A method, device, medium and equipment for detecting burrs based on an eddy current sensor

Abstract

The application relates to the technical field of intelligent manufacturing control, in particular to a chip jamming detection method and device based on an eddy current sensor, a medium and equipment. The method comprises the following steps: collecting an eddy current signal generated when a to-be-detected tool shank rotates in real time; dividing the eddy current signal by adopting a period segmentation strategy to obtain eddy current detection information of multiple detection periods; calculating a detection runout change rate of each detection period according to the eddy current detection information corresponding to each detection period and a runout reference value obtained in a calibration stage; and determining the number of detection periods whose detection runout change rates exceed a preset runout rate threshold value to determine whether the to-be-detected tool shank has chip jamming. The application calculates the detection runout change rate of each detection period by taking the eddy current detection information of each detection period and the runout reference value determined under the condition of no chip jamming as parameters, which is used to represent the signal difference under the condition of no chip jamming, and then whether there is chip jamming is determined according to the number of periods whose detection runout change rates exceed the preset runout rate threshold value, so that the detection precision is improved.

Description

Technical Field

[0001] This application relates to the field of intelligent manufacturing control technology, and in particular to a chip detection method, device, medium and equipment based on an eddy current sensor. Background Technology

[0002] In modern precision manufacturing, especially in CNC machining, the tooling system, as a key component connecting the machine tool spindle and the cutting tool, directly affects machining quality, equipment lifespan, and production efficiency due to its assembly accuracy and reliability. The tool holder (using standard interfaces such as BT, HSK, and CAT) achieves a high-rigidity connection by mating its tapered surface with the spindle's inner bore. It typically incorporates pull studs or clamping mechanisms to maintain stable tool clamping during high-speed rotation. However, in actual machining processes, the generation and accumulation of chips such as metal shavings, coolant residue, and dust often lead to chip jamming at the tapered surface of the tool holder. Chip jamming results in uneven contact between the tool holder and the spindle, leading to reduced machining accuracy. Furthermore, chip jamming can prevent the tool holder from fully inserting into the spindle or cause insufficient clamping force, resulting in tool loosening or even falling off during high-speed rotation, significantly impacting equipment lifespan.

[0003] In related technologies, besides methods with low detection and cleaning accuracy such as manual visual inspection and air blowing to remove stuck chips, vibration detection is a common method for detecting the presence of stuck chips. This method monitors vibration signals or acoustic emission characteristics during tool changing to indirectly determine the presence of stuck chips. However, this vibration detection method is an indirect inference, lacks specificity, and is easily affected by other interference factors (such as tool wear and spindle imbalance), making it difficult to achieve direct and reliable identification of stuck chips. Summary of the Invention

[0004] In view of the above-mentioned shortcomings and deficiencies of the prior art, this application provides a chip detection method, device, medium and equipment based on an eddy current sensor, the main purpose of which is to solve the problems of low automation, insufficient detection accuracy and poor adaptability in the current chip detection methods.

[0005] To achieve the above objectives, the main technical solutions adopted in this application include: In a first aspect, embodiments of this application provide a chip detection method based on an eddy current sensor, comprising: The eddy current signal generated when the tool holder under test rotates is acquired in real time, wherein the eddy current signal is used to characterize the distance between the eddy current sensor probe and the conical surface of the tool holder; The eddy current signal is divided using a period segmentation strategy to obtain eddy current detection information for multiple detection periods; Based on the eddy current detection information corresponding to each detection cycle and the oscillation reference value obtained during the calibration stage, the detection oscillation change rate for each detection cycle is calculated; the detection oscillation change rate is used to characterize the degree of deviation between the eddy current signal of the tool holder under test and the eddy current signal under chip-free conditions. The number of detection cycles in which the detected oscillation rate exceeds a preset oscillation rate threshold is determined in order to determine whether the tool holder under test has chips stuck.

[0006] Optionally, a period segmentation strategy is used to divide the eddy current signal, including: acquiring the target rotational speed of the tool holder under test and the sampling frequency of the eddy current sensor; calculating the theoretical period length based on the target rotational speed and the sampling frequency; truncating a pre-acquired segment of the chip-free eddy current signal based on the theoretical period length to obtain an initial reference period; using the initial reference period as a template, dividing the eddy current signal using a template matching method to obtain eddy current detection information for multiple detection periods.

[0007] Optionally, the initial reference period is used as a template to divide the eddy current signal using a template matching method, including: calculating the cross-correlation sequence between the initial reference period and the eddy current signal, and identifying multiple peaks in the cross-correlation sequence; taking the position of each peak as the starting point of a detection period, and based on adjacent starting points, extracting corresponding signal segments from the eddy current signal to obtain eddy current detection information for the multiple detection periods.

[0008] Optionally, after dividing the eddy current signal using template matching, the method further includes: calculating the Pearson correlation coefficient between each detection cycle and the initial reference cycle; removing detection cycles with correlation coefficients lower than a preset correlation threshold; using the RANSAC robust regression algorithm to identify the eddy current detection information corresponding to each removed detection cycle to obtain each region to be compensated; and using an interpolation method to compensate each region to be compensated.

[0009] Optionally, the step of calculating the detection oscillation change rate for each detection cycle based on the eddy current detection information corresponding to each detection cycle and the oscillation reference value obtained during the calibration phase includes: calculating the peak-to-peak value of the eddy current detection information corresponding to each compensated detection cycle to obtain the detection oscillation amount corresponding to each detection cycle; and calculating the detection oscillation change rate for each detection cycle based on the detection oscillation amount corresponding to each detection cycle and the oscillation reference value obtained during the calibration phase.

[0010] Optionally, the oscillation reference value obtained in the calibration stage is obtained based on the following steps: using a periodic segmentation strategy, a pre-acquired segment of a chip-free eddy current signal is periodically divided to obtain multiple chip-free eddy current signals for calibration periods; using the RANSAC robust regression algorithm, the chip-free eddy current signals for the multiple calibration periods are identified to obtain the compensation region corresponding to each calibration period; using an interpolation method, the compensation region corresponding to each calibration period is compensated to obtain a set of compensated chip-free eddy current signals; based on each calibration period in the set of chip-free eddy current signals, the peak-to-peak value of each calibration period is calculated to obtain the calibration oscillation amount corresponding to each calibration period; based on the set of chip-free eddy current signals, the median calibration period is calculated, and the calibration oscillation amount of the median calibration period is determined as the oscillation reference value obtained in the calibration stage.

