Encoder signal analysis method for mechanical fault diagnosis under variable working conditions

Abstract

An encoder signal analysis method for mechanical fault diagnosis under variable working conditions comprises the following steps: firstly, acquiring a planetary gear box encoder signal with a planetary gear fault, and dividing the planetary gear box encoder signal by an encoder line number to obtain an instantaneous angular displacement signal; and processing the instantaneous angular displacement signal by using primary difference and secondary difference to respectively obtain an instantaneous angular velocity signal and an instantaneous angular acceleration signal. The instantaneous angular acceleration signal is segmented according to the displacement information, it is ensured that each segment of signal only contains one fault pulse, the kurtosis average value of each segment of signal is calculated, and the filter is updated in a self-adaptive iteration mode with the maximum average kurtosis as the direction. And finally, giving a gearbox fault diagnosis result according to the processed encoder signal.

Description

Technical Field

[0001] The present invention relates to the technical field of mechanical equipment fault diagnosis, and in particular to an encoder signal analysis method for mechanical fault diagnosis under variable working conditions. Background Art

[0002] Gearboxes are the core transmission components in rotating machinery. Their failure is one of the main reasons for machinery downtime and even major accidents. Therefore, how to effectively diagnose gearbox failures at an early stage and take preventive measures in advance has become one of the most important and popular research topics in the field of mechanical fault diagnosis.

[0003] Traditional gearbox fault diagnosis methods mainly rely on vibration analysis technology. Vibration analysis technology is based on the rule that mechanical equipment failures are usually manifested as changes or abnormalities in vibration signals. It detects the vibration information of the gearbox and its related components to identify potential faults. However, with the increasing degree of integration of mechanical equipment, the interference of vibration signals in complex structures and the complexity of transmission paths have brought great challenges to traditional vibration analysis technology. Therefore, the application of vibration analysis in gearbox fault diagnosis has gradually exposed its limitations, and there is an urgent need to find a new test method that is more efficient, accurate and adaptable to changing working conditions.

[0004] In this context, encoders, as a sensor widely used in equipment control, have begun to receive more and more attention. Encoders are usually installed close to the gearbox system, so their signals contain rich mechanical dynamic information, especially closely related to the working status and fault behavior of the gearbox. Compared with traditional vibration analysis methods, analysis methods based on encoder signals have many advantages in gearbox fault diagnosis methods, which are mainly reflected in the following aspects:

[0005] High fault sensitivity: Since the encoder directly reflects the change in torque, the encoder signal is more sensitive to stiffness changes than the vibration signal.

[0006] Short transmission path: The encoder is built into the device, the signal transmission path is short and the energy attenuation during the transmission process is small.

[0007] Wide range of applications: The encoder is usually built into the equipment as a core component, without considering wiring and safety issues.

[0008] Low testing cost: Encoders, as common control components in equipment, have been widely used in various mechanical equipment.

[0009] The collected information is rich in dimensions: the encoder signal is the position signal of the rotating shaft, which can obtain speed and acceleration information while obtaining vibration information.

[0010] The analysis method of encoder signals for mechanical fault diagnosis under variable operating conditions can not only overcome the limitations of traditional vibration analysis methods, such as requiring speed signals at the same time as vibration signals, but also provide a more efficient, sensitive and adaptable solution to variable speed problems for gearbox fault diagnosis.

[0011] Therefore, researching and developing encoder signal analysis methods for mechanical fault diagnosis under variable operating conditions has important practical significance for realizing fault diagnosis under variable speed conditions, extending equipment service life, and reducing maintenance costs. Summary of the invention

[0012] The object of the present invention is to provide an encoder signal analysis method for mechanical fault diagnosis under variable working conditions.

[0013] To achieve the above object, the technical solution adopted by the present invention is:

[0014] An encoder signal analysis method for mechanical fault diagnosis under variable working conditions comprises the following steps:

[0015] Step 1: Collect the signal n collected by the encoder = [n(1) n(2) … n(N)], where the signal length is N. Encoder number of revolutions n r The information is calculated as follows:

[0016]

[0017] Where n 0 is the number of encoder lines.

[0018] Step 2: Use the first difference and the second difference to convert the instantaneous angular displacement signal n r Converted into instantaneous angular velocity signal v n =[v n (1) v n (2) … v n (N)] and the instantaneous angular acceleration signal x n =[x n (1) x n (2) … x n (N)].

[0019] Step 3: Based on the gearbox fault characteristic order (FCO), the adjacent fault cycles are phase-divided. The calculation formula is as follows:

[0020]

[0021] Where θ(n s ) represents the angular displacement corresponding to the starting sampling point of the sth fault cycle; ns represents the starting sampling point of the sth fault cycle, s=1,2,3,...,Nt, Nt represents the total number of fault cycles contained in the system.

[0022] Step 4: Use the filtered signal y n The average kurtosis (AK) is used as the objective function, and the average kurtosis of each signal is calculated to maximize the average kurtosis:

[0023]

[0024] K s is the kurtosis value of the sth segment, y n is the input signal x n The signal convolved with the filter is calculated as follows:

[0025]

[0026] Step 5: Optimize and solve formula (3) through the objective function method to construct the adaptive iterative expression of the filter coefficient:

[0027]

[0028] make

[0029] The above formula can be transformed into:

[0030]

[0031] Step 6: The formula f=A obtained by the above calculation -1 bUpdate the filter coefficients.

[0032] Step 7: Determine whether the stopping condition is met. This method uses the number of iterative calculations as the stopping condition. If not, return to step 5. If it is met, the iterative algorithm terminates. Output the final result.

