Sealing ring trimming equipment and self-adaptive control system thereof

Abstract

The invention discloses sealing ring trimming equipment and a self-adaptive control system thereof. The sealing ring trimming equipment comprises a combined cutting device, a detection device and the self-adaptive control system. The combined type cutting device integrates a rough cutting assembly and a mechanical fine cutting assembly, and rough cutting and fine trimming of a sealing ring blank can be completed in sequence. The detection device adopts a multispectral confocal sensor, and the self-adaptive control system comprises a rough cutting module which controls a rough cutting assembly to reserve a set flash; the fine cutting control module dynamically generates a tool path correction parameter and a cutting depth compensation value according to the real-time scanning data; the edge feature analysis module extracts key quality indexes such as residual flash contours and surface roughness; and the correction module dynamically adjusts the reserved flash amount and the fine cutting parameters through the fuzzy logic controller. The equipment is further provided with a cutter resistance monitoring module and an abrasion prediction compensation module, closed-loop control from rough cutting to fine cutting is achieved, the edge machining precision and production efficiency of the sealing ring are remarkably improved, and the equipment is suitable for the field of high-precision sealing element manufacturing.

Description

Technical Field

[0001] The present invention relates to the field of sealing ring trimming equipment, in particular to a sealing ring trimming equipment and an adaptive control system thereof. Background Art

[0002] In the field of sealing ring manufacturing, the trimming process is crucial to product quality. Traditional sealing ring trimming equipment usually adopts a single cutting method or a segmented rough and fine cutting process, which has many technical bottlenecks: First, the rough cutting and fine cutting links are independent of each other, lacking a real-time feedback mechanism, making it difficult to dynamically adjust the reserved flash amount and cutting parameters according to the actual processing status, resulting in unstable edge accuracy; second, the detection methods are lagging behind, mostly relying on manual sampling or offline detection, and it is impossible to obtain real-time edge three-dimensional morphology data during the processing process, making it difficult to effectively control surface roughness and micro-defects; third, the tool status monitoring capability is insufficient, and there is a lack of real-time compensation for key parameters such as tool wear and resistance changes, which easily leads to cutting depth deviation and path error; fourth, the influence of environmental factors (such as temperature changes and mechanical vibrations) on the deformation of the sealing ring material is not effectively controlled, resulting in greater external interference in the processing accuracy. The shortcomings of existing technologies have led to low efficiency in sealing ring trimming and difficulty in improving the yield rate. There is an urgent need for a high-precision trimming equipment with intelligent detection, adaptive control and multi-factor compensation. Summary of the Invention

[0003] The purpose of this application is to provide a sealing ring trimming device and its adaptive control system, which have the advantages of improving the sealing ring trimming accuracy, stability and yield rate.

[0004] In order to solve the above technical problems, the present invention is solved by the following technical solutions: a sealing ring trimming device and an adaptive control system thereof, comprising:

[0005] Combined cutting device, rough cutting component and mechanical fine cutting component integrated on the same working platform;

[0006] A detection device, comprising a multispectral confocal sensor for detecting the sealing ring on the workbench;

[0007] The adaptive control system is electrically connected to the combined cutting device and the detection device, and includes:

[0008] Rough cutting module controls the rough cutting component to perform rough cutting on the sealing ring blank and reserve a set amount of flash on the edge;

[0009] The precision cutting control module dynamically generates the tool path correction parameters and cutting depth compensation values ​​for mechanical precision cutting based on the real-time scanning data of the sealing ring edge after rough cutting by the detection device, and controls the mechanical precision cutting component to precisely trim the reserved flash edge;

[0010] The edge feature analysis module cooperates with the detection device to obtain the three-dimensional morphology data of the edge of the sealing ring after precision cutting, and extracts at least two of the residual flash profile, surface roughness, and micro-defect density as key quality indicators;

[0011] The correction module dynamically adjusts the reserved burr amount setting value of the next sealing ring rough cutting operation and / or the cutting parameters of the fine cutting control module according to the key quality indicators.

[0012] By adopting the above technical solution, the claim defines the core architecture of the equipment, and realizes the integrated operation of rough cutting and fine cutting through a combined cutting device, avoiding the positioning error of traditional segmented processing; the non-contact detection of the multi-spectral confocal sensor can obtain the three-dimensional edge morphology data in real time, solving the lag problem of offline detection; the adaptive control system constructs a closed-loop control logic of "rough cutting-detection-fine cutting-feedback correction", and through dynamic adjustment of the reserved flash amount and cutting parameters, the edge accuracy control is transformed from "experience-driven" to "data-driven", significantly improving the processing consistency, which is especially suitable for the precision trimming of elastic material sealing rings.

[0013] The present invention is further configured such that the rough cutting component is cut by laser or mechanical tool.

[0014] By adopting the above technical solutions, it is clear that rough-cut components can be cut by laser cutting or mechanical tool cutting. Laser cutting is suitable for fast rough cutting of high-precision, complex-contour blanks to avoid material deformation caused by mechanical stress; mechanical tool cutting is suitable for efficient removal of excess of high-hardness or large-size sealing rings. The compatibility design of the two methods enables the equipment to adapt to the processing requirements of different materials (such as rubber, silicone, and metal composite sealing rings), enhancing the versatility of the equipment.

[0015] The present invention is further configured as follows: the adaptive control system includes a tool resistance monitoring module, and the tool resistance monitoring module is configured with a torque sensor integrated in the mechanical fine cutting component; when the rough cutting component adopts a mechanical tool for cutting, the torque sensor is also integrated in the rough cutting component.

