Method for optimizing machining parameters based on risk prediction of electroplated diamond tool falling off

Abstract

The application discloses a processing parameter optimization method based on a falling risk prediction of electroplated diamond tools, and comprises the following steps: acquiring simulation use state parameters of the electroplated diamond tools; determining a grit bonding state vector based on the simulation use state parameters; determining a grit falling risk index based on the grit bonding state vector and the simulation use state parameters; constructing a grit falling trend prediction model based on the grit falling risk index, and determining a grit falling prediction risk value at a future time; and determining and executing optimal process control parameters based on the grit falling prediction risk value at the future time. The application can improve the accuracy of local grit falling risk identification and prediction, reduce the blindness and frequent fluctuations of parameter adjustment, reduce the grit falling and damage risk, and improve the processing stability of the electroplated diamond tools.

Description

Technical Field

[0001] This invention relates to the field of electroplated diamond technology, and more specifically to a method for optimizing processing parameters based on the prediction of the risk of electroplated diamond tool detachment. Background Technology

[0002] Electroplated diamond tools typically use a metal coating to bond diamond abrasive grains to the tool substrate surface. They are characterized by large abrasive grain protrusion height, strong cutting ability, and high machining accuracy, and are widely used in grinding, cutting, and precision machining of hard and brittle materials. During processing, the diamond abrasive grains are continuously subjected to cutting loads, frictional heat, and vibration impacts, making them prone to wear, breakage, or detachment from the coating. This can lead to a decrease in tool cutting ability, fluctuations in machining quality, and in severe cases, premature tool failure.

[0003] Existing technologies typically rely on operator experience or preset process parameters for processing control, and use single operating parameters such as cutting load, contact temperature, or vibration amplitude to determine the tool's working status. However, the abrasive grain protrusion status, damage degree, and bonding stability vary in different areas of the tool surface. Relying solely on overall operating parameters or fixed thresholds makes it difficult to accurately reflect the risk of localized abrasive grain detachment, and also makes it difficult to predict the risk trends of abrasive grain changes during future processing.

[0004] Furthermore, existing machining parameter adjustment methods typically employ fixed adjustment ranges or adjust a single parameter, making it difficult to coordinate the adjustment of spindle drive current settings, coolant flow rate, feed rate, and spindle speed based on different risk sources corresponding to cutting load, thermal effects, and vibration shock. They also cannot compare the impact of different candidate machining parameters on the future risk of abrasive grain shedding in advance, which can easily lead to problems such as parameter adjustment lag, unreasonable adjustment direction, or frequent fluctuations in process parameters. Summary of the Invention

[0005] To address the aforementioned problems, the first aspect of this invention provides a method for optimizing processing parameters based on the prediction of the risk of electroplated diamond tool detachment, comprising the following steps: S1. Obtain the simulated usage parameters of the electroplated diamond tool; S2. Determine the abrasive particle bonding state vector based on simulated usage state parameters; S3. Determine the abrasive detachment risk index based on the abrasive bonding state vector and simulated usage state parameters; S4. Construct an abrasive detachment trend prediction model based on the abrasive detachment risk index to determine the predicted risk value of abrasive detachment at future times; S5. Determine and execute the optimal process control parameters based on the predicted risk value of abrasive grain shedding at future time.

[0006] Preferably, the simulated usage parameters include abrasive grain protrusion height, abrasive grain damage area, cutting load, contact temperature, and vibration amplitude; The determination of the abrasive grain bonding state vector includes: S201. Determine the stable component of abrasive protrusion based on the abrasive protrusion height; S202. Determine the stable component of abrasive damage based on the abrasive damage area; S203. Determine the comprehensive bonding stability coefficient of abrasive particles based on the abrasive particle protrusion stability component and the abrasive particle damage stability component. S204. Construct the abrasive particle bonding state vector.

[0007] Preferably, determining the abrasive grain shedding risk index includes: S301. Determine the external action index based on simulated usage state parameters; S302. Determine the risk value of foundation shedding based on the comprehensive stability coefficient of abrasive particles and the external action index; S303. Determine the damage growth correction factor based on the abrasive damage area at adjacent acquisition times; S304. Determine the abrasive particle detachment risk index based on the basic detachment risk value and the damage growth correction factor.

[0008] Preferably, determining the predicted risk value of abrasive grain shedding at future times includes: S401. Construct a time series of risk in the abrasive particle observation area; S402. Construct a region prediction input sequence containing processing parameters; S403. Construct a predictive model for abrasive particle shedding trend based on the regional prediction input sequence; S404. Determine the predicted risk value of abrasive particle shedding at future times.

[0009] Preferably, determining and executing the optimal process control parameters includes: S501. Determine the intensity of process parameter adjustment and the dominant risk factors; S502. Generate candidate process control parameters based on the dominant risk factors and determine the optimal process control parameters; S503, Execute the optimal process control parameters and perform closed-loop updates.

[0010] Preferably, the abrasive grain protrusion stabilization component is: ; in, Indicates the first The monitored abrasive particles were in the first The stable component of abrasive particles at each acquisition moment; Indicates the first The monitored abrasive particles were in the first The height of abrasive protrusion at each acquisition moment; This indicates the critical protrusion height corresponding to the abrasive grain size range, coating material, and coating thickness of the current electroplated diamond tool; The stable component of abrasive damage is: ; in, Indicates the first The monitored abrasive particles were in the first Stable component of abrasive damage at each acquisition time Indicates the first The monitored abrasive particles and their surrounding coating were in the first... Abrasive damage area at each sampling time. This indicates the critical damage area before the abrasive grains loosen or fall off. The overall stability coefficient of the abrasive particles is: ; in, Indicates the first The monitored abrasive particles were in the first The overall stability coefficient of abrasive particles at each acquisition time. This represents the weighting coefficient corresponding to the stable component of abrasive grain protrusion. This represents the weighting coefficient corresponding to the stable component of abrasive damage.

[0011] Preferably, the external action index is: ; in, Indicates the first The external influence index corresponding to each data collection moment. Indicates the relationship with the first The cutting load corresponding to each acquisition time. Indicates the relationship with the first The equivalent contact temperature corresponding to each data acquisition moment. Indicates the relationship with the first The vibration amplitude corresponding to each acquisition time. This indicates the reference value for the cutting load. Indicates the reference temperature. This indicates the reference value for temperature rise. This indicates the reference value for vibration amplitude. , and These represent the external action weighting coefficients corresponding to cutting load, contact temperature, and vibration amplitude, respectively. The basic shedding risk value is: ; in, Indicates the first The monitored abrasive particles were in the first The baseline shedding risk value corresponding to each data collection moment Indicates the first The external influence index corresponding to each data collection moment; Indicates the first The monitored abrasive particles were in the first The overall stability coefficient of abrasive particles corresponding to each acquisition time; This represents the preset minimum stability coefficient to prevent the denominator from being zero; The damage growth correction factor determined based on the abrasive damage area at adjacent acquisition times includes: The damage growth correction factor is determined based on the abrasive damage growth rate. The abrasive damage growth rate is: ; in, Indicates the first The monitored abrasive particles were in the first The abrasive damage growth rate corresponding to each acquisition time. Indicates the first The monitored abrasive particles were in the first The abrasive damage area corresponding to each acquisition moment. This indicates the abrasive damage area corresponding to the same monitored abrasive grain at the previous data acquisition time. This indicates the actual time interval between the current data acquisition time and the previous data acquisition time; The damage growth correction factor is: ; in, Indicates the first The monitored abrasive particles were in the first Damage growth correction factor corresponding to each acquisition time point This represents the damage growth correction factor. Indicates the critical damage growth rate; The abrasive grain shedding risk index is: , ; in, Indicates the first The monitored abrasive particles were in the first The abrasive grain shedding risk index corresponding to each data collection moment. Indicates the risk value of base shedding. Indicates the damage growth correction factor; The maximum abrasive detachment risk index within each abrasive observation area is taken as the regional detachment risk value. The risk value of detachment in the area is: ; in, ; Indicates the first The first abrasive observation area in the first The risk value of regional detachment corresponding to each data collection time. Indicates the first One abrasive observation area, Indicates the first The abrasive grain at the first The spatial location at each acquisition time.