[0011] Optionally, determining whether the tool holder under test has jammed chips includes: calculating the proportion of the number of detection cycles in which the detection oscillation rate exceeds the preset oscillation rate threshold to the total number of cycles; if the proportion exceeds the preset proportion, determining that the tool holder under test has jammed chips; wherein, the preset oscillation rate threshold is determined based on the calibration oscillation rate, which is calculated based on the calibration oscillation amount of each calibration cycle in the calibration stage and the oscillation reference value obtained in the calibration stage.

[0012] Secondly, embodiments of this application provide a chip detection device based on an eddy current sensor, comprising: The acquisition unit is configured to acquire in real time the eddy current signal generated when the tool holder under test rotates, wherein the eddy current signal is used to characterize the distance between the eddy current sensor probe and the conical surface of the tool holder; The segmentation unit is configured to segment the eddy current signal using a periodic segmentation strategy to obtain eddy current detection information for multiple detection periods. The calculation unit is configured to calculate the detection oscillation change rate for each detection cycle based on the eddy current detection information corresponding to each detection cycle and the oscillation reference value obtained during the calibration stage; the detection oscillation change rate is used to characterize the degree of deviation between the eddy current signal of the tool holder under test and the eddy current signal under chip-free conditions. The detection unit is configured to determine the number of detection cycles in which the rate of change of the detected oscillation exceeds a preset oscillation rate threshold, so as to determine whether there is chip jamming in the tool holder under test.

[0013] Thirdly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the chip detection method based on an eddy current sensor described in the first aspect.

[0014] Fourthly, this application provides an electronic device, including a storage medium, a processor, and a computer program stored on the storage medium and executable on the processor, wherein the processor executes the computer program to implement the chip detection method based on an eddy current sensor as described in the first aspect.

[0015] Using the above technical solution, this application provides a chip detection method, apparatus, medium, and device based on an eddy current sensor. First, the eddy current signal generated during the rotation of the tool holder under test is acquired in real time. This eddy current signal characterizes the distance between the eddy current sensor probe and the conical surface of the tool holder. A periodic segmentation strategy is used to divide the eddy current signal, obtaining eddy current detection information for multiple detection cycles. Then, based on the eddy current detection information corresponding to each detection cycle and the oscillation reference value obtained during the calibration phase, the detection oscillation rate of change for each detection cycle is calculated. This oscillation rate characterizes the degree of deviation between the eddy current signal of the tool holder under test and the eddy current signal under chip-free conditions. Finally, the number of detection cycles in which the detection oscillation rate exceeds a preset oscillation rate threshold is determined to indicate whether chip jamming exists in the tool holder under test. Compared with related technologies, this paper presents a chip detection method based on an eddy current sensor. First, the detection period of the eddy current signal is strictly divided using a period segmentation strategy. Then, the eddy current detection information of each detection period is compared with the oscillation reference value representing a chip-free condition obtained during the calibration phase. The deviation rate of the oscillation change is calculated to determine the degree of deviation between the eddy current signal used for the tool holder under test and the eddy current signal under chip-free conditions. By detecting the oscillation change rate, the chip-related characteristics during the detection phase are compared with the waveform of the eddy current signal under chip-free conditions, highlighting the deviation caused by abnormal chip jamming and improving detection accuracy. Furthermore, when determining whether chip jamming exists, the proportion of cycles with oscillation change rates exceeding the limit is used as a criterion, effectively avoiding misjudgments caused by random noise or transient interference within a single cycle, thus improving the robustness of the detection. Attached Figure Description

[0016] Figure 1 A schematic flowchart of a chip detection method based on an eddy current sensor provided in this application embodiment; Figure 2 A flowchart illustrating the steps of the calibration stage provided in this application embodiment; Figure 3 This is a flowchart illustrating the steps of the detection stage provided in an embodiment of this application; Figure 4 This is a schematic diagram of a chip detection device based on an eddy current sensor provided in an embodiment of this application. Detailed Implementation

[0017] To better understand the above technical solutions, exemplary embodiments of this application will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of this application are shown in the drawings, it should be understood that this application can be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this application can be understood more clearly and thoroughly, and that the scope of this application can be fully conveyed to those skilled in the art.

[0018] As mentioned in the background, in modern precision manufacturing, especially in CNC machining, the tool system, as a key component connecting the machine tool spindle and the cutting tool, directly affects machining quality, equipment lifespan, and production efficiency due to its assembly accuracy and reliability. The tool holder achieves a high-rigidity connection through a tapered surface mating with the spindle's inner bore. It typically contains pull studs or clamping mechanisms to maintain stable tool clamping during high-speed rotation. However, in actual machining processes, the generation and accumulation of chips (metal chips, coolant residue, dust, etc.) often leads to chip jamming at the tapered surface of the tool holder. If chip jamming is not detected and addressed in a timely manner, it will lead to a series of serious consequences: First, chip jamming will disrupt the precision fit between the tool holder and the spindle, resulting in uneven contact surfaces, reduced system rigidity and dynamic balance performance, and consequently causing machining vibration, deterioration of surface roughness, and even dimensional deviations; Second, chip jamming may prevent the tool holder from being fully inserted into the spindle or cause insufficient clamping force, resulting in the tool loosening or even falling off during high-speed rotation, seriously threatening equipment safety and operator safety; In addition, long-term chip jamming will accelerate the wear of the spindle and tool holder, shorten the service life of critical components, and increase maintenance costs and downtime.

[0019] Current technologies employ methods such as manual visual inspection, air blowing to remove stuck debris, and vibration acoustic testing. However, these methods generally suffer from drawbacks such as low automation, insufficient detection accuracy, poor adaptability, blind spots, and difficulty in achieving online real-time monitoring. In particular, they fail to meet the development trends of intelligent manufacturing and unmanned workshops.

