FOA-VMD-Based Parameter Optimization Bearing Fault Diagnosis Method and System

Abstract

The present invention relates to a parameter optimization bearing fault diagnosis method and system based on FOA-VMD, which includes: using the fruit fly algorithm to perform a global optimal combination search on the VMD decomposition parameters of the penalty factor and the number of decompositions, selecting the information entropy increment combined with the kurtosis index as the objective function to obtain the optimal parameter combination; decomposing the original fault vibration signal by VMD after obtaining the optimal parameter combination to obtain several intrinsic mode components, and selecting the best component; performing envelope analysis processing on the best component, and further processing the best component by using the 1.5D spectrum to eliminate the interference of other components and highlight the fault characteristic frequency and its multiple frequencies, so as to obtain the fault type of the bearing. The present invention can effectively obtain the fault characteristics of the rolling bearing and improve the recognition accuracy of the fault characteristics. The present invention can be applied in the technical field of mechanical equipment fault diagnosis.

Description

Technical Field

[0001] The present invention relates to the technical field of mechanical equipment fault diagnosis, and particularly to a bearing fault diagnosis method and system with parameter optimization based on FOA-VMD. Background Technique

[0002] The health of rolling bearings is particularly important for the operation and maintenance of the system. It is necessary to timely discover the root cause of the fault and take preventive measures. However, when an accident occurs, the fault information interferes with each other, and it is very difficult to effectively extract it. Therefore, how to strengthen and extract more obvious features is particularly important and is the core of the diagnosis process.

[0003] Empirical mode decomposition is a common and practical signal processing method studied in recent years. Since its proposal, it has been widely used in the diagnosis of bearings. Its basic idea is to decompose a complex sequence into several components, which have different characteristic scales and show regularity compared with before decomposition. After decomposition, problems such as endpoint effect, envelope fitting, and mode mixing will occur. How to solve the deficiencies of EMD decomposition is the key to further research on signal processing.

[0004] Moreover, the periodic impact characteristics corresponding to the fault characteristic frequencies of different parts of the rolling bearing are also different, and different objective functions and optimal component processing methods have a great impact on the diagnosis results of VMD decomposition. Summary of the Invention

[0005] Aiming at the problem that it is difficult to extract the frequency characteristics when the rolling bearing fails, the purpose of the present invention is to provide a bearing fault diagnosis method and system with parameter optimization based on FOA-VMD, which can effectively obtain the fault characteristics of the rolling bearing and improve the recognition accuracy of the fault characteristics.

[0006] To achieve the above purpose, the present invention adopts the following technical solutions: A bearing fault diagnosis method with parameter optimization based on FOA-VMD, which includes: using the fruit fly algorithm to perform a global optimal combination search on the penalty factor and the number of decompositions of the VMD decomposition parameters, selecting the information entropy increment combined with the kurtosis index as the objective function to obtain the optimal parameter combination; decomposing the original fault vibration signal by VMD after obtaining the optimal parameter combination to obtain several intrinsic mode components, and selecting the best component; performing envelope analysis processing on the best component, and further processing the best component using a 1.5D spectrum to eliminate the interference of other components and highlight the fault characteristic frequency and its multiple frequencies to obtain the fault type of the bearing.

[0007] Further, the taste concentration value function of the fruit fly algorithm is the objective function S:

[0008] S = H i + 1 / abs(qd i - 3)

[0009] Among them, H i represents the information entropy increment of the i-th fruit fly, and qd i represents the kurtosis index of the i-th fruit fly, and abs represents the absolute value.

[0010] Furthermore, the information entropy increment combined with the kurtosis index is selected as the objective function, and the optimization objective is determined according to the objective function, including:

[0011] When the i-th fruit fly in the population is at different positions, find all the taste concentration values at this time. The component corresponding to the taste concentration value is the optimal component, and take as the optimization objective.

[0012] Furthermore, the fruit fly algorithm is used to perform a global optimal combination search on the penalty factor and the number of decompositions of the VMD decomposition parameters. The information entropy increment combined with the kurtosis index is selected as the objective function to obtain the optimal parameter combination, including:

[0013] Initialize the maximum number of iterations of the fruit fly algorithm, the initial position coordinate parameters [X_axis, Y_axis], and this position coordinate represents the parameter combination [α, k], where α and k are the penalty factor and the number of decompositions respectively;

[0014] Obtain the random direction and distance between the fruit fly individual and finding food;

[0015] When the individual is at different positions, first estimate its distance D from the origin, take the reciprocal of the distance D as the taste concentration judgment value, use the judgment value as the input of the concentration judgment function, obtain the fitness values of the fruit fly individual at different positions according to the objective function, and compare the magnitudes of the fitness values;

[0016] Iteratively update to obtain the fruit fly with the lowest fitness value in the fruit fly population. If its taste concentration value is smaller than the best taste concentration value of the previous generation, then retain the taste concentration value and its coordinates, and all fruit flies move closer to this fruit fly;

[0017] When the number of iterations reaches the preset maximum number of iterations, stop the iteration and output the best position and the best taste concentration value of the individual.

