A Multi-parameter Collaborative Monitoring and Control Method for Cutting Conditions of Metal Parts
By analyzing cutting torque and vibration data, the causes of load at corners within the cavity are identified, solving the problem of misjudgment in existing monitoring methods and improving the processing efficiency of metal parts.
Patent Information
- Application Number
- CN202611452715.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-09-18
- Publication Date
- 2026-11-17
AI Technical Summary
Existing cutting condition monitoring methods are unable to distinguish the various load causes at the corners inside the cavity, leading to normal load changes being misjudged as abnormal, which affects the processing efficiency of metal parts.
By acquiring the actual cutting torque and vibration data when the tool passes through the corner inside the cavity, and combining the cutting mechanics simulation model and fast Fourier transform, the torque deviation and vibration characteristic spectrum are analyzed to distinguish the changes in contact wrap angle, chip retention and abnormal cutting edge, and corresponding control measures are taken.
It reduces misjudgments of normal load changes, improves the processing efficiency of metal parts, and avoids unnecessary processing interruptions and ineffective control.
Smart Images

Figure CN123113576A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of industrial control system technology, specifically relating to a multi-parameter collaborative monitoring and control method for metal parts cutting conditions. Background Technology
[0002] In industrial control systems for cavity milling of metal parts, the cutting load is affected by factors such as tool path, cutting parameters, chip removal status, and tool condition. Abnormal cutting loads may lead to decreased machining quality or tool damage. Therefore, it is necessary to monitor the cutting conditions during cavity milling and adjust the machining process based on the monitoring results.
[0003] Existing cutting condition monitoring methods typically utilize sensors installed in the machining equipment to collect cutting force, spindle current, and vibration signals, comparing the amplitude of these signals with corresponding preset thresholds. When the amplitude of any signal exceeds the preset threshold, the CNC machining system determines that there is an abnormal load in the current machining state and reduces the feed rate, pauses machining, or performs a tool retraction operation according to a preset control strategy.
[0004] However, when the tool passes through the corner of the cavity, the contact angle between the tool and the workpiece changes with the tool path position. The normal load change caused by the increased contact angle, the additional cutting resistance caused by chip retention in the inner corner area, and the changes in cutting resistance caused by tool edge wear or damage can all manifest as increased cutting force, increased spindle current, and enhanced vibration. Simultaneously, the feed deceleration when the tool passes through the inner corner also alters the amplitude of these signals, causing different load causes to present similar monitoring results. Existing monitoring methods based on amplitude thresholds to determine whether the monitoring signal exceeds a preset range are unable to distinguish between normal inner corner load changes, chip retention, and tool edge abnormalities. This can easily lead to unnecessary feed deceleration, machining pauses, or tool retraction operations for normal load changes, thus affecting the machining efficiency of metal parts. Summary of the Invention
[0005] The purpose of this invention is to solve the problem that existing cutting condition monitoring methods rely on amplitude thresholds, which make it difficult to distinguish the various load causes at the corners inside the cavity. This can easily lead to misjudging normal changes in wrap angle as abnormalities and causing excessive machine shutdowns, thus affecting the processing efficiency of metal parts. Therefore, this invention proposes a multi-parameter collaborative monitoring and control method for the cutting conditions of metal parts.
[0006] This invention proposes a multi-parameter collaborative monitoring and control method for metal parts cutting conditions, the method comprising: The actual cutting torque and vibration data of the tool when passing through the corner inside the cavity are obtained, and the machining setting data of the tool machining the corner inside the cavity is read. The machining setting data is then input into the cutting mechanics simulation model to obtain the theoretical cutting torque. The theoretical cutting torque and the actual cutting torque are compared position by position to obtain the torque deviation curve. The torque deviation curve is then analyzed to obtain the degree of torque deviation and torque attenuation characteristics. The vibration data is subjected to a fast Fourier transform to obtain the vibration characteristic spectrum. Based on the vibration characteristic spectrum, a matching analysis is performed on the torque deviation and torque attenuation characteristics to determine the load cause, which includes changes in contact wrap angle, chip retention, or abnormal cutting edge. When the load is caused by a change in the contact wrap angle, the target feed rate is calculated based on the degree of torque deviation and a feed control command is generated. The feed control command is used to control the tool to continue executing the current tool path according to the target feed rate. When the load is caused by chip retention, the tool retraction distance is determined based on the torque deviation curve and a tool retraction and chip removal command is generated. The tool retraction and chip removal command is used to control the tool to exit the chip retention area. When the load is caused by an abnormal cutting edge, a machining pause command is generated. The machining pause command is used to stop the feed axis movement and pause the execution of the current toolpath.
[0007] Optionally, the machining setting data includes tool diameter, tool machining path, feed rate, metal part material parameters, and cavity corner geometry parameters; the step of reading the machining setting data for machining the cavity corners and inputting the machining setting data into the cutting mechanics simulation model to obtain the theoretical cutting torque includes: The tool machining path is used to obtain multiple path positions when the tool passes through the corner inside the cavity, and the contact wrap angle between the tool and the metal part at each path position is calculated based on the tool diameter and the geometric parameters of the corner inside the cavity. Calculate the effective cutting action at each path position based on the contact wrap angle and feed rate at each path position; Based on the material parameters of the metal parts, the corresponding cutting force coefficient is obtained. The effective cutting action and cutting force coefficient corresponding to each path position are input into the cutting mechanics simulation model to obtain the tangential cutting force corresponding to each path position. Calculate the theoretical cutting torque corresponding to each path position based on the tangential cutting force and tool diameter.
[0008] Optionally, feature analysis can be performed on the torque deviation curve to obtain the degree of torque deviation and torque attenuation characteristics, including: The torque deviation values in the torque deviation curve are read sequentially according to the path position of the tool machining path, and the path interval where the torque deviation value continuously exceeds the preset deviation threshold is determined as the torque deviation interval. Obtain the torque deviation value corresponding to each path position within the torque deviation interval, and calculate the degree of torque deviation based on each torque deviation value and the path length of the torque deviation interval; the degree of torque deviation is used to characterize the overall deviation of the actual cutting torque from the theoretical cutting torque; The path position with the largest torque deviation value within the torque deviation range is determined as the deviation peak position. Along the tool machining direction, starting from the deviation peak position, the path range where the torque deviation value shows a decreasing trend is determined as the torque decay range. The attenuation rate is calculated based on the decrease in torque deviation within the torque attenuation range and the path length of the torque attenuation range, and the attenuation rate is used as the torque attenuation characteristic. The torque attenuation characteristic is used to characterize the degree of attenuation of the deviation between the actual cutting torque and the theoretical cutting torque along the machining path after the tool passes the deviation peak position.