[0016] By adopting the above technical solution, the tool resistance monitoring module collects tool load data in real time during rough cutting / fine cutting through the torque sensor integrated in the cutting component: when mechanical tools are used for rough cutting, tool overload caused by excessive blank flash can be prevented; during fine cutting, the sudden change in resistance caused by uneven material hardness can be dynamically identified, and the cutting depth can be adjusted in conjunction with the adaptive control system to avoid edge cracking or incomplete cutting due to abnormal resistance, thereby improving the safety of the processing process.

[0017] The present invention is further configured as follows: the adaptive control system includes a tool wear prediction and compensation module, which predicts the wear state of mechanical precision cutting tools based on historical cutting data, current cutting parameters, key quality indicator change trends and real-time monitoring data of tool resistance, and adds wear compensation when generating tool path correction parameters and cutting depth compensation values.

[0018] The present invention is further configured such that: the tool wear prediction and compensation module includes a prediction unit for establishing and continuously updating a historical cutting torque variation curve of a fine cutting tool and a rough cutting component cut by a mechanical tool.

[0019] By adopting the above technical solutions, the tool wear prediction and compensation module builds a tool wear mathematical model based on historical cutting data, real-time resistance monitoring and quality index change trends:

[0020] By continuously updating the historical cutting torque change curve, the early stages of tool edge wear can be accurately identified. Compared with the traditional periodic tool change mode, this not only avoids the accuracy degradation caused by excessive wear, but also reduces tool waste.

[0021] Wear compensation is automatically taken into account when generating fine cutting parameters to compensate for insufficient cutting depth or path deviation caused by tool blunting, ensuring long-term machining accuracy stability.

[0022] The present invention is further configured as follows: the tool wear prediction and compensation module includes an early warning unit, which generates a fine cutting tool wear early warning signal when the current value or growth rate of the torque of the mechanical fine cutting component exceeds a dynamic threshold calculated based on a historical model; calculates the real-time difference between the rough cutting torque and the fine cutting torque in the processing of the same workpiece or the same batch of materials, establishes and updates a difference baseline model, and generates a tool or rough cutting parameter abnormality early warning when the deviation between the real-time difference and the baseline model exceeds a set tolerance.

[0023] The present invention is further configured as follows: the tool wear prediction and compensation module integrates wear warning, abnormal state warning and edge quality changes monitored by multi-spectral confocal monitoring to generate tool replacement suggestions or process parameter adjustment instructions.

[0024] By adopting the above technical solutions, the alarm unit improves equipment reliability through a dual early warning mechanism:

[0025] Dynamic threshold warning: Set torque growth thresholds based on historical models to capture sudden tool wear such as chipping in real time, avoiding batch scrapping.

[0026] Difference baseline model: By analyzing the real-time difference in torque between rough cutting and fine cutting in the same batch of processing, it can identify abnormal rough cutting parameters such as fluctuations in reserved flash amount or tool installation deviation. Combined with edge quality monitoring data, it generates tool replacement suggestions or process adjustment instructions, realizing preventive maintenance and reducing downtime for inspection.

[0027] The present invention is further configured such that: the multispectral confocal sensor performs non-contact three-dimensional contour scanning on the edge of the sealing ring through the principle of multi-wavelength light interference, thereby realizing real-time topography data collection.

[0028] By adopting the above technical solution, the multi-spectral confocal sensor uses the principle of multi-wavelength light interference to achieve non-contact three-dimensional contour scanning of the edge of the sealing ring: compared with traditional contact probe detection, it avoids the deformation error of the elastic material caused by contact pressure; compared with single spectrum detection, multi-wavelength fusion can adapt to high-precision measurement of surfaces of different colors and roughness, ensuring the reliability of morphology data in complex scenarios such as irregular burrs after rough cutting and changes in surface reflectivity after fine cutting, providing precise input for adaptive control.

[0029] The present invention is further configured as follows: the surface of the working platform is provided with a shock-absorbing buffer layer and an integrated temperature compensation device for maintaining the temperature of the sealing ring material within a set constant range during the cutting process to reduce thermal deformation errors caused by temperature changes.

[0030] By adopting the above technical solutions, the shock-absorbing buffer layer and temperature compensation device solve two major environmental interference factors:

[0031] The vibration-damping design reduces the impact of table vibration on precision cutting, such as micron-level cutting depth control, and avoids tool chatter or edge ripple defects caused by vibration;

[0032] Temperature compensation maintains constant material temperature and eliminates thermal expansion errors caused by temperature changes in polymer materials such as rubber. For example, a temperature fluctuation of ±5°C can lead to a linear dimensional change of 0.3%-0.5%, ensuring that the reserved flash amount during fine cutting accurately matches the actual cutting amount.

[0033] The present invention is further configured such that: the correction module has a built-in fuzzy logic controller, which dynamically optimizes the matching parameters of the rough cutting reserved burr amount and the fine cutting feed speed according to the coupling relationship between the residual burr profile deviation and the surface roughness.

[0034] By adopting the above technical solution, the fuzzy logic controller targets the coupling relationship between the "residual flash contour deviation" and the "surface roughness" in the trimming of the sealing ring. For example, insufficient reserved flash amount can easily lead to contour deviation, and excessive feed speed will aggravate the roughness. The matching parameters of the rough cutting reserve amount and the fine cutting feed speed are dynamically optimized through the fuzzy algorithm: Compared with traditional PID control, it can handle multi-variable nonlinear relationships. Under complex working conditions such as material hardness fluctuations and different degrees of tool wear, it can automatically optimize the optimal process parameter combination to achieve Pareto optimality of edge quality and processing efficiency.