[0012] Preferably, the construction of the risk time series for the abrasive observation region includes: The same abrasive observation area in continuous The regional detachment risk values ​​at each collection time are arranged in chronological order to obtain the regional risk time series: ; in, Indicates the first Abrasive observation area in The time series of regional risks corresponding to each moment. This indicates the number of sampling points included within the time window, and ; The constructed region prediction input sequence containing processing parameters includes: By combining the regional detachment risk value at the same acquisition time with the processing parameter vector, the regional operating status vector is obtained: ; The same abrasive observation area in continuous The regional operational state vectors at each acquisition time are arranged in chronological order to obtain the regional prediction input sequence: ; The abrasive grain shedding trend prediction model is as follows: ; in, Indicates the first Each abrasive observation area uses candidate machining parameters. Afterwards, in the future Predicted risk value at any given time. This represents a model for predicting the trend of abrasive grain shedding. Indicates model parameters, Indicates the prediction time steps; The determination of the predicted risk value for abrasive grain shedding at future times includes: Input the current region prediction input sequence and candidate processing parameters into the trained abrasive grain shedding trend prediction model to obtain the future region shedding prediction risk value: ; in, Indicates the first After selecting candidate processing parameters for the first abrasive observation region, in the future... Predicted risk value of abrasive grain shedding at any given time. This represents the parameters of the model after training is complete. This represents the standardized current region prediction input sequence. This represents the standardized candidate processing parameter vector; Based on the predicted risk values ​​for different abrasive grain observation areas, determine the maximum predicted risk value of future detachment of the electroplated diamond tool under the aforementioned candidate processing parameters: ; in, Indicates the candidate machining parameters used for electroplated diamond tools. Afterwards, in the future The maximum predicted risk value of future shedding at that time. This indicates the number of abrasive grain observation areas.

[0013] Preferably, the determination and execution of optimal process control parameters based on the predicted risk value of abrasive grain shedding at future times includes: The intensity of process parameter adjustment is determined based on the degree to which the predicted maximum future shedding risk value exceeds the preset safety risk value. The predicted maximum future shedding risk value while maintaining the current process control parameters is denoted as: ; The intensity of the process parameter adjustment is: ; in, Indicates the intensity of process parameter adjustment. This indicates the preset security risk value. This indicates the preset risk warning value; The dominant risk factors are determined based on the three components of the external effects index in S3: ; ; ; in, , and These represent the risk contribution corresponding to cutting load, contact temperature, and vibration amplitude, respectively. The dominant risk factors are: ; in, Indicates the current dominant risk factors; The step of generating candidate process control parameters and determining the optimal process control parameters based on the dominant risk factors includes: When cutting load is the dominant risk factor, candidate parameters for load adjustment can be generated: ; When contact temperature is the dominant risk factor, candidate parameters for temperature regulation can be generated: ; When the vibration amplitude is the dominant risk factor, two candidate vibration adjustment parameters can be generated, one to reduce the spindle speed and the other to increase the spindle speed. ; ; in, , , and These represent the preset maximum single adjustment amount for the corresponding process parameters; This indicates that the adjusted process parameters are limited to the allowable range of the simulated processing device; Each candidate process control parameter in the candidate parameter set is input into the abrasive grain shedding trend prediction model in S4 to obtain the corresponding future maximum shedding prediction risk value, and the candidate parameter with the smallest future maximum shedding prediction risk value is selected: ; in, Indicates the optimal process control parameters; This represents the set of candidate process control parameters generated at the current moment; The optimal process control parameters are: ; in, , , , These represent the optimal spindle drive current setting, optimal coolant flow rate, optimal feed rate, and optimal spindle speed, respectively.

[0014] By adopting the above technical solution, the present invention mainly has the following technical effects: 1. By selecting monitored abrasive grains within a fixed abrasive grain observation area, and considering factors such as abrasive grain protrusion height, abrasive grain damage area, cutting load, contact temperature, and vibration amplitude, the abrasive grain bonding state and abrasive grain detachment risk are determined. Furthermore, based on the detachment risk of monitored abrasive grains within each observation area, the regional detachment risk is determined, which can prevent local high-risk abrasive grains from being masked by the overall average state. In addition, by combining the regional risk time series to predict future detachment risk, the accuracy of identifying and predicting local abrasive grain detachment risk is improved.

[0015] 2. By using the spindle drive current setting, coolant flow rate, feed rate, and spindle speed as online adjustable process control parameters, the adjustment intensity is determined based on the predicted risk value of the maximum future abrasive shedding, and the adjustment direction is determined based on the risk contribution of cutting load, contact temperature, and vibration amplitude. Parameters with smaller predicted future abrasive shedding risk values ​​are selected from candidate process control parameters for closed-loop control. This reduces the blindness and frequent fluctuations in parameter adjustments, lowers the risk of abrasive shedding and breakage, and improves the processing stability of electroplated diamond tools. Attached Figure Description

[0016] Figure 1 This is a flowchart of a processing parameter optimization method based on the prediction of the risk of electroplated diamond tool detachment according to the present invention; Figure 2 This is a continuous spatial distribution cloud map of the abrasive grain detachment risk index in a machining parameter optimization method based on the prediction of detachment risk of electroplated diamond tools in this invention. Figure 3 This is a distribution map of abrasive particle observation area a machining parameter optimization method based on the prediction of abrasive particle abrasion risk in this invention. Figure 4 SEM images of electroplated diamond tools after completing a pre-set simulated use test using the process control parameters before optimization; Figure 5 SEM images of electroplated diamond tools after completing the same simulated use test using optimal process control parameters. Detailed Implementation

[0017] To enable those skilled in the art to better understand the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0019] Please see Figures 1-5 The first aspect of this invention provides a method for optimizing processing parameters based on the prediction of the risk of detachment from electroplated diamond tools, comprising the following steps: S1. Obtain the simulated usage parameters of the electroplated diamond tool; In some embodiments, the electroplated diamond tool can be installed in a simulation processing device, and the electroplated diamond tool can be rotated by a drive mechanism to make the electroplated diamond tool contact the simulated workpiece, so as to construct a simulated use scenario corresponding to the actual processing conditions of the electroplated diamond tool.

[0020] In some embodiments, an industrial microscope camera, a three-dimensional morphology detection device, a temperature sensor, a vibration sensor, and a force sensor installed on the simulated processing device can be used to monitor the abrasive morphology and processing operation status of the electroplated diamond tool during simulated use, so as to obtain simulated use status parameters.

[0021] In some embodiments, the simulated usage parameters include abrasive protrusion height, abrasive damage area, cutting load, contact temperature, and vibration amplitude.

[0022] In some embodiments, the abrasive protrusion height and abrasive damage area are used to characterize the microstructure of the abrasive grains on the surface of the electroplated diamond tool; the cutting load, contact temperature and vibration amplitude are used to characterize the overall load state, thermal state and dynamic impact state of the electroplated diamond tool during the contact process with the simulated workpiece.

[0023] In some embodiments, considering that the surface of an electroplated diamond tool contains a large number of abrasive grains, the effective working surface of the electroplated diamond tool can be divided into multiple abrasive grain observation areas, and a preset number of abrasive grains can be selected from each abrasive grain observation area as monitored abrasive grains to construct a monitored abrasive grain set, without having to monitor the status of all abrasive grains on the surface of the electroplated diamond tool one by one.

[0024] In some embodiments, the abrasive observation area can be divided along the circumferential, radial, or axial direction of the electroplated diamond tool, according to the tool's shape. The position and extent of each abrasive observation area remain fixed so that the same abrasive observation area can be repeatedly detected at different acquisition times.

[0025] In some embodiments, abrasive grains with large protrusion height, damaged abrasive grains, and a preset number of ordinary abrasive grains can be selected from each abrasive grain observation area as monitored abrasive grains, so that the monitored abrasive grain set simultaneously includes potentially high-risk abrasive grains and normal abrasive grains.

[0026] In some embodiments, the set of monitored abrasive particles is: ; in, Indicates the first The set of monitored abrasive particles corresponding to each acquisition time. These represent the abrasive grain identifiers corresponding to different monitored abrasive grains. Indicates the number of abrasive particles being monitored; In some embodiments, corresponding abrasive identifiers can be set for different monitored abrasive particles based on the observation area where the abrasive particle is located, the center position of the abrasive particle, and the image characteristics of the abrasive particle.

[0027] In some embodiments, the electroplated diamond tool can be made to return to the same detection position at different detection times by adjusting the tool installation position and spindle angle position, and position registration and feature matching can be performed on the abrasive images acquired at different detection times to identify the state changes of the same monitored abrasive at different detection times.

