Frequency domain feature extraction method and system for vibration signal of complex transmission system

Abstract

The present invention provides a method and system for extracting frequency domain features of vibration signals of a complex transmission system, including a vibration sensor, a multi-channel data acquisition device, a computer and data analysis software; the vibration sensor is installed on the complex transmission device to measure the vibration signal and convert it into an electrical signal input; the multi-channel data acquisition device performs anti-aliasing filtering on the vibration electrical signal and converts it into a digital signal for transmission to the computer; the data analysis software on the computer extracts frequency domain features of the signal. The frequency domain feature extraction method based on prior knowledge of structural parameters of complex transmission systems and component decomposition provided by the present invention can accurately extract the frequency domain features corresponding to each mechanical component of a transmission system with a complex structure, completely retain the information transmitted by the vibration signal, and then realize component-level and system-level status monitoring of the complex transmission system by comparing with historical operating data.

Description

Technical Field

[0001] The present invention relates to the technical field of vibration signal feature extraction, and in particular, to a method and system for extracting frequency domain features of vibration signals of a complex transmission system, and in particular, to a method and system for extracting frequency domain features of vibration signals of a complex transmission system based on component decomposition and a state detection method. Background Art

[0002] The complex transmission system is a complex mechanical-electrical-hydraulic composite gear-shifting and speed-changing system that can realize important functions such as steering, gear shifting and speed change of tracked vehicles in an integrated manner. It is the core power component of modern tracked vehicles. The complex transmission device has a complex structure and is mainly composed of rotating machinery such as shafts, gears, and bearings. Due to the harsh and changeable service conditions of the complex transmission device, these mechanical components are prone to degradation failures, which in turn poses a threat to the operating safety of the complex transmission device. Therefore, it is of great significance and value to carry out condition monitoring of these mechanical components in the rotating structure.

[0003] Since vibration signals can reflect rich information about the operating status of rotating machinery, some rotating machinery status features based on vibration signals are extracted. These features are mainly for the overall frequency band of the signal or some sub-frequency bands.

[0004] Patent document CN112161807A (application number: CN202011000937.0) discloses a fault diagnosis method for a speed change gearbox, which includes: collecting data from multiple vibration sensors to obtain multiple vibration signals of different shaft speeds, and performing low-pass filtering; creating a set of Gaussian reference signals related to the speed, and creating an adaptive noise filter based on the Gaussian reference signal; extracting vibration signal features based on the adaptive noise filter and the least mean square algorithm; dividing the extracted features into a training data set and a test data set, dividing the training data set into training set data and validation set data by a cross-validation method, and training a support vector machine model based on the training set data and the validation set data, and using the trained model to classify faults in the test data set to obtain classification results.

[0005] Patent document CN114091523A (application number: CN202111191225.6) discloses a gray fault diagnosis method for key rotating parts of a vehicle driven by signal frequency domain characteristics. According to the different characteristics of the frequency domain characteristics of vibration signals caused by different faults, the frequency response of Q-modulated wavelet transform is used to analyze the vibration signal of the rotating parts of the vehicle or its envelope signal, and a feature vector with the energy proportion of each sub-band in different frequency bands as the element is constructed; in view of the problem that the gray proximity correlation degree has different change trends of the two sequences when processing two sets of intersecting sequences, but the original cumulative difference is small, resulting in excessive correlation, a gray absolute proximity correlation degree model is proposed; driven by the constructed frequency domain feature vector, the gray absolute proximity correlation degree with the standard mode is calculated to identify the fault state of the key rotating parts of the vehicle.

[0006] Similar methods are mainly applied to gearboxes with simple structures. Due to the complex structure of complex transmission systems and the large number of mechanical components involved, the above-mentioned rough frequency band method cannot extract accurate and targeted features involving each component.

[0007] A large number of studies have shown that when a rotating mechanical component fails, the vibration signal excited by its periodic motion can convey rich information, which can be accurately reflected in the frequency domain. For example, when the shaft is misaligned, cracked, or other faults occur, a series of harmonics that are multiples of the rotation frequency will appear. When a gear fails, a modulation component of the meshing frequency and its higher harmonic components will appear in the frequency domain. When a bearing fails, a series of harmonic clusters separated by characteristic passing frequencies will appear. When the sampling time is long enough, the signal has a good resolution in the frequency domain, and the frequency domain features of each component can be well separated. However, there is currently a lack of technology that can extract accurate frequency domain features related to each rotating component.