[0020] To address the aforementioned issues, this application proposes a chip detection method based on an eddy current sensor. This method can be applied to or integrated into a chip detection system, such as in factory workshops, CNC machining centers, and tool clamping status detection of high-precision machine tools. During operation, it can execute any of the chip detection methods based on eddy current sensors mentioned below. Figure 1 As shown, the method includes: S101, which collects the eddy current signal generated when the tool holder under test rotates in real time.

[0021] In actual testing, chip jamming detection is usually performed immediately after tool change and initial activation. At this time, no actual operation is performed; instead, the system is checked for chip jamming. For example, after tool change, the tool is allowed to idle for 3 to 5 seconds, and continuous eddy current signals are acquired during this period. By detecting these eddy current signals, it is determined whether chip jamming is present.

[0022] In S101, the eddy current signal characterizes the distance between the eddy current sensor probe and the conical surface of the tool holder. That is, the eddy current sensor is used to collect the real-time distance between itself and the tool holder's conical surface. The specific installation location is not limited; it can be installed in a fixed position such as the machine tool spindle or the outer wall of the protective cover. As an optional installation method, the eddy current sensor can be fixed to the end face of the machine tool spindle, with the probe facing the outer conical surface of the tool holder. Thus, when the tool holder rotates, the sensor can continuously and non-contactly monitor the distance to every point on the tool holder's conical surface.

[0023] It should also be noted that no other distance sensor was selected in this embodiment because the method proposed in this embodiment is specific to specific detection scenarios such as high-precision machine tools. Other sensors that can provide distance signals are not suitable for this type of scenario. For example, laser sensors are contaminated by dust and debris during operation, and acoustic sensors require echo time detection, which has low accuracy and slow speed, and therefore are not suitable for this specific industrial application scenario.

[0024] S102, the eddy current signal is divided using a periodic segmentation strategy to obtain eddy current detection information for multiple detection cycles.

[0025] In S102, the acquired eddy current signal is a continuous signal used to characterize distance relationships; therefore, periodic division is necessary for subsequent detection processes. A preset periodic division strategy is used to accurately and adaptively divide the continuous eddy current signal into periods to facilitate subsequent chip detection. Specifically, the eddy current detection information refers to the eddy current signal to be detected in each detection period after being divided according to a standard period. This information may include waveform, peak and trough values, etc.

[0026] S103, calculate the rate of change of vibration in each detection cycle based on the eddy current detection information corresponding to each detection cycle and the vibration reference value obtained during the calibration stage.

[0027] The oscillation reference value obtained during the calibration phase is calculated based on the eddy current signal generated by the rotation of the tool holder under chip-free conditions. The oscillation amount is the difference between the peak and trough within a cycle, i.e., the peak-to-peak value.

[0028] Then, based on the eddy current detection information corresponding to each detection cycle and the oscillation reference value obtained during the calibration phase, the detection oscillation change rate for each detection cycle is calculated. This rate characterizes the deviation between the eddy current signal of the tool holder under test and the eddy current signal under chip-free conditions. The detection oscillation change rate is obtained by comparing the oscillation amount (called the detection oscillation amount) corresponding to each detection cycle with the oscillation reference value from the calibration phase during the detection phase. Because the signal waveform of the eddy current signal in each cycle characterizes the distance relationship with the tapered surface of the tool holder, which includes the characteristics of whether chips are stuck, calculating the oscillation amount for each cycle helps to extract the chip-sticking characteristics and compare them with the reference value (i.e., the oscillation reference value) under chip-free conditions during the calibration phase, thereby improving the detection accuracy.

[0029] S104, determine the number of detection cycles in which the rate of change of vibration exceeds the preset vibration rate threshold, in order to determine whether there is chip jamming in the tool holder under test.

[0030] In S104, the preset oscillation rate threshold is actually calculated based on the oscillation amount during the calibration phase. Then, the number of detection cycles where the oscillation rate exceeds the preset threshold is statistically analyzed, and the proportion of these cycles is used to determine if chip jamming exists. By using the statistical proportion of cycles where the oscillation rate exceeds the limit as a criterion, misjudgments caused by random noise or instantaneous interference within a single cycle are effectively avoided.

[0031] In this embodiment, the eddy current signal generated during the rotation of the tool holder under test is first acquired in real time. This eddy current signal characterizes the distance between the eddy current sensor probe and the conical surface of the tool holder. A periodic segmentation strategy is used to divide the eddy current signal, obtaining eddy current detection information for multiple detection cycles. Then, based on the eddy current detection information corresponding to each detection cycle and the oscillation reference value obtained during the calibration phase, the detection oscillation rate of change for each detection cycle is calculated. This oscillation rate characterizes the degree of deviation between the eddy current signal of the tool holder under test and the eddy current signal under chip-free conditions. Finally, the number of detection cycles in which the detection oscillation rate exceeds a preset oscillation rate threshold is determined to indicate whether chip jamming exists in the tool holder under test. Compared with related technologies, this paper presents a chip detection method based on an eddy current sensor. First, the detection period of the eddy current signal is strictly divided using a period segmentation strategy. Then, the eddy current detection information of each detection period is compared with the oscillation reference value representing a chip-free condition obtained during the calibration phase. The deviation rate of the oscillation change is calculated to determine the degree of deviation between the eddy current signal used for the tool holder under test and the eddy current signal under chip-free conditions. By detecting the oscillation change rate, the chip-related characteristics during the detection phase are compared with the waveform of the eddy current signal under chip-free conditions, highlighting the deviation caused by abnormal chip jamming and improving detection accuracy. Furthermore, when determining whether chip jamming exists, the proportion of cycles with oscillation change rates exceeding the limit is used as a criterion, effectively avoiding misjudgments caused by random noise or transient interference within a single cycle, thus improving the robustness of the detection.