[0009] Optionally, a fast Fourier transform is performed on the vibration data to obtain the vibration characteristic spectrum, including: The vibration data are arranged according to the path position of the tool machining path to obtain the vibration variation curve; Based on the starting and ending path positions of the torque deviation interval, the target vibration curve segment corresponding to the torque deviation interval is extracted from the vibration variation curve. The target vibration curve segment is divided into multiple sequentially arranged vibration sub-curve segments according to the preset path length, and a fast Fourier transform is performed on each vibration sub-curve segment to obtain the vibration spectrum curve corresponding to each vibration sub-curve segment. The vibration spectrum curve includes multiple frequency positions and the vibration amplitude corresponding to each frequency position. Establish the correspondence between each vibration spectrum curve and the path position of the corresponding vibration sub-curve segment, and arrange each vibration spectrum curve according to the position order of the tool machining path to obtain the vibration characteristic spectrum.
[0010] Optionally, a matching analysis is performed on the torque deviation and torque attenuation characteristics based on the vibration characteristic spectrum to determine the load cause, including: Extract the spectral peaks with vibration amplitudes greater than a preset amplitude threshold from the vibration spectrum curves corresponding to each vibration sub-curve segment, and obtain the spectral peak frequency and spectral peak amplitude corresponding to each spectral peak. When the torque deviation is greater than a preset deviation threshold and the torque attenuation characteristic is greater than a preset attenuation threshold, the change in torque deviation and the change in spectral peak amplitude of adjacent vibrating sub-curve segments are calculated according to the path position order of each vibrating sub-curve segment; the change in spectral peak amplitude refers to the difference in the largest spectral peak amplitude among adjacent vibrating sub-curve segments. The number of vibrating sub-curve segments whose changes in torque deviation and spectral peak amplitude have the same sign is used as the amplitude matching quantity. If the number of amplitude matching is greater than or equal to the preset matching number threshold, the load is determined to be caused by a change in the contact wrap angle; If the amplitude matching quantity is less than the preset matching quantity threshold, or the torque attenuation characteristic is less than or equal to the preset attenuation threshold, the frequency duration is determined based on the frequency of each spectral peak, and the load cause is determined based on the frequency duration to be chip retention or abnormal cutting edge.
[0011] Optionally, the frequency duration is determined based on the spectral peak frequency, and the cause of the load is determined based on the frequency duration, including whether it is due to chip retention or abnormal cutting edge. According to the path position order of each sub-curve segment, the spectral peak frequencies in adjacent sub-curve segments are compared to determine the matching spectral peaks whose frequency difference is less than or equal to the preset frequency tolerance. The matching spectrum peaks are continuously analyzed along the tool machining path. When the frequency difference between the matching spectrum peak in the next vibrating sub-curve segment and the matching spectrum peak in the previous vibrating sub-curve segment is less than or equal to the preset frequency tolerance, the corresponding spectrum peak frequency is determined to be continuously present. The number of vibrating sub-curve segments with continuously present spectrum peak frequencies is counted to obtain the frequency duration. When the frequency duration is less than the preset duration threshold, the load is determined to be caused by chip retention. When the frequency duration is greater than or equal to the preset duration threshold, the load is determined to be caused by an abnormal blade edge.
[0012] Optionally, the target feed rate includes a first target feed rate and a second target feed rate; the step of calculating the target feed rate based on the degree of torque deviation and generating feed control commands includes: Obtain the current feed rate and calculate the feed rate reduction based on the difference between the torque deviation and the preset deviation threshold. Based on the current feed rate and the amount of feed rate reduction, calculate the first target feed rate corresponding to the starting position of the torque decay interval; Based on the torque decay characteristics and the path length from each path position to the end path position within the torque decay interval, the predicted residual deviation between the actual cutting torque and the theoretical cutting torque is calculated. Based on the ratio of the predicted residual deviation to the torque deviation, the feed rate reduction is corrected to obtain the feed rate correction amount corresponding to each path position. Calculate the second target feed rate corresponding to each path position based on the first target feed rate and the feed rate correction. Write the first target feed rate and the second target feed rate into the current toolpath according to their corresponding path positions to generate feed control commands.
[0013] Optionally, the retraction distance is determined based on the torque deviation curve, and a retraction and chip removal command is generated, including: Read the torque deviation value in the torque deviation curve along the tool machining direction, and take the path position where the torque deviation value first exceeds the preset deviation threshold as the deviation start position; Obtain the current toolpath position and calculate the path distance along the current toolpath between the current toolpath position and the deviation from the starting position. Obtain the torque deviation value corresponding to each path position between the deviation from the starting position and the current toolpath position, and calculate the tool retraction correction distance based on the difference between each torque deviation value and the preset deviation threshold and the path length between the corresponding path positions; The sum of the path distance and the tool retraction correction distance is used as the tool retraction distance, and a tool retraction path is generated in the opposite direction of the current tool path, starting from the current tool path position. Generate a retraction and chip removal command based on the retraction path.
[0014] The beneficial effects of this invention are: 1. This application compares the actual cutting torque with the theoretical cutting torque position by position along the tool path to obtain a torque deviation curve, and extracts the degree of torque deviation and torque attenuation characteristics. Then, it combines this with vibration characteristic spectrum analysis to differentiate between changes in contact wrap angle, chip retention, and abnormal cutting edges. By comprehensively considering the overall change in torque deviation, the attenuation process, and vibration frequency characteristics, it can reduce confusion when different load causes have similar monitoring results, providing a basis for taking corresponding control measures for different load causes. 2. After determining the cause of the load, this application adopts corresponding response measures according to different load states. When the load is caused by a change in the contact wrap angle, the feed rate is adjusted and the current toolpath continues to be executed. When chip retention exists, the retraction distance is determined according to the torque deviation curve and retraction and chip removal are performed. When an abnormal cutting edge exists, the current machining is paused. This reduces unnecessary retraction or pause operations for changes in the contact wrap angle, avoids using incompatible control methods for chip retention and abnormal cutting edges, reduces ineffective control and unnecessary machining interruptions, and improves the machining efficiency of metal parts. Attached Figure Description
[0015] The invention will now be further described with reference to the accompanying drawings.
[0016] Figure 1 A flowchart of a multi-parameter collaborative monitoring and control method for metal parts cutting conditions provided in an embodiment of the present invention; Figure 2 This is a flowchart for determining the cause of load according to an embodiment of the present invention. Detailed Implementation
[0017] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.
[0018] It should be noted that all formula calculations in the scheme are purely numerical calculations.
[0019] 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.
[0020] This invention provides a multi-parameter collaborative monitoring and control method for the cutting conditions of metal parts. See also... Figure 1 The method includes the following steps: S1, acquire the actual cutting torque and vibration data when the tool passes through the corner inside the cavity, read the machining setting data of the tool machining the corner inside the cavity, and input the machining setting data into the cutting mechanics simulation model to obtain the theoretical cutting torque; In S1, an internal corner refers to the area within the cavity contour of a metal part where adjacent machining contour segments change direction and form an angle within the cavity. For example, when milling along the cavity contour using an end mill, as the tool moves from a straight contour segment to a contour segment in another direction, the contact range between the tool and the metal part changes, thus forming the corresponding internal corner machining area.