[0008] The existing vibration signal feature extraction methods mainly extract features from the entire frequency band or certain sub-bands of the signal, and are mainly used in gearboxes with simple structures. Due to the complex structure of complex transmission systems and the large number of mechanical components involved, the above-mentioned rough frequency band method cannot extract accurate and targeted features involving each component. Summary of the invention

[0009] In view of the defects in the prior art, the object of the present invention is to provide a method and system for extracting frequency domain features of vibration signals of a complex transmission system.

[0010] The method for extracting frequency domain features of vibration signals of a complex transmission system provided by the present invention comprises:

[0011] Step 1: Define the structural parameters of the complex transmission system;

[0012] Step 2: Calculate the characteristic frequency set of the shaft and gear according to the structural parameters;

[0013] Step 3: The complex transmission system operates under a specific speed condition to obtain a discrete vibration signal;

[0014] Step 4: Based on the discrete vibration signal, extract each characteristic frequency component in the characteristic frequency set using component decomposition method;

[0015] Step 5: Calculate the frequency domain index set based on the decomposed components;

[0016] Step 6: Compare the extracted frequency domain component indicators with the historical operating data for transmission system status evaluation.

[0017] Preferably, the rotating parts of the complex transmission are numbered, and each axis is numbered Each gear is numbered For a given engine speed of f s In the case of , the characteristic frequencies of each rotating component are calculated, where the characteristic frequency of the shaft is the frequency set f_Set(S i ), the characteristic frequencies of the gears are the meshing frequency and its multiples, as well as the modulation frequency f_Set(G j );

[0018] Get the overall frequency set Assume that the interval between the closest components in f_Set_total is Δf, then the sampling duration should be greater than 1 / Δf;

[0019] Among them, the subscript n S Indicates the total number of axis numbers; subscript n G Indicates the total number of gear numbers; i and j represent the subscripts of the shaft and gear, respectively.

[0020] Preferably, the complex transmission is operated under a specific gear and loading condition, and the engine speed is set to f s , using vibration sensors to collect signals, and obtaining discrete sampling s(t n ),n=0,1,...,N-1,where t n represents the sampling moment; N represents the number of discrete signal sampling points; n represents the discrete signal sampling time subscript.

[0021] Preferably, the nonlinear frequency modulation component is used to decompose and extract each characteristic frequency component, and for all elements f in f_Set_total, i , respectively extract the corresponding signal components from the vibration signal, and the corresponding signal components are expressed as:

[0022] s i (t) = u (i)(t)cos[2πf i (t)]+v (i) (t)sin[2πf i (t)]

[0023] Where: s i (t) represents the corresponding signal component; u (i) (t) represents the instantaneous amplitude of the cosine part of the signal; v (i) (t) represents the instantaneous amplitude of the sin part of the signal; f i (t) represents the frequency to be decomposed and extracted; u (i) Indicates the instantaneous amplitude of the cosine part of the signal; v (i) Indicates the instantaneous amplitude of the sin part of the signal;

[0024] The discrete form of this component is: i =H i y i , forming a linear system, where:

[0025]

[0026]

[0027]

[0028]

[0029] Where, F is the Fourier model matrix; F0 is twice the sampling frequency; L is the order of the given Fourier series, and the choice of L corresponds to the bandwidth of the decomposed signal, which is BW=2LF0, and BW represents the bandwidth of the decomposed signal;

[0030] is the magnitude vector matrix, where Using Tikhonov regularization, the amplitude matrix is ​​solved as Where λ is the regularization coefficient, and the reconstructed signal is expressed as Then decompose the signal components corresponding to all frequencies in f_Set_total; H i The linear system matrix representing the definition of the given frequency and Fourier model; Represents the corresponding part of the cosine function in the linear system matrix; Represents the corresponding part of the sin function in the linear system matrix.

[0031] Preferably, the narrowband energy of each frequency domain component is calculated to form a numerical sequence x n ,n=1,2…N, the selected indicators are:

[0032] First indicator: Geometric mean energy

[0033] Second indicator: Modulation intensity geometric mean where x n Represents meshing frequency and high-order harmonics, x n-fm Sidebands representing the meshing frequency and its multiples;

[0034] The third indicator: energy arithmetic mean

[0035] The fourth indicator: the arithmetic mean of modulation intensity

[0036] The frequency domain indexes are calculated for each mechanical component of the complex transmission system. i ,f_Set(S i ) contains each axis frequency and its multiple, calculates the first and third indexes, and forms Index_Set1(S i ) and Index_Set3(S i ) ; for gear G j ,f_Set(G j ) includes each meshing frequency and its multiple frequency, calculates the first to fourth indexes, and forms Index_Set1(G j )、Index_Set2(G j )、Index_Set3(G j ) and Index_Set4(G j ); finally, each frequency domain index is combined into a frequency domain index feature library of complex transmission system.

