A machining status monitoring method based on synchronous tracking of cutting teeth
By using dual-microphone positioning and synchronous tracking frequency band technology, the problems of sound source differentiation and motion law correlation in CNC milling machine machining status monitoring have been solved, achieving high-precision machining status identification and equipment adaptation.
Patent Information
- Application Number
- CN202610964982.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-06-30
- Publication Date
- 2026-11-17
AI Technical Summary
Existing CNC milling machine machining status monitoring methods struggle to accurately distinguish between tool cutting noise and environmental noise in complex machining environments, and lack correlation analysis with tool motion patterns, resulting in insufficient accuracy and stability of monitoring results.
By synchronously acquiring sound signals with dual microphones, locating the sound source and filtering the target sound signal using the time difference, calculating the tooth passage frequency by combining the spindle speed and the number of tool teeth, establishing a synchronous tracking frequency band, extracting the tooth impact pulse sequence, and calculating the cycle consistency parameter, real-time monitoring of the machining status is achieved.
It significantly improves the purity of extracting cutting sound from complex environments, accurately reflects changes in cutting state, and can identify four states: normal cutting, chatter, chipping, and air cutting. It reduces installation and maintenance costs and is suitable for retrofitting existing CNC milling machines.
Abstract
Description
Technical Field
[0001] This application relates to the field of CNC milling machine machining status monitoring technology, and in particular to a machining status monitoring method based on synchronous tracking of cutting teeth. Background Technology
[0002] CNC milling machines are essential machining equipment widely used in mechanical manufacturing, mold making, and aerospace component manufacturing. During machining, continuous cutting action occurs between the tool and the workpiece, and the machining status directly affects the workpiece's machining accuracy, surface quality, tool life, and equipment operational stability. Therefore, real-time monitoring of the machining status is crucial for improving machining quality and preventing equipment failures.
[0003] Currently, the main methods for monitoring the machining status of CNC milling machines include vibration monitoring, cutting force monitoring, current monitoring, acoustic emission monitoring, and sound monitoring. Among these, vibration monitoring typically involves collecting vibration signals by installing accelerometers for analysis, but the sensor's installation location has a significant impact on the monitoring results, and the installation and maintenance costs are high. Cutting force monitoring usually requires a dedicated force measuring device, which is expensive and not easily adopted on existing machine tools. Acoustic emission monitoring, while highly sensitive, suffers from expensive sensors and complex signal processing. Current monitoring, although simple to implement, has limited ability to identify sudden abnormal conditions such as tool chipping and chatter.
[0004] Compared to the above solutions, sound signal-based monitoring methods have advantages such as being non-contact, easy to install, and low in modification costs, and are therefore gradually gaining attention. Existing sound monitoring technologies typically use microphones to collect and process sound, and then perform spectral analysis, energy analysis, or feature extraction on the sound signals to achieve processing status identification.
[0005] However, in actual machining environments, the machining area is usually accompanied by multiple sources of interference, such as spindle fan noise, coolant spray noise, chip impact noise, and noise from surrounding equipment. Existing sound monitoring methods typically analyze the overall collected sound signal directly, making it difficult to effectively distinguish between tool cutting noise and environmental noise, thus affecting the accuracy of the monitoring results.
[0006] In addition, most existing sound monitoring methods rely on sound spectrum or energy characteristics for state judgment, lacking correlation analysis with the actual cutting motion of the tool. As a result, the monitoring results are easily affected by changes in the machining conditions under conditions such as tool wear, chatter, chipping, and empty cutting, and the stability and reliability of the identification need to be improved.
[0007] Therefore, how to accurately extract cutting sounds from complex machining environments and establish the correlation between sound signals and tool motion patterns to improve the accuracy of machining status monitoring has become a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention
[0008] The purpose of this application is to provide a machining status monitoring method based on synchronous tracking of cutting teeth, which realizes real-time monitoring and anomaly identification of machining status by correlating and analyzing the tool motion parameters with machining sound signals.
