TWI653410B - Diagnosis method of pre-pressure of ball transmission assembly - Google Patents

Abstract

The invention provides a method for diagnosing the pre-pressure of a ball transmission component, which includes the following steps: when the ball transmission component is actuated, a vibration sensing unit provided in the ball transmission component generates a vibration signal; by multiple signal parts of the vibration signal, respectively Perform time-frequency analysis to generate corresponding multiple test samples, where each test sample includes multiple frequency domain characteristics related to the ball pass frequency of the ball transmission assembly; calculate the similarity of the test samples with the pre-prepared multiple reference samples Degrees, where each reference sample includes the frequency domain characteristics corresponding to other ball transmission components with pre-pressure; when the similarity is greater than a predetermined threshold, it is determined that the ball transmission component has pre-pressure, otherwise it is determined that the ball transmission component is no longer Pre-pressure.

Description

Diagnosis method of pre-pressure of ball transmission assembly

The invention relates to a method for diagnosing the preload of a transmission component, in particular to a method for diagnosing the preload of a transmission component that is driven by balls.

The ball screw has the advantages of high positioning accuracy and long service life, so it has recently been widely used in machine tools that require precise positioning. Generally speaking, the ball screw is formed by screwing the nut through the rolling ball and the screw, and the nut moves linearly relative to the screw through the rolling of the ball in the nut.

However, the ball screw gradually generates backlash after being used for a period of time; relatively, the pre-pressure of the ball screw gradually decreases. If the backlash is too large and the ball screw does not have preload, the nut will easily vibrate back and forth during the reciprocating movement; this will accelerate the damage to the service life of the ball screw and will also reduce the positioning accuracy of the ball screw, so how It is an important issue to automatically diagnose whether the ball screw still has preload.

For example, Patent Publication No. I482919 of the Republic of China discloses a method for detecting the pre-pressure of a ball screw by obtaining the vibration signal of the ball screw during operation, and estimating the actual ball pass frequency order of the ball screw based on the vibration signal, and then borrowing Judging whether the ball screw still has the pre-pressure by comparing the difference between the estimated actual ball-pass frequency order and the theoretical ball-pass frequency order of the ball screw.

Therefore, the purpose of the present invention is to provide a method for diagnosing the pre-pressure of a ball transmission assembly.

Therefore, the method for diagnosing the pre-pressure of the ball transmission assembly of the present invention is implemented by a diagnosis system. The diagnosis system includes a calculation unit and a vibration sensing unit. The vibration sensing unit is disposed on the ball transmission assembly. The diagnosis method includes one step (a), one step (b), one step (c), one step (d) and one step (e).

The step (a) is that when the ball transmission assembly is actuated, the vibration sensing unit generates a vibration signal.

The step (b) is that the calculation unit generates corresponding test samples by performing time-frequency analysis on the signal parts of the vibration signal, and each test sample includes a plurality of test samples related to the ball transmission assembly. Frequency domain characteristics of the ball pass frequency.

In step (c), the calculation unit calculates a similarity between the test samples and a plurality of pre-prepared reference samples, where each reference sample includes the frequency domain characteristics corresponding to other ball transmission components with pre-pressure.

The step (d) is that when the calculation unit determines that the similarity is greater than a threshold value, the calculation unit determines that the ball transmission assembly has a pre-pressure.

The step (e) is that when the calculation unit determines that the similarity is not greater than the threshold value, the calculation unit determines that the ball transmission assembly has no preload.

The effect of the present invention lies in that it can automatically judge whether the ball screw has pre-pressure.

Referring to FIG. 1, the method for diagnosing the pre-pressure of the ball transmission assembly of the present invention is implemented by a diagnostic system. The diagnostic system includes a computing unit 7 and a vibration sensing unit 1 electrically connected to the computing unit 7. Preferably, the ball transmission component is a ball screw 2 and is provided on a machine tool 8, the vibration sensing unit 1 is an accelerometer and is provided on the nut 21 of the ball screw 2, and the calculation unit 7 is A computer device.

Referring to FIG. 2, the steps included in the method for diagnosing the pre-pressure of the ball transmission assembly of the present invention will be described in detail below.