[0028] Abrasive grain protrusion height is used to characterize the height at which diamond abrasive grains protrude from the coating surface. In some embodiments, three-dimensional topographic data of the area where the monitored abrasive grains are located can be obtained by using a laser confocal microscope, a white light interferometric three-dimensional profilometer, a focus-changing three-dimensional microscope, or a scanning electron microscope with three-dimensional reconstruction function; a coating reference plane is fitted based on the height data of the coating surface around the abrasive grains, and the vertical distance between the top of the abrasive grains and the coating reference plane is calculated to determine the abrasive grain protrusion height.

[0029] Abrasive damage area is used to characterize the degree of damage to the diamond abrasive grains and the surrounding coating. In some embodiments, the abrasive damage area is the cumulative union projection area of ​​the abrasive wear area, abrasive breakage area, abrasive crack equivalent area, and coating peeling area around the abrasive identified at the current acquisition time, after removing overlapping portions. For abrasive cracks represented by linear contours, the crack contour can be converted into an equivalent crack area according to a preset equivalent width.

[0030] In some embodiments, an industrial microscope camera can be used to acquire surface images of the monitored abrasive particles. The surface images of the abrasive particles at the current acquisition time are then registered and differentially processed with the surface images at the initial acquisition time or the previous acquisition time to identify newly added abrasive wear areas, abrasive breakage areas, abrasive crack equivalent areas, and coating peeling areas at the current acquisition time relative to the initial acquisition time or the previous acquisition time. The newly added damage areas are then incorporated into the cumulative damage area corresponding to the previous acquisition time. After removing the overlapping parts, the actual projected area of ​​the current cumulative damage area is calculated according to the image scale to obtain the abrasive damage area at the current acquisition time.

[0031] Cutting load is used to characterize the overall stress state during the contact process between the electroplated diamond tool and the simulated workpiece. In some embodiments, the cutting load can be continuously acquired by a force sensor mounted on the simulated machining device, and at least one of the average load, peak load, or root mean square load can be determined as the cutting load based on the load signal within a preset sampling time range before the current morphology detection time.

[0032] Contact temperature is used to characterize the thermal state of the contact area between the electroplated diamond tool and the simulated workpiece. In some embodiments, the contact temperature can be continuously collected by an infrared temperature measuring device or a temperature sensor, and the average temperature or the highest temperature can be determined as the equivalent contact temperature based on the temperature signal within a preset sampling time range before the current morphology detection time.

[0033] Vibration amplitude is used to characterize the dynamic impact state during the simulated machining process of electroplated diamond tools; In some embodiments, the vibration amplitude can be determined by continuously collecting vibration signals through a vibration sensor, filtering the vibration signals, and then determining the vibration peak value or root mean square value within a preset sampling time range before the current morphology detection time.

[0034] In some embodiments, the abrasive grain protrusion height and abrasive grain damage area can be obtained through periodic detection. Specifically, after the electroplated diamond tool completes the simulated processing for a preset time, the simulated processing device can be paused, allowing the electroplated diamond tool to return to a preset angle position, and then the observation areas of each abrasive grain can be detected by an industrial microscope camera and a three-dimensional morphology detection device.

[0035] In some embodiments, the cutting load, contact temperature, and vibration amplitude can be continuously collected during the simulated machining process. The cutting load, contact temperature, and vibration amplitude within a preset sampling time range before each morphology detection moment are correlated with the abrasive grain protrusion height and abrasive grain damage area obtained from the current morphology detection to avoid using the load, temperature, and vibration data after the simulated machining device has stopped as machining status parameters.

[0036] In some embodiments, for the same monitored abrasive grain and the same acquisition cycle, the abrasive grain protrusion height, abrasive grain damage area, cutting load, contact temperature, and vibration amplitude can be time-aligned to construct a simulated usage state vector: ; in, Indicates the first The monitored abrasive particles were in the first Simulated usage state vector corresponding to each acquisition time. Indicates the first The monitored abrasive particles were in the first The height of abrasive protrusion at each sampling moment. Indicates the first The monitored abrasive particles and their surrounding coating were in the first... Abrasive damage area at each sampling time. Indicates the relationship with the first The cutting load corresponding to each acquisition moment; Indicates the relationship with the first The equivalent contact temperature corresponding to each data acquisition moment. Indicates the relationship with the first Vibration amplitude values ​​corresponding to each acquisition time; In some embodiments, the abrasive grain protrusion height and abrasive grain damage area are obtained for different monitored abrasive grains; the cutting load, equivalent contact temperature and vibration amplitude are the overall state parameters of the contact area between the electroplated diamond tool and the simulated workpiece within the corresponding acquisition period, and are combined with the abrasive grain protrusion height and abrasive grain damage area of ​​each monitored abrasive grain within the acquisition period.

[0037] In some embodiments, the set of monitored abrasive particles can be periodically updated based on subsequently acquired abrasive particle images. When the protrusion height or damage area of ​​an unmonitored abrasive particle reaches a preset screening threshold, the corresponding abrasive particle can be added to the set of monitored abrasive particles; when a monitored abrasive particle has detached, the detachment status of the corresponding abrasive particle is recorded, and new abrasive particles are selected from the abrasive particle observation area where it is located to supplement the set of monitored abrasive particles.

[0038] By performing zonal testing on the surface of electroplated diamond tools and periodically updating the monitored abrasive grain set, it is possible to continuously monitor potentially high-risk abrasive grains in different abrasive grain observation areas while reducing the number of abrasive grains detected and the amount of data processing. This provides a data foundation for subsequently determining the abrasive grain bonding state vector, abrasive grain damage growth state, and abrasive grain detachment risk.

[0039] S2. Determine the abrasive particle bonding state vector based on simulated usage state parameters; In some embodiments, the abrasive bonding state vector is used to characterize the current bonding stability between the monitored abrasive and the coating.

[0040] In some embodiments, based on the abrasive protrusion height and abrasive damage area obtained in step S1, the abrasive protrusion stable component and the abrasive damage stable component can be determined respectively, and the abrasive protrusion stable component and the abrasive damage stable component can be fused to determine the abrasive comprehensive bonding stability coefficient, so as to construct the abrasive bonding state vector.

[0041] In some embodiments, determining the abrasive grain bonding state vector includes: S201. Determine the stable component of abrasive protrusion based on the abrasive protrusion height; In some embodiments, the abrasive protrusion stability component characterizes the effect of abrasive protrusion height on the bonding stability between the abrasive grain and the coating. For electroplated diamond tools with the same abrasive grain size range, coating material, and coating thickness, a larger abrasive protrusion height results in more abrasive grains protruding from the coating surface, leading to a longer lever arm under cutting loads and a higher likelihood of abrasive grain loosening or pull-out. Therefore, the abrasive protrusion stability component decreases as the abrasive protrusion height increases.

[0042] In some embodiments, the first The monitored abrasive particles were in the first The stable component of abrasive protrusion corresponding to each acquisition moment is: ; in, Indicates the first The monitored abrasive particles were in the first The stable component of abrasive particles at each acquisition moment; Indicates the first The monitored abrasive particles were in the first The height of abrasive protrusion at each acquisition moment; This indicates the critical protrusion height corresponding to the abrasive grain size range, coating material, and coating thickness of the current electroplated diamond tool; In some embodiments, when the abrasive protrusion height is 0, the abrasive protrusion stability component is 1; as the abrasive protrusion height approaches the critical protrusion height, the abrasive protrusion stability component gradually approaches 0; when the abrasive protrusion height reaches or exceeds the critical protrusion height, the abrasive protrusion stability component is set to 0.

[0043] In some embodiments, the critical protrusion height can be calibrated based on historical abrasive pull-out tests or simulated usage tests. Specifically, the protrusion heights of multiple abrasive grains before loosening or detachment can be statistically analyzed, and the critical protrusion height can be determined based on the average value of the protrusion heights, a preset quantile, or a preset safety factor.

[0044] S202. Determine the stable component of abrasive damage based on the abrasive damage area; In some embodiments, the abrasive damage stability component is used to characterize the influence of abrasive wear, abrasive breakage, abrasive cracks, and peeling of the coating around the abrasive on the abrasive bonding stability.

[0045] The larger the abrasive damage area, the more severe the damage to the abrasive itself or the surrounding coating, the smaller the effective bonding area between the abrasive and the coating, and the higher the likelihood of abrasive detachment. Therefore, the stable component of abrasive damage decreases as the abrasive damage area increases.