[0009] A machining status monitoring method based on synchronous tracking of cutting teeth includes the following steps: S1. Obtain the spindle speed information, tool tooth count information, and tool motion trajectory information of the CNC machine tool; S2. Processing sound signals are simultaneously acquired by a first microphone and a second microphone set at different locations in the processing area; S3. Determine the location of the sound source based on the time difference between the sound signals received by the two microphones, and filter the target sound signals from the contact area between the tool and the workpiece. S4. Calculate the tooth passing frequency based on the spindle speed information and the number of tool teeth information, establish a synchronous tracking frequency band, and extract the synchronous sound signal corresponding to the tooth passing frequency from the target sound signal; S5. Extract the blade impact pulse sequence from the synchronous sound signal; S6. Calculate the period consistency parameter based on the blade impact pulse sequence; S7. Determine the current processing status based on the cycle consistency parameter and output the monitoring results.
[0010] Furthermore, In step S3, the location of the sound source is determined based on the arrival time difference of the same sound event received by the two microphones. When the sound source is located within a preset spatial range centered on the contact point between the tool and the workpiece, the corresponding sound signal is retained as the target sound signal. When the sound source is located outside the preset spatial range, the corresponding sound signal is suppressed.
[0011] Furthermore, In step S4, the frequency Ft of the cutting teeth is calculated according to the following formula: Ft = N × Z ÷ 60; in: Ft is the passing frequency of the cutting teeth; N is the spindle speed; Z represents the number of teeth on the cutting tool; The synchronous tracking frequency band includes a fundamental frequency band centered at the blade passing frequency Ft and at least one corresponding harmonic frequency band, wherein the bandwidth of the fundamental frequency band and the harmonic frequency band is 1% to 10% of the corresponding center frequency.
[0012] Furthermore, In step S5, the synchronous sound signal is envelope demodulated and the blade impact pulse is identified by combining a peak detection algorithm to obtain the blade impact pulse sequence. The envelope demodulation process includes: performing a Hilbert transform on the synchronized audio signal to obtain an envelope curve; performing a low-pass filter on the envelope curve; and using a peak detection algorithm to identify the blade impact pulse. The cutoff frequency of the low-pass filter is 0.5 to 2 times the blade passage frequency.
[0013] Furthermore, In step S6, an actual periodic sequence is established based on the time interval between adjacent cutter impact pulses, and a periodic consistency parameter is calculated based on the deviation between the actual periodic sequence and the theoretical cutter period. The period consistency parameter is the average deviation rate of each actual pulse interval relative to the theoretical cutter tooth period. The larger the deviation between the actual pulse interval and the theoretical cutter tooth period, the larger the value of this parameter.
[0014] Furthermore, In step S7, when the period consistency parameter exceeds the first preset threshold, the current processing state is determined to be a chattering state. The first preset threshold is calibrated as follows: Under the cutting state where the tool and the workpiece are in normal contact and there is no chatter, the sound signals of multiple tooth cycles are collected and the cycle consistency parameters of each cycle are calculated. The first preset threshold is determined based on the statistical value of the calculated cycle consistency parameters.
[0015] Furthermore, In step S7, when a series of missing pulses appear in the tooth impact pulse sequence, the current machining state is determined to be a tool breakage state; when the tool breakage state is determined to be a tool breakage state, a stop command is sent to the CNC system. The term "continuous missing pulses" refers to the situation where, within multiple consecutive theoretical blade cycles, the number of valid pulses detected is less than half the number of pulses that should occur in that cycle.
[0016] Furthermore, In S7, when the sound energy in the synchronous tracking frequency band is lower than the second preset threshold, the current processing state is determined to be an empty cutting state. The second preset threshold is calibrated as follows: in the air cutting state where the tool and the workpiece are not in contact, the machining sound signal is collected and the sound energy in the synchronous tracking frequency band is extracted, and the second preset threshold is determined based on the sound energy value.