[0037] According to the historical data of the transmission system, the mean and variance of each frequency domain index are calculated. k , assuming that the mean and variance of its historical operating data are μ k and σ k , if the indicator exceeds μ in a certain measurement k +3σ k , it is considered that the component corresponding to the indicator has a suspected fault, an alarm is issued and component information is provided. i If the corresponding first or third indicator exceeds the limit, the shaft is considered to have a suspected fault, an alarm is issued and component information is provided. j If one of the corresponding first to fourth indicators exceeds the limit, it is considered that the gear has a suspected fault, an alarm is issued and component information is provided.

[0038] The vibration signal frequency domain feature extraction system of the complex transmission system provided by the present invention comprises:

[0039] Module M1: define the structural parameters of the complex transmission system;

[0040] Module M2: Calculate the characteristic frequency set of shaft and gear according to the structural parameters;

[0041] Module M3: The complex transmission system operates under a specific speed condition to obtain discrete vibration signals;

[0042] Module M4: Based on the discrete vibration signal, the component decomposition method is used to extract each characteristic frequency component in the characteristic frequency set;

[0043] Module M5: Calculate the frequency domain index set based on the decomposed components;

[0044] Module M6: Compare with historical operating data and issue alarms for components corresponding to over-limit indicators.

[0045] Preferably, the rotating parts of the complex transmission device are numbered, and each axis is numbered S1, S2, ..., S i ,...,S nS , each gear is numbered G1, G2, ..., G j ,...,G nG , given the engine speed f s In the case of , the characteristic frequencies of each rotating component are calculated, where the characteristic frequency of the shaft is the frequency set f_Set(S i ), the characteristic frequencies of the gears are the meshing frequency and its multiples, as well as the modulation frequency f_Set(G j );

[0046] Get the overall frequency set Assume that the interval between the closest components in f_Set_total is Δf, then the sampling duration should be greater than 1 / Δf;

[0047] Among them, the subscript nS represents the total number of shaft numbers; the subscript nG represents the total number of gear numbers; i and j represent the subscripts of the shaft and gear, respectively.

[0048] Preferably, the complex transmission is operated under a specific gear and loading condition, and the engine speed is set to f s , using vibration sensors to collect signals, and obtaining discrete sampling s(t n ),n=0,1,...,N-1,where t n represents the sampling moment; N represents the number of discrete signal sampling points; n represents the discrete signal sampling time subscript.

[0049] Preferably, the nonlinear frequency modulation component is used to decompose and extract each characteristic frequency component, and for all elements f in f_Set_total, i , respectively extract the corresponding signal components from the vibration signal, and the corresponding signal components are expressed as:

[0050] s i (t) = u (i) (t)cos[2πf i (t)]+v (i) (t)sin[2πf i (t)]

[0051] Where: s i (t) represents the corresponding signal component; u (i) (t) represents the instantaneous amplitude of the cosine part of the signal; v (i) (t) represents the instantaneous amplitude of the sin part of the signal; f i (t) represents the frequency to be decomposed and extracted; u (i) Indicates the instantaneous amplitude of the cosine part of the signal; v (i) Indicates the instantaneous amplitude of the sin part of the signal;

[0052] The discrete form of this component is: i =H i y i , forming a linear system, where:

[0053]

[0054]

[0055]

[0056]

[0057] Where, F is the Fourier model matrix; F0 is twice the sampling frequency; L is the order of the given Fourier series, and the choice of L corresponds to the bandwidth of the decomposed signal, which is BW=2LF0, and BW represents the bandwidth of the decomposed signal;

[0058] is the magnitude vector matrix, where Using Tikhonov regularization, the amplitude matrix is ​​solved as Where λ is the regularization coefficient, and the reconstructed signal is expressed as Then decompose the signal components corresponding to all frequencies in f_Set_total; H i The linear system matrix representing the definition of the given frequency and Fourier model; Represents the corresponding part of the cosine function in the linear system matrix; Represents the corresponding part of the sin function in the linear system matrix.

[0059] Preferably, the narrowband energy of each frequency domain component is calculated to form a numerical sequence x n ,n=1,2…N, the selected indicators are:

[0060] First indicator: Geometric mean energy

[0061] Second indicator: Modulation intensity geometric mean where x n Represents meshing frequency and high-order harmonics, x n-fm Sidebands representing the meshing frequency and its multiples;

[0062] The third indicator: energy arithmetic mean

[0063] The fourth indicator: the arithmetic mean of modulation intensity

[0064] The frequency domain indexes are calculated for each mechanical component of the complex transmission system. i ,f_Set(S i ) contains each axis frequency and its multiple, calculates the first and third indexes, and forms Index_Set1(S i ) and Index_Set3(S i ) ; for gear G j ,f_Set(G j ) includes each meshing frequency and its multiple frequency, calculates the first to fourth indexes, and forms Index_Set1(G j )、Index_Set2(G j )、Index_Set3(G j ) and Index_Set4(G j ); finally, each frequency domain index is combined into a frequency domain index feature library of complex transmission system.