[0032] As a feasible implementation method, the period segmentation strategy in S102 is used to divide the eddy current signal, including: acquiring the target rotational speed of the tool holder under test and the sampling frequency of the eddy current sensor; calculating the theoretical period length based on the target rotational speed and the sampling frequency; truncating a pre-acquired segment of the chip-free eddy current signal based on the theoretical period length to obtain an initial reference period; using the initial reference period as a template, dividing the eddy current signal by template matching to obtain eddy current detection information for multiple detection periods.

[0033] In this embodiment, the target rotational speed and sampling frequency of the tool holder under test are first obtained through an integrated system. Then, a period of theoretical cycle length is extracted from a pre-acquired, chip-free eddy current signal to obtain the initial reference cycle. The purpose of pre-acquired, chip-free eddy current signal is to ensure the accuracy of the initial reference cycle and avoid introducing errors caused by chip jamming. Then, using the initial reference cycle as a template, the eddy current signal acquired during the detection phase is periodically divided into multiple detection cycles. The theoretical cycle length is calculated using the target rotational speed and sampling frequency, and template matching is performed based on the initial reference cycle, achieving automated and high-precision period division of the eddy current signal. This method effectively overcomes the problem of inconsistent cycle lengths caused by small fluctuations in rotational speed, providing a stable and aligned periodic signal basis for subsequent oscillation calculations, and improving the reliability and consistency of the entire detection process.

[0034] As a feasible implementation method, the initial reference period is used as a template, and the eddy current signal is divided by template matching. This includes: calculating the cross-correlation sequence between the initial reference period and the eddy current signal, and identifying multiple peaks in the cross-correlation sequence; taking the position of each peak as the starting point of a detection period, and extracting corresponding signal segments from the eddy current signal based on adjacent starting points, so as to obtain eddy current detection information for multiple detection periods.

[0035] In this embodiment, the cross-correlation sequence is used to find the similarity with the initial reference period. For example, there is an initial reference period signal (i.e., template) p[n] of length L, and a sequence of length N (N The process involves dividing the continuous eddy current signal s[n] into L and calculating the cross-correlation sequence Rsp[m]. This can be intuitively understood as: using the template signal p[n] as a sliding window, sliding it from beginning to end on the continuous signal s[n], and calculating their similarity point by point. The m-th value of the cross-correlation sequence Rsp[m] is calculated using the following formula: (Formula 1) Here, m represents the initial displacement point of the template signal on the continuous eddy current signal. Cross-correlation sequences and peak detection are used to determine the starting point of each cycle, achieving sub-sampling level cycle positioning accuracy. This method exhibits good robustness to noise and minute deformations in the signal, ensuring accurate cycle synchronization and segmentation even under poor signal quality conditions, laying the foundation for subsequent precise comparisons.

[0036] Furthermore, after dividing the eddy current signal using template matching, the method also includes: calculating the Pearson correlation coefficient between each detection cycle and the initial reference cycle; removing detection cycles with correlation coefficients lower than a preset correlation threshold; using the RANSAC robust regression algorithm to identify the eddy current detection information corresponding to each removed detection cycle to obtain each region to be compensated; and using an interpolation method to compensate each region to be compensated.

[0037] In this embodiment, in order to initially screen out complete periodic segments, the Pearson correlation coefficient of each of the remaining periods is calculated based on the first period; only periods with correlation coefficients higher than the preset correlation threshold are retained, and periods with abnormal morphology caused by local debris, noise or instantaneous interference are removed.

[0038] The area to be compensated, also known as the groove area, refers to structures such as the transmission key or the edge of the pull stud hole on the outer conical surface of the tool holder used to transmit torque. These structures can cause some deviations in the acquired signal in some cases. Therefore, in practice, each detection cycle is considered to have a groove area, and compensation is performed using interpolation methods. Thus, this embodiment uses Pearson correlation coefficient to screen and remove abnormal cycles, and combines the RANSAC algorithm to identify and compensate the area to be compensated, forming a dual signal enhancement mechanism. This mechanism can not only effectively filter out invalid cycles caused by transient interference or severe chip jamming, but also repair local signal distortion caused by small chips or sensor noise, thereby significantly improving the signal quality used for final judgment and reducing the risk of misjudgment and missed judgment.

[0039] As a feasible implementation method, in S103, the detection oscillation change rate of each detection cycle is calculated based on the eddy current detection information corresponding to each detection cycle and the oscillation reference value obtained in the calibration stage. This includes: calculating the peak-to-peak value of the eddy current detection information corresponding to each compensated detection cycle to obtain the detection oscillation amount corresponding to each detection cycle; and calculating the detection oscillation change rate of each detection cycle based on the detection oscillation amount corresponding to each detection cycle and the oscillation reference value obtained in the calibration stage.

[0040] In this embodiment, the process of calculating the detected oscillation amount and the detected oscillation rate is explained. First, the detected oscillation amount is defined as the peak-to-peak value of the eddy current detection information corresponding to the detection cycle after compensation processing, i.e., the difference between the peak and trough, thus better reflecting the characteristics of chip jamming. Then, based on the detected oscillation amount corresponding to each detection cycle and the oscillation reference value obtained during the calibration phase, the detected oscillation change rate for each detection cycle is calculated. This allows for a comparison between the chip jamming characteristics during the detection phase and the eddy current signal waveform under chip-free conditions, highlighting the deviation caused by abnormal chip jamming.

[0041] As a feasible implementation method, the oscillation reference value obtained in the calibration stage is obtained based on the following steps: A periodic segmentation strategy is used to periodically divide a pre-acquired segment of a chip-free eddy current signal to obtain multiple calibration cycles of chip-free eddy current signals; a robust RANSAC regression algorithm is used to identify the chip-free eddy current signals in multiple calibration cycles to obtain the compensation region corresponding to each calibration cycle; an interpolation method is used to compensate the compensation region corresponding to each calibration cycle to obtain a compensated set of chip-free eddy current signals; based on each calibration cycle in the chip-free eddy current signal set, the peak-to-peak value of each calibration cycle is calculated to obtain the calibration oscillation amount corresponding to each calibration cycle; based on the chip-free eddy current signal set, the median calibration cycle is calculated, and the calibration oscillation amount of the median calibration cycle is determined as the oscillation reference value obtained in the calibration stage.