[0017] Furthermore, A two-stage denoising architecture is constructed by spatial domain filtering in S3 and time-frequency domain filtering in S4. In S3, sound source filtering based on time difference localization serves as the first stage of denoising, filtering out environmental noise outside the processing area. In S4, synchronous tracking filtering based on the tooth passing frequency serves as the second stage of denoising, filtering out interference signals unrelated to tooth cutting within the processing area. The synchronous sound signal after two stages of denoising is used for pulse extraction and state determination in S5 to S7.
[0018] The advantages of this application over the prior art include: 1. This application constructs a two-level noise reduction architecture through spatial domain filtering and time-frequency domain filtering, which eliminates environmental noise outside the processing area and further filters out interference signals unrelated to cutting within the processing area. The two form a causal progressive relationship rather than a simple superposition, which significantly improves the purity of extracting cutting sound from complex processing environments.
[0019] 2. This application synchronizes the sound signal with the cutting teeth through frequency tracking, so that the monitoring signal corresponds precisely with the actual movement law of the cutting teeth. The processing state is judged by calculating the average deviation rate of the impact pulse interval of each cutting tooth relative to the theoretical cycle. This establishes an intrinsic relationship between the sound signal and the physical movement of the cutting tool. Compared with the indirect monitoring method in the prior art that only relies on the sound spectrum or energy characteristics, it can more accurately reflect the changes in the cutting state of the cutting tool.
[0020] 3. This application combines cycle consistency parameter determination, pulse missing detection, and synchronous frequency band energy analysis to simultaneously identify four processing states: normal cutting, chatter, chipping, and air cutting. It is significantly superior to traditional sound monitoring methods under all working conditions.
[0021] 4. This application uses a non-contact microphone to collect signals, which does not require structural modifications to the machine tool, has low installation and maintenance costs, is suitable for upgrading various existing CNC milling machines, and has broad industrial application prospects. Detailed Implementation
[0022] The terminology used in the following embodiments is for the purpose of describing particular embodiments only and is not intended to be limiting of this application. As used in the specification and appended claims of this application, the singular expressions “a,” “an,” “the,” “the,” “the,” and “this” are intended to also include expressions such as “one or more,” unless the context clearly indicates otherwise. It should also be understood that in the following embodiments of this application, “at least one” and “one or more” refer to one, two, or more than two. The term “and / or” is used to describe the relationship between related objects, indicating that three relationships may exist; for example, A and / or B can indicate: A alone, A and B simultaneously, or B alone, where A and B can be singular or plural. The character “ / ” generally indicates that the preceding and following related objects are in an “or” relationship.
[0023] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.
[0024] The following provides a further detailed description of this application. Example
[0025] This embodiment provides a machining status monitoring method based on synchronous tracking of cutting teeth.
[0026] S1. Obtain the spindle speed information, tool tooth count information, and tool motion trajectory information of the CNC machine tool.
[0027] The spindle speed, tool number, and tool trajectory are read in real time through the CNC system interface.
[0028] For example: The spindle speed N is 6000 rpm; The number of teeth Z of the cutting tool is 4.
[0029] S2. Processing sound signals are simultaneously acquired by a first microphone and a second microphone located at different positions in the processing area.
[0030] The first and second microphones are installed on both sides of the machine tool processing area.
[0031] Two microphones simultaneously acquire sound signals generated during the processing and convert them into digital sound signals.
[0032] Preferably, the distance between the first microphone and the second microphone is 10cm to 50cm, and the synchronous sampling frequency is not less than twice the highest frequency of the processed sound signal.
[0033] S3. Determine the location of the sound source based on the time difference between the sound signals received by the two microphones, and filter the target sound signals from the contact area between the tool and the workpiece.
[0034] The system calculates the time difference between the arrival of the same sound event at the first microphone and the second microphone in real time.
[0035] Determine the direction of the sound source based on the time difference.