[0035] The present invention has significant technical effects due to the adoption of the above technical solutions: BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 This is a schematic diagram of the mechanical structure of the sealing ring trimming equipment;

[0037] Figure 2 It is a schematic diagram of the adaptive control system structure.

[0038] The parts indicated by the numerical symbols in the above drawings are as follows: 1. Combined cutting device; 2. Rough cutting component; 3. Mechanical fine cutting component. DETAILED DESCRIPTION

[0039] The present invention is further described in detail below with reference to the accompanying drawings and embodiments.

[0040] Example:

[0041] The sealing ring trimming process has long relied on a single cutting method or a segmented rough and fine cutting process. The rough cutting and fine cutting steps are independent of each other and lack a real-time feedback mechanism, resulting in unstable edge accuracy. Traditional equipment detection methods lag behind, relying on manual sampling or offline testing. They are unable to obtain three-dimensional morphology data in real time, making it difficult to effectively control surface roughness and micro-defects. The tool status monitoring capability is insufficient, and there is a lack of real-time compensation for wear and resistance changes, which can easily lead to cutting depth deviations. Environmental factors such as temperature changes and mechanical vibrations are not effectively controlled, and processing accuracy is easily affected by external interference.

[0042] The separation of roughing and finishing in traditional processes prevents dynamic adjustment of machining parameters, necessitating the establishment of a real-time feedback mechanism. Offline inspection cannot promptly capture machining defects, necessitating the introduction of online 3D scanning technology. Tool wear and environmental interference are not systematically compensated for, necessitating the development of a multi-parameter collaborative control model. Closed-loop control is achieved by integrating the roughing and finishing processes, combining real-time topography detection with dynamic path correction. Multispectral confocal technology is used to acquire edge data online, providing a basis for parameter adjustment. Temperature compensation and vibration reduction designs are introduced to reduce the impact of environmental interference on machining accuracy.

[0043] The present application proposes a sealing ring trimming device, including a combined cutting device, a rough cutting component and a mechanical fine cutting component integrated on the same work platform; the detection device includes a multi-spectral confocal sensor for detecting the sealing ring on the workbench; the adaptive control system is electrically connected to the combined cutting device and the detection device, including a rough cutting module that controls the rough cutting component to rough cut the sealing ring blank and reserve a set amount of burrs on the edge, a fine cutting control module that dynamically generates tool path correction parameters and cutting depth compensation values ​​based on real-time scanning data, and controls the mechanical fine cutting component to precisely trim the reserved burrs, an edge feature analysis module cooperates with the detection device to obtain the three-dimensional morphology data of the sealing ring edge after fine cutting and extracts key quality indicators, and a correction module dynamically adjusts the rough cutting reserved burr amount setting value and cutting parameters based on the key quality indicators.

[0044] A combined cutting device refers to a device that integrates the rough cutting and fine cutting functions on the same work platform. Specifically, it can be achieved by a structure in which a laser cutting head and a mechanical tool are installed side by side, and positioning errors are reduced by a shared motion mechanism. A multispectral confocal sensor refers to a detection device based on the principle of multi-wavelength light interference. Specifically, it can be achieved by a multi-channel optical probe and a high-speed data processing unit, and is used for non-contact three-dimensional contour scanning. An adaptive control system refers to an electronic control system with real-time data processing and parameter adjustment capabilities. Specifically, an industrial computer and a motion control card can work together to achieve the setting of rough cutting flash allowance, dynamic correction of fine cutting path, and feedback adjustment of quality indicators. Key quality indicators include residual flash profile, surface roughness, and micro-defect density. Specifically, the edge feature analysis module can be used to perform gradient calculation and defect identification and extraction on three-dimensional morphology data.

[0045] When the rough cutting component performs preliminary cutting on the sealing ring blank, the reserved flash amount is set through the adaptive control system to avoid excessive cutting that leads to insufficient fine cutting allowance. The multi-spectral confocal sensor performs a three-dimensional scan of the edge of the sealing ring after rough cutting, generates real-time morphology data and transmits it to the fine cutting control module. The mechanical fine cutting component dynamically adjusts the tool path and cutting depth according to the morphology data to eliminate the shape error generated in the rough cutting stage. After the fine cutting is completed, the edge feature analysis module scans the edge of the sealing ring again to extract the residual flash contour deviation and surface defect distribution. The correction module compares the quality index with the preset threshold and automatically updates the rough cutting reserve and fine cutting parameters of the next workpiece. The shock-absorbing buffer layer and temperature compensation device of the working platform continuously suppress external vibration and thermal deformation interference to ensure the environmental stability of the detection and cutting process.

[0046] When traditional equipment uses independent roughing and finishing processes, the roughing parameters are fixed, resulting in fluctuations in the finishing allowance. This solution, however, uses real-time topography scanning to dynamically adjust the roughing allowance. Existing technologies rely on manual spot checks and cannot provide timely feedback on processing defects. This solution uses a multi-spectral confocal sensor to acquire three-dimensional data online and achieve full-process quality monitoring. Traditional tool compensation relies solely on preset wear models. This solution combines real-time torque monitoring with historical data to predict tool status and achieve precise wear compensation. Existing equipment does not take temperature and vibration interference into account. This solution reduces the impact of environmental factors on processing accuracy by integrating a shock-absorbing layer and a temperature compensation device.

[0047] This application realizes closed-loop control of rough cutting and fine cutting processes, dynamically corrects processing parameters through real-time morphology data to ensure edge size accuracy; online three-dimensional detection effectively identifies surface roughness and micro-defects, and improves product qualification rate; tool state prediction is combined with environmental interference suppression to reduce cutting depth deviation and path error, and improve processing stability.

