Rapid milling chatter detection method based on correlation coefficient

Abstract

The invention discloses a rapid milling chatter detection method based on correlation coefficients, which comprises the following steps of: determining the mounting positions of a vibration sensor and a rotating speed sensor based on a processing workpiece, and determining the data volume acquired per revolution of a main shaft, the data length of a time window and a specified chatter threshold value; acquiring periodic synchronous multi-point sampling data; standardizing the sampled data; expanding the data into multi-dimensional data; calculating a correlation coefficient between adjacent data in the multi-dimensional data to obtain coefficient data, and calculating an average value of the coefficient data; and determining whether the chatter is generated based on the average value and a chatter threshold value. According to the method, the average correlation coefficient of two adjacent rows of data after expansion standardization is used as a fault index, the influence of signal noise on a flutter index is reduced, and the misdiagnosis rate and missed diagnosis rate of flutter are further reduced.

Description

Technical Field

[0001] The invention relates to the technical field of state detection, and in particular to a rapid milling chatter detection method based on correlation coefficient. Background Art

[0002] Milling is an advanced manufacturing technology with the characteristics of high efficiency, high precision and high surface quality. Milling has been widely used in mold manufacturing, aerospace and automobile manufacturing. In the milling process, inappropriate combination of processing parameters usually causes chatter. Chatter is a common abnormal milling vibration phenomenon. Chatter not only affects the processing quality and efficiency of the workpiece, but also accelerates tool wear and even causes damage to the machine tool. Therefore, in order to improve the quality and efficiency of milling, it is necessary to perform chatter detection on the milling process so as to detect chatter early and prevent the harm caused by chatter.

[0003] At present, researchers and engineers have proposed a variety of chatter detection technologies. The invention patent with application number 201910656987.5 discloses a chatter detection method based on signal processing and artificial intelligence. The chatter detection method first uses the empirical mode decomposition method to decompose the chatter signal into a set of intrinsic mode function signals, and then selects several intrinsic mode function signals containing more chatter information for analysis, extracts chatter features, and constructs a multi-dimensional chatter feature observation space. Then, the manifold learning algorithm is used to reduce the dimension of the multi-dimensional chatter feature observation space, and then the information of multiple chatter features is integrated, and finally the chatter is identified using a support vector machine. Although the above chatter detection methods can reliably detect chatter, the computational cost is relatively high and may not be suitable for real-time detection of high-speed milling chatter. The invention patent with application number 202210619937.1 discloses a fast and reliable milling chatter detection method, which uses a vibration sensor to obtain multiple sets of periodic synchronous single-point sampling data within each rotation period of the spindle, and fuses the chatter information collected by multiple vibration sensors, thereby shortening the data collection time, increasing the amount of data collected, and improving the real-time performance of chatter detection. Although this method can realize real-time chatter detection, the chatter index of this method is easily affected by noise, resulting in relatively low reliability of chatter detection. In addition, this method uses multiple vibration sensors, which not only increases the cost of chatter detection, but also increases the false alarm rate of chatter detection due to sensor failure. Summary of the invention

[0004] Purpose of the invention: In order to overcome the shortcomings of the background technology, the present invention requests to disclose a rapid milling chatter detection method based on correlation coefficient. The method adopts the average correlation coefficient of two adjacent rows of data after expansion and normalization as the fault indicator, which reduces the influence of signal noise on the chatter indicator, thereby reducing the false alarm rate and missed alarm rate of chatter detection.

[0005] Technical solution: The rapid milling chatter detection method based on correlation coefficient disclosed in the present invention comprises the following steps:

[0006] S1. Determine the installation positions of the vibration sensor and the speed sensor based on the workpiece to be processed, determine the amount of data collected per spindle revolution, the data length of the time window, and specify the chatter threshold;

[0007] S2, obtaining periodic synchronous multi-point sampling data;

[0008] S3, periodic synchronous multi-point sampling data obtained by standardization processing;

[0009] S4, expanding the standardized periodic synchronous multi-point sampling data into multi-dimensional data;

[0010] S5, calculating the correlation coefficient between adjacent data in the multidimensional data to obtain coefficient data, and calculating the average value of the coefficient data;

[0011] S6. Determine whether chattering occurs based on the average value and the chattering threshold.