[0018] Furthermore, the original fault vibration signal is decomposed by VMD after obtaining the optimal parameter combination to obtain several intrinsic mode components, and the best component is selected, including:

[0019] Calculate the information entropy of each IMF component, normalize the information entropy of each IMF component, obtain the information entropy increment value of each IMF component according to the normalized value, and obtain the objective function with the information entropy increment value and the kurtosis index;

[0020] The objective function values of each of the IMF components are obtained by solving the said objective function, and the IMF component corresponding to the minimum objective function value is taken as the optimal component; the smaller the objective function value of the IMF component signal, the easier it is to extract the fault frequency.

[0021] Further, the information entropy increment H of each of the IMF components i is calculated as follows:

[0022] H i =Δq n =-p(n)log2(p(n))

[0023] where n = 1, 2, …, N, and N represents the total number of IMF components obtained after VMD decomposition; p(n) is the value after information entropy normalization processing.

[0024] Further, after performing envelope analysis on the optimal component, the 1.5D spectrum is used to further process the optimal component, including:

[0025] Obtain the third-order cumulant R 3X (τ1, τ2) slice of the optimal component, and perform one-dimensional Fourier transform on the third-order cumulant R 3X (τ1, τ2) slice to obtain the 1.5D spectrum.

[0026] A bearing fault diagnosis system based on FOA-VMD for parameter optimization, which includes: an optimization module that uses the fruit fly algorithm to perform a global optimal combination search on the penalty factor and the number of decompositions of the VMD decomposition parameters, selects the information entropy increment combined with the kurtosis index as the objective function, and obtains the optimal parameter combination; an optimal component selection module that decomposes the original fault vibration signal by VMD after obtaining the optimal parameter combination to obtain several intrinsic mode components, and selects the optimal component; an analysis and processing module that performs envelope analysis on the optimal component and further processes the optimal component using the 1.5D spectrum to eliminate the interference of other components and highlight the fault characteristic frequency and its multiples to obtain the fault type of the bearing.

[0027] A computer-readable storage medium storing one or more programs, characterized in that the one or more programs include instructions that, when executed by a computing device, cause the computing device to execute any of the above methods.

[0028] A computing device, which includes: one or more processors, a memory, and one or more programs, where the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the one or more programs include instructions for executing any of the above methods.

[0029] Due to the above technical solutions adopted by the present invention, it has the following advantages:

[0030] In order to better enhance the accuracy of the diagnosis process and diagnose different fault types of bearings, this invention uses the fruit fly optimization algorithm to optimize the VMD decomposition parameters, improves the smell concentration determination function (objective function) in the fruit fly algorithm, performs 1.5D spectrum processing on the optimal components, and eliminates the interference of some redundant components and other factors. Based on the theory, actual test signals are selected for verification, effectively obtaining the fault characteristics of the rolling bearing and enhancing the recognition accuracy of the fault characteristics. Description of the Drawings

[0031] Figure 1 It is a schematic flow chart of the bearing fault diagnosis method based on FOA-VMD parameter optimization in an embodiment of this invention;

[0032] Figure 2 It is a flow chart of using the fruit fly algorithm to optimize the VMD decomposition parameters in an embodiment of this invention;

[0033] Figure 3a It is a time domain diagram of the bearing outer ring fault signal after adding noise in an embodiment of this invention;

[0034] Figure 3b It is an envelope diagram of the bearing outer ring fault signal after adding noise in an embodiment of this invention;

[0035] Figure 4 It is an iteration result diagram of the optimal target value (concentration value) in an embodiment of this invention;

[0036] Figure 5a It is a time domain diagram of the IMF4 component in an embodiment of this invention;

[0037] Figure 5b It is an envelope diagram of the IMF4 component in an embodiment of this invention;

[0038] Figure 6 It is a 1.5D spectrum processing diagram of the bearing outer ring fault signal in an embodiment of this invention;

[0039] Figure 7a It is a time domain diagram of the rolling element fault signal after adding noise in an embodiment of this invention;

[0040] Figure 7b It is an envelope diagram of the rolling element fault signal after adding noise in an embodiment of this invention;

[0041] Figure 8a It is a time domain diagram of the IMF5 component in an embodiment of this invention;

[0042] Figure 8b It is an envelope diagram of the IMF5 component in an embodiment of this invention;

[0043] Figure 9 It is the 1.5D spectrum processing diagram of the rolling element fault signal in an embodiment of the present invention. Specific Embodiments

[0044] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the described embodiments of the present invention fall within the scope of protection of the present invention.

