to obtain microscopic information characterizing the microscopic state of the process. Subsequently, the microscopic information is processed using preset deviation information to obtain the upstream process deviation value. Based on this upstream process deviation value, the system can generate process compensation parameters for downstream processes and obtain the process parameters for downstream processes accordingly. During the actual operation of downstream processes, the system monitors their operating parameters and adjusts the process parameters of downstream processes based on the deviation between the actual operating parameters and the expected state parameters, ultimately obtaining the adjusted process parameters. This application effectively solves the technical problem in the prior art of connecting seemingly unrelated subtle changes between different processes to detect potential risks in advance. This application can effectively correlate these subtle deviations in upstream processes with small, high-frequency data fluctuations in downstream processes (such as welding) that also do not exceed independently set abnormal ranges. This cross-process correlation analysis capability enables this application to predict and avoid potential quality problems that may occur after long-term exposure to the external environment due to factors such as minute stress concentration inside the weld before the product leaves the factory. Therefore, this application can effectively avoid quality problems, thereby saving production costs and economic losses for the factory, and significantly improving the quality control level and production efficiency of door and window manufacturing. Attached Figure Description

[0009] Figure 1 A schematic diagram illustrating a data analysis method for door and window manufacturing processes provided in this application.

[0010] Figure 2 A schematic diagram of a door and window manufacturing process data analysis system provided in this application. Detailed Implementation

[0011] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments. The components of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0012] Reference Figure 1 The diagram illustrates a data analysis method for door and window manufacturing processes according to an embodiment of the present invention, which may specifically include the following steps: S101, Obtain the equipment operating parameters of the upstream process; S102, Extract and process the equipment operating parameters to obtain microscopic information, wherein the microscopic information represents the microscopic state of the process; S103, by processing the microscopic information through preset deviation information, the deviation value of the upstream process is obtained; S104, Based on the deviation value of the upstream process, generate the process compensation parameters for the downstream process, and obtain the process parameters of the downstream process through the process compensation parameters; S105, monitors the actual operating parameters of downstream processes; S106. Based on the deviation between the actual operating parameters and the desired state parameters, the process parameters of the downstream process are adjusted to obtain the adjusted process parameters of the downstream process.

[0013] This application can effectively identify and warn of potential quality risks by performing correlation analysis on subtle data changes across processes, thereby improving product quality and production efficiency.

[0014] To better understand the data analysis method for door and window manufacturing processes proposed in this application, it is necessary to explain some of the key terms involved.

[0015] "Equipment operating parameters" refer to the raw data recorded by various sensors or the equipment itself during the door and window manufacturing process, such as current signals, vibration signals, load data, temperature distribution images, acoustic emission signals, and vibration signals of the cutting edge of the tool. These parameters directly reflect the working status of the equipment at a certain moment.

[0016] "Microscopic information" refers to the deep characteristics extracted from the original equipment operating parameters that characterize the microscopic state of the process. It is not simply raw data, but rather more representative and interpretable information obtained after processing and analysis, such as current fluctuation characteristics, high-frequency vibration characteristics, low-frequency vibration characteristics, load abrupt changes, temperature gradient distribution characteristics, frequency band energy characteristics, and vibration mode characteristics. This information can reveal subtle changes within the process that may be difficult to perceive macroscopically.

[0017] "Preset deviation information" refers to pre-defined reference information used to identify and quantify process deviations, such as vibration spectrum deviation. It provides a benchmark for judging whether microscopic information deviates from the normal state.

[0018] "Upstream process deviation value" refers to a numerical value that quantifies the deviation of an upstream process from its normal state, obtained by processing microscopic information and preset deviation information. This deviation value can reflect subtle and potential problems existing in the upstream process.

[0019] "Process compensation parameters" refer to correction values ​​calculated based on upstream process deviations, used to adjust downstream process parameters. Their purpose is to offset or mitigate the adverse effects of upstream deviations on downstream processes.

[0020] "Process parameters" refer to the key settings that control the downstream processing steps, such as the preheating time of the welding heating plate, the instantaneous temperature rise rate, and the head push pressure curve during the pressure holding stage.

[0021] "Actual operating parameters" refer to real-time data recorded by sensors or equipment during the actual operation of downstream processes.

[0022] "Desired state parameters" refer to the parameter values ​​that downstream processes should achieve under ideal or normal operating conditions.

[0023] "Deviation value" refers to the difference between actual operating parameters and expected state parameters, and is used to measure whether the operating status of downstream processes meets expectations.

[0024] The core of the data analysis method for door and window processing proposed in this application lies in perceiving and analyzing the subtle state of upstream processes, and then making forward-looking process compensations and adjustments to downstream processes to avoid potential quality problems.

[0025] First, this method requires acquiring the operating parameters of the upstream process equipment. These parameters are raw data reflecting the equipment's operating status. For example, various sensors can be used to acquire these parameters. One implementation involves deploying current sensors, vibration sensors, and load sensors in the milling unit to acquire the current signal of the milling spindle, the vibration signal of the milling tool, and the load data of the milling motor, respectively. These sensors can monitor changes in physical quantities during the milling process in real time. Another implementation involves setting up an optical sensor array, a high-frequency acoustic emission sensor array, and a micro-vibration sensor in the milling unit. The optical sensor array can acquire a temperature distribution image of the milling cutting area, reflecting the distribution of cutting heat; the high-frequency acoustic emission sensor array can acquire acoustic emission signals of the milling cutting area, characterizing microscopic events in the material removal process; and the micro-vibration sensor can acquire vibration signals of the tool cutting edge, reflecting the interaction between the tool and the workpiece. These sensors can capture the operating status of the upstream process from different dimensions.

[0026] Secondly, the acquired equipment operating parameters are extracted and processed to obtain microscopic information. Microscopic information refers to the deep characteristics that characterize the microscopic state of the process. For example, for current signals, vibration signals, and load data, the following processing can be performed: high-frequency components of the current signal are extracted to obtain current fluctuation characteristics, which can reflect subtle changes in motor load; multi-band spectral decomposition of the vibration signal is performed to obtain high-frequency and low-frequency vibration characteristics, which helps to identify vibration modes in different frequency ranges; transient response analysis of the load data is performed to obtain load abrupt change characteristics, which can reveal sudden changes in force during processing. Finally, these features are weighted and calculated to obtain comprehensive microscopic information. As another implementation method, when acquiring temperature distribution images, acoustic emission signals, and vibration signals, it is first necessary to perform time synchronization calibration on the optical sensor array, high-frequency acoustic emission sensor array, and micro vibration sensor, and to timestamp the temperature distribution images, acoustic emission signals, and vibration signals. Then, the data are time-aligned according to the timestamps to ensure the temporal correspondence of different types of data. Next, the time-aligned temperature distribution image is segmented to obtain the temperature gradient distribution characteristics of the cutting area; the time-aligned acoustic emission signal is subjected to time-frequency analysis to obtain frequency band energy characteristics; and the time-aligned vibration signal is decomposed using wavelet decomposition to obtain vibration mode characteristics. Finally, a correlation mapping relationship is established between the temperature gradient distribution characteristics, frequency band energy characteristics, and vibration mode characteristics, and these characteristics are weighted and combined according to this correlation mapping relationship to obtain the microscopic information of the upstream process status.

[0027] Next, the upstream process deviation value is obtained by processing the microscopic information through preset deviation information. The preset deviation information provides a benchmark for judging whether the microscopic information deviates from the normal state. For example, the preset deviation information may include vibration spectrum deviation. Specifically, the high-frequency and low-frequency vibration characteristics of the profile can be acquired first, and the material acoustic characteristics can be extracted from them. Then, the milling operation signal during the milling process of the milling unit is collected. Next, based on the current material acoustic characteristics of the profile, the material's inherent signal is extracted from the milling operation signal. This signal represents the signal component caused by the material's inherent characteristics. Finally, the material's inherent signal is filtered through vibration spectrum deviation to obtain the filtered material's inherent signal, and then converted based on the filtered material's inherent signal to obtain the upstream process deviation value. As another implementation method, the aging and wear information of the milling unit can be acquired and compared with the initial benchmark value to obtain the equipment error information caused by the overall aging of the equipment. Then, the preset deviation information is dynamically corrected based on the equipment error information to obtain the corrected deviation information. Finally, the microscopic information is processed through the corrected deviation information to obtain the upstream process deviation value.