In step 31, when the ball screw 2 is actuated, the vibration sensing unit 1 generates a vibration signal.

1 and FIG. 3 together, when the ball screw 2 is actuated, the nut 21 of the ball screw 2 moves back and forth between a first end 221 and a second end 222 of the screw 22. In the initial stage of moving from the first end 221 to the second end 222, the nut 21 moves toward the second end 222 with a first acceleration, so during this period a signal with a higher amplitude corresponding to the vibration signal 9 part S _11; then at the mid movement, the velocity nut 21 to move the second end 222, therefore correspond to the period of the vibration signal 9 has a lower amplitude portion of the signal S _12; then at the end of the movement, The nut 21 moves toward the second end 222 with a second acceleration opposite to the first acceleration in a direction, so during this period corresponds to a signal portion of the vibration signal 9 having a larger amplitude and a direction opposite to the signal portion S ₁₁ S ₁₃ .

Similarly, the nut 21 moves at the second acceleration, the constant velocity, and the first acceleration at the initial, middle, and final stages of the movement from the second end 222 to the first end 221, respectively. The signal portion S ₂₁ having a higher amplitude, the signal portion S ₂₂ having a lower amplitude, and the signal portion S ₂₃ having a higher amplitude should vibrate the signal 9.

Then in step 32, the calculation unit 7 captures a plurality of signal portions of the vibration signal 9 corresponding to the movement of the nut 21 at the same speed. For example, referring to Figure 3, by detecting a signal component corresponding to the peak point S ₁₁ P ₁₁ S ₁₃ corresponding to the valley portion of the signal point P _12, to define the point P ₁₁ is located between the peaks and valleys of the point P ₁₂ velocity of the nut 21 should be shifted signal portion S _12; Similarly, by detecting the valley portion ₂₁ a signal corresponding to the signal portion S P ₂₁ S ₂₃ corresponding to the peak point P _22, positioned to define between the signal portions ₂₁ and the valleys of peaks P ₂₂ P S _22.

Next, in step 33, for each signal portion captured in step 32, the calculation unit 7 filters the signal portion using a band pass filter, where the center frequency of the band pass filter This is the theoretical ball passing frequency of the ball screw 2.

Next, in step 34, for each signal portion that has been filtered, the calculation unit 7 generates a corresponding test sample by performing time-frequency analysis on the signal portion, where the test sample includes multiple The frequency domain characteristics generated by the time-frequency analysis and related to the ball passing frequency of the ball screw 2, wherein, referring to FIG. 1, the ball passing frequency is the return tube of the ball screw 2 per unit time 23 (return tube) The number of balls 24 passing through a certain position; preferably, the frequency domain features include the actual ball pass frequency order corresponding to the signal part, and the ball pass corresponding to the actual ball pass frequency order At least two of the vibration signal amplitude, the ball-pass vibration variation corresponding to the actual ball-pass frequency order, and the degree of dispersion of the ball-pass frequency order.

In detail, for a signal part corresponding to the constant speed movement of the nut 21, let the matrix [X _ti ] R ^{m × n} is the spectrum obtained after time-frequency analysis of the signal part, where t represents time, i represents frequency, and X _ti is the amplitude of frequency i at time t; the order corresponding to the spectrum The spectrum is [X _tj ] R ^{m × n} , where j = i / ω represents the frequency order, and ω is the rotation speed (revolutions / second) of the ball screw 2.

Referring to Fig. 4, for each frequency order j, calculate all corresponding frequency amplitudes, that is, {X _tj | t = 1,2,…, m}, the root mean square (RMS) r _j = And let h = r _j , then h is the estimated order of the actual spherical frequency, _{rh is} the amplitude of the vibration signal of the spherical pass, and the variance of {X _th | t = 1,2,…, m} is The variation of the ball pass vibration.

In addition, , And , Then the degree of dispersion of the ball pass frequency order is .

Next in step 35, the calculation unit 7 calculates the similarity between the test samples and a plurality of reference samples prepared in advance, where the reference samples and the test samples have the same frequency domain characteristics; for example, each test sample and each The reference samples are composed of the actual frequency order of the ball pass and the amplitude of the vibration signal of the ball pass.