[0065] According to the historical data of the transmission system, the mean and variance of each frequency domain index are calculated. k , assuming that the mean and variance of its historical operating data are μ k and σ k , if the indicator exceeds μ in a certain measurement k +3σ k , it is considered that the component corresponding to the indicator has a suspected fault, an alarm is issued and component information is provided. iIf the corresponding first or third indicator exceeds the limit, the shaft is considered to have a suspected fault, an alarm is issued and component information is provided. j If one of the corresponding first to fourth indicators exceeds the limit, it is considered that the gear has a suspected fault, an alarm is issued and component information is provided.

[0066] Compared with the prior art, the present invention has the following beneficial effects:

[0067] The frequency domain feature extraction method based on prior knowledge of complex transmission system structural parameters and component decomposition provided by the present invention can accurately extract the frequency domain features corresponding to each mechanical component in a transmission system with a complex structure, completely retain the information transmitted by the vibration signal, and realize component-level status monitoring by comparing with historical data. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments made with reference to the following drawings:

[0069] Figure 1 This is an overall flow chart of the frequency domain index extraction system according to an embodiment of the present invention;

[0070] Figure 2 A flow chart of characteristic frequency generation in an embodiment of the present invention;

[0071] Figure 3 A pseudo code flow chart for extracting each frequency component by component decomposition according to an embodiment of the present invention;

[0072] Figure 4 The present invention is a flowchart of calculating a frequency domain indicator set according to an embodiment of the present invention. DETAILED DESCRIPTION

[0073] The present invention is described in detail below in conjunction with specific embodiments. The following embodiments will help those skilled in the art to further understand the present invention, but are not intended to limit the present invention in any form. It should be noted that, for those of ordinary skill in the art, several changes and improvements can also be made without departing from the concept of the present invention. These all belong to the protection scope of the present invention.

[0074] Example:

[0075] The present invention proposes a method for extracting frequency domain features of vibration signals based on component decomposition. According to the parameter information and prior knowledge of rotating mechanical parts of a complex transmission system, a series of frequency components to be monitored are determined, the components are decomposed by a nonlinear frequency modulation component decomposition method, the frequency domain features are calculated separately, and the results are integrated into a frequency domain feature vector.

[0076] The present invention provides a frequency component extraction method based on the component decomposition method. Given a characteristic frequency, the linear system is solved using Tikhonov regularization to obtain the instantaneous amplitude of each frequency component, and then each frequency component is reconstructed. According to the reconstructed frequency components, a series of frequency domain indicators can be calculated for state monitoring.

[0077] The present invention proposes a vibration signal frequency domain feature extraction system based on component decomposition, comprising a vibration sensor, a multi-channel data acquisition device, a computer and data analysis software; the vibration sensor is installed and arranged on a complex transmission device, and is used to measure the vibration signal and convert it into an electrical signal input; the multi-channel data acquisition device performs anti-aliasing filtering on the vibration electrical signal and converts it into a digital signal for transmission to the computer; the data analysis software on the computer extracts the frequency domain features of the signal.

[0078] like Figure 1 As shown, the present invention provides a frequency domain extraction method for vibration signals of complex transmission devices based on component decomposition, and the specific steps are as follows:

[0079] Step 1, such as Figure 2 As shown in the figure, the rotating parts of a certain type of complex transmission device are numbered, and each axis is numbered as Each gear is numbered Given an engine speed of 1700 RPM, calculate the characteristic frequencies of each rotating component, where the characteristic frequency of the shaft is the frequency set f_Set (S i )The rotation frequency of the input shaft is 28.33Hz. The rotation frequencies of the other shafts can be calculated based on the gear meshing relationship in the transmission system. The characteristic frequencies of the gears are the meshing frequency and its multiples, as well as the modulation frequency f_Set(G j ), the meshing frequency is the product of the rotation frequency of the shaft and the number of teeth. The overall frequency set is further obtained Assume that the interval between the closest components in f_Set_total is Δf, then the sampling duration should be greater than 1 / Δf, and the selected sampling duration of 10s meets the condition.

[0080] Step 2: Run the complex transmission device under 4th gear and 50% load condition, set the engine speed to 1700RPM, collect vibration signals for 10s using a vibration sensor, and obtain discrete sampling s(t m ),m=1,2,...,M, where t m Indicates the sampling time.