[0028] Next, based on the upstream process deviation value, process compensation parameters for the downstream process are generated, and the process parameters for the downstream process are obtained through these compensation parameters. For example, upon receiving the upstream process deviation value, the current equipment operating status information of the welding unit can be obtained, including the wear level of the welding equipment and the ambient temperature. Then, the upstream process deviation value, the wear level of the welding equipment, and the ambient temperature are analyzed to identify the first parameter of these factors on the weld quality. This first parameter refers to the parameter that identifies and quantifies the influence of multiple factors on the weld quality. Finally, based on the first parameter, the preheating time of the welding heating plate, the instantaneous temperature rise rate, and the die head propulsion pressure curve during the holding pressure stage are adjusted to obtain the process parameters for the downstream process.

[0029] Subsequently, the actual operating parameters of downstream processes are monitored. For example, in the welding process, the actual temperature curve of the welding heating plate, the die head propulsion pressure curve, and the acoustic emission signal of the weld can be monitored in real time.

[0030] Finally, based on the deviation between the actual operating parameters and the desired parameters, the process parameters of the downstream processes are adjusted to obtain the adjusted process parameters for the downstream processes. For example, if a deviation is detected between the actual temperature curve of the welding heating plate and the desired curve, the preheating time or instantaneous temperature rise rate of the heating plate can be fine-tuned according to the deviation to ensure that the weld quality meets expectations.

[0031] The data analysis method for door and window manufacturing processes proposed in this application achieves effective identification, compensation, and adjustment of cross-process and subtle deviations in door and window manufacturing through a series of closely related steps. First, by acquiring equipment operating parameters from upstream processes, such as current signals, vibration signals, load data, temperature distribution images, acoustic emission signals, and vibration signals from the cutting edge of the tool, a comprehensive raw data foundation is provided for subsequent analysis. These parameters can reflect the real-time status of upstream processes from different dimensions.

[0032] Secondly, these raw equipment operating parameters are extracted and processed in depth to obtain microscopic information characterizing the micro-state of the process. For example, current fluctuation characteristics are obtained by extracting high-frequency components from current signals, high-frequency and low-frequency vibration characteristics are obtained by multi-band spectral decomposition of vibration signals, and load mutation characteristics are obtained by transient response analysis of load data. Alternatively, temperature distribution images, acoustic emission signals, and vibration signals are fused through time synchronization calibration, timestamp marking, and time alignment to extract temperature gradient distribution characteristics, frequency band energy characteristics, and vibration mode characteristics. This microscopic information can reveal subtle changes in upstream processes that are difficult to detect with the naked eye, such as slight wear of cutting tools and changes in the microstructure of profiles. These changes are the root cause of potential quality problems in downstream processes.

[0033] Next, the microscopic information is processed using preset deviation information to obtain the upstream process deviation value. Preset deviation information, such as vibration spectrum deviation, provides a benchmark for judging whether the microscopic information deviates from the normal state. By filtering the material's inherent signal against the vibration spectrum deviation, or by dynamically correcting the preset deviation information using equipment aging and wear information, the subtle deviations of the upstream process can be accurately quantified. This deviation value is the core of this method; it presents the potential problems of the upstream process in a quantitative form.

[0034] Then, based on the deviation values ​​of the upstream processes, process compensation parameters for the downstream processes are generated, and the process parameters of the downstream processes are obtained through these parameters. For example, after receiving the deviation values ​​of the upstream processes, the influence parameters of these factors on weld quality are identified by combining equipment operating status information such as the wear level of the welding equipment and ambient temperature. Based on these parameters, the process parameters of the downstream welding processes can be dynamically adjusted, such as the preheating time of the welding heating plate, the instantaneous temperature rise rate, and the die head propulsion pressure curve during the holding pressure stage. This step is the key to achieving proactive compensation in this method, enabling the downstream processes to self-adjust according to potential problems in the upstream processes, thereby avoiding the occurrence of quality problems.

[0035] Subsequently, the actual operating parameters of downstream processes are monitored, and the process parameters of downstream processes are further adjusted based on the deviation between the actual operating parameters and the expected parameters, resulting in adjusted process parameters. This feedback mechanism ensures that the process parameters of downstream processes are always in an optimal state, and even if there are still slight deviations after compensation, they can be corrected in a timely manner.

[0036] Through the aforementioned series of steps, the method of this application closely links subtle deviations in upstream processes with process adjustments in downstream processes, forming a closed-loop, intelligent data analysis and control system. It effectively solves the problem of traditional methods failing to identify subtle cross-process correlations, thereby predicting and avoiding potential quality issues before products leave the factory, significantly improving the overall quality and production efficiency of door and window products.

[0037] The data analysis method for door and window processing proposed in this application demonstrates significant innovation and advantages in solving existing technical problems. Traditional door and window processing data analysis methods primarily rely on monitoring individual devices or single parameters, which is insufficient to meet the current higher requirements for product quality and production efficiency. Especially in the multi-process, discrete production environment of door and window processing, existing data analysis methods mainly identify anomalies by setting fixed normal ranges for each processing unit. This method struggles to effectively identify and warn of situations where hidden deviations in upstream processes cause specific changes in downstream process data, leading to potential product quality problems.

[0038] The core innovation of this application lies in constructing a cross-process intelligent analysis and compensation mechanism based on microscopic information and deviation value transmission. Unlike existing technologies that only focus on independent alarms for a single process, this application can perform in-depth processing of equipment operating parameters in upstream processes (such as milling) to extract microscopic information characterizing the microscopic state of the process. For example, by performing multi-dimensional analysis of current signals, vibration signals, and load data, or by fusing temperature distribution images, acoustic emission signals, and vibration signals, this application can capture subtle changes that are difficult to detect using traditional methods, such as slight tool wear or changes in the microstructure of the profile.

[0039] Furthermore, this application processes this microscopic information using preset deviation information to obtain upstream process deviation values. For example, by filtering the inherent signals of the material through vibration spectrum deviation, or by dynamically correcting the preset deviation information using equipment aging and wear information, this application can accurately quantify the subtle deviations of the upstream process. This deviation value is the key breakthrough of this application, as it transmits potential problems from the upstream process to the downstream process in a quantifiable form.

[0040] Most importantly, this application generates process compensation parameters for downstream processes based on upstream process deviation values, and obtains the process parameters for downstream processes through these parameters. For example, after receiving upstream process deviation values, it identifies parameters affecting weld quality by combining information such as the wear level of welding equipment and ambient temperature, and dynamically adjusts the preheating time of the welding heating plate, the instantaneous temperature rise rate, and the die head propulsion pressure curve during the holding pressure stage accordingly. This proactive process compensation mechanism allows downstream processes to self-adjust before potential problems in upstream processes manifest, thereby fundamentally avoiding quality problems caused by the accumulation of minor deviations across processes.

[0041] Compared to the closest existing technology, the technical advantages of this application are: it achieves cross-process data correlation analysis, overcoming the limitations of traditional single-process monitoring; by extracting microscopic information and calculating upstream process deviation values, it can identify and quantify subtle deviations that traditional methods cannot detect; and it introduces a forward-looking process compensation mechanism, enabling downstream processes to proactively adjust based on potential problems in upstream processes, thereby effectively avoiding potential quality issues. Through these innovations, this application can significantly improve the overall quality of door and window products, reduce scrap rates, lower production costs, and increase production efficiency.