[0036] When the sound source is located within a preset spatial range centered on the contact point between the tool and the workpiece, the corresponding sound signal is retained as the target sound signal.
[0037] When the sound originates from the coolant nozzle area, the spindle motor area, or the external area of the machine tool, the corresponding sound signal is suppressed.
[0038] The preset spatial range can be set according to the size of the processing area, for example, a three-dimensional spatial area with a radius of 50mm to 150mm centered on the theoretical contact point between the tool and the workpiece.
[0039] This allows us to obtain the target sound signal.
[0040] In this step, spatial domain filtering of sound signals is achieved based on dual-microphone time difference positioning, which can effectively eliminate environmental interference noise outside the processing area.
[0041] S4. Calculate the tooth passing frequency based on the spindle speed information and the number of tool teeth information, establish a synchronous tracking frequency band, and extract the synchronous sound signal corresponding to the tooth passing frequency from the target sound signal.
[0042] The pass frequency of the cutting teeth is calculated according to the following formula: Ft = N × Z ÷ 60 in: Ft is the passing frequency of the cutting teeth, in Hz; N is the spindle speed, measured in rpm; Z represents the number of teeth on the cutting tool.
[0043] When the spindle speed N is 6000 rpm and the number of tool teeth Z is 4: Ft = 6000 × 4 ÷ 60 = 400Hz.
[0044] The system establishes a synchronous tracking frequency band centered at 400Hz.
[0045] At the same time, harmonic frequency bands of 800Hz, 1200Hz and 1600Hz were established.
[0046] The synchronous tracking frequency band includes a fundamental frequency band centered on the serration passing frequency Ft and at least one corresponding harmonic frequency band, with the bandwidth of each frequency band being 1% to 10% of its respective center frequency.
[0047] For example, the bandwidth of the fundamental frequency band can be 400Hz × 2% = 8Hz, that is, from 396Hz to 404Hz; the bandwidth of the second harmonic frequency band can be 800Hz × 2% = 16Hz, that is, from 792Hz to 808Hz.
[0048] Then, the synchronization sound signal corresponding to the above frequency band is extracted from the target sound signal.
[0049] In this step, the time-frequency domain filtering of sound signals is achieved by synchronously tracking the frequency of the cutting teeth, which can effectively eliminate interference signals unrelated to cutting of the cutting teeth in the machining area, such as coolant spray noise and chip collision noise.
[0050] S5. Extract the blade impact pulse sequence from the synchronized sound signal.
[0051] The system first performs envelope demodulation on the synchronized audio signal.
[0052] Then, a peak detection algorithm is used to identify the impact pulses generated when the cutting teeth pass through the workpiece.
[0053] As an optional implementation, the envelope demodulation process includes: performing a Hilbert transform on the synchronized audio signal to obtain an envelope curve, performing a low-pass filter on the envelope curve, and then using a peak detection algorithm to identify the blade impact pulse, wherein the cutoff frequency of the low-pass filter is 0.5 to 2 times the blade passage frequency.
[0054] For example, when the cutting tooth passes through at a frequency of 400Hz, the cutoff frequency of the low-pass filter can be set to 400Hz to filter out high-frequency fluctuations above 400Hz in the envelope curve and retain the envelope profile related to the impact of the cutting tooth.
[0055] On the filtered envelope curve, local maxima points where the envelope amplitude exceeds K times the envelope mean are identified as blade impact pulses, where K ranges from 1.5 to 3.0, for example, K=2.0.
[0056] Finally, a continuous sequence of blade impact pulses is obtained.
[0057] S6. Calculate the periodic consistency parameter based on the blade impact pulse sequence.
[0058] The system calculates the time interval between adjacent impact pulses to form an actual periodic sequence.
[0059] At the same time, the theoretical cutter period T0 = 1 / Ft is calculated based on the frequency of the cutter teeth.