[0021] Actual cutting torque refers to the torque corresponding to the cutting load exerted by the tool on the spindle rotation direction during actual cutting. In one embodiment, the actual cutting torque can be directly obtained by a torque sensor installed on the machine tool spindle; alternatively, the operating current of the spindle drive motor can be collected, and the spindle current can be converted into actual cutting torque according to a pre-established correspondence between spindle current and output torque.
[0022] Vibration data is used to characterize the mechanical vibration state of the cutting tool during the cutting process. It can be collected by vibration acceleration sensors installed on the spindle, tool holder, or near the machining area.
[0023] It should be noted that a correspondence can be established between the acquisition time of vibration data and the tool path position in the CNC system, so that the vibration data and the actual cutting torque can be aligned with the tool machining path, thereby providing a data basis for the subsequent matching analysis between torque deviation characteristics and vibration spectrum characteristics.
[0024] The machining settings include tool diameter, tool path, feed rate, metal part material parameters, and internal corner geometry parameters.
[0025] In one embodiment, machining setting data is input into a cutting mechanics simulation model to obtain theoretical cutting torque. Specifically, this involves: obtaining multiple path positions of the tool when it passes through the corner inside the cavity based on the tool machining path; calculating the contact wrap angle between the tool and the metal part at each path position based on the tool diameter and the geometric parameters of the corner inside the cavity; calculating the effective cutting action corresponding to each path position based on the contact wrap angle and feed rate; obtaining the corresponding cutting force coefficient based on the material parameters of the metal part; inputting the effective cutting action and cutting force coefficient corresponding to each path position into the cutting mechanics simulation model to obtain the tangential cutting force corresponding to each path position; and calculating the theoretical cutting torque corresponding to each path position based on the tangential cutting force and tool diameter.
[0026] Here, path position refers to the discrete position point selected by the tool during its movement along the tool machining path to characterize the tool's spatial position. The tool machining path corresponding to the corner within the cavity can be discretized according to the interpolation cycle of the CNC system, a preset path interval, or a preset sampling distance to obtain multiple path positions arranged sequentially according to the machining direction.
[0027] One way to calculate the contact wrap angle is to determine the overlap range between the tool circumference and the area to be machined based on the tool center coordinates, tool radius, and the profile of the cavity to be machined at each path position, and to determine the tool's entry boundary and exit boundary. The central angle formed by the entry and exit boundaries relative to the tool center is then used as the contact wrap angle for the corresponding path position.
[0028] The effective cutting action is used to characterize the degree of cutting action formed by the tool contact range and feed state at the current path position; the effective cutting ratio of the tool is determined according to the contact wrap angle, and the corresponding effective cutting action is calculated according to the effective cutting ratio and feed rate.
[0029] The cutting force coefficient is used to characterize the difference in cutting force response of different metal materials under the same cutting action conditions. It can be used to conduct calibrated cutting experiments on different metal materials in advance, record the cutting force under known cutting parameters, and determine the corresponding cutting force coefficient according to the relationship between the cutting force and the corresponding cutting action quantity, forming a correspondence table between material parameters and cutting force coefficient.
[0030] The cutting mechanics simulation model is a pre-trained cutting force prediction model used to establish a nonlinear mapping relationship between the effective cutting action, cutting force coefficient, and tangential cutting force. The model employs a multi-layer neural network, including an input layer, hidden layers, and an output layer. The input layer receives the effective cutting action and cutting force coefficient, the hidden layer performs feature mapping on the input parameters, and the output layer outputs the corresponding tangential cutting force.
[0031] In the training phase of the cutting mechanics simulation model, calibration cutting experiments are conducted under different contact wrap angles, feed rates, and metal material conditions. The actual cutting force generated during the cutting process is collected using a cutting force measuring device. Based on the tool's rotational position and cutting direction at the corresponding sampling time, the actual cutting force is decomposed into the tangential direction of the tool's circumference to obtain the corresponding actual tangential cutting force. The effective cutting action is calculated based on the contact wrap angle and feed rate corresponding to each calibration cutting experiment, and the cutting force coefficient is determined based on the metal material parameters. The effective cutting action and cutting force coefficient are used as training inputs, and the corresponding actual tangential cutting force is used as training labels to train a multi-layer neural network, obtaining the cutting mechanics simulation model. Therefore, the trained cutting mechanics simulation model can output the tangential cutting force corresponding to the corresponding cutting state based on the input effective cutting action and cutting force coefficient.
[0032] Tangential cutting force refers to the resultant force of the cutting forces generated by each cutting edge at the current path position, acting along the tangential direction of the tool circumference, which generates a resistance torque on the rotation of the tool.
[0033] It should be noted that since the lever arm of the tangential cutting force on the tool's rotation axis can be characterized by the tool radius, the theoretical cutting torque at the corresponding path position can be calculated based on the tangential cutting force and the tool diameter. Specifically: ,in, This represents the theoretical cutting torque corresponding to the i-th path position. This represents the tangential cutting force corresponding to the i-th path position. Indicates the diameter of the cutting tool.
[0034] Theoretical cutting torque is used to characterize the cutting load theoretically generated under the current toolpath, feed rate, cavity corner geometry, and material conditions, and serves as a benchmark for subsequent actual cutting torque comparison.
[0035] It should be noted that this application calculates the contact wrap angle at each path location and determines the corresponding theoretical cutting torque by combining the feed rate and material properties. This ensures that the theoretical cutting torque reflects the normal load changes at the corners within the cavity caused by variations in the tool contact range. Therefore, subsequent position-by-position comparisons between the actual cutting torque and the theoretical cutting torque can reduce the likelihood of normal load increases caused by cavity geometric changes being identified as abnormal loads, and provide a benchmark for industrial control systems to distinguish between contact wrap angle changes, chip retention, and abnormal cutting edges.
[0036] S2 compares the theoretical cutting torque with the actual cutting torque position by position to obtain the torque deviation curve, and performs feature analysis on the torque deviation curve to obtain the degree of torque deviation and torque attenuation characteristics.
[0037] In S2, the torque deviation curve is used to characterize the deviation of the actual cutting torque from the theoretical cutting torque during the tool's movement along the tool machining path. Specifically, the actual cutting torque corresponding to each path position is compared with the theoretical cutting torque, the corresponding torque deviation value is calculated, and the torque deviation values are arranged in the order of the tool machining path positions to obtain the torque deviation curve.
[0038] Torque deviation refers to the overall deviation of the actual cutting torque from the theoretical cutting torque within the torque deviation range, and is used to characterize the degree of additional cutting load within the corresponding path range.
[0039] Torque decay characteristics refer to the changes in torque deviation as the machining path decreases after the tool passes the peak position of torque deviation. This characteristic is used to characterize the reduction of additional cutting load in subsequent machining processes.