[0042] Specifically, this application further proposes the following steps for processing microscopic information through preset deviation information to obtain the upstream process deviation value: The high-frequency vibration characteristics and low-frequency vibration characteristics of the profile are obtained, and the acoustic characteristics of the material are extracted from the high-frequency vibration characteristics and low-frequency vibration characteristics. Collect milling operation signals during the milling process of the milling unit; Based on the acoustic characteristics of the current profile material, the inherent material signal is extracted from the milling operation signal; the inherent material signal represents the signal component caused by the material's own properties. The material's inherent signals are filtered by the vibration spectrum deviation to obtain the filtered material inherent signals; The upstream process deviation value is obtained by converting the inherent signal of the screened material.

[0043] Specifically, the preset deviation information can be understood as a benchmark or model used to evaluate the difference between the process state and the ideal state. In a preferred embodiment, the preset deviation information includes vibration spectrum deviation. Vibration spectrum deviation refers to the difference between the actual measured vibration signal spectrum and the ideal or benchmark vibration signal spectrum under specific processing conditions. This deviation can effectively reflect subtle changes caused by various factors such as tool wear, material property changes, and abnormal equipment conditions during processing.

[0044] When acquiring high-frequency and low-frequency vibration characteristics of profiles, data can be collected by installing miniature vibration sensors on the profiles or using non-contact measurement devices such as laser Doppler vibrometers. High-frequency vibration characteristics are usually related to factors such as the material's internal microstructure, grain boundary properties, and phonon scattering, while low-frequency vibration characteristics may be related to the material's macroscopic elastic modulus, density, and overall structural stiffness. The acoustic characteristics of the material extracted from these vibration characteristics refer to parameters that characterize the material's response to sound waves or vibration waves, such as sound velocity, damping coefficient, and acoustic impedance. The purpose is to establish the correlation between the material's physical properties and vibration response.

[0045] When collecting milling operation signals during the milling process in a milling unit, accelerometers, acoustic emission sensors, or force sensors installed on the milling unit can be used to monitor in real time the vibration, sound waves, or cutting force signals generated when the milling tool interacts with the profile. These signals contain a wealth of machining status information.

[0046] Specifically, based on the acoustic characteristics of the current profile, the inherent material signal is extracted from the milling operation signal. The inherent material signal can be understood as the signal component caused by the material's inherent properties (such as hardness, toughness, internal defects, etc.) during processing, excluding the influence of external factors such as equipment vibration and tool geometry. The purpose of extracting the inherent material signal is to isolate the influence of the material's inherent properties on the processing process, in order to more accurately assess the material condition.

[0047] In practical applications, the inherent signals of the material are filtered based on the vibration spectrum deviation. The filtering process can employ signal processing techniques such as spectrum analysis, wavelet transform, and Fourier transform to compare the spectrum of the material's inherent signal with a preset vibration spectrum deviation. For example, a vibration spectrum deviation threshold can be set; when the energy of a certain frequency band or a specific frequency component of the material's inherent signal exceeds this threshold, a deviation is considered to exist, and this is used as the basis for filtering. The aim is to identify signal components with significant spectral characteristics that are truly caused by material properties or processing anomalies.

[0048] Finally, the inherent signals of the screened materials are converted to obtain the upstream process deviation values. The conversion process maps the screened spectral characteristics, energy values, or specific frequency components to a quantified deviation value. For example, a mathematical model or lookup table can be established to convert the screened signal characteristics into a deviation index from 0 to 100, or specific physical quantity deviations, such as material hardness deviations or internal stress deviations, can be directly output. The purpose is to transform complex signal characteristics into easily understood and applied quantifiable indicators, providing a basis for subsequent process compensation.

[0049] Through the above technical solution, this application overcomes the shortcomings of traditional methods in identifying subtle deviations caused by material properties or processing dynamics. By introducing vibration spectrum deviation and combining it with the material's acoustic characteristics to filter inherent signals, the accuracy and sensitivity of upstream process deviation identification can be significantly improved. This enables the system to detect potential processing problems earlier and more accurately, such as changes in the microstructure inside the profile or early tool wear, thereby providing more precise data support for downstream process compensation and effectively improving the overall processing quality and stability of door and window products.

[0050] In some preferred embodiments, a specific example is given below. Suppose that in the milling process of door and window profiles, precise control of the surface quality and dimensional accuracy of the profiles is required. First, high-frequency and low-frequency vibration sensors are installed on the profiles to acquire their high-frequency and low-frequency vibration characteristics in their unprocessed state. Acoustic analysis algorithms are then used to extract the material acoustic characteristics of this batch of profiles, such as their natural frequency response curves. Next, during milling in the milling unit, an accelerometer installed near the milling cutter collects the milling operation signal in real time. Then, using the previously extracted material acoustic characteristics, signal decoupling technology is used to separate the material's inherent signal from the milling operation signal. This signal primarily reflects the material's response during milling. Subsequently, the spectrum of this material's inherent signal is compared with a preset vibration spectrum deviation model. For example, if the energy of the material's inherent signal within a specific frequency range is significantly higher or lower than the allowable range of the reference vibration spectrum deviation, a deviation is considered to exist. For example, when an abnormal increase in the energy of the material's inherent signal is detected in the 20kHz-30kHz frequency band, this may indicate the presence of microcracks or uneven hardness within the profile. Finally, based on the spectral characteristics of this selected material inherent signal, it is converted into a quantified upstream process deviation value, such as "material hardness deviation +5%" or "internal defect index 0.8," using a pre-trained machine learning model or empirical formula. This deviation value is then used to guide adjustments to process parameters in downstream welding processes, such as increasing preheating time or adjusting welding pressure, to ensure the quality of the final weld.

[0051] In practical applications, this application further proposes that the aforementioned equipment operating parameters include current signals, vibration signals, and load data; the steps for extracting and processing the aforementioned equipment operating parameters to obtain microscopic information include: High-frequency components are extracted from the current signal to obtain current fluctuation characteristics; The vibration signal is subjected to multi-band spectral decomposition to obtain high-frequency vibration characteristics and low-frequency vibration characteristics; Transient response analysis was performed on the load data to obtain load mutation characteristics; The current fluctuation characteristics, high-frequency vibration characteristics, low-frequency vibration characteristics, and load change characteristics are weighted and calculated to obtain microscopic information.

[0052] Specifically, the equipment operating parameters are set to include current signals, vibration signals, and load data. The current signal reflects the operating status of the equipment motor and load changes; its high-frequency components are often related to subtle processes such as tool wear and changes in cutting resistance. The vibration signal directly characterizes the dynamic behavior of mechanical components; vibration characteristics at different frequency bands can reveal information such as bearing wear, tool chatter, and material properties. The load data reflects the stress conditions of the equipment during processing; its transient response is crucial for identifying sudden events (such as tool chipping or material hardening).

[0053] Furthermore, when extracting and processing the above-mentioned equipment operating parameters to obtain microscopic information, the specific operations are as follows: First, high-frequency components are extracted from the current signal to obtain current fluctuation characteristics. High-frequency components can more sensitively capture subtle and rapidly changing load conditions during machining, such as instantaneous impacts or frictional changes when the tool comes into contact with the material. These changes are often difficult to detect with traditional average current.

[0054] Secondly, the vibration signal is subjected to multi-band spectral decomposition to obtain high-frequency and low-frequency vibration characteristics. Multi-band decomposition allows for refined analysis of the vibration signal. High-frequency vibration characteristics may be related to the microscopic state of the tool cutting edge, changes in the internal structure of the material, etc., while low-frequency vibration characteristics may reflect the overall stability of the equipment, the condition of the bearings, or the workpiece clamping condition.

[0055] Next, transient response analysis is performed on the load data to obtain the characteristics of load abrupt changes. Transient response analysis can identify drastic changes in load within a very short time, which is crucial for detecting abnormal events in the machining process (such as material defects or sudden tool obstruction).