[0012] Furthermore, the vibration sensor in S1 is an acceleration sensor, a velocity sensor or a displacement sensor.

[0013] Furthermore, the workpiece processed in S1 is a non-thin-wall workpiece or a thin-wall workpiece;

[0014] For milling of non-thin-walled workpieces, vibration sensors and speed sensors are installed on the high-speed spindle to obtain vibration and speed data; for milling of thin-walled workpieces, vibration sensors are installed on thin-walled workpieces or high-speed spindles to obtain vibration data; speed sensors are installed on high-speed spindles to obtain speed data, and the number of spindle revolutions within a time window is m.

[0015] Furthermore, the periodic synchronous multi-point sampling data X in S2 is expressed as:

[0016] X=[x(1) x(2) x(3) … x(i) … x(nm)];

[0017] Where: nm represents the data length of the time window.

[0018] Furthermore, in S3, the Z-score standardization method is used to standardize the acquired periodic synchronous sampling data X, and the processed data is expressed as:

[0019]

[0020] In the formula, represents the mean value of data X, and σ represents the standard deviation of data X.

[0021] Further, S4 is specifically: according to the amount of data n collected in each revolution, the standardized periodic synchronous multi-point sampling data X' is expanded into n-dimensional data Y, and the expanded n-dimensional data Y is expressed as:

[0022]

[0023] make The expanded n-dimensional data Y is expressed as:

[0024]

[0025] Furthermore, S5 is specifically as follows: calculating the adjacent data Y in the n-dimensional data Y i and data Y i+1 The correlation coefficient r(i) is used to obtain data C, and the mean value M of data C is calculated;

[0026] Data C is represented by:

[0027] C=[r(1) r(2) r(3) … r(i) … r(n-1)];

[0028] Among them, data Y i and data Y i+1 The correlation coefficient r(i) is expressed as:

[0029]

[0030] In the formula, cov(Y i ,Y i+1 ) is the data Y i and data Y i+1 The covariance of Yi and σ Yi+1 Represents data Y i and data Y i+1 The standard deviation of

[0031] The mean value M of data C is expressed as:

[0032]

[0033] Further, S6 is specifically as follows: if the average value is greater than or equal to the chatter threshold, it is determined that chatter has occurred; otherwise, it is considered that chatter has not occurred, and the process returns to S2 to continue chatter detection.

[0034] Beneficial effects: Compared with the prior art, the advantages of the present invention are:

[0035] 1. The method of the present invention has low calculation cost and can adopt a control unit with a slower calculation speed, thereby reducing the cost of the control unit; and the detection time delay is short, so the chatter can be detected quickly.

[0036] 2. The chatter index used in the method of the present invention is not affected by the noise in the signal, which reduces the false alarm rate and missed alarm rate of chatter and improves the reliability of chatter detection.

[0037] 3. The method of the present invention can be applied to a variable speed milling system. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 is a flow chart of the chatter detection method of the present invention;

[0039] Figure 2 It is the displacement data without chatter and the periodic synchronous multi-point sampling displacement data;

[0040] Figure 3 The displacement data when there is vibration and the periodic synchronous multi-point sampling displacement data;

[0041] Figure 4 It is the periodic synchronous multi-point sampling displacement data after normalization when there is no chatter;

[0042] Figure 5 It is the periodic synchronous multi-point sampling displacement data after normalization when there is chatter;

[0043] Figure 6 The graphs are the mean standard deviation and the mean correlation coefficient as the noise changes. DETAILED DESCRIPTION

[0044] like Figure 1 The rapid milling chatter detection method based on correlation coefficient shown includes the following steps:

[0045] Step 1: Based on the workpiece to be processed, determine the amount of data n collected synchronously per spindle revolution, the data length nm of the time window, specify the vibration threshold T, and select the vibration sensor and speed sensor and their installation locations.

[0046] The amount of data collected synchronously per spindle revolution is determined to be n=100, the data length of each time window is nm=1000, the vibration detection threshold T is specified to be 0.01, and the selected vibration sensor is a displacement sensor, which can also be an acceleration sensor and a velocity sensor.