[0046] In some embodiments, the first The monitored abrasive particles were in the first The abrasive damage stability component corresponding to each acquisition time is: ; in, Indicates the first The monitored abrasive particles were in the first Stable component of abrasive damage at each acquisition time Indicates the first The monitored abrasive particles and their surrounding coating were in the first... Abrasive damage area at each sampling time. This indicates the critical damage area before the abrasive grains loosen or fall off. In some embodiments, when no identifiable damage is observed in the abrasive grains, the abrasive grain damage area is 0 and the abrasive grain damage stability component is 1; as the abrasive grain damage area approaches the critical damage area, the abrasive grain damage stability component gradually approaches 0; when the abrasive grain damage area reaches or exceeds the critical damage area, the abrasive grain damage stability component is set to 0.

[0047] In some embodiments, the critical damage area can be obtained based on historical abrasive pull-out tests, simulated use tests, or statistical analysis of abrasive detachment samples. Specifically, the abrasive damage area corresponding to multiple abrasive particles before detachment can be statistically analyzed, and the critical damage area can be determined based on the average value of the abrasive damage area, a preset quantile, or a preset safety factor.

[0048] S203. Determine the comprehensive bonding stability coefficient of abrasive particles based on the abrasive particle protrusion stability component and the abrasive particle damage stability component. In some embodiments, the abrasive protrusion stability component and the abrasive damage stability component can be weighted and fused to determine the overall abrasive bonding stability coefficient: ; in, Indicates the first The monitored abrasive particles were in the first The overall stability coefficient of abrasive particles at each acquisition time. This represents the weighting coefficient corresponding to the stable component of abrasive grain protrusion. This represents the weighting coefficient corresponding to the stable component of abrasive damage; In some embodiments, the weighting coefficients satisfy: ,and , ; In some embodiments, the weighting coefficients can be calibrated based on historical abrasive grain detachment samples. Specifically, the abrasive grain protrusion height and abrasive grain damage area corresponding to the historical samples can be used as inputs, and whether the corresponding abrasive grains have detached can be used as the calibration result. The weighting coefficients can then be determined through regression analysis or parameter optimization methods.

[0049] In some embodiments, the abrasive grain bonding stability coefficient ranges from 0 to 1. The closer the abrasive grain bonding stability coefficient is to 1, the more stable the bonding state between the monitored abrasive grain and the coating is; the closer the abrasive grain bonding stability coefficient is to 0, the lower the bonding stability between the monitored abrasive grain and the coating is.

[0050] S204. Construct the abrasive particle bonding state vector; In some embodiments, the stability coefficient of abrasive protrusion, abrasive damage, and abrasive combined stability can be constructed. The monitored abrasive particles were in the first The abrasive grain bonding state vector corresponding to each acquisition time: , ; in, Indicates the first The monitored abrasive particles were in the first The abrasive grain bonding state vector corresponding to each acquisition moment. Indicates the stable component of abrasive grains. Indicates the stable component of abrasive damage. Indicates the overall stability coefficient of abrasive particle bonding; In some embodiments, step S2 characterizes the current static bonding state of the monitored abrasive particles based on the abrasive particle protrusion height and abrasive particle damage area at the current acquisition time; the change in abrasive particle damage area between adjacent acquisition times is used in subsequent steps to determine the abrasive particle damage growth rate, so as to characterize the dynamic development state of abrasive particle damage.

[0051] By analyzing the current static bonding state of abrasive grains and the dynamic development state of subsequent abrasive grain damage separately, we can avoid repeatedly calculating the abrasive grain damage area and provide a data basis for determining the risk of abrasive grain detachment based on the cutting load, contact temperature, and vibration amplitude.

[0052] S3. Determine the abrasive detachment risk index based on the abrasive bonding state vector and simulated usage state parameters; In some embodiments, the abrasive grain detachment risk index is used to characterize the risk of the monitored abrasive grains loosening or detaching under the combined effects of their own bonding stability, external processing conditions, and damage dynamic development.

[0053] In some embodiments, the abrasive bonding stability coefficient can be extracted from the abrasive bonding state vector determined in step S2, and the external action index can be determined based on the cutting load, contact temperature and vibration amplitude obtained in step S1; at the same time, the abrasive damage growth rate can be determined based on the abrasive damage area of ​​the same monitored abrasive at adjacent acquisition times, and then the abrasive shedding risk index can be determined.

[0054] In some embodiments, determining the abrasive grain shedding risk index includes: S301. Determine the external action index based on simulated usage state parameters; In some embodiments, the external action index is used to comprehensively characterize the cutting action, thermal action, and vibration and shock action experienced by the electroplated diamond tool during simulated machining.

[0055] In some embodiments, the cutting load, contact temperature and vibration amplitude can be dimensionlessly processed separately, and the dimensionless parameters can be weighted and fused to determine the external action index.

[0056] The specific method is as follows: ; in, Indicates the first The external influence index corresponding to each data collection moment. Indicates the relationship with the first The cutting load corresponding to each acquisition time. Indicates the relationship with the first The equivalent contact temperature corresponding to each data acquisition moment. Indicates the relationship with the first The vibration amplitude corresponding to each acquisition time. This indicates the reference value for the cutting load. Indicates the reference temperature. This indicates the reference value for temperature rise. This indicates the reference value for vibration amplitude. , and These represent the external action weighting coefficients corresponding to cutting load, contact temperature, and vibration amplitude, respectively. In some embodiments, the external action weighting coefficient satisfies: ,and , , ; In some embodiments, the cutting load reference value, temperature rise reference value, and vibration amplitude reference value can be calibrated based on historical stable machining samples and abrasive grain shedding samples, or determined based on the allowable operating range of the simulated machining device.

[0057] In some embodiments, the external action weighting coefficient can be determined based on historical simulated machining test data through regression analysis or parameter optimization methods. A larger external action index indicates a stronger overall external action experienced by the electroplated diamond tool under the current machining condition.

[0058] In some embodiments, the cutting load, contact temperature, and vibration amplitude are the overall tool state parameters within the corresponding acquisition period. Therefore, each monitored abrasive grain within the same acquisition period can use the same external action index. The risk differences between different monitored abrasive grains are mainly determined by the bonding stability and damage development state of each abrasive grain.

[0059] S302. Determine the risk value of foundation shedding based on the comprehensive stability coefficient of abrasive particles and the external action index; In some embodiments, the baseline detachment risk value is used to characterize the relative relationship between the degree of external influence and the degree of self-bonding stability of the monitored abrasive particles.

[0060] In some embodiments, the abrasive bonding stability coefficient can be extracted from the abrasive bonding state vector determined in step S2, and the base detachment risk value can be determined according to the following formula: ; in, Indicates the first The monitored abrasive particles were in the first The baseline shedding risk value corresponding to each data collection moment Indicates the first The external influence index corresponding to each data collection moment; Indicates the first The monitored abrasive particles were in the first The overall stability coefficient of abrasive particles corresponding to each acquisition time; This represents the preset minimum stability coefficient to prevent the denominator from being zero; In some embodiments, when the external action index remains constant, the smaller the abrasive grain bonding stability coefficient, the greater the risk of foundation detachment; when the abrasive grain bonding stability coefficient remains constant, the larger the external action index, the greater the risk of foundation detachment.

[0061] In some embodiments, when the overall stability coefficient of abrasive particles is not greater than a preset minimum stability coefficient, the corresponding monitored abrasive particles can be directly identified as abrasive particles with a high risk of shedding.

[0062] S303. Determine the damage growth correction factor based on the abrasive damage area at adjacent acquisition times; In some embodiments, step S2 characterizes the current static damage level of the abrasive particles based on the abrasive damage area at the current acquisition time; this step characterizes the dynamic development state of abrasive damage based on the change in abrasive damage area of ​​the same monitored abrasive particles at adjacent acquisition times, thereby using the abrasive damage area and its growth rate to characterize the static damage level and dynamic damage trend respectively, avoiding the repeated superposition of the same absolute damage amount as two independent risk items.