[0056] Finally, the obtained current fluctuation characteristics, high-frequency vibration characteristics, low-frequency vibration characteristics, and load change characteristics are weighted and calculated to obtain comprehensive microscopic information. The purpose of the weighted calculation is to adjust the importance of different characteristics in representing the microscopic state of the process, so as to more comprehensively and accurately reflect the actual microscopic state of the current process.

[0057] Through the above technical solution, this application can obtain more comprehensive, detailed, and accurate microscopic information about the process. By conducting multi-dimensional and in-depth analysis of current signals, vibration signals, and load data, this application significantly improves the ability to perceive subtle changes during processing. Specifically, the comprehensive application of current fluctuation characteristics, high-frequency / low-frequency vibration characteristics, and load abrupt change characteristics enables the system to more sensitively capture potential problems such as tool wear, material defects, and equipment malfunctions, thereby effectively avoiding misjudgments or omissions caused by incomplete or inaccurate information. As a result, the generated microscopic information can more realistically and meticulously reflect the actual operating status of upstream processes, providing more reliable data support for subsequent deviation calculations and downstream process compensation, ultimately helping to improve the processing quality and production stability of door and window products.

[0058] In some preferred embodiments, the milling process of door and window profiles is used as an example for illustration. In the milling unit, corresponding sensors can be deployed to acquire equipment operating parameters. For example, a high-frequency current sensor can be installed on the spindle motor to collect current signals in real time; a multi-axis vibration sensor can be installed on the milling spindle or tool clamping area to collect vibration signals; simultaneously, a force sensor or torque sensor can be integrated into the feed mechanism or spindle drive system to acquire load data.

[0059] Specifically, when the milling unit is machining profiles, the current signal collected by the high-frequency current sensor is sent to the signal processing unit. Through a bandpass filter and a high-frequency component extraction algorithm, the current fluctuation characteristics related to the cutting process are separated. For example, when the cutting resistance of the tool changes slightly, the high-frequency components of the current signal will change accordingly. Simultaneously, the raw vibration signal collected by the vibration sensor undergoes multi-band spectral decomposition using Fourier transform or wavelet transform to obtain the energy distribution within different frequency ranges. For example, energy enhancement in a specific high-frequency band may indicate tool wear or chatter, while anomalies in the low-frequency band may be related to loose equipment structure. Furthermore, the load data collected by the force sensor or torque sensor undergoes transient response analysis. For example, by setting thresholds and time windows, sudden increases or decreases in load within a very short time can be identified, which may correspond to tool entry into hard spots or chipping.

[0060] Ultimately, these extracted current fluctuation features, high-frequency vibration features, low-frequency vibration features, and load abrupt change features are weighted according to a preset weighting model. For example, in tool wear monitoring, the weight of high-frequency vibration features may be set higher, while in material defect detection, the weight of load abrupt change features may be more important. Through this weighted combination, the system can generate a comprehensive microscopic information vector that can fully and accurately characterize the microscopic state of the current milling process, such as indicating the degree of tool wear, material uniformity, or the operational health of the equipment.

[0061] This application further proposes a data analysis method for door and window manufacturing processes, wherein the steps of processing microscopic information by pre-setting deviation information to obtain the upstream process deviation value include: Obtain aging and wear information of the milling unit; The aging and wear information is compared with the initial reference value to obtain the equipment error information caused by the overall aging of the equipment; The preset deviation information is dynamically corrected based on the equipment error information to obtain the corrected deviation information; The upstream process deviation value is obtained by processing the microscopic information with the corrected deviation information.

[0062] Specifically, acquiring aging and wear information of a milling unit can be understood as obtaining data reflecting the wear degree and aging state of the milling unit in real time or periodically through various sensors or data acquisition methods. For example, this can be achieved by monitoring the cutting force of the tool, changes in vibration spectrum, surface roughness, trends in motor current changes, or by acquiring tool edge wear images through a vision inspection system. This information can characterize the performance degradation or structural changes of the milling unit due to factors such as friction and fatigue during long-term operation.

[0063] The process of comparing aging and wear information with initial baseline values ​​to obtain equipment error information caused by overall equipment aging involves comparing the currently acquired aging and wear information with the performance parameters of the equipment in a brand-new or ideal state (i.e., the initial baseline values). The initial baseline values ​​can be performance test data from when the equipment left the factory or after a major overhaul. By comparing these values, the performance deviation caused by aging and wear can be quantified, such as a decrease in cutting efficiency or a reduction in machining accuracy, thus obtaining equipment error information. This equipment error information reflects the impact of changes in the equipment's own condition on the machining process.

[0064] In practical applications, dynamically correcting preset deviation information based on equipment error information to obtain corrected deviation information refers to adjusting and updating the original, potentially static, preset deviation information using the aforementioned quantified equipment error information. For example, if the equipment error information indicates that wear on the milling unit has caused changes in its vibration characteristics, the vibration-related thresholds or model parameters in the preset deviation information can be adjusted accordingly. This dynamic correction ensures that the preset deviation information can adapt to the actual operating conditions of the equipment in real time, making it more targeted and accurate.

[0065] Therefore, processing microscopic information with corrected deviation information to obtain upstream process deviation values ​​refers to applying preset deviation information, after dynamic correction, to the microscopic information processing. The corrected deviation information can more accurately identify and quantify deviations in microscopic information caused by process anomalies themselves, rather than "false" deviations caused by equipment aging and wear, thus obtaining more realistic and accurate upstream process deviation values.

[0066] Through the above technical solution, this application can significantly improve the accuracy and robustness of upstream process deviation identification. By considering the aging and wear factors of the equipment itself, it avoids misjudging inherent changes in the equipment's state as process deviations, thereby making the generated downstream process compensation parameters more accurate and effective. This not only helps improve the overall quality and efficiency of door and window processing, but also extends the service life of the equipment, reduces maintenance costs, and enhances the adaptability and intelligence of the entire production process.

[0067] In some preferred embodiments, a specific example is given below. Suppose that on a door and window manufacturing production line, a milling unit is responsible for the precision cutting of profiles. As this milling unit operates for an extended period, its cutting tools gradually wear down, and the spindle bearings also show slight aging. In conventional methods, the preset vibration spectrum deviation information may be set based on the new equipment status. When tool wear causes a change in the milling vibration mode, the system may incorrectly identify this wear-induced vibration change as a process deviation, thereby generating unnecessary or inaccurate process compensation parameters.

[0068] According to the solution of this application, the wear condition of the cutting tool and the operating condition of the spindle bearing are continuously monitored by sensors (e.g., vibration sensors, current sensors) installed on the milling unit to obtain aging and wear information of the milling unit. For example, by analyzing the long-term trend of the cutting force or vibration signal of the cutting tool, the degree of tool wear can be determined; by monitoring the current fluctuation or temperature change of the spindle motor, the aging condition of the bearing can be assessed.

[0069] Next, this real-time acquired aging and wear information is compared with the initial reference values ​​of the milling unit in a brand-new state. For example, if the energy peak of the vibration spectrum of a new tool in a specific frequency range is X, while the energy peak of the currently worn tool becomes Y, the equipment error information caused by tool wear can be calculated.

[0070] Then, based on this quantified equipment error information, the original preset deviation information is dynamically corrected. For example, if the equipment error information indicates that tool wear leads to an increase in vibration energy at certain frequencies, then when processing microscopic information, the threshold or weight of the corresponding frequency in the preset deviation information can be adjusted to reduce its impact on the deviation value calculation, or a new compensation factor can be introduced.

[0071] Ultimately, this corrected deviation information is used to process the microscopic information extracted from the equipment operating parameters to obtain the upstream process deviation value. This way, the obtained upstream process deviation value will more accurately reflect deviations caused by genuine process anomalies such as material defects and operational errors, rather than "noise" caused by the aging and wear of the equipment itself. For example, when the hardness of the profile material undergoes a slight change, the corrected system can more sensitively and accurately identify the deviation caused by this material characteristic and generate corresponding process compensation parameters. This ensures that downstream welding processes can be precisely adjusted according to the actual process deviation, avoiding misjudgments and overcompensation caused by equipment aging.