[0048] The present application further proposes that the rough-cut components are cut by laser cutting or mechanical tool cutting.

[0049] Laser cutting involves the non-contact cutting of seal ring blanks using a high-energy laser beam. This can be achieved using a CO2 laser or fiber laser. It is suitable for seal ring materials with high thermal stability and can reduce the impact of mechanical stress on the material edge morphology. Mechanical tool cutting involves the contact cutting of seal ring blanks using rotating or punching tools. This can be achieved using carbide or diamond-coated tools. It is suitable for seal ring materials with a low elastic modulus and can form a regular flash profile through physical cutting.

[0050] The rough-cutting component uses either laser cutting or mechanical tool cutting, depending on the physical properties of the seal material. When processing materials with high thermal stability, laser cutting achieves rapid rough cutting by precisely controlling laser power and scanning speed. This creates a uniform heat-affected zone along the cut edge, providing a stable flash foundation for subsequent fine cutting. When processing materials with a low elastic modulus, mechanical tool cutting achieves controlled cutting by adjusting tool speed and feed rate, avoiding fluctuations in flash size caused by elastic deformation of the material. The selection of cutting method is automatically performed by the material recognition module of the adaptive control system, ensuring that the rough-cutting process matches the material properties.

[0051] Traditional equipment uses only a single rough-cutting method and cannot dynamically adjust the cutting mode based on material properties. For example, cutting heat-sensitive materials with mechanical tools can easily lead to carbonization of the edges, while laser cutting of highly elastic materials can produce irregular flash. This solution offers two optional rough-cutting configurations, enabling the optimal cutting mode to be matched to the physical properties of different materials, thus controlling the stability of flash morphology from the source.

[0052] This application solves the problem of poor material adaptability of traditional equipment due to a single cutting method, effectively avoids defects such as abnormal burr size and edge damage caused by the mismatch between the cutting mode and the material properties, and provides a uniform and consistent burr foundation for subsequent fine cutting processes, thereby improving overall processing accuracy and yield rate.

[0053] The present application further proposes that the adaptive control system includes a tool resistance monitoring module, which is configured with a torque sensor integrated in the mechanical fine cutting component; when the rough cutting component adopts a mechanical tool for cutting, the torque sensor is also integrated in the rough cutting component.

[0054] The tool resistance monitoring module dynamically assesses cutting resistance by collecting real-time torque data from the tool during mechanical fine or rough cutting. This module can be implemented using an embedded data acquisition card and signal processing algorithms. Its purpose is to promptly detect tool anomalies or material hardness fluctuations by monitoring changes in cutting resistance. A torque sensor is a sensing device used to measure the torque applied to rotating components. It can be implemented using a strain gauge or magnetoelastic sensor. Its purpose is to convert the dynamic torque during mechanical cutting into an electrical signal, providing quantitative data for resistance monitoring.

[0055] A torque sensor is installed on the main shaft or drive shaft of the mechanical fine-cutting component to collect the cutting torque data of the fine-cutting tool in real time and transmit it to the tool resistance monitoring module through the data interface. When the rough-cutting component adopts the mechanical tool cutting method, the drive shaft of the rough-cutting tool is also integrated with a torque sensor to realize the synchronous torque monitoring of the rough-cutting and fine-cutting links. During the cutting process, the torque sensor continuously outputs a dynamic torque signal. The tool resistance monitoring module determines abnormal torque fluctuations based on preset thresholds or historical data models. For example, when the torque exceeds the set range or a sudden change occurs, the adaptive control system is triggered to adjust the cutting parameters or issue a warning signal.

[0056] Traditional equipment typically relies solely on offline testing or manual experience to determine tool status, failing to obtain real-time cutting resistance data. This makes it difficult to promptly correct machining errors caused by tool wear or material anomalies. This solution integrates torque sensors in both the fine-cutting and mechanical tool-based rough-cutting assemblies, enabling real-time online monitoring of cutting resistance. Combined with the dynamic adjustment capabilities of the adaptive control system, this effectively suppresses cutting depth deviations caused by tool wear or material hardness variations.

[0057] This application solves the problem that traditional equipment lacks the ability to monitor changes in tool resistance in real time. It can dynamically sense changes in tool status and material properties during the cutting process, and promptly correct cutting parameters to avoid residual burrs or surface damage caused by abnormal resistance, thereby improving the accuracy and stability of sealing ring trimming processing.

[0058] This application further proposes an adaptive control system including a tool wear prediction and compensation module, which predicts the wear status of mechanical precision cutting tools based on historical cutting data, current cutting parameters, key quality indicator change trends and real-time monitoring data of tool resistance, and adds wear compensation when generating tool path correction parameters and cutting depth compensation values.

[0059] The tool wear prediction and compensation module is a control unit that integrates multi-dimensional machining data to establish a prediction model and generate compensation values. This can be achieved through a machine learning algorithm combined with real-time sensor data fusion. It is used to dynamically offset the cutting force attenuation and path deviation caused by tool wear during the cutting process. Historical cutting data refers to a dataset that correlates cutting parameters and wear status of the tool during different machining cycles. This can be achieved by storing the data in a database and associating timestamps with machining batch information, providing a training foundation for the prediction model. Key quality indicator trends refer to the fluctuation patterns of quality parameters such as residual flash profile and surface roughness during continuous machining. This can be achieved by statistically analyzing time-series data collected by a multispectral confocal sensor and extracting characteristic curves to reflect the degree of tool performance degradation. Real-time tool resistance monitoring data refers to the dynamic cutting resistance signal collected by a torque sensor during machining. This can be achieved through high-frequency sampling and filtering to reduce noise, providing instant feedback on the contact status between the tool and the workpiece. Wear compensation refers to the conversion of the equivalent tool wear output from the prediction model into path offset or cutting depth increment. This can be calculated using a geometric error compensation algorithm combined with a material removal rate model to correct the tool's actual cutting position.