Then in step 36, the calculation unit 7 determines whether the similarity between the test samples and the pre-prepared reference samples is greater than a predetermined threshold; when the determination result is positive, it is determined that the ball screw 2 has a pre-pressure ( Step 37); and when the judgment result is negative, it is determined that the ball screw 2 has no pre-pressure (Step 38).

In particular, the reference samples are generated in a similar way to the test samples, the only difference is that the reference samples are for other ball screws with pre-pressure, such as just shipped / new ball screws To perform the samples generated in steps 31 to 34 above. Although the reference samples need to be prepared in advance, the pre-preparation work only needs to be performed once; after the reference samples are available, the reference samples can be used to diagnose whether any ball screws that have been used for a period of time still have pre-pressure .

Regarding calculating the similarity between the test samples and the reference samples, in one embodiment, a first probability distribution of the test samples may be estimated from the test samples, and the reference samples may be estimated from the reference samples. Reference a second probability distribution of the sample, and then calculate the similarity of the first probability distribution and the second probability distribution as the similarity of the test samples and the reference samples. For example, both the first probability distribution and the second probability distribution are normal distributions, and the Bhattacharyya distance between the first probability distribution and the second probability distribution is calculated. Is smaller, the higher the similarity between the first probability distribution and the second probability distribution, and the higher the similarity between the test samples and the reference samples, that is, the opposite number of the Pap The similarity of the test samples and the reference samples.

Referring to FIG. 5, in another embodiment, k-means (k-means) clustering method, fuzzy C-means (FCM) clustering method or other existing clustering methods may be used to refer to these reference samples 4 Divide into multiple reference sample groups 5 in advance, then calculate the Pap distance of the test samples 6 and each reference sample group 5, and take the opposite of the average value of the Pap distances as the test samples 6 and the Wait for the similarity of reference sample 4.

Alternatively, the distance between each test sample 6 and each reference sample group 5 is calculated, and the opposite of the average value of the distances is used as the similarity between the test samples 6 and the reference samples 4. The distance between the test sample 6 and the reference sample group 5 may be the Euclidean distance of the group center of the test sample 6 and the reference sample group 5, or the test sample 6 and The smallest of the Euclidean distances of each reference sample 4 in the reference sample group 5.

In summary, the method for diagnosing the pre-pressure of the ball transmission assembly of the present invention generates the vibration signal by using the vibration sensing unit to sense the vibration of the ball transmission assembly, and the vibration signal corresponds to the ball transmission assembly, etc. Quickly actuating the signal parts to perform feature extraction to generate the test samples with a plurality of frequency domain characteristics related to the ball pass frequency of the ball transmission assembly, and calculate the test samples and the pre-prepared reference samples Based on the similarity, it can be judged whether the ball transmission assembly has a preload, so it can indeed achieve the purpose of the invention.

However, the above are only examples of the present invention, and should not be used to limit the scope of implementation of the present invention, any simple equivalent changes and modifications made in accordance with the scope of the patent application of the present invention and the contents of the patent specification are still Within the scope of the invention patent.