[0045] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present application. As used herein, unless the context clearly indicates otherwise, the singular forms are also intended to include the plural forms. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they specify the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0046] The method and system for parameter-optimized bearing fault diagnosis based on FOA-VMD provided by the present invention can effectively solve the problem of difficult extraction of frequency characteristics when rolling bearings fail. It uses the fruit fly algorithm to optimize the extraction of fault characteristics of VMD parameters and simultaneously improves the selection of the objective function. The present invention includes: First, use the fruit fly algorithm (FOA) to perform a global optimal combination search on the penalty factor and the number of decompositions of the VMD decomposition parameters, select the information entropy increment combined with the kurtosis index as the objective function, and after obtaining the optimal parameter combination, perform VMD processing on the signal to obtain several intrinsic mode components. Then perform envelope analysis on the best component. To reduce the interference of redundant components and noise, finally, use the 1.5D spectrum to further study the best component, thereby diagnosing the fault type of the bearing. The measured fault signals are used to verify the present invention, and the results show that the present invention can effectively extract the frequency characteristics of the fault signals, proving that the present invention has a certain degree of accuracy and research value.

[0047] In an embodiment of the present invention, as Figure 1 shown, a method for parameter-optimized bearing fault diagnosis based on FOA-VMD is provided. In this embodiment, the method includes the following steps:

[0048] 1) Use the fruit fly algorithm to perform a global optimal combination search on the penalty factor and the number of decompositions of the VMD decomposition parameters, select the information entropy increment combined with the kurtosis index as the objective function, and obtain the optimal parameter combination;

[0049] 2) Decompose the original fault vibration signal by VMD after obtaining the optimal parameter combination to obtain several intrinsic mode components, and select the best component;

[0050] 3) After performing envelope analysis on the optimal component, further process the optimal component using the 1.5D spectrum to eliminate interference from other components and highlight the fault characteristic frequency and its multiples, so as to obtain the fault type of the bearing; among them, the signal diagram obtained through the 1.5D spectrum processing can more clearly highlight the fault characteristic signal frequency and its multiples, and the fault type can be judged according to the values of the fault frequency and its multiples.

[0051] In the above steps, the essence of VMD decomposition is to reconstruct the fault signal through variational constraints, and the obtained intrinsic mode function U k (t) is:

[0052] U k (t) = A k (t)cos(Φ k (t)) (1)

[0053] In the formula, A k (t) is the instantaneous amplitude of U k (t), A K (t) ≥ 0; ω k (t) is the instantaneous frequency of U k (t); Φ k (t) represents the instantaneous phase, and t represents time.

[0054] The constrained variational model formula of the component is:

[0055]

[0056] In the formula, represents the partial derivative, f represents the input signal x(t), and δ is the Dirac distribution.

[0057] The above formula (2) is used to describe the process of demodulating the signal through Gaussian smoothing (the square root of the L2 norm gradient) to obtain the bandwidth of each mode function.

[0058] VMD decomposition includes the following steps:

[0059] (1) Initialize {u k 1}, {ω k 1}, λ 1 and n to 0;

[0060] (2) n = n + 1, and loop sequentially;

[0061] (3) According to {u k} and {ω kUpdate according to the update formula of {} until the number of decompositions reaches K, then stop the inner loop;

[0062] (4) According to Update λ;

[0063] (5) Given the precision ε, if the stopping condition is satisfied, the update ends; otherwise, continue to loop under step (2), and output to obtain K components.

[0064] Among them, the stopping condition is:

[0065] In the formula, {u k} are the K IMF signal components obtained by decomposition; {ω k} are the central frequencies of the IMF components; f is the input signal; k is the number of decompositions.