[0063] In some embodiments, for the same monitored abrasive grain that exists simultaneously in both the current monitored abrasive grain set and the monitored abrasive grain set at the previous acquisition time, the abrasive grain damage growth rate can be determined based on its abrasive grain damage area at adjacent acquisition times: ; in, Indicates the first The monitored abrasive particles were in the first The abrasive damage growth rate corresponding to each acquisition time. Indicates the first The monitored abrasive particles were in the first The abrasive damage area corresponding to each acquisition moment. This indicates the abrasive damage area corresponding to the same monitored abrasive grain at the previous data acquisition time. This indicates the actual time interval between the current data acquisition time and the previous data acquisition time; In some embodiments, when the current abrasive damage area is less than the abrasive damage area at the previous acquisition time, the difference can be considered to be caused by image recognition error, position registration error or measurement fluctuation, and the abrasive damage growth rate can be set to 0.

[0064] In some embodiments, the damage growth correction factor can be determined based on the abrasive damage growth rate: ; in, Indicates the first The monitored abrasive particles were in the first Damage growth correction factor corresponding to each acquisition time point This represents the damage growth correction factor. Indicates the critical damage growth rate; In some embodiments, the damage growth correction coefficient and critical damage growth rate can be calibrated based on historical abrasive grain detachment samples. Specifically, the abrasive grain damage growth rate can be statistically analyzed over multiple sampling cycles before abrasive grain detachment, and the damage growth correction coefficient and critical damage growth rate can be determined through regression analysis or parameter optimization methods.

[0065] In some embodiments, when the abrasive damage area does not increase, the damage growth correction factor is 1; as the abrasive damage growth rate increases, the damage growth correction factor gradually increases.

[0066] In some embodiments, for a newly added abrasive grain set at the current acquisition time but lacking the damage area from the previous acquisition time, its damage growth correction factor can be set to 1 at the current acquisition time, and its damage growth rate can be calculated based on the abrasive grain damage area obtained at the next acquisition time. The damage state already formed by the newly added abrasive grain is characterized by the abrasive grain damage stability component in step S2.

[0067] S304. Determine the abrasive particle detachment risk index based on the basic detachment risk value and the damage growth correction factor; In some embodiments, the abrasive grain shedding risk index can be determined based on the base shedding risk value and the damage growth correction factor: , ; in, Indicates the first The monitored abrasive particles were in the first The abrasive grain shedding risk index corresponding to each data collection moment. Indicates the risk value of base shedding. Indicates the damage growth correction factor; In some embodiments, the external action reference value, external action weight coefficient, combined stability parameter and damage growth correction parameter can be calibrated based on historical stable abrasive samples and abrasive detachment samples, so that the critical value of the abrasive detachment risk index is 1.

[0068] In some embodiments, when When the corresponding monitored abrasive grain is identified as a stable abrasive grain; when When the corresponding monitored abrasive grain is identified as the critical abrasive grain for shedding, then... At that time, the corresponding monitored abrasive particles are identified as abrasive particles with a high risk of shedding.

[0069] In some embodiments, when the monitored abrasive grains have detached, the actual detachment time and detachment state of the abrasive grains can be recorded, and the detachment state can be used as the training label for a subsequent abrasive grain detachment trend prediction model.

[0070] In some embodiments, the regional shedding risk value corresponding to different abrasive observation areas can be determined based on the abrasive observation area where each monitored abrasive grain is located.

[0071] For the The risk value of abrasive particle observation area can be expressed as: ; in, ; Indicates the first The first abrasive observation area in the first The risk value of regional detachment corresponding to each data collection time. Indicates the first One abrasive observation area, Indicates the first The abrasive grain at the first The spatial location at each acquisition time.

[0072] By using the maximum abrasive shedding risk index within each abrasive observation area as the regional shedding risk value, we can avoid the local high-risk abrasive particles being masked by the average risk of other stable abrasive particles in the same area.

[0073] like Figure 2 As shown, spatial mapping can be performed based on the spatial location of different monitored abrasive grains and their abrasive grain detachment risk index to form a continuous spatial distribution cloud map of the abrasive grain detachment risk index on the surface of electroplated diamond tools. The color gradually transitions from blue to red, representing abrasive grain detachment risk from low to high; when the abrasive grain detachment risk index is greater than or equal to the detachment threshold of 1, it can be uniformly mapped to red.

[0074] Figure 2The regional risk differences are mainly due to the different comprehensive stability coefficients and abrasive damage growth rates of the different monitored abrasive particles. When the tool cutting load, equivalent contact temperature, or vibration amplitude increases within the corresponding acquisition period, the overall risk of abrasive particle detachment in each observation area increases.

[0075] It should be noted that, Figure 2 The continuous risk distribution can be obtained by spatial interpolation based on the spatial location of the discrete monitored abrasive particles and the abrasive particle detachment risk index. It is used to intuitively display the risk change trend between adjacent abrasive particles. The spatial interpolation result is not used as newly added abrasive particle measured risk data.

[0076] like Figure 3 As shown, the abrasive particle detachment risk index of each monitored abrasive particle can be summarized according to the pre-divided fixed abrasive particle observation areas, and the maximum abrasive particle detachment risk index in each observation area can be determined as the regional detachment risk value of the corresponding area. Furthermore, color mapping is performed based on the spatial location of each observation area and the regional detachment risk value to form a distribution map of abrasive particle detachment risk in the observation areas.

[0077] Figure 3 Each annular or sector-shaped partition corresponds to a fixed abrasive grain observation area. Red areas indicate areas with a shedding risk value greater than or equal to the shedding threshold of 1, classifying them as high-risk areas; yellow or green areas indicate areas with a moderate shedding risk; and blue areas indicate areas with a low shedding risk. By using the maximum abrasive grain shedding risk index within the area as the regional shedding risk value, it is possible to avoid localized high-risk abrasive grains being masked by the average risk of other stable abrasive grains within the same area.

[0078] Figure 2 It is mainly used to reflect the continuous spatial variation and local risk concentration of the risk of a single monitored abrasive grain falling off on the tool surface; Figure 3 It is mainly used to convert continuous risk distributions into risk states corresponding to fixed observation areas, so as to construct regional risk time series and predict the future abrasive detachment risk of different regions. By combining... Figure 2 and Figure 3 It can not only intuitively display the continuous spatial distribution of abrasive grain shedding risk, but also form regional risk characterization results suitable for trend prediction and closed-loop control of processing parameters.

[0079] S4. Construct an abrasive detachment trend prediction model based on the abrasive detachment risk index to determine the predicted risk value of abrasive detachment at future times; In some embodiments, the predicted risk value of abrasive grain shedding at a future time is used to characterize the degree of risk of abrasive grain shedding at a preset future time for different abrasive grain observation areas of the electroplated diamond tool after adopting candidate processing parameters.

[0080] In some embodiments, determining the predicted risk value of abrasive grain shedding at future times includes: S401. Construct a time series of risk in the abrasive particle observation area; In some embodiments, the regional shedding risk value determined in S3 can be continuously obtained according to the fixed abrasive observation area divided in S1.

[0081] In some embodiments, the same abrasive observation area can be continuously The regional detachment risk values ​​at each collection time are arranged in chronological order to obtain the regional risk time series: ; in, Indicates the first Abrasive observation area in The time series of regional risks corresponding to each moment. This indicates the number of sampling points included within the time window, and ; In some embodiments, since the location of the abrasive observation area is fixed, even if part of the monitored abrasive in the area is detached or replaced by a new representative abrasive, the regional risk time series can continue to be updated according to the maximum detachment risk index of the currently monitored abrasive in that area, thereby avoiding interruption of the risk time series due to the detachment of a single abrasive or changes in abrasive identification.

[0082] In some embodiments, the detachment risk values ​​for each region are collected at the same sampling time interval. When the sampling time intervals are inconsistent, they can be converted into an equal time interval sequence through time alignment and interpolation.

[0083] S402. Construct a region prediction input sequence containing processing parameters; In some embodiments, while acquiring the risk value of regional detachment, the controller of the simulated processing device can record the process control parameter settings that are issued and actually executed at the corresponding acquisition time.

[0084] In some embodiments, the process control parameters include spindle drive current setting, coolant flow rate, feed rate, and spindle speed.

[0085] In some embodiments, the above-mentioned process control parameters are selected because they can all be adjusted online by the simulated machining device and respectively affect the main channels through which the risk of abrasive grain detachment occurs. Specifically, the spindle drive current setting is used to adjust the spindle output torque or the upper limit of torque, thereby affecting the tool cutting load; the coolant flow rate is used to adjust the heat exchange and chip removal capacity of the machining area, thereby affecting the contact temperature; the feed rate is used to adjust the amount of material removed per unit time and the cutting impact borne by the abrasive grains; and the spindle speed is used to adjust the contact frequency between the abrasive grains and the workpiece, and affects the frictional heat and vibration state. Therefore, the above-mentioned process control parameters can adjust the risk of abrasive grain detachment from the aspects of mechanical load, thermal effect, and dynamic impact, and are suitable as closed-loop adaptive control variables.