[0047] The installation position of the vibration sensor is selected according to the type of workpiece to be processed. When the workpiece is not a thin-walled workpiece, the vibration sensor and the speed sensor are installed on the spindle of the milling machine to measure the spindle vibration and speed. For the milling of thin-walled workpieces, the displacement sensor can be installed on the thin-walled workpiece or the high-speed spindle to obtain vibration data. In addition, the speed sensor is installed on the spindle to obtain the speed data; m can also be regarded as the number of spindle revolutions within a time window. The workpiece to be processed in this embodiment is selected as a thin-walled workpiece. Therefore, the vibration sensor can be installed on the thin-walled workpiece or on the spindle. In this embodiment, the displacement sensor is installed on the thin-walled workpiece to measure the displacement of the thin-walled workpiece perpendicular to the processing direction.

[0048] Step 2: Obtain periodic synchronous multi-point sampling data.

[0049] According to the vibration sensor and the rotation speed sensor described in step 1, periodic synchronous multi-point sampling data is obtained, and the data can be expressed as:

[0050] X=[x(1) x(2) x(3) … x(100) … x(1000)].

[0051] In this embodiment, the displacement sensor installed on the thin-walled workpiece is used to measure the vibration displacement signal of the thin-walled workpiece perpendicular to the processing direction. The speed sensor installed on the spindle is used to collect spindle speed information. The displacement data without chatter and the periodic synchronous multi-point displacement data are as follows: Figure 2 The displacement data during vibration and the periodic synchronous multi-point displacement data are shown in Figure 3 shown.

[0052] Step 3: Standardize the acquired periodic synchronous multi-point sampling data X.

[0053] The Z-score standardization method is used to standardize the acquired periodic synchronous multi-point sampling data X. The standardized data can be expressed as:

[0054]

[0055] In the formula, represents the mean value of data X, and σ represents the standard deviation of data X.

[0056] In this embodiment, the periodic synchronous multi-point sampling signal after the standardized processing without chattering is as follows: Figure 4 The periodic synchronous multi-point sampling signal after the standardized processing during the vibration is as shown in Figure 5 As shown in the figure, through standardization, the periodic synchronous multi-point sampling data is standardized to a uniform scale. This makes it easier to perform subsequent vibration detection.

[0057] Step 4: Expand the periodically synchronized multi-point sampling data after the extended normalization process into multi-dimensional data Y.

[0058] According to the amount of data collected per revolution of 100, the standardized periodic synchronous multi-point sampling data is expanded into 100-dimensional data. The expanded 100-dimensional data Y can be expressed as:

[0059]

[0060] For the convenience of expression, The expanded 100-dimensional data Y is expressed as:

[0061]

[0062] The purpose of extending the periodic synchronous multi-point sampling data is to calculate the correlation coefficient later. The periodic synchronous multi-point sampling data after extension and standardization can well display the chatter information. Compared with the periodic synchronous single-point sampling data, the periodic synchronous multi-point sampling data has more data volume and contains more chatter information.

[0063] In this embodiment, since 100 data are collected in each rotation, the acquired periodic synchronous multi-point sampling data can be expanded into a group of 100 rows and 10 columns of data.

[0064] Step 5: Calculate the adjacent data Y in the 100-dimensional data Y i and data Y i+1 The correlation coefficient r(i) is used to obtain data C, and the mean value M of data C is calculated.

[0065] In step 5, the adjacent data Y in the 100-dimensional data Y is calculated i and data Y i+1 The data C obtained by the correlation coefficient can be expressed as:

[0066] C=[r(1) r(2) r(3) … r(i) … r(99)],

[0067] Among them, data Y i and data Y i+1 The correlation coefficient r(i) can be expressed as:

[0068]

[0069] In the formula, cov(Y i ,Y i+1 ) is the data Y i and data Y i+1 The covariance of Yi and σ Yi+1 Represents data Y i and data Yi+1 The standard deviation of .

[0070] The mean value M of data C can be expressed as:

[0071]

[0072] Step 6: Determine whether chattering occurs.

[0073] The calculated average value M is compared with a predetermined threshold value T. If M is greater than or equal to the predetermined threshold value T, it can be determined that chatter has occurred. Otherwise, it is considered that chatter has not occurred, and the process returns to step 2 to continue chatter detection.