[0066] In this embodiment, the Fruit Fly Optimization Algorithm (FOA) evolves the foraging actions of fruit flies into a global optimization method, and the basic parameters are as follows:

[0067] The size of the fruit fly tribe is Sizepop;

[0068] The maximum number of iterations is Maxgen;

[0069] The initial position of the fruit fly tribe is:

[0070] X_axis and Y_axis (3) The random distance and direction between the fruit fly individual and finding food are:

[0071] X i = X_axis + RandomValue (4)

[0072] Y i = Y_axis + RandomValue

[0073] The distance D isti between the fruit fly individual and the origin and the taste concentration judgment value S i are respectively:

[0074]

[0075] S i = 1 / Disti

[0076] The taste concentration value Smell of the fruit fly individual position is:

[0077] Smell = Function(S i ) (6)

[0078] The Drosophila with the best concentration in the Drosophila tribe is:

[0079] Smellbest = bestSmell

[0080] X_axis = X(bestlndex)

[0081] Y_axis = Y(bestlndex) (7)

[0082] In the above step 1), the taste concentration value function of the Drosophila algorithm is the objective function S, which is:

[0083] S = H i + 1 / abs(qd i - 3)

[0084] Among them, H i represents the information entropy of the i-th Drosophila, and qd i represents the kurtosis index of the i-th Drosophila, and abs represents the absolute value.

[0085] In the above step 1), when performing VMD decomposition, the setting of the decomposition number and the penalty factor parameter has a relatively obvious impact on the decomposition. To obtain the best parameter combination [K, α], in this embodiment, the Drosophila global optimization algorithm is used to optimize it.

[0086] Select the information entropy increment combined with the kurtosis index as the objective function, and determine the optimization objective according to the objective function. Specifically: when the i-th Drosophila in the population is at different positions, calculate all the taste concentration values at this time, and the component corresponding to the taste concentration value is the optimal component, and take as the optimization objective.

[0087] In the above step 1), the Drosophila algorithm is used to perform a global optimal combination search for the VMD decomposition parameters, the penalty factor and the decomposition number. Select the information entropy increment combined with the kurtosis index as the objective function to obtain the optimal parameter combination, as Figure 2 shown, including the following steps:

[0088] 1.1) Initialize the maximum number of iterations Maxgen of the Drosophila algorithm, and the initial position coordinate parameters [X_axis, Y_axis]. This position coordinate represents the parameter combination [α, k], where α and k are the penalty factor and the decomposition number respectively;

[0089] 1.2) Obtain the random direction and distance between the Drosophila individual and finding food;

[0090] 1.3) When the individual is at different positions, first estimate its distance D from the origin, take the reciprocal of the distance D as the taste concentration judgment value, use the judgment value as the input of the concentration judgment function, and obtain the fitness value of the fruit fly individual at different positions according to the objective function. And compare the magnitudes of the fitness values.

[0091] 1.4) Iteratively update to obtain the fruit fly with the lowest fitness value in the fruit fly population. If its taste concentration value is smaller than the best taste concentration value of the previous generation, then retain the taste concentration value and its coordinates, and all fruit flies move closer to this fruit fly.

[0092] 1.5) Repeat the above steps 1.2) to 1.4). When the number of iterations reaches the preset maximum number of iterations, stop the iteration and output the best position and the best taste concentration value of the individual.

[0093] In the above step 2), decompose the original fault vibration signal by VMD after obtaining the optimal parameter combination to obtain several intrinsic mode functions, and select the best component, including the following steps:

[0094] 2.1) Calculate the information entropy of each IMF component, normalize the information entropy of each IMF component, and obtain the information entropy increment value of each IMF component according to the normalized value. Use the information entropy increment value and the kurtosis index to obtain the objective function; since the objective function is composed of the information entropy increment combined with the kurtosis index, it can better reflect the signal fault characteristics.

[0095] 2.2) Solve the objective function value of each IMF component according to the objective function, and take the IMF component corresponding to the smallest objective function value as the best component; the smaller the objective function value of the IMF component signal, the easier it is to extract the fault frequency; since both the information entropy increment and the kurtosis value are evaluation standard functions and are measures reflecting the complexity of the signal, combining them makes the diagnosis more accurate.

[0096] In the above step 2.1), the information entropy increment value H i is calculated as follows:

[0097] H i = Δq n = -p(n)log2(p(n))

[0098] In the formula, n = 1, 2,..., N, where N represents the total number of IMF components obtained after VMD decomposition; p(n) is the value after information entropy normalization.

[0099] In this embodiment, when the fruit fly algorithm is optimized, a smell concentration function, i.e., the objective function, needs to be determined. In the initial stage of a fault, different fault points will bring corresponding changes in the signal amplitude and period, and will also have different effects on the complexity of the vibration signal. Since the fault signal will obtain a certain number of IMF components after VMD decomposition, each component has its corresponding characteristic information. From the arrangement of the central frequencies to the magnitude of the energy, they all reflect the complexity of a component system. The noisy signal contains not only the characteristics of the signal itself but also the characteristic frequencies corresponding to the faults. To strengthen the selection of the characteristics of the noisy signal, a signal system with low complexity and few interference components is required.