[0086] In some embodiments, the first The processing parameter vector at time t is: ; in, This indicates the spindle drive current setting value. Indicates coolant flow rate. Indicates the feed rate. Indicates the spindle speed; In some embodiments, the regional detachment risk value at the same acquisition time can be combined with the processing parameter vector to obtain the regional operating status vector: ; Furthermore, the same abrasive observation area is continuously... The regional operational state vectors at each acquisition time are arranged in chronological order to obtain the regional prediction input sequence: ; in, Indicates the first Abrasive observation area in The region prediction input sequence corresponding to the given time; By using historical regional risks and corresponding historical processing parameters as inputs, the prediction model can learn the correspondence between changes in processing parameters and changes in regional abrasive shedding risk.

[0087] S403. Construct a predictive model for abrasive particle shedding trend based on the regional prediction input sequence; In some embodiments, the abrasive shedding trend prediction model is used to predict the future shedding risk of the region after adopting the candidate processing parameters, based on the historical risk changes in the abrasive observation area, historical processing parameters, and candidate processing parameters.

[0088] In some embodiments, the candidate processing parameter vector is: ; in, , , and These represent the candidate spindle drive current setting, candidate coolant flow rate, candidate feed rate, and candidate spindle speed, respectively. In some embodiments, a long short-term memory neural network can be used to construct the abrasive grain shedding trend prediction model. The long short-term memory neural network first performs temporal encoding on the regional prediction input sequence, then combines the encoding result with the candidate processing parameter vector, and outputs the future regional shedding prediction risk value through a fully connected layer.

[0089] In some embodiments, the training set, validation set, and test set can be divided in proportions of 70%, 15%, and 15%, respectively. When dividing the dataset, time series samples generated by the same electroplated diamond tool or the same continuous processing experiment are assigned to the same data subset to avoid data leakage between adjacent time windows.

[0090] In some embodiments, the abrasive grain shedding trend prediction model is: ; in, Indicates the first Each abrasive observation area uses candidate machining parameters. Afterwards, in the future Predicted risk value at any given time. This represents a model for predicting the trend of abrasive grain shedding. Indicates model parameters, Indicates the prediction time steps; In some embodiments, the prediction time steps can be... Set a processing parameter adjustment cycle, and keep the candidate processing parameters unchanged within the processing parameter adjustment cycle.

[0091] In some embodiments, the abrasive grain shedding trend prediction model can be trained using historical simulated processing data. Specifically, the historical region prediction input sequence and the actual processing parameters used within the corresponding prediction time period are used as training inputs, and the future... The region shedding risk value determined by S3 is used as the training label at all times.

[0092] In some embodiments, when the monitored abrasive particles in the corresponding abrasive particle observation area are in the future... If a dropout has already occurred before a certain time, the corresponding training label can be set as a preset dropout risk label. ,and .

[0093] In some embodiments, the loss function used for model training is: ; in, This represents the mean squared error loss function. Indicates the number of training samples. Indicates the first The predicted risk value corresponding to each training sample This indicates the corresponding actual area detachment risk value or preset detachment risk label; In some embodiments, by unifying feature data of different dimensions to the same scale to facilitate model training, the regional risk index and processing parameters can be standardized based on the mean and standard deviation in the training dataset. ; in, This represents the original parameters to be standardized. and These represent the mean and standard deviation of the parameter in the training dataset, respectively. This represents the standardized parameters; In some embodiments, the mean and standard deviation determined during the training phase are used in model validation and actual prediction processes to ensure the authenticity of the model evaluation and the consistency of online prediction.

[0094] S404. Determine the predicted risk value of abrasive particle shedding at future time points; In some embodiments, the current region prediction input sequence and candidate processing parameters can be input into the trained abrasive grain shedding trend prediction model to obtain the future region shedding prediction risk value: ; in, Indicates the first After selecting candidate processing parameters for the first abrasive observation region, in the future... Predicted risk value of abrasive grain shedding at any given time. This represents the parameters of the model after training is complete. This represents the standardized current region prediction input sequence. This represents the standardized candidate processing parameter vector; Furthermore, based on the predicted risk values ​​of different abrasive grain observation areas, the predicted maximum future detachment risk value of the electroplated diamond tool under the aforementioned candidate processing parameters can be determined: ; in, Indicates the candidate machining parameters used for electroplated diamond tools. Afterwards, in the future The maximum predicted risk value of future shedding at that time. This indicates the number of abrasive grain observation areas.

[0095] In some embodiments, the risk status of a corresponding region can be determined based on the predicted risk value of future regional shedding corresponding to each abrasive observation region. When At that time, the first Each abrasive grain observation area was determined as the predicted stable region; when At that time, the first The abrasive particle observation area was determined as the critical region for predicting shedding; when At that time, the first The abrasive observation area was identified as a region with a high risk of shedding.

[0096] Furthermore, when When, the electroplated diamond tool as a whole is determined as the predicted stable state; when At that time, the electroplated diamond tool as a whole was identified as being in a state with a predicted high risk of detachment.

[0097] In some embodiments, when the historical data collected has not yet reached the number of sampling points required for the time window, the current area detachment risk value can be temporarily used as the future predicted risk value for that area, and the abrasive detachment trend prediction model can be activated after the number of sampling points reaches the preset requirement.

[0098] By using the maximum abrasive shedding risk index of a fixed abrasive observation area as the prediction target, the prediction sequence can be kept continuous even when there are a large number of monitored abrasive particles, abrasive particle shedding occurs, or the monitored abrasive particles are dynamically updated. Furthermore, by directly incorporating candidate processing parameters into the abrasive shedding trend prediction model, the impact of different candidate processing parameters on the future regional shedding risk can be predicted. This allows subsequent steps to select processing parameters with lower maximum shedding prediction risk values ​​from multiple candidate processing parameters, achieving adaptive optimization of processing parameters.

[0099] S5. Determine and execute the optimal process control parameters based on the predicted risk value of abrasive particle shedding at future moments; In some embodiments, the adjustment range of process parameters can be determined based on the predicted maximum future detachment risk value of the electroplated diamond tool, the adjustment direction of process parameters can be determined based on the contribution of cutting load, contact temperature and vibration amplitude to the risk of abrasive grain detachment, and the adjustment results can be verified using the abrasive grain detachment trend prediction model in S4 to determine the optimal process control parameters.

[0100] In some embodiments, determining and executing the optimal process control parameters includes: S501. Determine the intensity of process parameter adjustment and the dominant risk factors; In some embodiments, the current process control parameters can be input into the abrasive grain shedding trend prediction model in S4 to obtain the future maximum shedding prediction risk value when maintaining the current process control parameters.

[0101] For ease of subsequent calculations, the predicted maximum future shedding risk value while maintaining the current process control parameters is denoted as: ; in, ; This represents the current process control parameter vector. This indicates the current spindle drive current setting. Indicates the current coolant flow rate. Indicates the current feed rate. Indicates the current spindle speed; In some embodiments, the intensity of process parameter adjustment can be determined based on the degree to which the predicted maximum future shedding risk value exceeds a preset safety risk value: ; in, Indicates the intensity of process parameter adjustment. This indicates the preset security risk value. This indicates a preset risk warning value, and: ; In some embodiments, it can be Set the abrasive grain shedding threshold to 1.

[0102] When the predicted maximum future detachment risk value is not greater than the preset safety risk value. This indicates that the current process control parameters will be maintained; when the predicted maximum future shedding risk value reaches or exceeds the risk warning value, This indicates that the parameter is adjusted according to the preset maximum single adjustment amount.

[0103] In some embodiments, the dominant risk factor can be determined based on the three components of the externality index in S3: ; ; ; in, , and These represent the risk contribution corresponding to cutting load, contact temperature, and vibration amplitude, respectively. The dominant risk factors are: ; in, This indicates the current dominant risk factors; S502. Generate candidate process control parameters based on the dominant risk factors and determine the optimal process control parameters; When cutting load is the dominant risk factor, candidate parameters for load adjustment can be generated: ; When contact temperature is the dominant risk factor, candidate parameters for temperature regulation can be generated: ; When the vibration amplitude is the dominant risk factor, two candidate vibration adjustment parameters can be generated, one to reduce the spindle speed and the other to increase the spindle speed. ; ; in, , , and These represent the preset maximum single adjustment amount for the corresponding process parameters; This indicates that the adjusted process parameters are limited to the allowable range of the simulated processing device; In some embodiments, by simultaneously setting candidate parameters for increasing and decreasing spindle speed, it is possible to avoid directly assuming that decreasing spindle speed will necessarily reduce vibration, and to select the speed adjustment direction with less vibration risk through subsequent prediction.