[0033] Step 8: Give the gearbox fault diagnosis result based on the processed encoder signal.

[0034] Compared with the prior art, the present invention has the following beneficial effects:

[0035] a) The present invention solves the problem of mechanical fault diagnosis under variable speed conditions.

[0036] b) The present invention can effectively extract the gearbox fault characteristics in the encoder signal under complex interference and has high robustness.

[0037] c) The vibration of the present invention solves the problem of synchronously obtaining diagnostic information and speed information in gearbox failure by only utilizing coded signals. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 The planetary gearbox test bench is an embodiment.

[0039] Figure 2 The figure is a flow chart of the method of the present invention.

[0040] Figure 3 is the encoder signal collected in the embodiment.

[0041] Figure 4 It is the instantaneous angular velocity signal after a difference in the embodiment.

[0042] Figure 5 is the instantaneous angular acceleration signal after secondary difference in the embodiment.

[0043] Figure 6 These are the results obtained using the proposed method in the examples. DETAILED DESCRIPTION

[0044] The present invention is described in detail below with reference to the accompanying drawings and embodiments.

[0045] Take the planetary gearbox fault simulation test bench as an example. The test bench consists of a motor, a planetary gearbox, a brake, etc. Figure 1 As shown. The encoders are mounted on the input and output shaft ends of the planetary gearbox. The planetary gear with prefabricated tooth surface spalling failure is placed in the planetary gearbox for testing.

[0046] like Figure 2 As shown, the encoder signal analysis method under variable speed includes the following steps:

[0047] Step 1: Collect the signal n = [n(1)n(2)…n(N)] collected by the encoder, where the signal length is N = 90000. Number of turns n r The information is calculated as follows:

[0048]

[0049] Where n 0 =5000 is the number of lines of the encoder.

[0050] Step 2: Use the first difference and the second difference to convert the instantaneous angular displacement signal n r Converted into instantaneous angular velocity signal v n =[v n (1)v n (2)…v n (N)] and the instantaneous angular acceleration signal xn =[x n (1) x n (2) … x n (N)].

[0051] Step 3: Based on the gearbox fault characteristic order, the adjacent fault cycles are phase-divided. The calculation formula is as follows:

[0052]

[0053] Where θ(n s ) represents the angular displacement corresponding to the starting sampling point of the sth fault cycle; n s , represents the starting sampling point of the sth fault cycle, s=1,2,3,...,7, and the signal is decomposed into 7 fault cycles in total.

[0054] Step 4: Use the filtered signal y n The average kurtosis of is used as the objective function, and the average kurtosis of each signal is calculated to maximize the average kurtosis:

[0055]

[0056] K s is the kurtosis value of the sth segment, y n is the input signal x n The signal convolved with the filter is calculated as follows:

[0057]

[0058] Step 5: Optimize and solve formula (3) through the objective function method to construct the adaptive iterative expression of the filter coefficient:

[0059]

[0060] make

[0061] The above formula can be transformed into:

[0062]

[0063] Step 6: The formula f=A obtained by the above calculation -1 bUpdate the filter coefficients.

[0064] Step 7: Determine whether the stopping condition is met. This article will use the number of iterations Iter = 50 as the stopping standard. If not, return to step 5. If it is met, the iterative algorithm terminates. Output the final result.

[0065] Step 8: Give the gearbox fault diagnosis result based on the processed encoder signal.

[0066] like Figure 6 As shown in the figure, the result of applying the method proposed in the present invention to the encoder signal is shown, and the impact component with an interval of 31 teeth can be clearly seen. Because the number of teeth of the planetary gear in the planetary gearbox is 31. That is, the result confirms that the stiffness fluctuation of the planetary gearbox will be reflected in the encoder signal every 31 teeth, thus indicating that the fault occurs on the planetary gear. It also confirms that the present invention has obvious advantages over the traditional method.

Claims

1. A method for analyzing encoder signals for mechanical fault diagnosis under variable working conditions, characterized in that: The following steps are involved: Step 1: Collect the signal n collected by the encoder = [n(1) n(2) … n(N)], where the signal length is N. Encoder number of revolutions n r The information is calculated as follows: Where n0 is the number of encoder lines. Step 2: Use the first difference and the second difference to convert the instantaneous angular displacement signal n r Converted into instantaneous angular velocity signal v n =[v n (1) v n (2) ... v n (N)] and the instantaneous angular acceleration signal x n =[x n (1) x n (2) ... x n (N)]. Step 3: Based on the gearbox fault characteristic order, the adjacent fault cycles are phase-divided. The calculation formula is as follows: Where θ(n s ) represents the angular displacement corresponding to the starting sampling point of the sth fault cycle; n s Represents the starting sampling point of the sth fault cycle, s = 1, 2, 3, ..., Nt, Nt represents the total number of fault cycles contained in the system. Step 4: Use the filtered signal y n The average kurtosis of is used as the objective function, and the average kurtosis of each signal is calculated to maximize the average kurtosis: K s is the kurtosis value of the sth segment, y n is the input signal x n The signal convolved with the filter is calculated as follows: Step 5: Optimize and solve formula (3) through the objective function method to construct the adaptive iterative expression of the filter coefficient: make The above formula can be transformed into: Step 6: The formula f=A obtained by the above calculation -1 bUpdate the filter coefficients. Step 7: Determine whether the stopping condition is met. This method uses the number of iterative calculations as the stopping condition. If not, return to step 5. If it is met, the iterative algorithm terminates. Output the final result. Step 8: Give the gearbox fault diagnosis result based on the processed encoder signal.

Images