[0042] In this embodiment, the oscillation reference value is selected as the median value of the calibration oscillation amount in each calibration cycle of the chip-free eddy current signal during the calibration phase. The specific cycle processing procedure is consistent with the cycle processing procedure in the aforementioned detection phase. Using the median method to generate the reference cycle signal has stronger noise resistance and robustness compared to the mean method. Even if there are minor anomalies that are not fully compensated in some calibration cycles, the median value can effectively suppress their influence, thereby obtaining a cleaner and more stable reference benchmark that better represents the "ideal chip-free" state, improving the accuracy of the calibration model.

[0043] As a feasible implementation method, the above-mentioned interpolation method includes using spline functions for compensation; and performing compensation processing on each region to be compensated using an interpolation method, specifically:

[0044] In this embodiment, The signal corresponding to the i-th period, The signal is the result of fitting with a spline function. Let represent the signal of the i-th cycle after compensation. This formula expresses that if a groove region exists within the i-th cycle, a spline function is used for fitting, replacing the signal of that cycle. As mentioned earlier, the region to be compensated is the groove region. By using spline function interpolation compensation, a smooth curve fitting that conforms to the natural shape of the region to be compensated can be achieved. This method eliminates outliers while preserving the structural characteristics of the tool holder to the maximum extent, avoiding additional distortion introduced by improper compensation, and ensuring the authenticity and comparability of the compensated signal.

[0045] As a feasible implementation method, determining whether there is chipping in the tool holder under test includes: calculating the proportion of the number of detection cycles in which the rate of change of vibration exceeds a preset vibration rate threshold to the total number of cycles; if the proportion exceeds the preset proportion, determining that there is chipping in the tool holder under test; wherein, the preset vibration rate threshold is determined based on the calibrated vibration rate, which is calculated based on the calibrated vibration amount of each calibration cycle in the calibration stage and the vibration reference value obtained in the calibration stage.

[0046] In this embodiment, the proportion of the total number of detection cycles in which the rate of change of vibration exceeds a preset vibration rate threshold is used as the basis for judgment. If it exceeds a certain proportion, it is judged as a chip jamming state; otherwise, it is a normal state.

[0047] A preset oscillation rate threshold is proposed, which is an adaptive criterion mechanism based on the calibration oscillation rate. The preset oscillation rate threshold is determined based on the calibration oscillation rate, which is calculated from the calibration oscillation amount in each calibration cycle and the oscillation reference value obtained during the calibration phase. This achieves an adaptive threshold setting mechanism. This allows the detection standard to be dynamically adjusted according to different tool holder models, sensor installation states, and machine tool operating environments, enhancing the method's versatility and engineering applicability, and avoiding the incompatibility of fixed thresholds under different working conditions.

[0048] It's important to clarify here that the purpose of calibrating the oscillation amount differs from that of detecting the oscillation amount. The purpose of the detection phase is to compare the eddy current signal under test with the eddy current signal without jamming (represented by the oscillation reference value in the calibration phase) to determine the presence of jamming. The purpose of the calibration phase is to obtain reference values ​​based on the eddy current signal without jamming (i.e., the ideal eddy current signal) (e.g., the initial reference period as a standard template for period division, and the oscillation reference value as a representation of the standard waveform without jamming), for use in the detection phase. Furthermore, the detected oscillation amount is used to highlight the jamming characteristics of the eddy current signal under test through peak-to-peak values, while the purpose of calibrating the oscillation amount is to obtain a preset oscillation rate threshold (the oscillation threshold is established based on the calibrated oscillation amount), thus facilitating the judgment of jamming during the detection phase.

[0049] The above embodiments are described from the perspective of the detection stage. Next, from the perspective of the completeness of the solution, a chip detection method based on an eddy current sensor provided in this embodiment will be further explained. The method includes a calibration stage and a detection stage. The purpose of the calibration stage is mainly to obtain reference values ​​under chip-free conditions, so that the parameters obtained in the detection stage can be compared with the reference values ​​to determine the chip-free condition. Specifically, the calibration stage processes the eddy current signal under chip-free conditions, and the steps are as follows: Figure 2 As shown, it includes: Adaptive period segmentation addresses the issue that the acquired voltage signal may contain inconsistent period lengths, local distortions, or incomplete periods due to mechanical vibration, feed fluctuations, or chip jamming interference. This embodiment first preprocesses the eddy current signal, converting it into a distance signal s[n]. Then, it introduces signal standardization preprocessing and performs multi-level abnormal period removal on the standardized signal. This process specifically includes S201-S205: S201, the system determines the target rotational speed of the tool holder based on the known speed. (Unit: r / min) and sensor sampling frequency (Unit: Hz), calculate the theoretical number of sampling points corresponding to a single cycle, using the following formula: (Formula 2) In S201, the sensor sampling frequency is a time-dimensional quantity, representing how many points are collected per second. Assume... =1000, collecting 1000 points per second. The toolholder rotation speed is a quantity representing the number of revolutions per minute. For example, 3000 revolutions per minute translates to 50 revolutions per second. Therefore, one revolution takes 0.02 seconds, and 20 points are collected in 0.02 seconds, meaning 20 points are collected per cycle. .

[0050] S202, subsequently, from the ideal original signal Specify the starting position The length of the cut-off point is A segment is used as the initial reference period, specifically as follows: 1) (Formula 3) S203, to eliminate the effects of amplitude offset and gain difference, zero-mean unit variance standardization is performed on the initial reference period and the ideal original signal, respectively, which can be combined with formula four: (Formula 4) In Formula 4, This is the initial reference period after standardization. The ideal original signal after standardization. and These represent the initial reference period and the mean and standard deviation of the original signal, respectively.