[0081] Step 3, such as Figure 3 As shown in the pseudo code, the characteristic frequency components are extracted by component decomposition, and for each element f in f_Set_total i, respectively extract the corresponding signal components from the vibration signal. The corresponding signal components can be expressed as s i (t) = u (i) (t)cos[2πf i (t)]+v (i) (t)sin[2πf i (t)], we only need to solve u (i) and v (i) The discrete form of this component is: s i =H i y i ;

[0082] in:

[0083]

[0084]

[0085]

[0086]

[0087] F0 is twice the sampling frequency, and L is the order of the given Fourier series. The choice of L corresponds to the bandwidth of the decomposed signal, which is BW=2LF0, where BW represents the bandwidth of the decomposed signal. Considering that the signal may have a certain speed fluctuation, it is selected as the bandwidth, which is 0.5% of the frequency. is the magnitude vector matrix, where The amplitude matrix can be solved using Tikhonov regularization as Where λ is the regularization coefficient, which is selected as 0.05. The reconstructed signal can be expressed as Then, the signal components corresponding to all frequencies in f_Set_total can be decomposed.

[0088] Step 4, such as Figure 4 As shown, the frequency domain index is calculated. The narrowband energy of each frequency domain component is calculated to form a numerical sequence x n , n = 1, 2...N, four typical indicators are selected: (1) Energy geometric mean: (2) Geometric mean of modulation intensity: where x n Represents meshing frequency and high-order harmonics, x n-fm Represents the sidebands of the meshing frequency and its multiples; (3) Energy arithmetic mean: (4) Arithmetic mean of modulation intensity: The frequency domain indicators are calculated for each mechanical component of the complex transmission system. i ,f_Set(S i) contains each axis frequency and its multiples, and can calculate (1) (3) indicators to form Index_Set1 (S i ) and Index_Set3(S i ). For gear G j ,f_Set(G j ) contains each meshing frequency and its multiples (including sidebands), which can be calculated (1)(2)(3)(4) to form Index_Set1(G j )、Index_Set2(G j )、Index_Set3(G j ) and Index_Set4(G j ). Various frequency domain indicators are combined into a frequency domain indicator feature library for complex transmission systems.

[0089] Step 5: Calculate the mean and variance of each frequency domain index based on the historical data of the transmission system. k , assuming that the mean and variance of its historical operating data are μ k and σ k , if the indicator exceeds μ in a certain measurement k +3σ k , it is considered that the component corresponding to the indicator has a suspected fault, an alarm is issued and component information is provided. i If the corresponding first or third indicator exceeds the limit, the shaft is considered to have a suspected fault, an alarm is issued and component information is provided. j If one of the corresponding first to fourth indicators exceeds the limit, it is considered that the gear has a suspected fault, an alarm is issued and component information is provided.

[0090] In step 2, in addition to using the three-axis acceleration sensor, a single-axis acceleration sensor or other vibration sensors such as a displacement sensor may also be used.

[0091] In step 2, in addition to conducting the experiment under constant speed conditions, the experiment can also be conducted under variable speed conditions. By using the key phase signal or the tachometer signal to perform order analysis on the signal, the vibration signal under stable working conditions can also be obtained.

[0092] The vibration signal frequency domain feature extraction system of the complex transmission system provided by the present invention includes: module M1: defining the structural parameters of the complex transmission system; module M2: calculating the characteristic frequency set of the shaft and the gear according to the structural parameters; module M3: the complex transmission system operates under a specific speed condition to obtain a discrete vibration signal; module M4: based on the discrete vibration signal, using the component decomposition method to extract each characteristic frequency component in the characteristic frequency set; module M5: calculating the frequency domain index set according to the decomposed components; module M6: comparing with the historical operation data, and alarming the components corresponding to the over-limit index.

[0093] The rotating parts of the complex transmission are numbered, and each axis is numbered Each gear is numbered For a given engine speed of f s In the case of , the characteristic frequencies of each rotating component are calculated, where the characteristic frequency of the shaft is the frequency set f_Set(S i ), the characteristic frequencies of the gears are the meshing frequency and its multiples, as well as the modulation frequency f_Set(G j );Get the overall frequency set Assuming that the interval between the closest components in f_Set_total is Δf, the sampling duration should be greater than 1 / Δf; wherein, the subscript nS represents the total number of axis numbers; the subscript nG represents the total number of gear numbers; i and j represent the subscripts of the axis and the gear, respectively.