[0040] It should be noted that when the tool passes through a corner within the cavity, the contact wrap angle changes with the path position, so even if the tool and chip removal are normal, the actual cutting torque may increase significantly. If the cutting state is judged solely based on whether the actual cutting torque exceeds a fixed threshold, it is easy to identify the normal load increase caused by changes in cavity geometry as an abnormal load. Therefore, this embodiment compares the actual cutting torque with the theoretical cutting torque obtained in S1 according to the path position, using the theoretical cutting torque as the reference load under the current geometric conditions and machining parameters. The deviation from the reference load is separated from the actual cutting load, thereby obtaining a torque deviation curve that reflects changes in additional cutting load.
[0041] In one implementation, feature analysis is performed on the torque deviation curve to obtain the degree of torque deviation and torque attenuation characteristics, specifically: The torque deviation values in the torque deviation curve are read sequentially according to the path position of the tool machining path, and the path interval where the torque deviation value continuously exceeds the preset deviation threshold is determined as the torque deviation interval. Obtain the torque deviation value corresponding to each path position within the torque deviation interval, and calculate the degree of torque deviation based on each torque deviation value and the path length of the torque deviation interval; The path position with the largest torque deviation value within the torque deviation range is determined as the deviation peak position. Along the tool machining direction, starting from the deviation peak position, the path range where the torque deviation value shows a decreasing trend is determined as the torque decay range. The attenuation rate is calculated based on the decrease in torque deviation within the torque attenuation range and the path length of the torque attenuation range, and the attenuation rate is used as a characteristic of torque attenuation.
[0042] Among them, the torque deviation value refers to the difference between the actual cutting torque and the theoretical cutting torque at the same path position.
[0043] A preset deviation threshold is used to distinguish between torque deviation fluctuations under normal cutting conditions and torque deviations exceeding the normal fluctuation range. In one embodiment, historical data on machining corners within the cavity when the tool is in a normal state and there are no chips is collected in advance. The torque deviation value between the actual cutting torque and the theoretical cutting torque is calculated according to each path position, and the average value and standard deviation of the torque deviation value are statistically analyzed. The preset deviation threshold is determined based on the average value and standard deviation.
[0044] In one implementation, the path positions within the torque deviation range are set at the same preset path intervals. Therefore, the average value of the torque deviation corresponding to each path position can be used as the degree of torque deviation, i.e.: ,in, Indicates the degree of torque deviation. This represents the torque deviation value corresponding to the i-th path position within the torque deviation interval. This indicates the number of times the torque deviates from the path position within the range.
[0045] Therefore, the degree of torque deviation comprehensively reflects the overall degree of the actual cutting torque being higher than the theoretical cutting torque within the torque deviation range, which can reduce the impact of instantaneous torque fluctuations at a single path position on subsequent judgment results.
[0046] A decreasing trend in torque deviation means that after the tool passes the deviation peak position, the torque deviation gradually decreases as the tool moves along the path. Specifically, the torque deviation values corresponding to consecutive path positions are read sequentially after the deviation peak position, and the change in torque deviation between adjacent path positions is calculated. When the change in torque deviation is less than or equal to a preset tolerance, and the overall torque deviation values corresponding to multiple consecutive path positions decrease, it is determined that the torque deviation value within the corresponding path range shows a decreasing trend. The preset tolerance allows for small local fluctuations in the torque deviation value during the overall decrease, and can be pre-calibrated based on the torque deviation change of adjacent path positions in the historical torque deviation curve under normal cutting conditions.
[0047] One way to calculate the decay rate is: Where R represents the decay rate, This indicates the torque deviation value corresponding to the peak deviation position. This indicates the torque deviation value corresponding to the end of the torque decay range. This indicates the path length of the torque decay range. A larger decay rate means that after the tool passes the peak deviation position, the deviation of the actual cutting torque from the theoretical cutting torque decreases more rapidly within a shorter path; a smaller decay rate means that the deviation continues for a longer distance along the subsequent machining path.
[0048] It should be noted that during the machining of corners within the cavity, different load causes can all lead to an increase in the actual cutting torque. However, the magnitude of the torque deviation and the process of change after the peak position differ, making it difficult to fully reflect this difference using a single torque deviation value. Therefore, this application reflects the overall deviation of the actual cutting torque from the theoretical cutting torque within the torque deviation range through the degree of torque deviation, and reflects the decrease of the torque deviation along the machining path after the deviation peak through torque decay characteristics. This allows for the description of the actual cutting load from both the overall degree of deviation and the subsequent change process, providing a basis for further differentiation of different load causes using vibration acceleration spectrum analysis.
[0049] S3: Perform a Fast Fourier Transform on the vibration data to obtain the vibration characteristic spectrum, and perform matching analysis on the torque deviation and torque attenuation characteristics based on the vibration characteristic spectrum to determine the cause of the load. See also Figure 2 The details are as follows: In S3, the vibration characteristic spectrum refers to the set of spectra obtained by dividing the vibration data within the torque deviation range according to the tool machining path and performing frequency domain analysis on the vibration data within different path ranges. It is used to characterize the main frequency components and corresponding vibration amplitudes of the vibration signal at different path positions.
[0050] Load causes refer to the main reasons that cause the actual cutting torque to deviate from the theoretical cutting torque, including at least changes in the contact wrap angle, chip retention, and abnormal cutting edge. Changes in the contact wrap angle refer to the change in the contact range between the tool and the metal part as the tool passes through a corner within the cavity, causing a corresponding change in the cutting load. Chip retention refers to the additional load caused by the failure of processed chips to be discharged in time, resulting in collisions or further cutting between the tool and the metal part. Abnormal cutting edge refers to the abnormal cutting load and vibration generated during continuous cutting due to wear, chipping, or localized damage to the tool's cutting edge.
[0051] In one implementation, a fast Fourier transform is performed on the vibration data to obtain the vibration characteristic spectrum. Specifically, the vibration data is arranged according to the path position of the tool machining path to obtain a vibration change curve; according to the start and end path positions of the torque deviation interval, a target vibration curve segment corresponding to the torque deviation interval is extracted from the vibration change curve; the target vibration curve segment is divided into multiple sequentially arranged vibration sub-curve segments according to a preset path length, and a fast Fourier transform is performed on each vibration sub-curve segment to obtain the vibration spectrum curve corresponding to each vibration sub-curve segment; a correspondence is established between each vibration spectrum curve and the path position of the corresponding vibration sub-curve segment, and the vibration spectrum curves are arranged according to the position order of the tool machining path to obtain the vibration characteristic spectrum.
[0052] The vibration variation curve refers to the data sequence formed by arranging the collected vibration data sequentially according to the corresponding toolpath positions. The target vibration curve segment refers to the local vibration data segment in the vibration variation curve corresponding to the torque deviation interval. The vibration sub-curve segment refers to the vibration data segment in the target vibration curve segment corresponding to a preset path length.