[0060] During the finishing process, the tool wear prediction and compensation module continuously receives key quality indicator data from the detection device, real-time resistance data from the torque sensor, and current cutting parameters. Historical cutting data is analyzed offline to establish a correlation model between tool wear, cutting parameters, and quality indicators, while real-time data is used to update the model parameters online. When the model predicts that the tool wear reaches the set threshold, the module converts the wear into a cutting depth compensation value and a path offset, which are added to the original parameters generated by the finishing control module. For example, when the cutting edge wear causes insufficient cutting depth, the compensation amount will dynamically increase the feed rate according to the material's elastic recovery coefficient to ensure that the flash is completely removed.

[0061] Traditional tool wear monitoring relies on periodic downtime for inspection or fixed-cycle tool replacement strategies, which fail to respond to tool status changes in real time, leading to dimensional deviations and surface defects later in the process. This solution, however, fuses multi-source data to establish a dynamic prediction model that automatically identifies wear trends and generates compensation parameters during continuous processing, avoiding batch-to-batch quality fluctuations caused by tool performance degradation.

[0062] This application can offset the impact of tool wear on machining accuracy in real time during the fine-cutting process, reducing residual flash or overcutting caused by tool blunting, while also extending the effective tool life and reducing production interruptions caused by frequent tool changes. Furthermore, by correlating quality indicators with wear prediction results, early warning of abnormal wear conditions can be provided, avoiding workpiece scrapping due to sudden tool failure.

[0063] The present application further proposes that the tool wear prediction and compensation module includes a prediction unit for establishing and continuously updating a historical cutting torque variation curve of a fine cutting tool and a rough cutting component cut by a mechanical tool.

[0064] The historical cutting torque change curve is a trend record formed by collecting torque data from the tool during different machining stages. Specifically, this data can be collected in real time using a torque sensor installed on the tool spindle, and the torque change information during historical machining tasks can be stored in a time series database. This curve is used to reflect the correlation between tool wear and cutting resistance. Continuous updating refers to comparing and analyzing the real-time torque data collected during the current machining process with the historical database. Specifically, a sliding window algorithm is used to weight the torque data from the most recent machining cycle and input the updated data into the prediction model. This process can dynamically track tool performance degradation trends.

[0065] As the finishing tool and the rough-cutting assembly of the mechanical tool perform cutting operations, the torque sensor continuously collects cutting resistance data and transmits it to the prediction unit. The prediction unit matches the torque fluctuation characteristics of the current machining phase with the data curves of the same tool type and material processing scenarios in the historical database, identifying states where the torque increases abnormally or the fluctuation pattern deviates from the standard curve. If the torque curve is detected to have a similar shape to the historical wear stage data, the prediction unit automatically marks the current tool status as the corresponding wear level and synchronizes this level information with the wear compensation algorithm.

[0066] Traditional tool wear monitoring typically relies on fixed-cycle replacement or manual judgment, making it impossible to quantitatively analyze the relationship between changes in cutting resistance and tool wear. This solution, by establishing a dynamically updated torque history curve database, accurately captures the changes in tool torque characteristics at different wear stages, providing multi-dimensional data support for predictive models.

[0067] This application enables quantitative assessment and trend prediction of tool wear, effectively resolving the misjudgment problem inherent in traditional methods due to a lack of historical data. By continuously updating torque curve comparison and analysis, it can proactively identify the risk of abnormal tool wear, providing accurate wear input parameters for the compensation algorithm, thereby reducing cutting accuracy deviations caused by tool performance degradation.

[0068] The present application further proposes that the tool wear prediction and compensation module includes an early warning unit, which generates a fine cutting tool wear early warning signal when the current value or growth rate of the torque of the mechanical fine cutting component exceeds the dynamic threshold calculated based on the historical model; calculates the real-time difference between the rough cutting torque and the fine cutting torque in the processing of the same workpiece or the same batch of materials, establishes and updates the difference baseline model, and generates a tool or rough cutting parameter abnormality early warning when the deviation between the real-time difference and the baseline model exceeds the set tolerance.

[0069] The dynamic threshold refers to the allowable fluctuation range derived from the torque change trend model established based on historical cutting data. Specifically, a machine learning algorithm can be used to perform regression analysis on historical torque data to generate a dynamic threshold curve that changes with the processing batch or material type, which is used to judge torque anomalies in real time. The difference baseline model refers to a reasonable difference range model for rough cutting and fine cutting torque. Specifically, the initial difference baseline can be established by statistically analyzing the correspondence between rough cutting and fine cutting torque during the processing of the same batch of materials, and the baseline range can be updated according to real-time data during continuous processing to monitor sudden changes in rough cutting parameters or tool status. The deviation refers to the standard deviation multiple of the real-time difference and the baseline model. Specifically, a sliding window algorithm can be used to calculate the degree of deviation of the current difference relative to the baseline mean, and an early warning will be triggered when it exceeds the preset standard deviation multiple.