[0100] In view of all the components obtained after VMD decomposition, the optimal component can best reflect the characteristics of the signal. Information entropy can express the measure of the complexity of the signal system. Based on this, the information entropy increment is proposed. If the information entropy increment is smaller, it means that the types of system changes are fewer, the number of different situations is less, and the fault frequency of the signal is easier to extract; on the contrary, if the information entropy increment is larger, the system complexity is higher, the interference components are more, and it is difficult to strengthen the clear fault characteristic frequency.

[0101] After the fault signal x(t) is decomposed by VMD, it obtains components x1(t), x2(t),..., x n (t) (n = 1, 2,..., N). The process of selecting the best component based on the complexity of the IMF information entropy increment is as follows:

[0102] Calculate the information entropy H(x n (t)) of each IMF component:

[0103]

[0104] Normalize the information entropy of each IMF component:

[0105]

[0106] Calculate the information entropy increment value of each IMF component:

[0107] H i =Δq n =-p(n)log2(p(n)) (10)

[0108] However, considering the disadvantage that the common objective function has an unclear feedback on the impact characteristics of the signal, in this embodiment, the kurtosis index is fused on the basis of the entropy value. It is extremely sensitive to the subtle changes in the fault signal and can be well applied to the early fault judgment. The kurtosis index is:

[0109]

[0110] where μ n is the mean value of IMF n (x(t)); σ n is the variance of IMF n (x(t)).

[0111] When the rolling bearing is in a healthy state, its vibration signal is normally distributed, and the kurtosis value is 3. When faults occur in different parts, the vibration signal is affected and shows a certain deviation from the normal distribution. The vibration signal reflects the fault information of the bearing. The larger the kurtosis value, the more fault information there is, and the more obvious the fault characteristics are. In summary, combining the kurtosis value and the information entropy increment value as a new objective function can more effectively extract the characteristic frequency of the fault signal. The flavor concentration value function (objective function S) is:

[0112] S = H i + 1 / abs(qd i - 3) (12)

[0113] In the above step 3), after performing envelope analysis on the optimal component, the 1.5D spectrum is used to further process the optimal component. Specifically: calculate the third-order cumulant R 3X (τ1, τ2) slice of the optimal component, and perform one-dimensional Fourier transform on the third-order cumulant R 3X (τ1, τ2) slice to obtain the 1.5D spectrum.

[0114] In this embodiment, in the complex domain, the 1.5D spectrum processing is performed on the vibration signal to obtain the 1.5D envelope spectrum generated by coupling.

[0115] The third-order cumulant R 3X (τ1, τ2) slice of the zero-mean stationary signal x(t) is defined as:

[0116] R 3x (τ, τ) = E{x(t)x(t + τ)x(t + τ)} (13)

[0117] R 3x (τ, τ) (τ1 = τ2 = τ)

[0118] where E{·} is the mathematical expectation, and the one-dimensional Fourier transform of R 3x (τ, τ) is the 1.5D spectrum B(ω) of x(t):

[0119]

[0120] After the signal is envelope-processed, the 1.5D spectrum method can eliminate the interference of other components. While suppressing white noise and strengthening the fundamental frequency component, it can also effectively remove non-coupled harmonic components, and can highlight the fault characteristic frequency more than only performing envelope demodulation processing.

[0121] Example 1:

[0122] Taking the fault analysis of the bearing outer ring as an example, the measured signal is analyzed. To prove the applicability of the method of the present invention, the bearing fault data provided by Case Western Reserve University is selected for verification. The model of the driving end bearing used is SKF6205, the fault size is 0.07 inches, the electronic load is 1 horsepower, the sampling frequency is 12 kHz, the rotational speed is 1797 r / min, the rotational frequency is 29.95 Hz, and a single-point damage is caused to the bearing outer ring by electrical discharge machining. By calculating the fault frequency f of the outer ring i = 107.18 Hz. Considering that the noise contained in the artificially implanted signal is too small, Gaussian white noise with a signal-to-noise ratio of -3 dB is added in this embodiment to make the vibration signal closer to the actual situation. The time-domain diagram of the signal after adding noise is as shown in Figure 3a shown, and the envelope spectrum obtained after envelope processing is as shown in Figure 3b shown.