[0053] The preset path length is used to determine the machining path range covered by each vibration sub-curve segment. It can be determined based on the vibration data sampling frequency, feed rate, and the frequency resolution required for frequency analysis. When the vibration sampling frequency is... An oscillator curve segment contains When there are 100 sampling points, the frequency resolution corresponding to the Fast Fourier Transform is: Therefore, the required frequency resolution can be determined based on the frequency components to be identified. Determine the number of sampling points, and then calculate the corresponding path length based on the feed rate. This ensures that each vibratory sub-curve segment has sufficient vibration sampling data for frequency identification, while also preserving the vibration variations between different path positions.
[0054] It should be noted that directly performing a Fast Fourier Transform on the vibration data within the entire torque deviation range only yields the overall frequency composition of the entire range, making it difficult to determine which specific segment of the machining path a change in frequency or amplitude occurs in. Therefore, dividing the target vibration curve segment into multiple sub-curve segments and performing frequency domain analysis on each allows for the establishment of a correspondence between spectral characteristics and tool path positions while preserving frequency information. This provides a foundation for subsequent determinations of whether vibration characteristics change synchronously with torque deviation and whether a particular frequency persists.
[0055] A vibration spectrum curve refers to the curve formed by the frequency components and their corresponding vibration amplitudes within a segment of an oscillator, including multiple frequency positions and the vibration amplitudes corresponding to each frequency position. Specifically, for a segment including... For each sampling point of the vibration sub-curve segment, a Fast Fourier Transform is performed on the vibration data to obtain the spectral values corresponding to multiple discrete frequency positions. The corresponding vibration amplitude is then determined based on the amplitude of the spectral values. Each frequency position can be represented as: By arranging multiple frequency positions and their corresponding vibration amplitudes in frequency order, the vibration spectrum curve can be obtained.
[0056] It should be noted that the vibration signal during tool cutting includes multiple vibration components such as spindle rotation, periodic cutting of the cutting teeth, and abnormal contact. It is difficult to distinguish the different sources based on the original vibration amplitude alone. By using Fast Fourier Transform, vibration components with different periodic characteristics can be distributed to different frequency positions, so that the main vibration components can be represented by spectral peak frequency and spectral peak amplitude, which facilitates the analysis of vibration changes corresponding to different load conditions.
[0057] It should be noted that by performing fast Fourier transform on vibration data of different path ranges within the torque deviation range, the different vibration components originally superimposed on the vibration change curve can be converted into frequency and amplitude features, and the path position corresponding to each spectral feature can be preserved, thus providing a data basis for comparing vibration changes with torque changes and identifying the persistence of vibration features.
[0058] In one implementation, a matching analysis is performed on the torque deviation and torque attenuation characteristics based on the vibration characteristic spectrum. Specifically, this involves: extracting spectral peaks with vibration amplitudes greater than a preset amplitude threshold from the vibration spectrum curves corresponding to each vibrating sub-curve segment, and obtaining the peak frequency and amplitude corresponding to each peak; when the torque deviation is greater than a preset deviation threshold and the torque attenuation characteristics are greater than a preset attenuation threshold, calculating the change in torque deviation and peak amplitude corresponding to adjacent vibrating sub-curve segments according to the path position order of each vibrating sub-curve segment; using the number of vibrating sub-curve segment groups with the same sign for the change in torque deviation and the change in peak amplitude as the amplitude matching quantity; if the amplitude matching quantity is greater than or equal to a preset matching quantity threshold, determining that the load is caused by a change in the contact wrap angle; if the amplitude matching quantity is less than the preset matching quantity threshold, or the torque attenuation characteristics are less than or equal to a preset attenuation threshold, determining the frequency duration based on the peak frequency corresponding to each peak, and determining that the load is caused by chip retention or abnormal cutting edge based on the frequency duration.
[0059] The preset amplitude threshold is used to exclude frequency components corresponding to background vibration and low amplitude noise. Historical vibration data under normal cutting conditions can be collected in advance and calibrated according to the vibration amplitude fluctuation range at each frequency position in the normal vibration spectrum.
[0060] A spectral peak refers to a local maximum point in the vibration spectrum curve where the vibration amplitude is greater than a preset amplitude threshold and higher than the vibration amplitude at adjacent frequency positions.
[0061] The preset deviation threshold is a judgment threshold set for the degree of torque deviation. It is used to determine whether the overall deviation of the actual cutting load has reached the preset level. It can be pre-calibrated based on the historical torque deviation under normal cutting conditions and different load conditions.
[0062] The preset attenuation threshold is a judgment threshold set for torque attenuation characteristics. It is used to distinguish different attenuation states after the torque deviation passes the peak position. It can be calibrated based on the attenuation rate corresponding to different load states in historical processing data.
[0063] The change in torque deviation refers to the difference between the torque deviation value corresponding to the subsequent oscillator curve segment and the torque deviation value corresponding to the previous oscillator curve segment; the change in spectral peak amplitude refers to the difference between the maximum spectral peak amplitude in the subsequent oscillator curve segment and the maximum spectral peak amplitude in the previous oscillator curve segment.
[0064] A preset matching quantity threshold is used to determine whether the torque deviation change and the main vibration amplitude change remain in the same direction within a sufficient number of paths. It can be calibrated based on the number of times the two changes remain in the same direction under historical contact wrap angle change conditions.
[0065] Frequency duration refers to the number of vibratory sub-curve segments with approximately the same spectral peak frequency that appear consecutively along the tool machining path. It is used to characterize the path range in which the corresponding abnormal vibration component continues to act.
[0066] It should be noted that when the load is mainly caused by changes in the contact wrap angle, the cutting load and corresponding vibration response usually change along with the path position as the contact range between the tool and the metal part changes. Therefore, by comparing the signs of the changes in torque deviation and the changes in spectral peak amplitude, it can be determined whether they have the same increasing or decreasing trend. Further statistical analysis of the number of amplitude matches can avoid making judgments directly based on occasional vibration fluctuations at a single path position, improving the reliability of identifying changes in contact wrap angle. When the number of amplitude matches does not reach the preset matching threshold, or when the torque deviation decays slowly after the peak, it indicates that the synergistic characteristics of the current load change and the contact wrap angle change are not obvious. Therefore, further differentiation between chip retention and cutting edge abnormalities should be made based on the persistence of the spectral peak frequency.
[0067] Furthermore, the frequency duration is determined based on the frequency of each spectral peak, and the load is determined to be caused by chip retention or cutting edge abnormality based on the frequency duration. Specifically: according to the path position sequence of each vibrating sub-curve segment, the spectral peak frequencies in adjacent vibrating sub-curve segments are compared to determine the matching spectral peaks whose frequency difference is less than or equal to the preset frequency tolerance; the matching spectral peaks are continuously analyzed along the tool machining path; when the frequency difference between the matching spectral peak in the subsequent vibrating sub-curve segment and the matching spectral peak in the previous vibrating sub-curve segment is less than or equal to the preset frequency tolerance, the corresponding spectral peak frequency is determined to be continuously present, and the number of vibrating sub-curve segments with continuously present spectral peak frequencies is counted to obtain the frequency duration; when the frequency duration is less than the preset duration threshold, the load is determined to be caused by chip retention; when the frequency duration is greater than or equal to the preset duration threshold, the load is determined to be caused by cutting edge abnormality.