[0070] During the sealing ring processing, the early warning unit continuously collects real-time torque data of the mechanical precision cutting components and compares it with the dynamic threshold generated by the historical model. When the instantaneous value of the torque or the growth rate per unit time exceeds the dynamic threshold, it indicates that the precision cutting tool may have abnormal wear, and the system automatically generates a warning signal. At the same time, for the same workpiece or the same batch of materials, the system calculates the torque difference between the rough cutting and fine cutting stages in real time, and determines whether the current difference is in a reasonable range based on the difference baseline model established based on historical data. If the real-time difference deviates from the baseline model by more than the set tolerance, a tool abnormality or rough cutting parameter mismatch warning is triggered. For example, when the rough cutting tool wears and the cutting resistance increases, the rough cutting torque increases. If the original parameters are still used in the fine cutting stage, the fine cutting torque may increase synchronously due to excessive flash allowance. At this time, the difference deviation will exceed the baseline range, and the system will immediately issue an abnormal signal.

[0071] Traditional tool monitoring methods rely solely on fixed thresholds or manual experience to determine tool status. They are unable to adapt to the dynamic changes in different material batches or processing stages, and are prone to misjudgments or missed judgments. However, this solution combines dynamic thresholds with a differential baseline model to track torque change trends during processing in real time and identify abnormal patterns based on a data-driven model, significantly improving early warning accuracy and response speed. For example, in existing technologies, rough cutting parameter anomalies may not be discovered until scrap appears in the fine cutting stage. However, this solution uses differential deviation analysis to identify parameter mismatches immediately after the rough cutting is completed.

[0072] This application effectively solves the problem of lagging tool status and process parameter anomalies in traditional seal ring trimming equipment. It provides real-time early warning of fine-cutting tool wear and can simultaneously identify processing deviations caused by improper rough-cutting parameter settings or tool anomalies. This avoids batch processing defects caused by accumulated tool wear or parameter mismatches, and improves the stability and yield rate of the seal ring trimming process.

[0073] This application further proposes a tool wear prediction and compensation module that integrates wear warning, abnormal state warning, and edge quality changes monitored by multi-spectral confocal monitoring to generate tool replacement suggestions or process parameter adjustment instructions.

[0074] Wear warning refers to a dynamic threshold model established based on the historical cutting torque change curve. When the real-time monitored tool torque exceeds the threshold, a warning signal is triggered. Specifically, it can be achieved by collecting cutting resistance data in real time through a torque sensor, combined with a machine learning algorithm to predict the remaining life of the tool. Abnormal state warning refers to a warning signal generated by comparing the real-time difference between the rough cutting and fine cutting torques with the difference of the baseline model. Specifically, it can be achieved by using a difference baseline model establishment module and combining a sliding window algorithm to calculate the torque difference deviation in real time. Edge quality changes monitored by multi-spectral confocal monitoring refer to obtaining three-dimensional morphological data of the sealing ring edge through the principle of multi-wavelength light interference. Specifically, it can be achieved by using a multi-spectral confocal sensor for non-contact scanning and combining an image processing algorithm to extract residual burr profile deviation and surface roughness parameters.

[0075] During the continuous processing of the sealing ring, the torque sensor collects the cutting resistance data of the mechanical precision-cut components in real time, and determines whether to trigger a wear warning through a dynamic threshold model trained with historical data. At the same time, the difference baseline model continuously calculates the real-time difference between the rough cutting and precision cutting torques, and triggers an abnormal state warning when the deviation exceeds the tolerance. The multi-spectral confocal sensor synchronously scans the edge of the sealing ring after precision cutting to obtain the residual flash contour deviation and surface roughness data. The tool wear prediction and compensation module inputs the above three types of data into the decision model. When it is detected that the tool wear reaches a critical value or the edge quality continues to deteriorate, a tool replacement instruction is generated; when an abnormal state is detected but the tool does not meet the replacement conditions, a parameter adjustment instruction for the cutting speed, feed rate or cutting depth is generated.

[0076] Traditional equipment relies solely on single torque monitoring or periodic manual spot checks to determine tool status, failing to correlate the causal relationship between abnormal cutting parameters and changes in edge quality. This solution integrates multi-source monitoring data to establish a dynamic correlation model between actual tool wear, sudden abnormalities, and machining quality results. This allows tool replacement decisions to take both equipment status and machining quality requirements into account, avoiding resource waste caused by premature replacements or batch defects caused by delayed replacements.

[0077] This application automatically optimizes maintenance strategies based on real-time machining status and quality feedback, reducing defects such as incomplete burr removal and surface scratches caused by tool wear or parameter mismatch, while also avoiding increased production costs due to excessive tool replacement. During continuous machining, the system can autonomously determine the root cause of abnormal operating conditions, distinguishing between natural tool wear and sudden parameter misalignment, and output targeted maintenance instructions to ensure the stability of the seal ring trimming quality.

[0078] This application further proposes a multispectral confocal sensor that performs non-contact three-dimensional contour scanning of the edge of the sealing ring through the principle of multi-wavelength light interference to achieve real-time morphology data collection.

[0079] A multispectral confocal sensor is a detection device based on the principle of optical confocality and a multi-wavelength light source. Specifically, it can be implemented using an integrated probe consisting of multiple independent wavelength laser emission units, an interferometer, and a high-resolution photodetector. Different wavelength beams form interference fringes on the seal ring surface, and the phase difference of the reflected light is analyzed to calculate the surface height change. The principle of multi-wavelength optical interference involves the interference of two or more different wavelength beams on an object's surface. Specifically, this can be achieved using a combination of red, blue, and green wavelengths in conjunction with a beam splitter. This multi-wavelength superposition eliminates the phase ambiguity associated with single-wavelength measurements, improving height measurement resolution. Non-contact 3D profile scanning is a scanning method that eliminates physical contact with the surface being measured. Specifically, an optical probe maintains a fixed distance from the edge of the seal ring being measured, and the beam focus is adjusted to achieve full coverage scanning, avoiding surface scratches or material deformation caused by mechanical contact. Real-time topography data acquisition involves the synchronous transmission of the scanned 3D coordinate information to a control system. Specifically, a high-speed data acquisition card combined with a parallel computing unit can complete point cloud data processing and generate a 3D topography model within a single scan cycle.