[0123] From Figure 3b it can be seen that it is difficult to see the fault frequency only by envelope analysis for the signal after adding white noise, and the original signal is greatly affected. Using the method proposed by the present invention for processing, the iterative results of the optimal target value (concentration value) are as shown in Figure 4 shown. The optimal parameter combination found by optimization is [k, α] = [2478, 6]. The VMD decomposition of the original signal after parameter optimization obtains the spectra of 6 IMF components. It can be seen by observation that, compared with other components, the waveform law in IMF4 is more obvious, its frequency band center is around 3000 Hz, near the resonance frequency, and this classification is the best component. The time-domain diagram of the IMF4 component is as shown in Figure 5a shown, and after envelope processing, it is as shown in Figure 5b shown.

[0124] From Figure 5b it can be seen that this component extracts the impact part with obvious amplitude in the original signal containing noise, and relatively clear and obvious peaks appear around the bearing outer ring fault frequency f0. At the same time, there are some corresponding relationships between the rotational frequency f r , the outer ring frequency f0: f o -f r = 77.64 Hz, f0 + f r = 137.7 Hz, the second harmonic (2f o = 215.3 Hz) and the third harmonic (3f0 = 323 Hz) of the outer ring fault frequency can also be found. From this, it can be seen that a fault has occurred in the rolling bearing outer ring.

[0125] Finally, it is processed by the 1.5D spectrum, and the results are as follows Figure 6As shown, it is clearly found that f0 (107.5 Hz), its second harmonic 2f0 = 107.5 × 2 = 215 Hz, and its third harmonic 3f0 = 107.5 × 3 = 322.5 Hz, effectively identifying the fault of the outer ring of the bearing.

[0126] Embodiment 2:

[0127] Taking the rolling element fault analysis as an example, the effectiveness of the method of the present invention when faults occur in different parts is further verified, as well as its practical usability. A test bench is built for diagnostic testing with a planetary gearbox as the carrier. Since the rolling bearing is an important component in the gearbox system, its diagnosis is also very important. The above embodiment verified the fault identification experiment of the outer ring of the bearing, and this experiment will verify the fault diagnosis of the rolling elements of the bearing. The faulty parts are installed in the gearbox.

[0128] The sampling frequency of this experiment is 12.50 kHz, the motor speed is 1980 r / min, and the time-domain diagram of the collected vibration signal is as Figure 7a shown, and the envelope diagram is as Figure 7b shown. The fault frequency characteristics cannot be found only from the envelope diagram, and then its optimized analysis is carried out.

[0129] First, optimize the parameters of VMD decomposition. The best parameter combination found after the optimization process is [k, α] = [3478, 5]. The original signal is decomposed into the spectra of 5 IMF components, and IMF5 is selected as the best component by observation. The time-domain diagram of the IMF5 component is as Figure 8a shown, and after envelope processing, it is as Figure 8b shown.

[0130] From Figure 8b it can be observed that a relatively clear and obvious peak appears at the rolling element fault frequency f d = 58.59 Hz. Observe the relationship between the bearing rotation frequency f r = 17 Hz and the rolling element fault frequency f d . f d ± f r are exactly the peaks appearing on the left and right sides, 41.75 Hz and 75.44 Hz, and it is found that the second harmonic (2f d = 117.18 ≈ 117.2) and the third harmonic (3f d = 175.77 Hz ≈ 175.8) of the rolling element fault frequency also coincide with the peaks appearing in the figure. Thus, it can be seen that the rolling elements of the rolling bearing have failed. Finally, process it with a 1.5D spectrum, and the result is as Figure 9 shown. It is clearly found that f d and its harmonics, effectively diagnosing the rolling element fault.

[0131] In summary, since the fault signals of rolling bearings are vulnerable to noise interference, the present invention extracts the fault features of signals by using VMD decomposition with optimized parameters, and successfully realizes fault judgment. The improved fruit fly algorithm is used to optimize the selection of the objective function, and the comprehensive function of the information entropy increment and kurtosis value is selected, which not only improves the accuracy of finding the best parameter combination by optimization, but also enriches the feature information of the best component. It is found through experimental verification that the results are accurate, effective and have significant advantages.

[0132] In an embodiment of the present invention, a bearing fault diagnosis system with parameter optimization based on FOA-VMD is provided, which includes:

[0133] An optimization module that uses the fruit fly algorithm to perform a global optimal combination search on the VMD decomposition parameters, namely the penalty factor and the number of decompositions, selects the information entropy increment combined with the kurtosis index as the objective function, and obtains the optimal parameter combination;

[0134] An optimal component selection module that decomposes the original fault vibration signal by VMD after obtaining the optimal parameter combination to obtain several intrinsic mode components, and selects the optimal component;

[0135] An analysis and processing module that performs envelope analysis on the optimal component and further processes the optimal component using a 1.5D spectrum to eliminate the interference of other components and highlight the fault characteristic frequency and its multiples, so as to obtain the fault type of the bearing.