[0094] Running a complex transmission in a specific gear and load condition, setting the engine speed to f s , using vibration sensors to collect signals, and obtaining discrete sampling s(t n ),n=0,1,...,N-1,where t n represents the sampling moment; N represents the number of discrete signal sampling points; n represents the discrete signal sampling time subscript.

[0095] The characteristic frequency components are extracted by nonlinear frequency modulation component decomposition. i , respectively extract the corresponding signal components from the vibration signal, and the corresponding signal components are expressed as: s i (t) = u (i) (t)cos[2πf i (t)]+v (i) (t)sin[2πf i (t)]; where: s i (t) represents the corresponding signal component; u (i) (t) represents the instantaneous amplitude of the cosine part of the signal; v (i)(t) represents the instantaneous amplitude of the sin part of the signal; f i (t) represents the frequency to be decomposed and extracted; u (i) Indicates the instantaneous amplitude of the cosine part of the signal; v (i) Represents the instantaneous amplitude of the sin part of the signal; the discrete form of this component is: s i =H i y i , forming a linear system, where: Where, F is the Fourier model matrix; F0 is twice the sampling frequency; L is the order of the given Fourier series, and the choice of L corresponds to the bandwidth of the decomposed signal, which is BW=2LF0, and BW represents the bandwidth of the decomposed signal; is the magnitude vector matrix, where Using Tikhonov regularization, the amplitude matrix is ​​solved as Where λ is the regularization coefficient, and the reconstructed signal is expressed as Then decompose the signal components corresponding to all frequencies in f_Set_total; H i The linear system matrix representing the definition of the given frequency and Fourier model; Represents the corresponding part of the cosine function in the linear system matrix; Represents the corresponding part of the sin function in the linear system matrix.

[0096] Calculate the narrowband energy of each frequency domain component to form a numerical sequence x n ,n=1,2…N, the selected indicators are: First indicator: energy geometric mean Second indicator: Modulation intensity geometric mean where x n Represents meshing frequency and high-order harmonics, x n-fm Indicates the sideband of the meshing frequency and its multiples; The third indicator: the arithmetic mean of energy The fourth indicator: the arithmetic mean of modulation intensity The frequency domain indexes are calculated for each mechanical component of the complex transmission system. i ,f_Set(S i ) contains each axis frequency and its multiple, calculates the first and third indexes, and forms Index_Set1(S i ) and Index_Set3(S i ) ; for gear G j ,f_Set(G j ) includes each meshing frequency and its multiple frequency, calculates the first to fourth indexes, and forms Index_Set1(G j)、Index_Set2(G j )、Index_Set3(G j ) and Index_Set4(G j ); finally, each frequency domain index is combined into a frequency domain index feature library of complex transmission system.

[0097] According to the historical data of the transmission system, the mean and variance of each frequency domain index are calculated. k , assuming that the mean and variance of its historical operating data are μ k and σ k , if the indicator exceeds μ in a certain measurement k +3σ k , it is considered that the component corresponding to the indicator has a suspected fault, an alarm is issued and component information is provided. i If the corresponding first or third indicator exceeds the limit, the shaft is considered to have a suspected fault, an alarm is issued and component information is provided. j If one of the corresponding first to fourth indicators exceeds the limit, it is considered that the gear has a suspected fault, an alarm is issued and component information is provided.

[0098] Those skilled in the art know that, in addition to implementing the system, device and its various modules provided by the present invention in a purely computer-readable program code, it is entirely possible to implement the same program in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers and embedded microcontrollers by logically programming the method steps. Therefore, the system, device and its various modules provided by the present invention can be considered as a hardware component, and the modules included therein for implementing various programs can also be considered as structures within the hardware component; the modules for implementing various functions can also be considered as both software programs for implementing the method and structures within the hardware component.

[0099] The above describes the specific embodiments of the present invention. It should be understood that the present invention is not limited to the above specific embodiments, and those skilled in the art can make various changes or modifications within the scope of the claims, which does not affect the essence of the present invention. In the absence of conflict, the embodiments of the present application and the features in the embodiments can be combined with each other arbitrarily.

Claims

1. A frequency domain feature extraction method for vibration signals of a complex transmission system, characterized in that: include: Step 1: Define the structural parameters of the complex transmission system; Step 2: Calculate the characteristic frequency set of the shaft and gear according to the structural parameters; Step 3: The complex transmission system operates under a specific speed condition to obtain a discrete vibration signal; Step 4: Based on the discrete vibration signal, extract each characteristic frequency component in the characteristic frequency set using component decomposition method; Step 5: Calculate the frequency domain index set based on the decomposed components; Step 6: Compare the extracted frequency domain component index with the historical operation data for transmission system status evaluation; The rotating parts of the complex transmission are numbered, and each axis is numbered Each gear is numbered For a given engine speed of f s In the case of , the characteristic frequencies of each rotating component are calculated, where the characteristic frequency of the shaft is the frequency set f_Set(S i ), the characteristic frequencies of the gears are the meshing frequency and its multiples, as well as the modulation frequency f_Set(G j ); Get the overall frequency set Assume that the interval between the closest components in f_Set_total is Δf, then the sampling duration should be greater than 1 / Δf; Among them, the subscript n S Indicates the total number of axis numbers; subscript n G Indicates the total number of gear numbers; i and j represent the subscripts of the shaft and gear, respectively.