[0072] In addition, this application proposes a more refined and comprehensive method for obtaining operating parameters of upstream process equipment by using multi-source heterogeneous sensors to work together to obtain more representative data.

[0073] In some embodiments described above in this application, a step for obtaining equipment operating parameters of an upstream process is proposed. However, in its implementation, relying solely on conventional methods for obtaining equipment operating parameters may be insufficient to comprehensively and accurately capture the microscopic state of the process, thereby affecting the accuracy of subsequent data analysis. Therefore, this application further proposes that the aforementioned step for obtaining equipment operating parameters of an upstream process includes: An optical sensor array, a high-frequency acoustic emission sensor array, and a micro vibration sensor are set in the milling unit. The temperature distribution image of the milling cutting area in the milling unit is obtained through the optical sensor array. Acoustic emission signals from the milling cutting area are acquired using a high-frequency acoustic emission sensor array. The vibration signal of the cutting edge of the tool is obtained by using a miniature vibration sensor; Temperature distribution images, acoustic emission signals, and vibration signals are used to determine the equipment operating parameters of the upstream process.

[0074] Specifically, a milling unit refers to the equipment unit used in door and window processing to cut and shape profiles, and its working state directly affects product quality. An optical sensor array can be understood as a collection of multiple optical sensors that can monitor the surface temperature distribution of the milling cutting area in real time in a non-contact manner, generating a temperature distribution image. This temperature distribution image reflects the heat distribution generated during cutting, the friction between the tool and the material, and the local thermal deformation of the material. A high-frequency acoustic emission sensor array is a collection of sensors that can capture transient elastic waves generated during material deformation or fracture under stress, acquiring acoustic emission signals from the milling cutting area. This acoustic emission signal carries information about microscopic events such as microcrack propagation, phase transitions, or tool wear within the material. A miniature vibration sensor is a small, highly sensitive vibration measurement device placed near the cutting edge of the tool to acquire the vibration signal of the cutting edge in real time. This vibration signal can characterize the tool wear state, cutting stability, and dynamic changes in cutting force. By identifying three types of multi-dimensional, high-precision real-time data—temperature distribution images, acoustic emission signals, and vibration signals—as equipment operating parameters for upstream processes, a richer and more reliable data foundation can be provided for subsequent microscopic information extraction and deviation identification.

[0075] Through the aforementioned technical solution, this application can obtain more comprehensive, detailed, and spatiotemporally correlated operating parameters of upstream process equipment. It can more effectively capture subtle and complex process deviations such as tool micro-wear, internal material defects, and abnormal cutting heat distribution, thereby providing more accurate and reliable raw data input for subsequent micro-information extraction, deviation identification, and downstream process compensation. This multi-dimensional data acquisition method greatly improves the accuracy and robustness of data analysis methods throughout the entire door and window manufacturing process, contributing to more precise process control and product quality improvement.

[0076] This application further proposes a method for extracting and processing equipment operating parameters. By performing time synchronization calibration, timestamp marking, time alignment, and fusion processing on multi-source sensor data, it ensures that the obtained microscopic information can accurately and comprehensively characterize the microscopic state of the upstream process.

[0077] In this regard, this application further proposes steps for extracting and processing the operating parameters of the aforementioned equipment to obtain microscopic information, including: Time synchronization calibration is performed on the optical sensor array, the high-frequency acoustic emission sensor array, and the miniature vibration sensor. The temperature distribution image, acoustic emission signal, and vibration signal are timestamped. The temperature distribution image, acoustic emission signal, and vibration signal are time-aligned according to the timestamp marker to obtain the time-aligned temperature distribution image, acoustic emission signal, and vibration signal; The time-aligned temperature distribution image, acoustic emission signal, and vibration signal are fused to obtain microscopic information about the upstream process status.

[0078] Specifically, time synchronization calibration for optical sensor arrays, high-frequency acoustic emission sensor arrays, and miniature vibration sensors refers to ensuring, through hardware or software means, that data collected by different sensors at the same time can be accurately correlated. For example, a unified clock source can be used to synchronize all sensors, or time synchronization can be achieved through a Network Time Protocol (NTP) to eliminate time discrepancies between different sensors. The aim is to provide a unified time reference for subsequent data processing, ensuring the temporal consistency of multi-source data.

[0079] The time-stamping operation for temperature distribution images, acoustic emission signals, and vibration signals can be understood as attaching a precise timestamp to each data point or each data frame when it is acquired. This timestamp records the precise moment of data acquisition, for example, accurate to the millisecond or even microsecond level. Its purpose is to provide the necessary timing information for subsequent time alignment operations, enabling data from different sensors to be accurately matched according to their acquisition time.

[0080] In practical applications, time alignment of temperature distribution images, acoustic emission signals, and vibration signals based on timestamps involves using this timestamp information to adjust data from different sensors onto a common time axis through techniques such as interpolation, resampling, or sliding windows. For example, if the sampling rate of the acoustic emission signal is higher than that of the temperature distribution image, interpolation methods can be used to extend the temperature data points to the same time resolution as the acoustic emission signal, or all data can be resampled to the lowest common frequency. The aim is to eliminate time offsets caused by differences in sampling frequencies or data transmission delays, ensuring that data from different sensors accurately reflect the same state of the process at the same point in time.

[0081] Furthermore, fusing the time-aligned temperature distribution image, acoustic emission signal, and vibration signal to obtain microscopic information about the upstream process status refers to integrating time-aligned heterogeneous data to form a more comprehensive and robust description of the process status. For example, various fusion strategies can be employed, such as feature-level fusion, decision-level fusion, or data-level fusion. Feature-level fusion extracts features from each signal (such as temperature gradient, frequency domain energy, and vibration amplitude) and then combines these features. Its purpose is to comprehensively utilize the advantages of multi-source data, compensate for the limitations of single-sensor information, thereby obtaining more accurate and comprehensive microscopic information and effectively reducing noise and uncertainty.

[0082] Through the above technical solution, this application effectively solves the problem of inaccurate microscopic information caused by time asynchrony and data heterogeneity in traditional multi-sensor data processing. By strictly calibrating time synchronization, timestamp marking, and time alignment, the temporal consistency of multi-source data is ensured, enabling precise correlation of data collected by different sensors at the same time. Furthermore, by fusing the time-aligned multi-source data, temperature, acoustic, and vibration information can be comprehensively utilized to form a more comprehensive and robust description of the upstream process status, thereby significantly improving the accuracy and reliability of the acquired microscopic information. This accurate microscopic information is crucial for subsequent calculation of upstream process deviations and generation of process compensation parameters for downstream processes, effectively improving the intelligence level of door and window manufacturing and product quality stability.

[0083] In some preferred embodiments, a specific example is given below. Suppose that during the milling process of door and window profiles, it is necessary to monitor the operating status of the milling unit in real time to identify potential anomalies. To this end, an optical sensor array is deployed in the milling unit to acquire temperature distribution images of the milling cutting area, a high-frequency acoustic emission sensor array is used to acquire acoustic emission signals from the milling cutting area, and a micro-vibration sensor is used to acquire vibration signals from the cutting edge of the tool.

[0084] To ensure that this heterogeneous data accurately reflects the true state of the milling process, all sensors are first time-synchronized. For example, a high-precision master clock can be used to send synchronization signals to all sensors via wired or wireless means, ensuring that they begin data acquisition at the same frequency and phase.

[0085] Next, during the data acquisition process, each sensor adds a precise timestamp when generating temperature distribution images, acoustic emission signals, and vibration signals. For example, when the optical sensor array acquires a frame of temperature image at t=10.000 seconds, the image is marked as t=10.000 seconds; when the high-frequency acoustic emission sensor array acquires a segment of acoustic emission signal at t=10.001 seconds, the signal is marked as t=10.001 seconds.