[0068] Among them, the matching spectral peak refers to the spectral peak located in the adjacent vibrating sub-curve segments and whose frequency difference is within the preset frequency tolerance. It is used to characterize that there are approximately the same main vibration frequencies in the adjacent path range.
[0069] The preset frequency tolerance is used to allow for small frequency drifts of the same vibration component between adjacent vibratory sub-curve segments. The preset frequency tolerance can be determined based on the frequency resolution of the Fast Fourier Transform (FFT) and the range of frequency changes caused by spindle speed fluctuations during normal machining. For example, using the FFT frequency resolution... Based on this, and combined with the frequency offset corresponding to normal spindle speed fluctuations, a preset frequency tolerance is determined. This avoids the same physical vibration component being incorrectly identified as different frequencies due to falling into adjacent frequency points or due to small changes in spindle speed.
[0070] The preset duration threshold is used to distinguish between abnormal frequency components that occur briefly and those that occur continuously. It can be predetermined based on the distribution of the duration of frequency under the calibration conditions of chip retention and abnormal cutting edge.
[0071] It should be noted that the additional contact caused by chip retention is affected by the chip location and the chip removal process, usually manifesting as localized, intermittent re-cutting or collision, and the corresponding abnormal vibration components are more likely to appear within a limited path range. In contrast, the abnormal cutting edge is fixed on the cutting edge of the tool, and as the tool rotates continuously, the abnormal cutting edge can repeatedly participate in cutting; therefore, the corresponding vibration components are more likely to persist across multiple consecutive paths. Based on calibration data under current machine tool, cutting tool, material, and machining parameter conditions, a correspondence between the frequency duration and the two aforementioned load causes can be established.
[0072] It should be noted that this application utilizes the degree of torque deviation and torque attenuation characteristics to determine the magnitude and path-dependent variation of the additional cutting load. Furthermore, it uses the amplitude variation and frequency continuity in the vibration spectrum to further differentiate the load sources, thus avoiding judgment based solely on a single load signal when different causes all manifest as torque increases. Therefore, in machining areas with complex load variations, such as corners within the cavity, it is possible to further distinguish between loads caused by geometric contact changes and loads caused by chip retention or abnormal cutting edges. This provides a basis for subsequent feed adjustments, tool retraction and chip removal, or machining pauses based on different load causes.
[0073] S4, when the load is caused by changes in the contact wrap angle, calculate the target feed speed based on the degree of torque deviation and generate feed control commands.
[0074] In S4, the feed control command refers to the control information used to adjust the speed of the tool along the current tool path. It includes at least the path position and the corresponding target feed rate, so as to control the tool to continue to execute the current tool path according to the target feed rate.
[0075] The target feed rate includes a first target feed rate and a second target feed rate. The first target feed rate is the initial deceleration value determined based on the current torque deviation, used to reduce the tool feed rate at the beginning of the torque decay range. The second target feed rate is the feed rate further adjusted based on the remaining load state during the torque decay process, used to gradually restore the feed rate as the additional cutting load decreases within the torque decay range.
[0076] In one implementation, generating a feed control command specifically involves: acquiring the current feed rate; calculating the feed rate reduction based on the difference between the torque deviation and a preset deviation threshold; calculating a first target feed rate corresponding to the starting position of the torque decay interval based on the current feed rate and the feed rate reduction; calculating the predicted residual deviation between the actual cutting torque and the theoretical cutting torque based on the torque decay characteristics and the path length between each path position and the ending path position within the torque decay interval; correcting the feed rate reduction based on the ratio of the predicted residual deviation to the torque deviation to obtain the feed rate correction amount corresponding to each path position; calculating a second target feed rate corresponding to each path position based on the first target feed rate and the feed rate correction amount; and writing the first target feed rate and the second target feed rate into the current tool path according to their corresponding path positions to generate a feed control command.
[0077] The feed rate reduction refers to the speed reduction required from the current feed rate. A pre-established relationship between the degree to which torque deviation exceeds a preset threshold and the feed rate reduction can be established based on historical machining data or calibration experiments. The greater the torque deviation exceeds the preset threshold, the greater the current additional cutting load, and the larger the corresponding feed rate reduction.
[0078] The first target feed rate is obtained by subtracting the feed rate reduction from the current feed rate. It is used to reduce the tool feed rate at the beginning of the torque decay range to suppress the current high cutting load.
[0079] Predicted residual deviation refers to the magnitude of the deviation between the actual cutting torque and the theoretical cutting torque at each path position, based on the decreasing state of the torque deviation within the torque decay interval. In one embodiment, the degree of torque deviation decay along the machining path is obtained based on the torque decay characteristics, and the corresponding predicted residual deviation is determined by combining this with the path length from each path position to the end position of the torque decay interval. The farther away from the end position of the torque decay interval, the larger the corresponding predicted residual deviation; as the tool gradually approaches the end position, the predicted residual deviation gradually decreases. The correspondence can be pre-calibrated based on historical torque decay data.
[0080] It should be noted that when the cause of the load has been determined to be the change in the contact wrap angle, the decrease in torque deviation reflects that the additional cutting load caused by the change in the tool contact range is gradually subsiding. Therefore, the torque attenuation characteristic reflects the degree of attenuation of the additional load, while the path length from the current position to the end of the attenuation interval reflects the path range that the attenuation process has not yet been completed. The combination of the two can characterize the additional cutting load that has not yet subsided at each path position.
[0081] The feed rate correction refers to the amount of feed rate that can be recovered from the first target feed rate based on the predicted residual deviation. Specifically, the ratio of the predicted residual deviation to the torque deviation represents the degree of residual load at the current path position, and the feed rate reduction is corrected accordingly. A larger ratio indicates more additional cutting load that has not yet subsided, corresponding to a smaller feed rate correction. A second target feed rate is obtained based on the first target feed rate and the feed rate correction, and this second target feed rate gradually increases as the predicted residual deviation decreases.
[0082] It should be noted that the first target feed rate corresponds to the starting position of the torque decay interval and is used to perform initial speed reduction after identifying a high additional cutting load; the second target feed rate corresponds to multiple path positions after the starting position within the torque decay interval and is used to gradually restore the feed rate as the predicted residual deviation decreases. Compared to continuously using a fixed low feed rate in the corner area throughout the cavity, the above method can maintain a larger speed reduction range at high load positions and promptly increase the feed rate after the load gradually subsides.
[0083] It should be noted that when the load is caused by a change in the contact wrap angle, it indicates that the current load increase is mainly caused by changes in the tool contact range at the corners within the cavity, and there is no need to directly retract the tool or pause the machining process. This application determines the initial speed reduction based on the degree of torque deviation and adjusts the feed rate at each path position according to the subsequent load decay state, thereby maintaining continuous execution of the current toolpath while taking into account cutting stability and machining efficiency.