2. The vibration signal frequency domain feature extraction method of a complex transmission system according to claim 1 is characterized in that: Running a complex transmission in a specific gear and load condition, setting the engine speed to f s , using vibration sensors to collect signals, and obtaining discrete sampling s(t n ),n=0,1,...,N-1,where t n represents the sampling moment; N represents the number of discrete signal sampling points; n represents the discrete signal sampling time subscript.

3. The method for extracting frequency domain features of vibration signals of complex transmission systems according to claim 2, characterized in that: The characteristic frequency components are extracted by nonlinear frequency modulation component decomposition. i , respectively extract the corresponding signal components from the vibration signal, and the corresponding signal components are expressed as: s i (t)=u (i) (t)cos[2πf i (t)]+v (i) (t)sin[2πf i (t)] Where: s i (t) represents the corresponding signal component; u (i) (t) represents the instantaneous amplitude of the cosine part of the signal; v (i) (t) represents the instantaneous amplitude of the sin part of the signal; f i (t) represents the frequency to be decomposed and extracted; u (i) Indicates the instantaneous amplitude of the cosine part of the signal; v (i) Indicates the instantaneous amplitude of the sin part of the signal; The discrete form of this component is: i =H i y i , forming a linear system, where: Where, F is the Fourier model matrix; F0 is twice the sampling frequency; L is the order of the given Fourier series, and the choice of L corresponds to the bandwidth of the decomposed signal, which is BW=2LF0, and BW represents the bandwidth of the decomposed signal; is the magnitude vector matrix, where Using Tikhonov regularization, the amplitude matrix is ​​solved as Where λ is the regularization coefficient, and the reconstructed signal is expressed as Then decompose the signal components corresponding to all frequencies in f_Set_total; H i The linear system matrix representing the definition of the given frequency and Fourier model; Represents the corresponding part of the cosine function in the linear system matrix; Represents the corresponding part of the sin function in the linear system matrix.

4. The method for extracting frequency domain features of vibration signals of a complex transmission system according to claim 3, characterized in that: Calculate the narrowband energy of each frequency domain component to form a numerical sequence x n ,n=1,2…N, the selected indicators are: First indicator: Geometric mean energy Second indicator: Modulation intensity geometric mean where x n Represents meshing frequency and high-order harmonics, x n-fm Sidebands representing the meshing frequency and its multiples; The third indicator: energy arithmetic mean The fourth indicator: the arithmetic mean of modulation intensity The frequency domain indexes are calculated for each mechanical component of the complex transmission system. i ,f_Set(S i ) contains each axis frequency and its multiple, calculates the first and third indexes, and forms Index_Set1(S i ) and Index_Set3(S i ) ; for gear G j ,f_Set(G j ) includes each meshing frequency and its multiple frequency, calculates the first to fourth indexes, and forms Index_Set1(G j )、Index_Set2(G j )、Index_Set3(G j ) and Index_Set4(G j ); finally, each frequency domain index is combined into a frequency domain index feature library of complex transmission system.

5. The method for extracting frequency domain features of vibration signals of a complex transmission system according to claim 4, characterized in that: According to the historical data of the transmission system, the mean and variance of each frequency domain index are calculated. For one of the indexes, k , assuming that the mean and variance of its historical operating data are μ k and σ k , if the indicator exceeds μ in a certain measurement k +3σ k , it is considered that the component corresponding to the indicator has a suspected fault, an alarm is issued and component information is provided; for example, if a shaft S i If the corresponding first or third indicator exceeds the limit, the shaft is considered to have a suspected fault, an alarm is issued and component information is provided; for example, a gear G j If one of the corresponding first to fourth indicators exceeds the limit, it is considered that the gear has a suspected fault, an alarm is issued and component information is provided.