[0080] During operation, the multispectral confocal sensor projects a combined beam of light from a multi-wavelength laser emission unit onto the edge of the seal ring. The reflected light forms an interference signal through an interferometer. A photodetector converts the optical signal into an electrical signal, and a phase demodulation algorithm is used to calculate the three-dimensional coordinates of each measurement point. During the scanning process, the sensor moves along the circumference of the seal ring, continuously collecting edge profile data, generating a three-dimensional model that includes height, curvature, and surface texture. This data is then transmitted in real time to the adaptive control system.

[0081] Traditional seal ring trimming equipment often uses contact probes or single-wavelength laser scanning, which can lead to slow measurement speeds, surface damage, and an inability to eliminate phase errors. This solution eliminates measurement ambiguity through multi-wavelength interferometry, combines it with non-contact scanning to avoid material damage, and enables real-time acquisition of high-precision 3D topography, effectively addressing technical shortcomings such as offline detection lag and incomplete surface quality data.

[0082] This application can continuously obtain three-dimensional morphological data of the edge of the sealing ring during the trimming process, providing a real-time basis for the correction of the precision cutting tool path and the adjustment of cutting parameters, avoiding batch processing errors caused by detection delays, and at the same time reducing the mechanical damage to the sealing ring surface caused by contact measurement, significantly improving the flash removal accuracy and surface quality control capabilities.

[0083] The present application further proposes that the surface of the working platform is provided with a shock-absorbing buffer layer and an integrated temperature compensation device to maintain the temperature of the sealing ring material within a set constant range during the cutting process, so as to reduce the thermal deformation error caused by temperature changes.

[0084] The shock-absorbing buffer layer is a layer of elastic material covering the surface of the work platform. It can be made of polyurethane foam or a rubber-based composite material. Its elastic modulus matches the hardness characteristics of the sealing ring material. By absorbing the vibration energy generated during the mechanical cutting process, it reduces the resonance amplitude between the tool and the workpiece. The temperature compensation device is a closed-loop temperature control unit embedded in the work platform. It can be implemented by combining a thermoelectric cooler and distributed temperature sensors. By real-time monitoring of the temperature distribution in the sealing ring contact area, it dynamically adjusts the cooling power to maintain the material in a thermally stable state.

[0085] The shock-absorbing buffer layer absorbs lateral vibrations generated by the mechanical precision-cut components through elastic deformation, preventing vibration energy from being transferred to the seal and causing edge distortion. The temperature compensation device actively regulates the temperature of the contact surface between the work platform and the seal using a thermoelectric cooler. For example, if the platform temperature rises during continuous processing, cooling mode is activated. This, combined with a temperature sensor, forms a closed-loop control circuit, ensuring that the seal material remains within a preset temperature range during the cutting process, thereby suppressing dimensional deviations caused by differences in thermal expansion coefficients.

[0086] In some specific embodiments, the thickness of the shock-absorbing buffer layer can be set to 3-5 mm, and its surface can be processed into a honeycomb microstructure to enhance energy dissipation efficiency. The thermoelectric cooler of the temperature compensation device can adopt a multi-zone independent control mode. For example, the work platform can be divided into four temperature control zones, each equipped with an independent sensor and cooling unit to achieve local temperature compensation.

[0087] Conventional equipment lacks a vibration-damping structure on the work platform and active temperature control, causing tool vibration to be directly transmitted to the workpiece. Furthermore, frictional heat generated during machining causes material expansion to be unresolved, leading to fluctuations in the cutting edge size. This solution utilizes a vibration-damping buffer layer and a temperature compensation device to mitigate both mechanical vibration and thermal deformation.

[0088] This application effectively suppresses the interference of mechanical vibration on the tool path accuracy, and at the same time eliminates the dimensional instability of the sealing ring material caused by temperature fluctuations, making the generation basis of the precision cutting tool path correction parameters more reliable, thereby improving the repeatability accuracy and edge morphology consistency of the flash trimming process.

[0089] This application further proposes a built-in fuzzy logic controller in the correction module, which dynamically optimizes the matching parameters of the rough cutting reserved burr amount and the fine cutting feed speed based on the coupling relationship between the residual burr profile deviation and the surface roughness.

[0090] A fuzzy logic controller refers to a control unit based on a fuzzy rule reasoning system. Specifically, it can be implemented using a multi-input and multi-output rule base structure to handle nonlinear and uncertain correlations between process parameters. The residual flash contour deviation refers to the geometric difference between the residual flash on the edge of the sealing ring after fine cutting and the target contour. Specifically, it can be calculated using a contour fitting algorithm for three-dimensional morphology data. The coupling relationship of surface roughness refers to the interaction between the microscopic roughness of the residual flash area and the cutting parameters. Specifically, it can be modeled by correlating the surface texture features collected by a multi-spectral confocal sensor with the historical data of the cutting parameters. Dynamic optimization refers to adjusting the proportional coefficient of the rough cutting reserve setting value and the fine cutting feed speed through fuzzy reasoning based on the deviation between the real-time detection data and the preset quality target, so that the cutting parameter combination is always in the optimal matching state.