[0086] Subsequently, based on these timestamps, the acquired temperature distribution images, acoustic emission signals, and vibration signals are time-aligned. For example, if the vibration signal has the highest sampling frequency, it can be used as a reference. Then, through methods such as linear interpolation or spline interpolation, the data points of the temperature image and acoustic emission signal are adjusted to the same time axis as the vibration signal, ensuring that at any given time point, temperature, acoustic, and vibration data corresponding to that moment can be obtained.

[0087] Finally, the time-aligned temperature distribution image, acoustic emission signal, and vibration signal are fused. For example, temperature gradient features of local hotspot areas in the temperature image, specific frequency band energy features in the acoustic emission signal, and principal vibration frequency and amplitude features in the vibration signal can be extracted. These features are then weighted and combined or fused using a machine learning model to obtain comprehensive microscopic information that accurately characterizes the tool wear, material cutting status, or the presence of abnormal vibrations in the current milling process. This approach provides a more comprehensive and accurate understanding of the upstream process status, offering a reliable data foundation for subsequent deviation identification and process compensation.

[0088] In some embodiments described above in this application, a method is proposed to fuse time-aligned temperature distribution images, acoustic emission signals, and vibration signals to obtain microscopic information about the upstream process status. Specifically, the step of fusing time-aligned temperature distribution images, acoustic emission signals, and vibration signals to obtain microscopic information about the upstream process status may include the following operations: The temperature distribution image after time alignment is segmented into regions to obtain the temperature gradient distribution characteristics of the cutting area. Time-frequency analysis was performed on the time-aligned acoustic emission signal to obtain the frequency band energy characteristics. Wavelet decomposition was performed on the time-aligned vibration signal to obtain vibration mode features; Establish the correlation mapping relationship between the temperature gradient distribution characteristics, frequency band energy characteristics, and vibration mode characteristics; Based on the aforementioned correlation mapping relationship, the temperature gradient distribution characteristics, frequency band energy characteristics, and vibration mode characteristics are weighted and combined to obtain the microscopic information of the upstream process status.

[0089] Specifically, region segmentation refers to dividing a time-aligned temperature distribution image into multiple regions with specific physical meanings. For example, separating the cutting region from the non-cutting region in the image allows for more precise focusing on temperature changes in key areas. This yields the temperature gradient distribution characteristics of the cutting region, which characterizes the rate and direction of temperature change in the cutting region spatially, reflecting the dynamic processes of heat generation, conduction, and dissipation during the cutting process.

[0090] Time-frequency analysis can be understood as the simultaneous analysis of the time-aligned acoustic emission signal in both the time and frequency domains. Methods such as short-time Fourier transform, wavelet transform, or Hilbert-Huang transform are used to reveal the frequency components and energy distribution of the signal at different times. Through time-frequency analysis, frequency band energy characteristics can be obtained, specifically the energy intensity of the acoustic emission signal within a specific frequency range. This effectively characterizes the energy release from microscopic events such as material deformation, friction, and fracture during the cutting process.

[0091] In practical applications, wavelet decomposition specifically involves multi-scale decomposition of time-aligned vibration signals, breaking them down into components of different frequencies (e.g., high-frequency detail components and low-frequency approximation components). This allows the capture of local features of the signal at different time scales. The aim is to obtain vibration mode features that reflect the vibration patterns and intensities of the tool or workpiece at different frequencies during the cutting process, such as specific vibration modes caused by tool wear, cutting chatter, etc.

[0092] Furthermore, establishing the correlation mapping relationship between the temperature gradient distribution characteristics, frequency band energy characteristics, and vibration mode characteristics refers to constructing a mathematical or logical model of the interaction between these multimodal characteristics through data-driven or model-driven methods. For example, machine learning algorithms (such as neural networks, support vector machines, decision trees, etc.) can be used to train historical data to learn the intrinsic relationship between different feature combinations and the micro-state of upstream processes.

[0093] Based on this, the temperature gradient distribution characteristics, frequency band energy characteristics, and vibration mode characteristics are weighted and combined according to the aforementioned correlation mapping relationship. The purpose is to comprehensively utilize the information from different characteristics to more comprehensively and accurately characterize the microscopic information of the upstream process status. The weight allocation can be preset or dynamically adjusted according to the sensitivity and reliability of each characteristic to the process status and actual application requirements, thereby achieving effective fusion of multi-source heterogeneous data.

[0094] Through the above technical solution, this application enables more refined and comprehensive extraction and characterization of microscopic information about the upstream process status. Compared to simple data fusion, this solution significantly improves the sensitivity and accuracy of microscopic information to changes in process status by selectively extracting features (such as temperature gradient, frequency band energy, and vibration modes) from different types of sensor data and further establishing weighted combinations of the correlation mapping relationships between these features. This allows for more effective identification of subtle deviations that are difficult to detect using traditional methods, such as early tool wear and changes in material microstructure, providing a more reliable and abundant data foundation for subsequent deviation identification and process compensation, thereby enhancing the intelligence level of the door and window manufacturing process and the stability of product quality.

[0095] This application further proposes an optimization scheme aimed at improving the ability to identify and process complex or unknown deviation types by dynamically adjusting the weight allocation in the correlation mapping relationship to more accurately obtain micro-information on the status of upstream processes.

[0096] The steps described above for weighted combination of the temperature gradient distribution characteristics, frequency band energy characteristics, and vibration mode characteristics based on the aforementioned correlation mapping relationship to obtain the microscopic information of the upstream process status include: When an unknown deviation type is detected, the weight allocation in the associated mapping relationship is adjusted based on the real-time changes in the current temperature gradient distribution characteristics, frequency band energy characteristics, and vibration mode characteristics, as well as the correlation between the temperature gradient distribution characteristics, frequency band energy characteristics, and vibration mode characteristics and the identified deviation type, to obtain the microscopic information of the upstream process status; the deviation type represents a complex subtle deviation of tool wear and profile microstructure changes.

[0097] Specifically, "unknown deviation types" refer to process anomalies not explicitly included in the system's pre-training or settings, or combinations thereof exceeding the recognition capabilities of existing models. These deviations may be caused by a combination of factors, such as the simultaneous existence of localized micro-chipping of the cutting tool and abnormal grain structure within the profile, resulting in complex characteristics that are difficult to categorize individually. "Real-time changes" refer to the continuous monitoring and dynamic analysis of temperature gradient distribution characteristics, frequency band energy characteristics, and vibration mode characteristics over time to capture their instantaneous or trend-based fluctuations. The real-time changes of these characteristics can be understood as the continuous acquisition of data through a sensor array during the machining process, followed by real-time processing and feature extraction to obtain the feature values ​​and their rates of change within the current moment or a short time window.

[0098] "Correlation between identified deviation types" refers to the mapping relationship or statistical correlation established by the system through historical data learning and expert experience accumulation between various known deviation types (e.g., single tool wear, single profile hardness anomaly, etc.) and temperature gradient distribution characteristics, frequency band energy characteristics, and vibration mode characteristics. For example, a specific vibration mode may be highly correlated with tool wear, while a certain temperature gradient change may be related to a decrease in material cutting performance.

[0099] "Adjusting the weight allocation in the associated mapping relationship" refers to dynamically modifying the proportion of each feature in the feature fusion process based on the real-time characteristics of the detected unknown deviation types and the similarity or deviation between these real-time features and the known deviation types. For example, if the real-time data shows a composite pattern similar to known tool wear characteristics but also has microstructural change characteristics of new materials, the system will adaptively adjust the weights of temperature gradient, frequency band energy, and vibration mode based on the original associated mapping relationship according to this composite pattern, so as to more accurately reflect the true state of the current process.

[0100] "Complex subtle deviations characterized by tool wear and profile microstructure changes" refer to the small and complex anomalies that affect machining quality, caused by the combined effects of tool condition (such as wear and chipping) and profile characteristics (such as internal defects and grain inhomogeneity), which this solution focuses on and can handle. These deviations are often difficult to accurately judge using a single indicator and require comprehensive analysis of multi-source micro-features.