[0136] The system provided in this embodiment is used to execute the above method embodiments. For the specific process and detailed content, please refer to the above embodiments and will not be elaborated here.

[0137] In an embodiment of the present invention, a computing device structure is provided. The computing device may be a terminal, which may include: a processor, a communications interface, a memory, a display screen, and an input device. Among them, the processor, the communications interface, and the memory complete communication with each other through a communication bus. The processor is used to provide computing and control capabilities. The memory includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. When the computer program is executed by the processor, a bearing fault diagnosis method is implemented; the internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communications interface is used to communicate with an external terminal in a wired or wireless manner. The wireless manner can be implemented through WIFI, a carrier network, NFC (Near Field Communication), or other technologies. The display screen may be a liquid crystal display screen or an electronic ink display screen. The input device may be a touch layer covering the display screen, or a button, a trackball, or a touchpad provided on the housing of the computing device, or an external keyboard, touchpad, or mouse, etc. The processor can call the logical instructions in the memory to execute the following method: use the fruit fly algorithm to perform a global optimal combination search on the penalty factor and the number of decompositions of the VMD decomposition parameters, select the information entropy increment combined with the kurtosis index as the objective function to obtain the optimal parameter combination; decompose the original fault vibration signal by VMD after obtaining the optimal parameter combination to obtain several intrinsic mode components, and select the best component; perform envelope analysis processing on the best component, and further process the best component using a 1.5D spectrum to eliminate the interference of other components and highlight the fault characteristic frequency and its multiples to obtain the fault type of the bearing.

[0138] In addition, when the logical instructions in the above-mentioned memory are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The foregoing storage medium includes: various media such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disc that can store program codes.

[0139] Those skilled in the art can understand that the structure of the above computing device is only a part of the structure related to the solution of this application, and does not constitute a limitation on the computing device to which the solution of this application is applied. The specific computing device may include more or fewer components, or combine some components, or have different component arrangements.

[0140] In an embodiment of the present invention, a computer program product is provided. The computer program product includes a computer program stored on a non-transitory computer-readable storage medium. The computer program includes program instructions. When the program instructions are executed by a computer, the computer can execute the methods provided in the above method embodiments. For example, it includes: using the fruit fly algorithm to perform a global optimal combination search on the penalty factor and the number of decompositions of the VMD decomposition parameters, selecting the information entropy increment combined with the kurtosis index as the objective function to obtain the optimal parameter combination; decomposing the original fault vibration signal by VMD after obtaining the optimal parameter combination to obtain several intrinsic mode components, and selecting the best component; after performing envelope analysis processing on the best component, and further processing the best component using the 1.5D spectrum to eliminate the interference of other components and highlight the fault characteristic frequency and its multiples to obtain the fault type of the bearing.

[0141] In an embodiment of the present invention, a non-transitory computer-readable storage medium is provided. The non-transitory computer-readable storage medium stores server instructions. The computer instructions enable the computer to execute the methods provided in the above embodiments. For example, it includes: using the fruit fly algorithm to perform a global optimal combination search on the penalty factor and the number of decompositions of the VMD decomposition parameters, selecting the information entropy increment combined with the kurtosis index as the objective function to obtain the optimal parameter combination; decomposing the original fault vibration signal by VMD after obtaining the optimal parameter combination to obtain several intrinsic mode components, and selecting the best component; after performing envelope analysis processing on the best component, and further processing the best component using the 1.5D spectrum to eliminate the interference of other components and highlight the fault characteristic frequency and its multiples to obtain the fault type of the bearing.

[0142] The implementation principle and technical effect of the computer-readable storage medium provided in the above embodiment are similar to those of the above method embodiment, and will not be elaborated here.

[0143] This application is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of this application. It should be understood that each process and / or block in the flowchart and / or block diagram, and the combination of processes and / or blocks in the flowchart and / or block diagram can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate for implementing in the processFigure 1 one or more processes and / or blocks Figure 1 means for the functions specified in one or more blocks

[0144] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to work in a particular manner, such that the instructions stored in the computer-readable memory produce a manufacture including an instruction means that implements the functions in the process Figure 1 one or more processes and / or blocks Figure 1 specified in one or more blocks

[0145] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus, such that a series of operational steps are performed on the computer or other programmable apparatus to produce a computer-implemented process, so that the instructions executed on the computer or other programmable apparatus provide steps for implementing the functions specified in the process Figure 1 one or more processes and / or blocks Figure 1 specified in one or more blocks