[0091] During the fine-cutting process after the rough-cutting stage, a multispectral confocal sensor scans the three-dimensional topography of the seal ring edge in real time, inputting the residual flash profile deviation and surface roughness data into the fuzzy logic controller. The controller's built-in fuzzy rule library contains empirical data on optimal parameter matching for different material properties and tool conditions. The input variables are converted into fuzzy quantities through membership functions. After rule reasoning, the controller outputs the adjustment coefficient for the rough-cutting reserve flash amount and the correction value for the fine-cutting feed rate. For example, if the residual flash profile deviation is detected to have increased but the surface roughness index is qualified, the fuzzy controller will automatically increase the rough-cutting reserve for the next workpiece; if the surface roughness exceeds the threshold but the flash amount is normal, the fine-cutting feed rate will be reduced to improve the surface quality.

[0092] Traditional trimming equipment relies on fixed parameter combinations or manual adjustments based on experience, making it ineffective for matching parameters across different material batches and tool wear conditions. This solution, however, uses a fuzzy logic controller to establish a dynamic correlation model between cutting parameters and quality indicators. This model automatically adapts to changing process conditions, ensuring trimming accuracy while avoiding response delays caused by manual intervention.

[0093] This application realizes the intelligent matching of rough cutting reserve and fine cutting parameters, effectively solves the problems of burr residue and substandard surface quality caused by fluctuations in material properties or tool wear, and significantly improves the processing consistency and quality stability of the sealing ring trimming process.

Claims

1. A sealing ring trimming device and its adaptive control system, characterized in that: include: Combined cutting device, rough cutting component and mechanical fine cutting component integrated on the same working platform; A detection device, comprising a multispectral confocal sensor for detecting the sealing ring on the workbench; The adaptive control system is electrically connected to the combined cutting device and the detection device, and includes: Rough cutting module controls the rough cutting component to perform rough cutting on the sealing ring blank and reserve a set amount of flash on the edge; The precision cutting control module dynamically generates the tool path correction parameters and cutting depth compensation values ​​for mechanical precision cutting based on the real-time scanning data of the sealing ring edge after rough cutting by the detection device, and controls the mechanical precision cutting component to precisely trim the reserved flash edge; The edge feature analysis module cooperates with the detection device to obtain the three-dimensional morphology data of the edge of the sealing ring after precision cutting, and extracts at least two of the residual flash profile, surface roughness, and micro-defect density as key quality indicators; The correction module dynamically adjusts the reserved burr amount setting value of the next sealing ring rough cutting operation and / or the cutting parameters of the fine cutting control module according to the key quality indicators.

2. A sealing ring trimming device and its adaptive control system according to claim 1, characterized in that: The rough cutting component is cut by laser cutting or mechanical tool cutting.

3. The sealing ring trimming device and the adaptive control system thereof according to claim 1, characterized in that: The adaptive control system further comprises a tool resistance monitoring module, which is configured with a torque sensor integrated in the mechanical fine cutting component; when the rough cutting component adopts a mechanical tool for cutting, the torque sensor is also integrated in the rough cutting component.

4. The sealing ring trimming device and the adaptive control system thereof according to claim 1, characterized in that: The adaptive control system also includes a tool wear prediction and compensation module, which predicts the wear status of mechanical precision cutting tools based on historical cutting data, current cutting parameters, key quality indicator change trends and real-time monitoring data of tool resistance, and adds wear compensation when generating tool path correction parameters and cutting depth compensation values.

5. The sealing ring trimming device and the adaptive control system thereof according to claim 4, characterized in that: The tool wear prediction and compensation module includes a prediction unit for establishing and continuously updating a historical cutting torque variation curve of a fine cutting tool and a rough cutting component cut by a mechanical tool.

6. The sealing ring trimming device and the adaptive control system thereof according to claim 4, characterized in that: The tool wear prediction and compensation module includes an early warning unit, which generates a fine cutting tool wear early warning signal when the current value or growth rate of the torque of the mechanical fine cutting component exceeds the dynamic threshold calculated based on the historical model; calculates the real-time difference between the rough cutting torque and the fine cutting torque in the processing of the same workpiece or the same batch of materials, establishes and updates the difference baseline model, and generates a tool or rough cutting parameter abnormality early warning when the deviation between the real-time difference and the baseline model exceeds the set tolerance.

7. The sealing ring trimming device and the adaptive control system thereof according to claim 6, characterized in that: The tool wear prediction and compensation module integrates wear warning, abnormal state warning and edge quality changes monitored by multi-spectral confocal monitoring to generate tool replacement suggestions or process parameter adjustment instructions.

8. The sealing ring trimming device and the adaptive control system thereof according to claim 1, characterized in that: The multispectral confocal sensor performs non-contact three-dimensional contour scanning on the edge of the sealing ring through the principle of multi-wavelength light interference, thereby realizing real-time shape data collection.

9. The sealing ring trimming device and the adaptive control system thereof according to claim 1, characterized in that: The surface of the working platform is provided with a shock-absorbing buffer layer and an integrated temperature compensation device for maintaining the temperature of the sealing ring material within a set constant range during the cutting process to reduce thermal deformation errors caused by temperature changes.

10. The sealing ring trimming device and the adaptive control system thereof according to claim 1, characterized in that: The correction module has a built-in fuzzy logic controller, which dynamically optimizes the matching parameters of the rough cutting reserved burr amount and the fine cutting feed speed according to the coupling relationship between the residual burr profile deviation and the surface roughness.

Images