[0101] This application's solution effectively addresses the limitations of traditional fixed-weight combination methods in handling complex or unknown process deviations by introducing a dynamic adaptation mechanism for unknown deviation types. Specifically, when the system detects a deviation that cannot be directly categorized into a known type, it no longer rigidly applies a preset correlation mapping relationship. Instead, it utilizes real-time changes in the current temperature gradient distribution characteristics, frequency band energy characteristics, and vibration mode characteristics, and analyzes the correlation between these real-time characteristics and the identified deviation types. For example, if the real-time vibration signal shows a pattern similar to minor tool wear, but the temperature gradient exhibits characteristics related to hard spots in a certain material, the system will identify this as a possible complex minor deviation.

[0102] It is precisely because of this comprehensive consideration of the correlation between real-time feature changes and known deviations that the system can intelligently adjust the weight allocation in the correlation mapping relationship. For example, in the aforementioned composite deviation scenario, the system may increase the weight of vibration mode features in reflecting tool condition, increase the weight of temperature gradient features in reflecting material condition, and adjust the weight of frequency band energy features accordingly. This allows the microscopic information of the upstream process state obtained by final fusion to more accurately characterize the composite subtle deviation of tool wear and profile microstructure changes. This dynamic adjustment mechanism enables the system to understand and quantify complex process anomalies from multiple dimensions and levels, avoiding misjudgments or information omissions caused by single features or fixed weights.

[0103] In some preferred embodiments, this application is implemented as follows: Assuming that during the milling process of door and window profiles, the system continuously collects data through an optical sensor array, a high-frequency acoustic emission sensor array, and a micro vibration sensor, and extracts temperature gradient distribution characteristics, frequency band energy characteristics, and vibration mode characteristics. At a certain moment, the system detects that the combination pattern of these characteristics does not perfectly match any known single tool wear or single profile defect pattern, and is identified as an "unknown deviation type".

[0104] Specifically, the vibration pattern characteristics show slight signs of tool cutting edge dulling, but its spectral energy distribution is slightly abnormal, and the temperature gradient distribution characteristics exhibit an unusual peak in a localized area of ​​the cutting region. The system performs correlation analysis based on these real-time changing characteristics with tool wear patterns and profile hard spot patterns identified in the historical database. For example, the system might find that the current vibration pattern has 80% similarity to a certain degree of tool wear, while the temperature peak has 70% similarity to localized overheating caused by inclusions inside a certain profile.

[0105] Based on this correlation analysis, the system dynamically adjusts the weights used when fusing these features. For example, if under normal circumstances, vibration mode features have a weight of 0.5 when determining tool condition and temperature gradient features have a weight of 0.3 when determining material condition, upon detecting a combined deviation, the system might adjust the weight of vibration mode features to 0.6 (to further emphasize tool condition), while simultaneously adjusting the weight of temperature gradient features to 0.4 (to further emphasize local material anomalies), and correspondingly adjusting the weight of frequency band energy features. Through this adaptive weight adjustment, the system can more accurately reflect the true process state under the combined effects of tool wear and profile microstructure changes, thereby generating more precise microscopic information and guiding subsequent process compensation.

[0106] This application further proposes a data analysis method for door and window processing, wherein the process compensation parameters include a first parameter; the process parameters of the downstream process include the preheating time of the welding heating plate, the instantaneous temperature rise rate, and the machine head propulsion pressure curve during the pressure holding stage; the step of generating process compensation parameters for the downstream process based on the deviation value of the upstream process, and obtaining the process parameters of the downstream process through the process compensation parameters includes: After receiving the deviation value from the upstream process, the equipment operating status information of the current welding unit is obtained, including the wear level of the welding equipment and the ambient temperature. The upstream process deviation value, the wear degree of the welding equipment, and the ambient temperature are analyzed to identify a first parameter affecting the weld quality; the first parameter refers to the parameter that identifies and quantifies the influence of multiple factors on the weld quality. Based on the first parameter, adjust the preheating time of the welding heating plate, the instantaneous temperature rise rate, and the head propulsion pressure curve during the pressure holding stage.

[0107] Specifically, the process compensation parameter is defined as the first parameter, which aims to identify and quantify the impact of various factors on weld quality. These factors include not only deviations from upstream processes but also equipment operating status information of the downstream welding unit itself, such as the wear and tear of the welding equipment and ambient temperature. By comprehensively considering these factors, the first parameter can more comprehensively and accurately reflect the potential variables affecting weld quality. Specifically, the downstream process parameters refer to the key control variables that directly affect weld formation and quality during door and window welding, including the preheating time of the welding heating plate, the instantaneous temperature rise rate, and the die head propulsion pressure curve during the holding pressure stage. Precise control of these parameters is crucial for obtaining high-quality welds.

[0108] In practical applications, after receiving deviation values ​​from upstream processes, the system further acquires information on the current operating status of the welding unit's equipment. This information specifically includes the wear level of the welding equipment, such as monitoring the wear or service life of key components through sensors, and the ambient temperature, such as real-time acquisition of the welding area temperature through environmental sensors. Subsequently, the upstream process deviation values, the wear level of the welding equipment, and the ambient temperature are analyzed in depth to identify and quantify their combined impact on weld quality, thereby obtaining the first parameter. Finally, based on the identified first parameter, the preheating time of the welding heating plate, the instantaneous temperature rise rate, and the head propulsion pressure curve during the holding pressure stage are precisely adjusted to achieve optimized control of the downstream welding process.

[0109] Through the above technical solution, this application can significantly improve the control precision and weld quality of downstream welding processes in door and window manufacturing. By fully considering various real-time influencing factors such as upstream process deviations, welding equipment wear, and ambient temperature, the generated process compensation parameters can more accurately reflect actual needs, thereby enabling precise adjustments to the preheating time of the welding heating plate, the instantaneous temperature rise rate, and the die head propulsion pressure curve during the pressure holding stage. This not only effectively compensates for the shortcomings of traditional methods under complex and variable working conditions but also enhances the robustness and adaptability of the entire processing, ultimately ensuring that the welding quality of door and window products reaches higher standards and reducing scrap rates and rework costs.

[0110] In some preferred embodiments, assuming that in the milling process of window and door profiles, upstream data analysis identifies a minor deviation in the surface roughness of the profile cutting surface caused by tool wear, with this deviation value determined to be 0.05 mm, the system first obtains the current equipment operating status information of the welding unit when transmitting this deviation value to the downstream welding unit. For example, a sensor detects that the welding heating plate has accumulated 5000 hours of operation, and its wear level has reached a preset threshold. Simultaneously, an ambient temperature sensor shows that the current workshop temperature is 15°C, which is 20°C lower than the standard operating temperature. At this point, the system inputs the upstream process deviation value (0.05 mm), the wear level of the welding equipment (5000 hours), and the ambient temperature (15°C) into a preset analysis model. This model, through machine learning algorithms or an expert system, comprehensively evaluates the impact of these factors on the final weld strength, appearance, and sealing performance, thereby identifying and quantifying the first parameter affecting weld quality. For example, the model calculates that due to the combined effects of upstream deviation, equipment wear, and low-temperature environment, the weld strength may decrease by 10%, and minor defects may appear in the appearance. Based on this first parameter, the system automatically adjusts the process parameters of the downstream welding process. Specifically, to compensate for these adverse effects, the preheating time of the welding heating plate may be extended by 10 seconds, the instantaneous temperature rise rate may be increased by 5°C / second, and the pressure curve of the die head during the pressure holding phase may be adjusted to increase the pressure by 5% within a specific time period. Through this dynamic, multi-factor comprehensive adjustment, even in the presence of upstream deviations, equipment wear, and changes in ambient temperature, the final quality of the door and window welds can be ensured to meet design requirements.