[0060] The period consistency parameter is the average deviation rate of each actual pulse interval relative to the theoretical cutter tooth period, and its specific calculation method is as follows: First, obtain the time interval between adjacent cutter tooth impact pulses, denoted as . , , ... , where m is the total number of pulses detected; Secondly, the theoretical cutter period T0 = 1 / Ft is calculated based on the cutter tooth passing frequency Ft; Then, the deviation rate of each actual pulse interval relative to the theoretical cutter tooth cycle is calculated one by one, i.e. ; Finally, the average of all deviation rates is taken to obtain the periodic consistency parameter C.
[0061] The physical meaning of the periodic consistency parameter C is: the average deviation of the tooth impact pulse interval from the theoretical uniform interval. The smaller the C value, the more precisely the impact time of each tooth follows the theoretical period, and the more stable the cutting process is; the larger the C value, the stronger the randomness of the tooth impact time, and the less stable the cutting process is.
[0062] When the number of pulses detected within a certain theoretical cutting tooth cycle is less than the number of cutting tooth Z, it is determined that there is a pulse missing in that cycle.
[0063] For example, suppose that in a certain detection, the actual pulse intervals are 2.4ms, 2.6ms, and 2.5ms, and the theoretical period is 2.5ms. Then the deviation rates of each pulse interval from the theoretical period are 4%, 4%, and 0%, respectively, and the average deviation rate is (4%+4%+0%) / 3≈2.7%, that is, the period consistency parameter is about 0.027.
[0064] For example, if a cutting tool has 4 teeth, but only 3 pulses are detected in a complete cutting tooth cycle, then it is determined that 1 pulse is missing.
[0065] The calibration method for the first preset threshold is as follows: Under normal cutting conditions with no chatter, the sound signals of multiple tooth cycles are collected, and the cycle consistency parameter of each cycle is calculated. The first preset threshold is determined based on the statistical value of the calculated cycle consistency parameter. For example, under normal cutting conditions without chatter, 200 tooth cycles are continuously collected, the cycle consistency parameter of each cycle is calculated, the mean μ and standard deviation σ are obtained, and the first preset threshold Th1 = μ + 3σ is taken.
[0066] The calibration method for the second preset threshold is as follows: Under a no-contact cutting state between the tool and the workpiece, the machining sound signal is collected, and the sound energy within the synchronous tracking frequency band is extracted. The second preset threshold is determined based on the sound energy value. For example, the machining sound signal is collected under a no-contact cutting state, the total sound energy value within the synchronous tracking frequency band is extracted, and this process is repeated 20 times to obtain the average energy value E_avg. The second preset threshold Th2 = E_avg × 1.5 is then used.
[0067] Criteria for determining consecutive missing pulses: Within multiple consecutive theoretical cutting tooth cycles, the number of valid pulses detected is less than half the number of pulses that should occur in that cycle. For example, if the cutting tool has 4 teeth, theoretically there should be 4 pulses per cutting tooth cycle. If, within 3 consecutive theoretical cutting tooth cycles, the number of pulses detected in each cycle is less than 2, it is determined to be consecutive missing pulses.
[0068] S7. Determine the current processing status based on the cycle consistency parameter and output the monitoring results.
[0069] Based on the first preset threshold, the second preset threshold, and the continuous missing pulse determination criteria specified in S6, the state is determined according to the following rules: When the cycle consistency parameter is lower than the first preset threshold, it is determined to be a normal cutting state; When the periodic consistency parameter exceeds the first preset threshold, it is determined to be a flutter state; When a continuous absence of cutting tooth impact pulses is detected, it is determined to be a tool chipping condition; The sound energy within the synchronous tracking frequency band is obtained by integrating the power of the synchronous sound signal within the fundamental frequency band and each harmonic frequency band. When the sound energy is lower than the second preset threshold, it is determined to be in an empty-cut state.
[0070] The system outputs the corresponding monitoring signal based on the judgment result.