6. A frequency domain feature extraction system for vibration signals of a complex transmission system, characterized in that: include: Module M1: define the structural parameters of the complex transmission system; Module M2: Calculate the characteristic frequency set of shaft and gear according to the structural parameters; Module M3: The complex transmission system operates under a specific speed condition to obtain discrete vibration signals; Module M4: Based on the discrete vibration signal, the component decomposition method is used to extract each characteristic frequency component in the characteristic frequency set; Module M5: Calculate the frequency domain index set based on the decomposed components; Module M6: Compare with historical operation data and alarm the corresponding components of the over-limit indicators; The rotating parts of the complex transmission are numbered, and each axis is numbered Each gear is numbered For a given engine speed of f s In the case of , the characteristic frequencies of each rotating component are calculated, where the characteristic frequency of the shaft is the frequency set f_Set(S i ), the characteristic frequencies of the gears are the meshing frequency and its multiples, as well as the modulation frequency f_Set(G j ); Get the overall frequency set Assume that the interval between the closest components in f_Set_total is Δf, then the sampling duration should be greater than 1 / Δf; Among them, the subscript n S Indicates the total number of axis numbers; subscript n G Indicates the total number of gear numbers; i and j represent the subscripts of the shaft and gear, respectively.

7. The vibration signal frequency domain feature extraction system for a complex transmission system according to claim 6, characterized in that: Running a complex transmission in a specific gear and load condition, setting the engine speed to f s , using vibration sensors to collect signals, and obtaining discrete sampling s(t n ),n=0,1,...,N-1,where t n represents the sampling moment; N represents the number of discrete signal sampling points; n represents the discrete signal sampling time subscript.

8. The vibration signal frequency domain feature extraction system for a complex transmission system according to claim 7, characterized in that: The characteristic frequency components are extracted by nonlinear frequency modulation component decomposition. i , respectively extract the corresponding signal components from the vibration signal, and the corresponding signal components are expressed as: s i (t)=u (i) (t)cos[2πf i (t)]+v (i) (t)sin[2πf i (t)] Where: s i (t) represents the corresponding signal component; u (i) (t) represents the instantaneous amplitude of the cosine part of the signal; v (i) (t) represents the instantaneous amplitude of the sin part of the signal; f i (t) represents the frequency to be decomposed and extracted; u (i) Indicates the instantaneous amplitude of the cosine part of the signal; v (i) Indicates the instantaneous amplitude of the sin part of the signal; The discrete form of this component is: i =H i y i , forming a linear system, where: Where, F is the Fourier model matrix; F0 is twice the sampling frequency; L is the order of the given Fourier series, and the choice of L corresponds to the bandwidth of the decomposed signal, which is BW=2LF0, and BW represents the bandwidth of the decomposed signal; is the magnitude vector matrix, where Using Tikhonov regularization, the amplitude matrix is ​​solved as Where λ is the regularization coefficient, and the reconstructed signal is expressed as Then decompose the signal components corresponding to all frequencies in f_Set_total; H i The linear system matrix representing the definition of the given frequency and Fourier model; Represents the corresponding part of the cosine function in the linear system matrix; Represents the corresponding part of the sin function in the linear system matrix.

9. The vibration signal frequency domain feature extraction system for a complex transmission system according to claim 8, characterized in that: Calculate the narrowband energy of each frequency domain component to form a numerical sequence x n ,n=1,2…N, the selected indicators are: First indicator: Geometric mean energy Second indicator: Modulation intensity geometric mean where x n Represents meshing frequency and high-order harmonics, x n-fm Sidebands representing the meshing frequency and its multiples; The third indicator: energy arithmetic mean The fourth indicator: the arithmetic mean of modulation intensity The frequency domain indexes are calculated for each mechanical component of the complex transmission system. i ,f_Set(S i ) contains each axis frequency and its multiple, calculates the first and third indexes, and forms Index_Set1(S i ) and Index_Set3(S i ) ; for gear G j ,f_Set(G j ) includes each meshing frequency and its multiple frequency, calculates the first to fourth indexes, and forms Index_Set1(G j )、Index_Set2(G j )、Index_Set3(G j ) and Index_Set4(G j ); finally, each frequency domain index is combined into a frequency domain index feature library of complex transmission system.

10. The vibration signal frequency domain feature extraction system for complex transmission system according to claim 9, characterized in that: According to the historical data of the transmission system, the mean and variance of each frequency domain index are calculated. For one of the indexes, k , assuming that the mean and variance of its historical operating data are μ k and σ k , if the indicator exceeds μ in a certain measurement k +3σ k , it is considered that the component corresponding to the indicator has a suspected fault, an alarm is issued and component information is provided; for example, if a shaft S i If the corresponding first or third indicator exceeds the limit, the shaft is considered to have a suspected fault, an alarm is issued and component information is provided; for example, a gear G j If one of the corresponding first to fourth indicators exceeds the limit, it is considered that the gear has a suspected fault, an alarm is issued and component information is provided.

Images