[0111] Secondly, referring to Figure 2 This application further proposes a data analysis system for the door and window manufacturing process, the system comprising: The parameter acquisition module 201 is used to acquire the equipment operating parameters of the upstream process. Information extraction module 202 is used to extract and process the operating parameters of the device to obtain microscopic information; The deviation identification module 203 is used to process the microscopic information through preset deviation information to obtain the upstream process deviation value; The compensation parameter module 204 is used to generate process compensation parameters for the downstream process based on the deviation value of the upstream process, and to obtain the process parameters of the downstream process through the process compensation parameters. The operation monitoring module 205 is used to monitor the actual operating parameters of the downstream process. The parameter adjustment module 206 is used to adjust the process parameters of the downstream process according to the deviation between the actual operating parameters and the desired state parameters, so as to obtain the adjusted process parameters of the downstream process.

[0112] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for analyzing data during the manufacturing process of doors and windows, characterized in that, Includes the following steps: Obtain the equipment operating parameters of upstream processes; The operating parameters of the equipment are extracted and processed to obtain microscopic information; the microscopic information represents the microscopic state of the process. The upstream process deviation value is obtained by processing the microscopic information through preset deviation information; Based on the deviation value of the upstream process, process compensation parameters for the downstream process are generated, and the process parameters of the downstream process are obtained through the process compensation parameters. Monitor the actual operating parameters of the downstream processes; Based on the deviation between the actual operating parameters and the desired state parameters, the process parameters of the downstream process are adjusted to obtain the adjusted process parameters of the downstream process.

2. The method for analyzing data during the door and window manufacturing process according to claim 1, characterized in that, The preset deviation information includes vibration spectrum deviation; the step of processing the microscopic information using the preset deviation information to obtain the upstream process deviation value includes: The high-frequency vibration characteristics and low-frequency vibration characteristics of the profile are obtained, and the acoustic characteristics of the material are extracted from the high-frequency vibration characteristics and low-frequency vibration characteristics. Collect milling operation signals during the milling process of the milling unit; Based on the acoustic characteristics of the current profile material, the inherent material signal is extracted from the milling operation signal; the inherent material signal represents the signal component caused by the material's own properties. The material's inherent signals are filtered by the vibration spectrum deviation to obtain the filtered material inherent signals; The upstream process deviation value is obtained by converting the inherent signal of the screened material.

3. The method for analyzing data during the door and window manufacturing process according to claim 1, characterized in that, The equipment operating parameters include current signals, vibration signals, and load data; the step of extracting and processing the equipment operating parameters to obtain microscopic information includes: High-frequency components are extracted from the current signal to obtain current fluctuation characteristics; The vibration signal is subjected to multi-band spectral decomposition to obtain high-frequency vibration characteristics and low-frequency vibration characteristics; Transient response analysis was performed on the load data to obtain load mutation characteristics; The current fluctuation characteristics, high-frequency vibration characteristics, low-frequency vibration characteristics, and load change characteristics are weighted and calculated to obtain microscopic information.

4. The method for analyzing data during the door and window manufacturing process according to claim 1, characterized in that, The step of processing the microscopic information using preset deviation information to obtain the upstream process deviation value includes: Obtain aging and wear information of the milling unit; The aging and wear information is compared with the initial reference value to obtain the equipment error information caused by the overall aging of the equipment; The preset deviation information is dynamically corrected based on the equipment error information to obtain the corrected deviation information; The upstream process deviation value is obtained by processing the microscopic information with the corrected deviation information.

5. The method for analyzing data during the manufacturing process of doors and windows according to claim 1, characterized in that, The acquisition of equipment operating parameters for upstream processes includes: An optical sensor array, a high-frequency acoustic emission sensor array, and a micro vibration sensor are set in the milling unit. The temperature distribution image of the milling cutting area in the milling unit is obtained through the optical sensor array. Acoustic emission signals from the milling cutting area are acquired using a high-frequency acoustic emission sensor array. The vibration signal of the cutting edge of the tool is obtained by using a miniature vibration sensor; Temperature distribution images, acoustic emission signals, and vibration signals are used to determine the equipment operating parameters of the upstream process.

6. The method for analyzing data during the manufacturing process of doors and windows according to claim 5, characterized in that, The step of extracting and processing the operating parameters of the device to obtain microscopic information includes: Time synchronization calibration is performed on the optical sensor array, the high-frequency acoustic emission sensor array, and the miniature vibration sensor. The temperature distribution image, acoustic emission signal, and vibration signal are timestamped. The temperature distribution image, acoustic emission signal, and vibration signal are time-aligned according to the timestamp markers to obtain the time-aligned temperature distribution image, acoustic emission signal, and vibration signal; The time-aligned temperature distribution image, acoustic emission signal, and vibration signal are fused to obtain microscopic information about the upstream process status.

7. The method for analyzing data during the door and window manufacturing process according to claim 6, characterized in that, The step of fusing the time-aligned temperature distribution image, acoustic emission signal, and vibration signal to obtain microscopic information about the upstream process status includes: The time-aligned temperature distribution image is segmented to obtain the temperature gradient distribution characteristics of the cutting area. Time-frequency analysis was performed on the time-aligned acoustic emission signal to obtain the frequency band energy characteristics. Wavelet decomposition was performed on the time-aligned vibration signal to obtain vibration mode features; Establish the correlation mapping relationship between the temperature gradient distribution characteristics, frequency band energy characteristics, and vibration mode characteristics; Based on the aforementioned correlation mapping relationship, the temperature gradient distribution characteristics, frequency band energy characteristics, and vibration mode characteristics are weighted and combined to obtain the microscopic information of the upstream process status.

8. The method for analyzing data during the manufacturing process of doors and windows according to claim 7, characterized in that, The step of weighting and combining the temperature gradient distribution characteristics, frequency band energy characteristics, and vibration mode characteristics according to the correlation mapping relationship to obtain the microscopic information of the upstream process status includes: When an unknown deviation type is detected, the weight allocation in the associated mapping relationship is adjusted based on the real-time changes in the current temperature gradient distribution characteristics, frequency band energy characteristics, and vibration mode characteristics, as well as the correlation between the temperature gradient distribution characteristics, frequency band energy characteristics, and vibration mode characteristics and the identified deviation type, to obtain the microscopic information of the upstream process status; the deviation type represents a complex subtle deviation of tool wear and profile microstructure changes.

9. The method for analyzing data during the manufacturing process of doors and windows according to claim 7, characterized in that, The process compensation parameters include a first parameter; the process parameters of the downstream process include the preheating time of the welding heating plate, the instantaneous temperature rise rate, and the die head propulsion pressure curve during the pressure holding stage; the step of generating the process compensation parameters of the downstream process based on the deviation value of the upstream process, and obtaining the process parameters of the downstream process through the process compensation parameters, includes: After receiving the deviation value from the upstream process, the equipment operating status information of the current welding unit is obtained, including the wear level of the welding equipment and the ambient temperature. The upstream process deviation value, the wear degree of the welding equipment, and the ambient temperature are analyzed to identify a first parameter affecting the weld quality; the first parameter refers to the parameter that identifies and quantifies the influence of multiple factors on the weld quality. Based on the first parameter, adjust the preheating time of the welding heating plate, the instantaneous temperature rise rate, and the head propulsion pressure curve during the pressure holding stage.

10. A data analysis system for door and window manufacturing processes, characterized in that, The system includes: The parameter acquisition module is used to acquire the operating parameters of the equipment in the upstream process. The information extraction module is used to extract and process the operating parameters of the device to obtain microscopic information; The deviation identification module is used to process the microscopic information through preset deviation information to obtain the deviation value of the upstream process; The compensation parameter module is used to generate process compensation parameters for the downstream process based on the deviation value of the upstream process, and to obtain the process parameters of the downstream process through the process compensation parameters. The operation monitoring module is used to monitor the actual operating parameters of the downstream process; The parameter adjustment module is used to adjust the process parameters of the downstream process based on the deviation between the actual operating parameters and the desired state parameters, so as to obtain the adjusted process parameters of the downstream process.

Images