[0051] S204 uses template matching to divide the calibration period. Specifically, it divides the ideal original signal... With template (initial reference period) Perform cross-correlation calculations to detect significant peak locations and calculate the order of intervals between adjacent peaks. This is to determine the starting point of two adjacent cycles. Multimodal analysis is used to select the main cycle length that best matches the target rotational speed. Optionally, this can also be used to estimate the actual rotational speed for verification. The specific formula is as follows: (Formula 5) S205, to initially screen out complete periodic segments, uses the first period as a benchmark and calculates the Pearson correlation coefficient between each of the remaining periods and the first period; only periods with correlation coefficients higher than a preset threshold are retained, while periods with abnormal morphology caused by local debris, noise, or transient interference are removed to further improve the consistency of periodic data, and finally output a highly consistent set of periods. .

[0052] It should be noted that the above process effectively solves problems such as unstable signal period, segment distortion, and speed drift, significantly improving the accuracy and robustness of period segmentation and laying a reliable data foundation for high-precision chip detection.

[0053] S206, for each calibration cycle For sequences considered to contain outliers (grooves), a robust RANSAC fitting method is used to identify the groove location signals, and interpolation methods, such as spline functions, can be applied. The compensation signal for the notch position is fitted, i.e., the notch region in the original period is replaced to obtain the compensated reference period. The specific formula is as follows: (Formula 6) in, The signal corresponding to the i-th period, The signal is the result of fitting with a spline function. This is the signal of the i-th cycle after compensation processing.

[0054] S207, calculate the calibrated oscillation amount for each cycle, using the following formula: (Formula 7) in, To calibrate the oscillation, This is the signal of the i-th cycle after compensation processing.

[0055] S208, Calculate the reference value of the oscillation during the calibration phase. The calculation formula is as follows: (Formula 8) in, The median calibration period is calculated at each sampling point n to obtain the median calibration period.

[0056] For example, .in, To determine the median calibration period within the calibration period after compensation processing.

[0057] S209. Based on the oscillation reference value during the calibration phase and the calibration oscillation amount corresponding to each calibration cycle, determine the oscillation rate threshold. The calculation formulas are shown in Formulas 9 and 10. (Formula 9) (Formula 10) In Formula Nine, The rate of change of oscillation for each calibration period is also the relative degree of deformation for each period. Furthermore, in Formula 10, The oscillation rate threshold, To determine the mean of the rate of change of the calibrated periodic oscillation, To calibrate the standard deviation of the rate of change of the periodic oscillation, This is a reference coefficient, with an empirical value of 3, which can be adjusted flexibly. It should be noted that distribution statistics are used to adaptively update the threshold, for example, when the threshold is sensitive to chipping. It can be set to 0.1; if a certain level of chipping is tolerated, the threshold is adjusted. The value can be relaxed to 0.2, which is empirically derived and can ensure field robustness and a low false alarm rate.

[0058] The next steps in the detection phase are as follows: Figure 3 As shown, S301 acquires the eddy current signal of the tool holder under test, and S302 obtains the standardized initial reference period through step S203. Multiple detection cycles are obtained by dividing the cycle using template matching. The steps for abnormal cycle filtering (S303) and groove compensation (S304) are the same as those for S205 and S206, and will not be repeated here. This process yields the compensated detection cycle. .

[0059] S305, calculate the peak-to-peak value of each post-filter detection cycle as the detection oscillation amount, as shown in Formula 11: (Formula Eleven) In Formula 11, To detect the amount of runout, which is the compensated detection cycle. The peak value.

[0060] S306, based on the detected oscillation amount and the oscillation reference value during the calibration phase, calculate the rate of change of detected oscillation for each cycle, as shown in Formula Twelve: (Formula 12) In formula twelve, To detect the rate of change of oscillation, To detect the amount of oscillation, This is the midpoint calibration oscillation amount, which is also the oscillation reference value during the calibration stage.

[0061] S307, determine the proportion of detection cycles in which the oscillation rate exceeds the preset oscillation rate threshold to the total number of cycles, in order to determine the chip jamming situation.

[0062] The chip detection method based on an eddy current sensor provided by any of the above embodiments has at least the following technical effects: (1) A chip detection paradigm based on eddy current sensing and periodic signal modeling By leveraging the high sensitivity, non-contact, and anti-contamination characteristics of eddy current sensors for conductive metal foreign objects, and combining this with the periodicity of the tool holder's rotation, a space-time aligned signal analysis framework is constructed to achieve direct and accurate identification of stuck chips.

[0063] (2) An integrated calibration method of "groove self-identification-compensation-calibration modeling" is proposed. The innovative RANSAC robust regression model is introduced to automatically distinguish between structural grooves (outer points) and normal responses (inner points) during the calibration stage, generate a smooth reference period after groove compensation, and calculate the point-by-point standard deviation based on the compensated signal to construct a calibration model that is consistent throughout the circumference, thereby fundamentally eliminating structural interference and avoiding detection blind spots.

[0064] (3) Design an adaptive anomaly criterion mechanism guided by oscillation amount An adaptive detection strategy based on oscillation amplitude is proposed, which can effectively identify overall deformation-type chipping. This mechanism significantly improves the detection coverage and robustness.

[0065] (4) Full-process automation, seamlessly integrated into the automatic tool change (ATC) process. From signal acquisition, cycle segmentation, groove processing to anomaly detection, the entire process requires no manual intervention, with processing latency lower than the interval between tool changes. It can be embedded into the CNC system for real-time execution, truly meeting the core requirements of unmanned factories for high cycle time, high reliability, and online detection.

[0066] Based on specific experimental data, this embodiment, through RANSAC groove compensation and calibration model modeling, achieves a chip detection accuracy of over 98% (actual measured data), with a false negative rate of less than 1% and a false alarm rate of less than 2%, far superior to manual visual inspection and traditional vision solutions. The chip detection criterion mechanism balances sensitivity and specificity: it can detect fine filamentous aluminum chips with a jump of more than 1 mil. It provides timely warnings before chip jamming causes spindle vibration, tool loosening, or machining deviations, preventing safety accidents such as tool fly-out and spindle damage, extending the service life of the spindle and tool holder by more than 30%, and reducing unplanned downtime by more than 50 hours per year.

[0067] Furthermore, as Figures 1 to 3 The specific implementation of the method shown in this embodiment provides a chip detection device based on an eddy current sensor, such as... Figure 4 As shown, the device includes: a data acquisition unit 401, a data division unit 402, a data calculation unit 403, and a data detection unit 404.