[0104] In some embodiments, when It can also generate comprehensive protection candidate parameters: ; The comprehensive protection candidate parameters are used to simultaneously reduce cutting load, improve cooling capacity, and reduce feed impact.

[0105] In some embodiments, the current process control parameters, the candidate adjustment parameters corresponding to the dominant risk factors, and the comprehensive protection candidate parameters can be used to form a candidate parameter set. .

[0106] The candidate parameter set may include at most the current process control parameters, two dominant risk adjustment candidate parameters, and one comprehensive protection candidate parameter.

[0107] Each candidate process control parameter in the candidate parameter set is input into the abrasive grain shedding trend prediction model in S4 to obtain the corresponding future maximum shedding prediction risk value, and the candidate parameter with the smallest future maximum shedding prediction risk value is selected: ; in, Indicates the optimal process control parameters; This represents the set of candidate process control parameters generated at the current moment; When the difference between the predicted maximum shedding risk values ​​of two candidate process control parameters is less than the preset prediction error, the candidate parameter with the smaller change relative to the current process control parameter is selected first to reduce frequent fluctuations in process parameters.

[0108] S503, Execute the optimal process control parameters and perform closed-loop updates; In some embodiments, the optimal process control parameter is: ; in, , , , These represent the optimal spindle drive current setting, optimal coolant flow rate, optimal feed rate, and optimal spindle speed, respectively.

[0109] In some embodiments, the optimal process control parameters can be sent to the controller of the simulated machining device, and the controller can adjust the spindle drive, coolant circulation system, feed mechanism and spindle rotation mechanism.

[0110] After the optimal process control parameters are continuously executed for one parameter adjustment cycle, S1 to S5 are re-executed to update the abrasive simulated usage state, abrasive bonding state, abrasive shedding risk, and optimal process control parameters, thereby forming a closed-loop adaptive control process.

[0111] When the predicted maximum risk of future shedding is not greater than the preset safety risk value, the current process control parameters can be maintained and processing can continue.

[0112] When the predicted risk value of the maximum future shedding corresponding to all candidate process control parameters is greater than 1, the comprehensive protection candidate parameters can be executed and a risk warning can be output; when the risk still does not decrease after the comprehensive protection candidate parameters are executed, the simulated processing device can be controlled to suspend processing.

[0113] Please see Figure 4 and Figure 5 , Figure 4 To obtain SEM images of electroplated diamond tools after simulated use tests using the pre-optimized process control parameters, Figure 5 SEM images of electroplated diamond tools after completing the same simulated use test using optimal process control parameters.

[0114] In some embodiments, Figure 4The red circles are used to mark abrasive grain detachment pits, abrasive grain damage areas, or areas where the coating around the abrasive grain has peeled off. Abrasive grain detachment pits are the pitted areas formed after the diamond abrasive grain detaches from the coating; abrasive grain damage areas are areas where the diamond abrasive grain has cracked or is partially missing; and coating peeling areas are areas where the coating around the abrasive grain has cracked, peeled off, or the material has detached.

[0115] Depend on Figure 4 and Figure 5 As can be seen from the field-of-view comparison shown, Figure 5 The number of intact abrasive grains is relatively large, while the number of obvious detachment pits, abrasive grain damage areas, and coating peeling areas is relatively small. This indicates that, in this embodiment, using the optimal process control parameters determined in this application helps to reduce abrasive grain detachment and damage, improve the integrity of abrasive grain morphology, and enhance the stability of the simulated processing.

[0116] A second aspect of the present invention provides a machining parameter optimization system based on the prediction of the risk of electroplated diamond tool detachment, for executing the aforementioned machining parameter optimization method based on the prediction of the risk of electroplated diamond tool detachment, the system comprising: The acquisition module is used to acquire simulated usage state parameters of the monitored abrasive grains, including abrasive grain protrusion height, abrasive grain damage area, tool cutting load, equivalent contact temperature, and vibration amplitude. The first determining module is used to determine the abrasive protrusion stability component, abrasive damage stability component, and abrasive overall bonding stability coefficient of the monitored abrasive based on the abrasive protrusion height and abrasive damage area, and to construct an abrasive bonding state vector. The second determining module is used to determine the shedding risk index of the monitored abrasive particles based on the abrasive particle bonding state vector, tool cutting load, equivalent contact temperature, vibration amplitude and abrasive particle damage area growth rate, and to determine the regional shedding risk value based on the shedding risk index of the monitored abrasive particles in each fixed abrasive particle observation area. The prediction module is used to construct a regional risk time series based on the regional shedding risk values ​​at different acquisition times, and combine historical process control parameters and candidate process control parameters to predict the future shedding prediction risk value of each abrasive observation area after adopting different candidate process control parameters. The adaptive control module is used to determine the intensity of process parameter adjustment based on the current predicted risk value of the maximum future shedding, determine the dominant risk factors based on the risk contribution of cutting load, equivalent contact temperature and vibration amplitude, generate candidate process control parameters based on the process parameter adjustment intensity and the dominant risk factors, and call the prediction module to determine the predicted risk value of the maximum future shedding corresponding to each candidate process control parameter. The adaptive control module is also used to determine the candidate process control parameter with the smallest predicted risk value of the maximum future shedding as the optimal process control parameter, and send the optimal process control parameter to the controller of the simulated machining device to adjust the spindle drive current set value, coolant flow rate, feed speed and spindle speed. In some embodiments, the adaptive control module is further configured to trigger the acquisition module to reacquire the simulated usage state parameters after the optimal process control parameters have been continuously executed for one parameter adjustment cycle, and sequentially update the abrasive bonding state vector, abrasive detachment risk index, future detachment prediction risk value, and optimal process control parameters to form a closed-loop adaptive control; when the future maximum detachment prediction risk value is not greater than the preset safety risk value, the adaptive control module maintains the current process control parameters and continues to execute; when the future maximum detachment prediction risk values ​​corresponding to all candidate process control parameters are greater than 1, the adaptive control module executes the comprehensive protection candidate parameters and outputs an abrasive detachment risk warning; when the future maximum detachment prediction risk value still does not decrease after the comprehensive protection candidate parameters are executed, the simulated processing device is controlled to suspend processing.

[0117] Finally, it should be noted that the embodiments disclosed in this invention are merely preferred embodiments of this invention and are only used to illustrate the technical solutions of this invention, not to limit it. Although this invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this invention.

Claims

1. A method for optimizing processing parameters based on the prediction of the risk of electroplated diamond tool detachment, characterized in that, Includes the following steps: S1. Obtain the simulated usage parameters of the electroplated diamond tool; S2. Determine the abrasive particle bonding state vector based on simulated usage state parameters; S3. Determine the abrasive detachment risk index based on the abrasive bonding state vector and simulated usage state parameters; S4. Construct an abrasive detachment trend prediction model based on the abrasive detachment risk index to determine the predicted risk value of abrasive detachment at future times; S5. Determine and execute the optimal process control parameters based on the predicted risk value of abrasive grain shedding at future time.

2. The method for optimizing processing parameters based on the prediction of the risk of electroplated diamond tool detachment according to claim 1, characterized in that, The simulated usage parameters include abrasive grain protrusion height, abrasive grain damage area, cutting load, contact temperature, and vibration amplitude. The determination of the abrasive grain bonding state vector includes: S201. Determine the stable component of abrasive protrusion based on the abrasive protrusion height; S202. Determine the stable component of abrasive damage based on the abrasive damage area; S203. Determine the comprehensive bonding stability coefficient of abrasive particles based on the abrasive particle protrusion stability component and the abrasive particle damage stability component. S204. Construct the abrasive particle bonding state vector.

3. The method for optimizing processing parameters based on the prediction of the risk of electroplated diamond tool detachment according to claim 2, characterized in that, The determination of the abrasive grain shedding risk index includes: S301. Determine the external action index based on simulated usage state parameters; S302. Determine the risk value of foundation shedding based on the comprehensive stability coefficient of abrasive particles and the external action index; S303. Determine the damage growth correction factor based on the abrasive damage area at adjacent acquisition times; S304. Determine the abrasive particle detachment risk index based on the basic detachment risk value and the damage growth correction factor.