[0074] It can be concluded that for the periodic synchronous multi-point sampling data when there is no chatter, the calculated average value M is equal to 0.004, which is less than the predetermined threshold value T=0.1, and the chatter detection program returns to the second step to continue the chatter detection; for the periodic synchronous multi-point sampling data when there is chatter, the calculated average value M is equal to 0.25, which is greater than the predetermined threshold value T=0.1, and it can be determined that chatter has occurred.

[0075] It should be pointed out that the average correlation coefficient used in the invention patent is not affected by disturbance noise, while the standard deviation used in the invention patent with application number 202210619937.1 is greatly affected by disturbance noise. Figure 6 The influence of disturbance noise on the average correlation coefficient and the average standard deviation is given. It can be seen from the figure that the average correlation coefficient has good disturbance noise robustness, which is significantly better than the average standard deviation.

Claims

1. A rapid milling chatter detection method based on correlation coefficient, characterized in that: The following steps are involved: S1. Determine the installation positions of the vibration sensor and the speed sensor based on the workpiece to be processed, determine the amount of data collected per spindle revolution, the data length of the time window, and specify the chatter threshold; S2, obtaining periodic synchronous multi-point sampling data; S3, periodic synchronous multi-point sampling data obtained by standardization processing; S4, expanding the standardized periodic synchronous multi-point sampling data into multi-dimensional data; S5, calculating the correlation coefficient between adjacent data in the multidimensional data to obtain coefficient data, and calculating the average value of the coefficient data; S6. Determine whether chattering occurs based on the average value and the chattering threshold.

2. The rapid milling chatter detection method based on correlation coefficient according to claim 1, characterized in that: The vibration sensor in S1 is an acceleration sensor, a velocity sensor or a displacement sensor.

3. The rapid milling chatter detection method based on correlation coefficient according to claim 1, characterized in that: The workpiece processed in S1 is a non-thin-wall workpiece or a thin-wall workpiece; For milling of non-thin-walled workpieces, vibration sensors and speed sensors are installed on the high-speed spindle to obtain vibration and speed data; for milling of thin-walled workpieces, vibration sensors are installed on thin-walled workpieces or high-speed spindles to obtain vibration data; speed sensors are installed on high-speed spindles to obtain speed data, and the number of spindle revolutions within a time window is m.

4. The rapid milling chatter detection method based on correlation coefficient according to claim 1, characterized in that: The periodic synchronous multi-point sampling data X in S2 is expressed as: X=[x(1) x(2) x(3) … x(i) … x(nm)]; Where: nm represents the data length of the time window.

5. The rapid milling chatter detection method based on correlation coefficient according to claim 1, characterized in that: In S3, the Z-score standardization method is used to standardize the acquired periodic synchronous sampling data X. The processed data is expressed as: In the formula, represents the mean value of data X, and σ represents the standard deviation of data X.

6. The rapid milling chatter detection method based on correlation coefficient according to claim 1, characterized in that: S4 is specifically: according to the amount of data n collected in each revolution, the standardized periodic synchronous multi-point sampling data X' is expanded into n-dimensional data Y, and the expanded n-dimensional data Y is expressed as: make The expanded n-dimensional data Y is expressed as:

7. The rapid milling chatter detection method based on correlation coefficient according to claim 1, characterized in that: S5 is specifically: calculating the adjacent data Y in the n-dimensional data Y i and data Y i+1 The correlation coefficient r(i) is used to obtain data C, and the mean value M of data C is calculated; Data C is represented by: C=[r(1) r(2) r(3) … r(i) … r(n-1)]; Among them, data Y i and data Y i+1 The correlation coefficient r(i) is expressed as: In the formula, cov(Y i ,Y i+1 ) is the data Y i and data Y i+1 The covariance of and Represents data Y i and data Y i+1 The standard deviation of The mean value M of data C is expressed as:

8. The rapid milling chatter detection method based on correlation coefficient according to claim 1, characterized in that: S6 is specifically as follows: if the average value is greater than or equal to the chatter threshold, it is determined that chatter has occurred; otherwise, it is considered that chatter has not occurred, and the process returns to S2 to continue chatter detection.

Images