[0068] The acquisition unit 401 is configured to acquire in real time the eddy current signal generated when the tool holder under test rotates, wherein the eddy current signal is used to characterize the distance between the eddy current sensor probe and the conical surface of the tool holder; The segmentation unit 402 is configured to segment the eddy current signal using a period segmentation strategy to obtain eddy current detection information for multiple detection periods. The calculation unit 403 is configured to calculate the detection oscillation change rate for each detection cycle based on the eddy current detection information corresponding to each detection cycle and the oscillation reference value obtained during the calibration stage; the detection oscillation change rate is used to characterize the degree of deviation between the eddy current signal of the tool holder under test and the eddy current signal under chip-free conditions. The detection unit 404 is configured to determine the number of detection cycles in which the detection oscillation rate exceeds a preset oscillation rate threshold, so as to determine whether there is chip jamming in the tool holder under test.

[0069] In a specific application scenario, the segmentation unit 402 is further configured to acquire the target rotational speed of the tool holder under test and the sampling frequency of the eddy current sensor; calculate the theoretical cycle length based on the target rotational speed and the sampling frequency; extract a pre-acquired segment of the chip-free eddy current signal based on the theoretical cycle length to obtain an initial reference cycle; and use the initial reference cycle as a template to segment the eddy current signal using a template matching method to obtain eddy current detection information for multiple detection cycles.

[0070] In a specific application scenario, the partitioning unit 402 is further configured to calculate the cross-correlation sequence between the initial reference period and the eddy current signal, and identify multiple peaks in the cross-correlation sequence; take the position of each peak as the starting point of a detection period, and extract the corresponding signal segment from the eddy current signal based on adjacent starting points, so as to obtain the eddy current detection information of the multiple detection periods.

[0071] In a specific application scenario, the partitioning unit 402 is further configured to calculate the Pearson correlation coefficient between each detection cycle and the initial reference cycle; remove detection cycles with correlation coefficients lower than a preset correlation threshold; use the RANSAC robust regression algorithm to identify the eddy current detection information corresponding to each removed detection cycle to obtain each region to be compensated; and use an interpolation method to compensate each region to be compensated.

[0072] In specific application scenarios, the calculation unit 403 is further configured to calculate the peak-to-peak value of the eddy current detection information corresponding to each compensated detection cycle, and obtain the detection oscillation amount corresponding to each detection cycle; based on the detection oscillation amount corresponding to each detection cycle and the oscillation reference value obtained in the calibration stage, the detection oscillation change rate of each detection cycle is calculated.

[0073] In specific application scenarios, the computing unit 403 is further configured to employ a periodic segmentation strategy to periodically divide a pre-acquired segment of a chip-free eddy current signal, obtaining multiple calibration periods of chip-free eddy current signals; employ the RANSAC robust regression algorithm to identify the multiple calibration periods of the chip-free eddy current signals, obtaining the compensation region corresponding to each calibration period; employ an interpolation method to compensate the compensation region corresponding to each calibration period, obtaining a compensated set of chip-free eddy current signals; calculate the peak-to-peak value of each calibration period based on each calibration period in the set of chip-free eddy current signals, obtaining the calibration oscillation amount corresponding to each calibration period; and calculate the median calibration period based on the set of chip-free eddy current signals, determining the calibration oscillation amount of the median calibration period as the oscillation reference value obtained in the calibration stage.

[0074] In a specific application scenario, the detection unit 404 is further configured to calculate the proportion of the number of detection cycles in which the detection oscillation rate exceeds the preset oscillation rate threshold to the total number of cycles; if the proportion exceeds the preset proportion, it is determined that the tool holder under test has chips stuck; wherein, the preset oscillation rate threshold is determined based on the calibration oscillation rate, which is calculated based on the calibration oscillation amount of each calibration cycle in the calibration stage and the oscillation reference value obtained in the calibration stage.

[0075] It should be noted that other corresponding descriptions of the functional units involved in the chip detection device based on an eddy current sensor provided in this embodiment can be found in [reference needed]. Figures 1 to 3 The corresponding description in [the document] will not be repeated here.

[0076] Based on the above, Figures 1 to 3 Accordingly, this embodiment also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-described method. Figures 1 to 3 The method shown.

[0077] Based on this understanding, the technical solution of this application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (such as CD-ROM, USB flash drive, mobile hard drive, etc.) and includes several instructions to cause a computer device (such as personal computer, server, or network device, etc.) to execute the methods of various implementation scenarios of this application.

[0078] Based on the above, Figures 1 to 3 The method shown, and Figure 4 To achieve the above objectives, this application also provides an electronic device, which can be configured on a computer side, etc. The device includes a storage medium and a processor; the storage medium is used to store a computer program; the processor is used to execute the computer program to achieve the above-described objectives. Figures 1 to 3 The method shown.

[0079] Optionally, the aforementioned physical devices may also include a user interface, a network interface, a camera, radio frequency (RF) circuitry, sensors, audio circuitry, a Wi-Fi module, etc. The user interface may include a display screen, input units such as a keyboard, etc., and optional user interfaces may also include USB interfaces, card reader interfaces, etc. The network interface may optionally include standard wired interfaces, wireless interfaces (such as Wi-Fi interfaces), etc.

[0080] Those skilled in the art will understand that the physical device structure provided in this embodiment does not constitute a limitation on the physical device, and may include more or fewer components, or combine certain components, or have different component arrangements.

[0081] The storage medium may also include an operating system and a network communication module. The operating system is a program that manages the hardware and software resources of the aforementioned physical device, supporting the operation of information processing programs and other software and / or programs. The network communication module is used to enable communication between the various components within the storage medium, as well as communication with other hardware and software in the information processing physical device.