1‧‧‧Vibration sensing unit

2‧‧‧ball screw

21‧‧‧ Nut

22‧‧‧screw

221‧‧‧The first end

222‧‧‧Second end

23‧‧‧Return tube

24‧‧‧ball

31 ~ 38‧‧‧Step

4‧‧‧Reference sample

5‧‧‧Reference sample group

6‧‧‧Test sample

7‧‧‧Calculation unit

8‧‧‧Tool machine

9‧‧‧ Vibration signal

S ₁₁ ~ S ₁₃ ‧‧‧ Signal part of vibration signal

S ₂₁ ~ S ₂₃ ‧‧‧ Signal part of vibration signal

P ₁₁ ‧‧‧ Peak point of vibration signal

P ₁₂ ‧‧‧ Valley of vibration signal

P ₂₁ ‧‧‧ Valley of vibration signal

P ₂₂ ‧‧‧Peak point of vibration signal

h‧‧‧ Actual pass frequency order

r _h ‧‧‧Ball vibration signal amplitude

Other features and functions of the present invention will be clearly presented in the embodiment with reference to the drawings, in which: FIG. 1 is a schematic diagram illustrating the placement of a vibration sensing unit on the nut of a ball screw; FIG. 2 is A flowchart illustrating the steps involved in the pre-pressure diagnosis method of the ball transmission assembly of the present invention; FIG. 3 is a schematic diagram illustrating a vibration signal generated by the vibration sensing unit; FIG. 4 is a schematic diagram illustrating a corresponding order spectrum The frequency order of each frequency and its corresponding frequency amplitude root mean square; and Figure 5 is a schematic diagram illustrating the division of multiple reference samples into three reference sample groups.

Claims (9)

A method for diagnosing the pre-pressure of a ball transmission assembly is implemented by a diagnosis system. The diagnosis system includes a calculation unit and a vibration sensing unit. The vibration-sensing unit is disposed on the ball transmission assembly. The method for pre-pressure diagnosis of the ball transmission assembly It includes the following steps: (a) When the ball drive assembly is actuated, the vibration sensing unit generates a vibration signal; (b) The calculation unit is generated by performing time-frequency analysis on multiple signal parts of the vibration signal, respectively Corresponding multiple test samples, where each test sample includes a plurality of frequency domain characteristics related to the ball pass frequency of the ball transmission assembly; (c) the calculation unit calculates the test samples and the pre-prepared reference samples A first similarity, where each reference sample includes the frequency domain characteristics corresponding to other ball transmission components with preload; (d) when the calculation unit determines that the first similarity is greater than a threshold, the calculation The unit determines that the ball transmission assembly has a preload; and (e) when the calculation unit determines that the first similarity is not greater than the When the threshold value, the computing unit determines that the ball has no drive assembly preload. The method for diagnosing the pre-pressure of the ball transmission assembly according to claim 1, wherein in this step (b), for each signal part, the calculation unit first filters the signal part using a band-pass filter and then filters the The signal part of is subjected to time-frequency analysis, where the center frequency of the band-pass filter is the theoretical ball-pass frequency of the ball drive assembly. The method for diagnosing the pre-pressure of the ball transmission assembly according to claim 1, wherein in the step (b), each signal part is extracted from a part of the vibration signal corresponding to the ball transmission assembly operating at a constant speed. The method for diagnosing the pre-pressure of the ball transmission assembly according to claim 1, wherein in the step (b), the frequency domain characteristics include the actual ball pass frequency order and the amplitude of the ball pass vibration signal corresponding to the actual ball pass frequency order . At least two of the variation of the ball pass vibration corresponding to the actual frequency order of the ball pass and the degree of dispersion of the order of the ball pass frequency. The method for diagnosing the pre-pressure of the ball transmission assembly according to claim 1, wherein in step (c), the reference samples have been previously divided into a plurality of reference sample groups, and for each reference sample group, the calculation unit calculates the Wait for a second similarity between the test sample and the reference sample group, and calculate the first similarity based on the second similarity. The method for diagnosing the pre-pressure of the ball transmission assembly according to claim 5, wherein in step (c), the calculation unit estimates a first probability distribution of the test samples based on the test samples, and for each reference sample group , Estimate a second probability distribution of the reference sample group according to the reference sample group, and calculate the similarity of the first probability distribution and the second probability distribution as the second similarity of the test sample and the reference sample group degree. The method for diagnosing the pre-pressure of the ball transmission assembly according to claim 1, wherein in step (c), the reference samples have been divided into at least one reference sample group, and for each test sample, the calculation unit calculates the test sample The distance to each reference sample group, and the similarity between the test samples and the reference samples is calculated according to the distances. The method for diagnosing the pre-pressure of the ball transmission assembly according to claim 7, wherein in step (c), the distance between the test sample and the reference sample group is the distance between the test sample and the group center of the reference sample group. The method for diagnosing the pre-pressure of the ball transmission assembly according to claim 7, wherein in step (c), the distance between the test sample and the reference sample group is the distance between the test sample and each reference sample in the reference sample group The smallest.