[0084] S5, when the load is caused by chip retention, determines the retraction distance based on the torque deviation curve and generates a retraction chip removal command.
[0085] In S5, the retraction and chip removal command includes at least the retraction start position, retraction direction, and retraction distance. It is used to control the tool to move from the current toolpath position in the opposite direction of the current toolpath, so that the tool exits the chip retention area and forms a chip removal space.
[0086] In one embodiment, the torque deviation value in the torque deviation curve is read along the tool machining direction, and the path position where the torque deviation value first exceeds a preset deviation threshold is taken as the deviation start position; the current tool path position is obtained, and the path distance along the current tool path between the current tool path position and the deviation start position is calculated; the torque deviation value corresponding to each path position between the deviation start position and the current tool path position is obtained, and the tool retraction correction distance is calculated based on the difference between each torque deviation value and the preset deviation threshold and the path length between the corresponding path positions; the sum of the path distance and the tool retraction correction distance is taken as the tool retraction distance, and a tool retraction path is generated in the opposite direction of the current tool path starting from the current tool path position; a tool retraction and chip removal command is generated based on the tool retraction path.
[0087] The deviation from the starting position is used to characterize the location where the actual cutting torque begins to continuously exceed the normal torque deviation range. When the load cause has been determined to be chip retention, this position can be used as a reference point where the chip begins to exert a significant additional load on the tool. The path distance between the current toolpath position and the deviation from the starting position is used as the basic tool retraction distance to ensure that the tool at least exits the identified abnormal load path range.
[0088] The tool retraction correction distance refers to the additional retraction length beyond the basic retraction distance. It is used to determine the extra retraction range after the tool deviates from its starting position based on the degree of additional load caused by chip retention.
[0089] In one implementation, the excess amount of each torque deviation value relative to a preset deviation threshold between the starting position and the current toolpath position is calculated, and accumulated in conjunction with the path length between adjacent path positions to obtain the degree of load excess within that path range. A tool retraction correction distance is determined based on the degree of load excess from a pre-established correspondence between tool retraction distances. Specifically, the greater the magnitude of the torque deviation exceeding the preset deviation threshold and the longer the continuous path range, the greater the corresponding degree of load excess, and a larger tool retraction correction distance is determined. The correspondence can be pre-established based on historical machining data or chip removal calibration experiments under different chip retention levels.
[0090] It should be noted that the additional cutting load caused by chip retention is related not only to the instantaneous torque deviation at a certain path position, but also to the path range through which the chip continuously affects the tool within the machining area. Therefore, by simultaneously utilizing the amount by which the torque deviation exceeds a preset deviation threshold and the corresponding path length, the magnitude and duration of the additional load caused by chip retention can be comprehensively reflected, avoiding the need to determine the tool retraction distance solely based on a single torque deviation value.
[0091] It should be noted that the advantage of using the sum of the path distance and the retraction correction distance as the retraction distance is that the position where the torque deviation first exceeds the preset deviation threshold reflects the location where chip retention has already significantly affected the cutting load. However, there may be a certain difference between the actual chip accumulation range and the position where the sensor detects the load change. If the tool is only retracted to a position deviating from the starting position, the tool may still be near the chip accumulation area, resulting in insufficient chip removal space. Therefore, increasing the retraction correction distance according to the degree of abnormal load, so that the tool retracts further when the chip retention is higher, can increase the space between the tool and the original cutting area, providing conditions for the retained chips to detach or be discharged.
[0092] It should be noted that when the cause of the load is determined to be chip retention, simply reducing the feed rate cannot directly eliminate the chips already retained in the machining area. This application determines the basic retraction range based on the position where the torque deviation first exceeds the normal range, and further adjusts the retraction distance according to the magnitude of the overload and the range of the continuous path, so that the retraction distance can be adjusted according to the degree of chip retention, thereby avoiding unnecessary large-scale retraction while providing sufficient clearance for the removal of retained chips.
[0093] S6 generates a machining pause command when the load is caused by an abnormal cutting edge.
[0094] In S6, the machining pause command is used to stop the feed axis movement and pause the execution of the current toolpath.
[0095] In one implementation, when the load is determined to be caused by an abnormal cutting edge, the current toolpath position is obtained, and a machining pause command is generated and sent to the CNC system to stop the feed axis and pause the execution of the toolpath after the current position.
[0096] It should be noted that abnormal cutting edge is an abnormality in the tool itself, and simply reducing the feed rate or retracting the tool to remove chips will not eliminate the abnormality. If machining continues, the abnormal cutting edge will still participate in cutting, which may further aggravate tool damage or affect machining quality.
[0097] The embodiments of the present invention have been described in detail above, but the content is only a preferred embodiment of the present invention and should not be considered as limiting the scope of the present invention. All equivalent changes and improvements made in accordance with the claims of the present invention should still fall within the scope of the claims of the present invention.
Claims
1. A multi-parameter collaborative monitoring and control method for metal parts cutting conditions, characterized in that, The method includes: The actual cutting torque and vibration data of the tool when passing through the corner inside the cavity are obtained, and the machining setting data of the tool machining the corner inside the cavity is read. The machining setting data is then input into the cutting mechanics simulation model to obtain the theoretical cutting torque. The theoretical cutting torque and the actual cutting torque are compared position by position to obtain the torque deviation curve. The torque deviation curve is then analyzed to obtain the degree of torque deviation and torque attenuation characteristics. The vibration data is subjected to a fast Fourier transform to obtain the vibration characteristic spectrum. Based on the vibration characteristic spectrum, a matching analysis is performed on the torque deviation and torque attenuation characteristics to determine the load cause, which includes changes in contact wrap angle, chip retention, or abnormal cutting edge. When the load is caused by a change in the contact wrap angle, the target feed rate is calculated based on the degree of torque deviation and a feed control command is generated. The feed control command is used to control the tool to continue executing the current tool path according to the target feed rate. When the load is caused by chip retention, the tool retraction distance is determined based on the torque deviation curve and a tool retraction and chip removal command is generated. The tool retraction and chip removal command is used to control the tool to exit the chip retention area. When the load is caused by an abnormal cutting edge, a machining pause command is generated. The machining pause command is used to stop the feed axis movement and pause the execution of the current toolpath.
2. The multi-parameter collaborative monitoring and control method for metal parts cutting conditions according to claim 1, characterized in that, The machining setting data includes tool diameter, tool path, feed rate, metal part material parameters, and cavity corner geometry parameters; the process of reading the machining setting data for machining the cavity corners and inputting the machining setting data into the cutting mechanics simulation model to obtain the theoretical cutting torque includes: The tool machining path is used to obtain multiple path positions when the tool passes through the corner inside the cavity, and the contact wrap angle between the tool and the metal part at each path position is calculated based on the tool diameter and the geometric parameters of the corner inside the cavity. Calculate the effective cutting action at each path position based on the contact wrap angle and feed rate at each path position; Based on the material parameters of the metal parts, the corresponding cutting force coefficient is obtained. The effective cutting action and cutting force coefficient corresponding to each path position are input into the cutting mechanics simulation model to obtain the tangential cutting force corresponding to each path position. Calculate the theoretical cutting torque corresponding to each path position based on the tangential cutting force and tool diameter at each path position.