[0071] When the tool is determined to be in a chipped state, a stop command is sent to the CNC system, and the stop response time is no more than 50ms. Example
[0072] This embodiment focuses on illustrating the two-level denoising architecture consisting of S3 and S4 and their synergistic effect.
[0073] In step S3, spatial domain filtering is achieved through dual-microphone time-difference localization. Specifically, there is a significant difference between the time difference of ambient noise outside the processing area (such as operating noise from adjacent equipment, workshop reflection noise, etc.) reaching the two microphones and the time difference of sound sources within the processing area reaching the two microphones. Based on this, noise outside the processing area is suppressed. This step solves the problem of "which sounds are not emanating from near the blade tip".
[0074] In step S4, time-frequency domain filtering is achieved through synchronous tracking of the cutting tooth passage frequency. Specifically, even if the sound originates from the machining area, it may still contain sound components unrelated to cutting, such as coolant spray and chip collision. Since the cutting sound strictly follows the cutting tooth passage frequency and its harmonic frequencies, while interference sounds from coolant, chips, etc., do not follow this pattern, non-cutting noise within the area is filtered out. This step solves the problem of "which sounds within the machining area are unrelated to cutting."
[0075] S3 and S4 constitute a two-stage denoising architecture, filtering the processed audio signal from both the spatial and time-frequency domains: the first stage addresses the "spatial selectivity" problem, while the second stage addresses the "time-frequency selectivity" problem. The former provides the latter with a relatively clean spatial range signal, and the latter further refines the extraction based on the former. The two stages form a causal progression in the signal processing flow, jointly providing a high signal-to-noise ratio synchronous audio signal for pulse extraction and state determination in stages S5 to S7. Example
[0076] This embodiment provides experimental data to illustrate the technical effects of this application.
[0077] Experimental conditions: Machining center: Maximum spindle speed 12000 rpm Workpiece material: Titanium alloy TC4 Cutting tool: 12mm diameter three-tooth carbide end mill Cooling method: Water-based coolant injection Spindle speed: 6000 rpm Sampling frequency: 100kHz Tests were conducted under four conditions: normal cutting, chattering, tool chipping, and air cutting.
[0078] The first preset threshold is calibrated in the manner described in Example 1, that is, the sound signals of multiple tooth cycles are collected under normal cutting conditions without chatter and the cycle consistency parameter is calculated, and the first preset threshold is determined based on the statistical value of this set of data.
[0079] The second preset threshold is calibrated in the manner described in Example 1, that is, the sound signal is collected and processed in the air-cut state and the sound energy in the synchronous tracking frequency band is extracted, and the second preset threshold is determined based on the energy value.
[0080] Comparative experiment: Operating conditions Full-band energy analysis Full-band spectrum analysis This application Normal cutting 72% 78% 96% Flutter state 58% 65% 94% Knife chipping 45% 52% 98% Air cut state 61% 70% 97% The experimental data above show that the present application, by combining a two-level denoising architecture with periodic consistency parameter determination, significantly outperforms traditional sound monitoring methods under all processing conditions, proving the effectiveness and advancement of the technical solution presented in this application.
Claims
1. A machining status monitoring method based on synchronous tracking of cutting teeth, characterized in that, Includes the following steps: S1. Obtain the spindle speed information, tool tooth count information, and tool motion trajectory information of the CNC machine tool; S2. Processing sound signals are simultaneously acquired by a first microphone and a second microphone set at different locations in the processing area; S3. Determine the location of the sound source based on the time difference between the sound signals received by the two microphones, and filter the target sound signals from the contact area between the tool and the workpiece. S4. Calculate the tooth passing frequency based on the spindle speed information and the number of tool teeth information, establish a synchronous tracking frequency band, and extract the synchronous sound signal corresponding to the tooth passing frequency from the target sound signal; S5. Extract the blade impact pulse sequence from the synchronous sound signal; S6. Calculate the period consistency parameter based on the blade impact pulse sequence; S7. Determine the current processing status based on the cycle consistency parameter and output the monitoring results.