[0082] Through the above description of the embodiments, those skilled in the art can clearly understand that this application can be implemented by means of software plus necessary general-purpose hardware platforms, or it can be implemented by hardware. By applying the solution of this embodiment, compared with related technologies, a chip detection method based on an eddy current sensor is provided. First, the detection period of the eddy current signal is strictly divided by a period segmentation strategy. Then, the eddy current detection information of each detection period is compared with the oscillation reference value representing the chip-free condition obtained in the calibration stage to calculate the detection oscillation change rate, which is the degree of deviation between the eddy current signal used for the tool holder under test and the eddy current signal under the chip-free condition. By detecting the oscillation change rate, the chip characteristics in the detection stage are compared with the waveform of the eddy current signal under the chip-free condition, highlighting the deviation caused by the abnormal part of the chip, thus improving the detection accuracy. At the same time, when judging whether there is chip, the proportion of cycles with the detection oscillation change rate exceeding the limit is used as a criterion, effectively avoiding misjudgment caused by random noise or instantaneous interference in a single cycle, thus improving the robustness of detection.

[0083] In the description of this application, it should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0084] In the description of this specification, the terms "one embodiment," "some embodiments," "embodiment," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0085] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make modifications, alterations, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A chip detection method based on an eddy current sensor, characterized in that, include: The eddy current signal generated when the tool holder under test rotates is acquired in real time, wherein the eddy current signal is used to characterize the distance between the eddy current sensor probe and the conical surface of the tool holder; The eddy current signal is divided using a period segmentation strategy to obtain eddy current detection information for multiple detection periods; Based on the eddy current detection information corresponding to each detection cycle and the oscillation reference value obtained during the calibration stage, the detection oscillation change rate for each detection cycle is calculated; the detection oscillation change rate is used to characterize the degree of deviation between the eddy current signal of the tool holder under test and the eddy current signal under chip-free conditions. The number of detection cycles in which the detected oscillation rate exceeds a preset oscillation rate threshold is determined in order to determine whether the tool holder under test has chips stuck.

2. The method according to claim 1, characterized in that, The eddy current signal is divided using a periodic segmentation strategy, including: The target rotational speed of the tool holder under test and the sampling frequency of the eddy current sensor are obtained; Calculate the theoretical period length based on the target rotational speed and the sampling frequency; The initial reference period is obtained by truncating a pre-acquired segment of a chip-free eddy current signal based on the theoretical period length. Using the initial reference period as a template, the eddy current signal is divided by template matching to obtain eddy current detection information for multiple detection periods.

3. The method according to claim 2, characterized in that, Using the initial reference period as a template, the eddy current signal is divided using a template matching method, including: Calculate the cross-correlation sequence between the initial reference period and the eddy current signal, and identify multiple peaks in the cross-correlation sequence; The position of each peak is taken as the starting point of a detection cycle, and based on adjacent starting points, corresponding signal segments are extracted from the eddy current signal to obtain eddy current detection information for the multiple detection cycles.

4. The method according to claim 3, characterized in that, After dividing the eddy current signal using template matching, the method further includes: Calculate the Pearson correlation coefficient between each detection cycle and the initial reference cycle; Detection cycles with correlation coefficients lower than a preset correlation threshold are excluded. The RANSAC robust regression algorithm is used to identify the eddy current detection information corresponding to each detection cycle after the elimination is completed, and to obtain each region to be compensated. Interpolation methods are used to compensate for each region to be compensated.

5. The method according to claim 4, characterized in that, The step of calculating the rate of change of vibration in each detection cycle based on the eddy current detection information corresponding to each detection cycle and the vibration reference value obtained during the calibration phase includes: Calculate the peak-to-peak value of the eddy current detection information corresponding to each compensated detection cycle to obtain the detection oscillation amount corresponding to each detection cycle. Based on the detection oscillation amount corresponding to each detection cycle and the oscillation reference value obtained during the calibration stage, the detection oscillation change rate for each detection cycle is calculated.

6. The method according to claim 1, characterized in that, The pendulum reference value obtained during the calibration phase is based on the following steps: A periodic segmentation strategy is adopted to periodically divide a pre-acquired segment of a chip-free eddy current signal to obtain multiple calibration periods of chip-free eddy current signals. The RANSAC robust regression algorithm is used to identify the chip-free eddy current signals of the multiple calibration cycles, and the compensation area corresponding to each calibration cycle is obtained. An interpolation method is used to compensate the area to be compensated for each calibration cycle, resulting in a set of chip-free eddy current signals after compensation. Based on each calibration period in the set of chip-free eddy current signals, the peak-to-peak value of each calibration period is calculated to obtain the calibration oscillation amount corresponding to each calibration period; Based on the set of chip-free eddy current signals, the median calibration period is calculated, and the calibration oscillation amount of the median calibration period is determined as the oscillation reference value obtained in the calibration stage.

7. The method according to claim 6, characterized in that, Determining whether the tool holder under test has stuck chips includes: Calculate the proportion of the total number of detection cycles in which the detected oscillation rate exceeds the preset oscillation rate threshold. If the ratio exceeds the preset ratio, it is determined that the tool holder under test has chips stuck in it; The preset oscillation rate threshold is determined based on the calibrated oscillation rate, which is calculated based on the calibrated oscillation amount of each calibration cycle in the calibration phase and the oscillation reference value obtained in the calibration phase.

8. A chip detection device based on an eddy current sensor, characterized in that, include: The acquisition unit is configured to acquire in real time the eddy current signal generated when the tool holder under test rotates, wherein the eddy current signal is used to characterize the distance between the eddy current sensor probe and the conical surface of the tool holder; The segmentation unit is configured to segment the eddy current signal using a periodic segmentation strategy to obtain eddy current detection information for multiple detection periods. The calculation unit is configured to calculate the detection oscillation change rate for each detection cycle based on the eddy current detection information corresponding to each detection cycle and the oscillation reference value obtained during the calibration stage; the detection oscillation change rate is used to characterize the degree of deviation between the eddy current signal of the tool holder under test and the eddy current signal under chip-free conditions. The detection unit is configured to determine the number of detection cycles in which the rate of change of the detected oscillation exceeds a preset oscillation rate threshold, so as to determine whether there is chip jamming in the tool holder under test.

9. An electronic device, comprising a storage medium, a processor, and a computer program stored on the storage medium and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method of any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1 to 7.

Images