3. The multi-parameter collaborative monitoring and control method for metal parts cutting conditions according to claim 1, characterized in that, The characteristic analysis of the torque deviation curve to obtain the degree of torque deviation and torque attenuation characteristics includes: The torque deviation values in the torque deviation curve are read sequentially according to the path position of the tool machining path, and the path interval where the torque deviation value continuously exceeds the preset deviation threshold is determined as the torque deviation interval. Obtain the torque deviation value corresponding to each path position within the torque deviation interval, and calculate the degree of torque deviation based on each torque deviation value and the path length of the torque deviation interval; the degree of torque deviation is used to characterize the overall deviation of the actual cutting torque from the theoretical cutting torque; The path position with the largest torque deviation value within the torque deviation range is determined as the deviation peak position. Along the tool machining direction, starting from the deviation peak position, the path range where the torque deviation value shows a decreasing trend is determined as the torque decay range. The attenuation rate is calculated based on the decrease in torque deviation within the torque attenuation range and the path length of the torque attenuation range, and the attenuation rate is used as the torque attenuation characteristic. The torque attenuation characteristic is used to characterize the degree of attenuation of the deviation between the actual cutting torque and the theoretical cutting torque along the machining path after the tool passes the deviation peak position.
4. The multi-parameter collaborative monitoring and control method for metal parts cutting conditions according to claim 3, characterized in that, The step of performing a fast Fourier transform on the vibration data to obtain the vibration characteristic spectrum includes: The vibration data are arranged according to the path position of the tool machining path to obtain the vibration variation curve; Based on the starting and ending path positions of the torque deviation interval, the target vibration curve segment corresponding to the torque deviation interval is extracted from the vibration variation curve. The target vibration curve segment is divided into multiple sequentially arranged vibration sub-curve segments according to the preset path length, and a fast Fourier transform is performed on each vibration sub-curve segment to obtain the vibration spectrum curve corresponding to each vibration sub-curve segment. The vibration spectrum curve includes multiple frequency positions and the vibration amplitude corresponding to each frequency position. Establish the correspondence between each vibration spectrum curve and the path position of the corresponding vibration sub-curve segment, and arrange each vibration spectrum curve according to the position order of the tool machining path to obtain the vibration characteristic spectrum.
5. The multi-parameter collaborative monitoring and control method for metal parts cutting conditions according to claim 4, characterized in that, The process of performing matching analysis on the torque deviation and torque attenuation characteristics based on the vibration characteristic spectrum to determine the load cause includes: Extract the spectral peaks with vibration amplitudes greater than a preset amplitude threshold from the vibration spectrum curves corresponding to each vibration sub-curve segment, and obtain the spectral peak frequency and spectral peak amplitude corresponding to each spectral peak. When the torque deviation is greater than a preset deviation threshold and the torque attenuation characteristic is greater than a preset attenuation threshold, the change in torque deviation and the change in spectral peak amplitude of adjacent vibrating sub-curve segments are calculated according to the path position order of each vibrating sub-curve segment; the change in spectral peak amplitude refers to the difference in the largest spectral peak amplitude among adjacent vibrating sub-curve segments. The number of vibrating sub-curve segments whose changes in torque deviation and spectral peak amplitude have the same sign is used as the amplitude matching quantity. If the number of amplitude matching is greater than or equal to the preset matching number threshold, the load is determined to be caused by a change in the contact wrap angle; If the amplitude matching quantity is less than the preset matching quantity threshold, or the torque attenuation characteristic is less than or equal to the preset attenuation threshold, the frequency duration is determined based on the frequency of each spectral peak, and the load cause is determined based on the frequency duration to be chip retention or abnormal cutting edge.
6. The multi-parameter collaborative monitoring and control method for metal parts cutting conditions according to claim 5, characterized in that, The process of determining the frequency duration based on the spectral peak frequency and determining whether the load is caused by chip retention or abnormal cutting edge based on the frequency duration includes: According to the path position order of each sub-curve segment, the spectral peak frequencies in adjacent sub-curve segments are compared to determine the matching spectral peaks whose frequency difference is less than or equal to the preset frequency tolerance. The matching spectrum peaks are continuously analyzed along the tool machining path. When the frequency difference between the matching spectrum peak in the next vibrating sub-curve segment and the matching spectrum peak in the previous vibrating sub-curve segment is less than or equal to the preset frequency tolerance, the corresponding spectrum peak frequency is determined to be continuously present. The number of vibrating sub-curve segments with continuously present spectrum peak frequencies is counted to obtain the frequency duration. When the frequency duration is less than the preset duration threshold, the load is determined to be caused by chip retention. When the frequency duration is greater than or equal to the preset duration threshold, the load is determined to be caused by an abnormal blade edge.
7. The multi-parameter collaborative monitoring and control method for metal parts cutting conditions according to claim 6, characterized in that, The target feed rate includes a first target feed rate and a second target feed rate; The step of calculating the target feed speed based on the degree of torque deviation and generating feed control commands includes: Obtain the current feed rate and calculate the feed rate reduction based on the difference between the torque deviation and the preset deviation threshold. Based on the current feed rate and the amount of feed rate reduction, calculate the first target feed rate corresponding to the starting position of the torque decay interval; Based on the torque decay characteristics and the path length from each path position to the end path position within the torque decay interval, the predicted residual deviation between the actual cutting torque and the theoretical cutting torque is calculated. Based on the ratio of the predicted residual deviation to the torque deviation, the feed rate reduction is corrected to obtain the feed rate correction amount corresponding to each path position. Calculate the second target feed rate corresponding to each path position based on the first target feed rate and the feed rate correction. Write the first target feed rate and the second target feed rate into the current toolpath according to their corresponding path positions to generate feed control commands.
8. The multi-parameter collaborative monitoring and control method for metal parts cutting conditions according to claim 1, characterized in that, The process of determining the retraction distance based on the torque deviation curve and generating the retraction and chip removal command includes: Read the torque deviation value in the torque deviation curve along the tool machining direction, and take the path position where the torque deviation value first exceeds the preset deviation threshold as the deviation start position; Obtain the current toolpath position and calculate the path distance along the current toolpath between the current toolpath position and the deviation from the starting position. Obtain the torque deviation value corresponding to each path position between the deviation from the starting position and the current toolpath position, and calculate the tool retraction correction distance based on the difference between each torque deviation value and the preset deviation threshold and the path length between the corresponding path positions; The sum of the path distance and the tool retraction correction distance is used as the tool retraction distance, and a tool retraction path is generated in the opposite direction of the current tool path, starting from the current tool path position. Generate a retraction and chip removal command based on the retraction path.