2. The processing status monitoring method according to claim 1, characterized in that: In step S3, the location of the sound source is determined based on the arrival time difference of the same sound event received by the two microphones. When the sound source is located within a preset spatial range centered on the contact point between the tool and the workpiece, the corresponding sound signal is retained as the target sound signal. When the sound source is located outside the preset spatial range, the corresponding sound signal is suppressed.
3. The processing status monitoring method according to claim 1, characterized in that: In step S4, the frequency Ft of the cutting teeth is calculated according to the following formula: Ft = N × Z ÷ 60; in: Ft is the passing frequency of the cutting teeth; N is the spindle speed; Z represents the number of teeth on the cutting tool; The synchronous tracking frequency band includes a fundamental frequency band centered at the blade passing frequency Ft and at least one corresponding harmonic frequency band, wherein the bandwidth of the fundamental frequency band and the harmonic frequency band is 1% to 10% of the corresponding center frequency.
4. The processing status monitoring method according to claim 1, characterized in that: In step S5, the synchronous sound signal is envelope demodulated and the blade impact pulse is identified by combining a peak detection algorithm to obtain the blade impact pulse sequence. The envelope demodulation process includes: performing a Hilbert transform on the synchronized audio signal to obtain an envelope curve; performing a low-pass filter on the envelope curve; and using a peak detection algorithm to identify the blade impact pulse. The cutoff frequency of the low-pass filter is 0.5 to 2 times the blade passage frequency.
5. The processing status monitoring method according to claim 1, characterized in that: In step S6, an actual periodic sequence is established based on the time interval between adjacent cutter impact pulses, and a periodic consistency parameter is calculated based on the deviation between the actual periodic sequence and the theoretical cutter period. The period consistency parameter is the average deviation rate of each actual pulse interval relative to the theoretical cutter tooth period. The larger the deviation between the actual pulse interval and the theoretical cutter tooth period, the larger the value of this parameter.
6. The processing status monitoring method according to claim 1, characterized in that: In step S7, when the period consistency parameter exceeds the first preset threshold, the current processing state is determined to be a chattering state. The first preset threshold is calibrated as follows: Under the cutting state where the tool and the workpiece are in normal contact and there is no chatter, the sound signals of multiple tooth cycles are collected and the cycle consistency parameters of each cycle are calculated. The first preset threshold is determined based on the statistical value of the calculated cycle consistency parameters.
7. The processing status monitoring method according to claim 1, characterized in that: In step S7, when a series of missing pulses appear in the tooth impact pulse sequence, the current machining state is determined to be a tool breakage state; when the tool breakage state is determined to be a tool breakage state, a stop command is sent to the CNC system. The term "continuous missing pulses" refers to the situation where, within multiple consecutive theoretical blade cycles, the number of valid pulses detected is less than half the number of pulses that should occur in that cycle.
8. The processing status monitoring method according to claim 1, characterized in that: In S7, when the sound energy in the synchronous tracking frequency band is lower than the second preset threshold, the current processing state is determined to be an empty cutting state. The second preset threshold is calibrated as follows: in the air cutting state where the tool and the workpiece are not in contact, the machining sound signal is collected and the sound energy in the synchronous tracking frequency band is extracted, and the second preset threshold is determined based on the sound energy value.
9. The processing status monitoring method according to claim 1, characterized in that: A two-stage denoising architecture is constructed by spatial domain filtering in S3 and time-frequency domain filtering in S4. In S3, sound source filtering based on time difference localization serves as the first stage of denoising, filtering out environmental noise outside the processing area. In S4, synchronous tracking filtering based on the tooth passing frequency serves as the second stage of denoising, filtering out interference signals unrelated to tooth cutting within the processing area. The synchronous sound signal after two stages of denoising is used for pulse extraction and state determination in S5 to S7.