4. The method for optimizing processing parameters based on the prediction of the risk of electroplated diamond tool detachment according to claim 3, characterized in that, The determination of the predicted risk value for abrasive grain shedding at future times includes: S401. Construct a time series of risk in the abrasive particle observation area; S402. Construct a region prediction input sequence containing processing parameters; S403. Construct a predictive model for abrasive particle shedding trend based on the regional prediction input sequence; S404. Determine the predicted risk value of abrasive particle shedding at future times.

5. The method for optimizing processing parameters based on the prediction of the risk of electroplated diamond tool detachment according to claim 4, characterized in that, The determination and execution of optimal process control parameters includes: S501. Determine the intensity of process parameter adjustment and the dominant risk factors; S502. Generate candidate process control parameters based on the dominant risk factors and determine the optimal process control parameters; S503, Execute the optimal process control parameters and perform closed-loop updates.

6. The method for optimizing processing parameters based on the prediction of the risk of electroplated diamond tool detachment according to claim 2, characterized in that, The abrasive grain protrusion stability component is: ； in, Indicates the first The monitored abrasive particles were in the first The stable component of abrasive particles at each acquisition moment; Indicates the first The monitored abrasive particles were in the first The height of abrasive protrusion at each acquisition moment; This indicates the critical protrusion height corresponding to the abrasive grain size range, coating material, and coating thickness of the current electroplated diamond tool; The stable component of abrasive damage is: ； in, Indicates the first The monitored abrasive particles were in the first Stable component of abrasive damage at each acquisition time Indicates the first The monitored abrasive particles and their surrounding coating were in the first... Abrasive damage area at each sampling time. This indicates the critical damage area before the abrasive grains loosen or fall off. The overall stability coefficient of the abrasive particles is: ； in, Indicates the first The monitored abrasive particles were in the first The overall stability coefficient of abrasive particles at each acquisition time. This represents the weighting coefficient corresponding to the stable component of abrasive grain protrusion. This represents the weighting coefficient corresponding to the stable component of abrasive damage.

7. The method for optimizing processing parameters based on the prediction of the risk of electroplated diamond tool detachment according to claim 5, characterized in that, The external action index is: ； in, Indicates the first The external influence index corresponding to each data collection time. Indicates the relationship with the first The cutting load corresponding to each acquisition time. Indicates the relationship with the first The equivalent contact temperature corresponding to each data acquisition moment. Indicates the relationship with the first The vibration amplitude corresponding to each acquisition time. This indicates the reference value for the cutting load. Indicates the reference temperature. This indicates the reference value for temperature rise. This indicates the reference value for vibration amplitude. , and These represent the external action weighting coefficients corresponding to cutting load, contact temperature, and vibration amplitude, respectively. The basic shedding risk value is: ； in, Indicates the first The monitored abrasive particles were in the first The baseline shedding risk value corresponding to each data collection moment Indicates the first The external influence index corresponding to each data collection moment; Indicates the first The monitored abrasive particles were in the first The overall stability coefficient of abrasive particles corresponding to each acquisition moment; This represents the preset minimum stability coefficient to prevent the denominator from being zero; The damage growth correction factor determined based on the abrasive damage area at adjacent acquisition times includes: The damage growth correction factor is determined based on the abrasive damage growth rate. The abrasive damage growth rate is: ； in, Indicates the first The monitored abrasive particles were in the first The abrasive damage growth rate corresponding to each acquisition time. Indicates the first The monitored abrasive particles were in the first The abrasive damage area corresponding to each acquisition moment. This indicates the abrasive damage area corresponding to the same monitored abrasive grain at the previous data acquisition time. This indicates the actual time interval between the current data acquisition time and the previous data acquisition time; The damage growth correction factor is: ； in, Indicates the first The monitored abrasive particles were in the first Damage growth correction factor corresponding to each acquisition time. This represents the damage growth correction factor. Indicates the critical damage growth rate; The abrasive grain shedding risk index is: ， ； in, Indicates the first The monitored abrasive particles were in the first The abrasive grain shedding risk index corresponding to each data collection moment. Indicates the risk value of base shedding. Indicates the damage growth correction factor; The maximum abrasive detachment risk index within each abrasive observation area is taken as the regional detachment risk value. The risk value of detachment in the area is: ； in, ; Indicates the first The first abrasive observation area in the first The risk value of regional detachment corresponding to each data collection time. This represents the i-th abrasive grain observation area. Indicates the first The abrasive grain at the first The spatial location at each acquisition time.

8. The method for optimizing processing parameters based on the prediction of the risk of electroplated diamond tool detachment according to claim 7, characterized in that, The constructed abrasive particle observation region risk time series includes: The same abrasive observation area in continuous The regional detachment risk values ​​at each collection time are arranged in chronological order to obtain the regional risk time series: ； in, Indicates the first Abrasive observation area in The time series of regional risks corresponding to each moment. This indicates the number of sampling points included within the time window, and ; The constructed region prediction input sequence containing processing parameters includes: By combining the regional detachment risk value at the same acquisition time with the processing parameter vector, the regional operating status vector is obtained: ； The same abrasive observation area in continuous The regional operational state vectors at each acquisition time are arranged in chronological order to obtain the regional prediction input sequence: ； The abrasive grain shedding trend prediction model is as follows: ； in, Indicates the first Each abrasive observation area uses candidate machining parameters. Afterwards, in the future Predicted risk value at any given time. This represents a model for predicting the trend of abrasive grain shedding. Indicates model parameters, Indicates the prediction time steps; The determination of the predicted risk value for abrasive grain shedding at future times includes: Input the current region prediction input sequence and candidate processing parameters into the trained abrasive grain shedding trend prediction model to obtain the future region shedding prediction risk value: ； in, Indicates the first After selecting candidate processing parameters for the first abrasive observation region, in the future... Predicted risk value of abrasive grain shedding at any given time. This represents the parameters of the model after training is complete. This represents the standardized current region prediction input sequence. This represents the standardized candidate processing parameter vector; Based on the predicted risk values ​​for different abrasive grain observation areas, determine the maximum predicted risk value of future detachment of the electroplated diamond tool under the aforementioned candidate processing parameters: ； in, Indicates the candidate machining parameters used for electroplated diamond tools. Afterwards, in the future The maximum predicted risk value of future shedding at that time. This indicates the number of abrasive grain observation areas.

9. The method for optimizing processing parameters based on the prediction of the risk of electroplated diamond tool detachment according to claim 8, characterized in that, The determination and execution of optimal process control parameters based on the predicted risk value of abrasive grain shedding at future times includes: The intensity of process parameter adjustment is determined based on the degree to which the predicted maximum future shedding risk value exceeds the preset safety risk value. The predicted maximum future shedding risk value while maintaining the current process control parameters is denoted as: ; The intensity of the process parameter adjustment is: ； in, Indicates the intensity of process parameter adjustment. This indicates the preset security risk value. This indicates a preset risk warning value; The dominant risk factors are determined based on the three components of the external effects index in S3: ； ； ； in, , and These represent the risk contribution corresponding to cutting load, contact temperature, and vibration amplitude, respectively. The dominant risk factors are: ； in, This indicates the current dominant risk factors; The step of generating candidate process control parameters and determining the optimal process control parameters based on the dominant risk factors includes: When cutting load is the dominant risk factor, candidate parameters for load adjustment can be generated: ； When contact temperature is the dominant risk factor, candidate parameters for temperature regulation can be generated: ； When the vibration amplitude is the dominant risk factor, two candidate vibration adjustment parameters can be generated, one to reduce the spindle speed and the other to increase the spindle speed. ； ； in, , , and These represent the preset maximum single adjustment amount for the corresponding process parameters; This indicates that the adjusted process parameters are limited to the allowable range of the simulated processing device; Each candidate process control parameter in the candidate parameter set is input into the abrasive grain shedding trend prediction model in S4 to obtain the corresponding future maximum shedding prediction risk value, and the candidate parameter with the smallest future maximum shedding prediction risk value is selected: ； in, Indicates the optimal process control parameters; This represents the set of candidate process control parameters generated at the current moment; The optimal process control parameters are: ； in, , , , These represent the optimal spindle drive current setting, optimal coolant flow rate, optimal feed rate, and optimal spindle speed, respectively.

Images