[0146] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A bearing fault diagnosis method for parameter optimization based on FOA-VMD, characterized in that Including: Using the fruit fly algorithm to perform a global optimal combination search on the penalty factor and the number of decompositions of the VMD decomposition parameters, selecting the information entropy increment combined with the kurtosis index as the objective function, and obtaining the optimal parameter combination; Performing VMD decomposition on the original fault vibration signal with the obtained optimal parameter combination to obtain several intrinsic mode components, and selecting the best component; After performing envelope analysis on the best component, further processing the best component using a 1.5D spectrum to eliminate the interference of other components and highlight the fault characteristic frequency and its multiples to obtain the fault type of the bearing; The smell concentration value function of the fruit fly algorithm is the objective function S: S = H i + 1 / abs(qd i - 3) Among them, H i represents the information entropy increment of the i-th fruit fly, and qd i represents the kurtosis index of the i-th fruit fly, where abs represents the absolute value; The information entropy increment combined with the kurtosis index is selected as the objective function, and the optimization objective is determined according to the objective function, including: When the $i$-th fruit fly in the population is at different positions, find all the taste concentration values at this time The component corresponding to the taste concentration value is the optimal component, and take as the optimization objective.

2. The parameter optimization bearing fault diagnosis method based on FOA-VMD according to claim 1, characterized in that, The global optimal combination search for the penalty factor and the number of decompositions of the VMD decomposition parameters using the fruit fly algorithm, selecting the information entropy increment combined with the kurtosis index as the objective function, and obtaining the optimal parameter combination, including: Initializing the maximum number of iterations of the fruit fly algorithm, the initial position coordinate parameters [X_axis, Y_axis], and this position coordinate represents the parameter combination [α, k], where α and k are the penalty factor and the number of decompositions respectively; Obtaining the random direction and distance between the fruit fly individuals and finding food; When the individuals are in different positions, first estimate their distance D from the origin, take the reciprocal of the distance D as the smell concentration judgment value, use the judgment value as the input of the concentration judgment function, obtain the fitness values of the fruit fly individuals at different positions according to the objective function, and compare the magnitudes of the fitness values; Iteratively update to obtain the fruit fly with the lowest fitness value in the fruit fly population. If its smell concentration value is smaller than the best smell concentration value of the previous generation, then retain this smell concentration value and its coordinates, and all fruit flies move closer to this fruit fly; When the number of iterations reaches the preset maximum number of iterations, stop the iteration and output the best position and the best smell concentration value of the individual.

3. The parameter-optimized bearing fault diagnosis method based on FOA-VMD according to claim 1, wherein The performing VMD decomposition on the original fault vibration signal with the obtained optimal parameter combination to obtain several intrinsic mode components, and selecting the best component, including: Calculating the information entropy of each IMF component, normalizing the information entropy of each IMF component, and obtaining the information entropy increment value of each IMF component according to the normalized value, and obtaining the objective function with the information entropy increment value and the kurtosis index; Solving the objective function value of each IMF component according to the objective function, and taking the IMF component corresponding to the smallest objective function value as the best component; the smaller the objective function value of the IMF component signal, the easier it is to extract the fault frequency.

4. The parameter-optimized bearing fault diagnosis method based on FOA-VMD according to claim 3, characterized in that The information entropy increment H of each IMF component i is calculated as follows: H i = Δq n = -p(n) log2(p(n)) Where n = 1, 2,..., N, and N represents the total number of IMF components obtained after VMD decomposition; p(n) is the value after information entropy normalization.

5. The parameter-optimized bearing fault diagnosis method based on FOA-VMD according to claim 1, wherein After performing envelope analysis on the best component, further processing the best component using a 1.5D spectrum, including: Obtain the third-order cumulant \(R\) of the optimal component 3X (\(\tau_1,\tau_2\)) slice, and for the third-order cumulant \(R\) 3X (\(\tau_1,\tau_2\)) slice is subjected to one-dimensional Fourier transform to obtain a 1.5D spectrum.

6. A bearing fault diagnosis system with parameter optimization based on FOA-VMD, characterized in that, Including: An optimization module that uses the fruit fly algorithm to perform a global optimal combination search on the penalty factor and the number of decompositions of the VMD decomposition parameters, selects the information entropy increment combined with the kurtosis index as the objective function, and obtains the optimal parameter combination; The optimal component selection module decomposes the original fault vibration signal by VMD after obtaining the optimal parameter combination, obtains several intrinsic mode components, and selects the optimal component; The analysis and processing module performs envelope analysis on the optimal component, and further processes the optimal component using the 1.5D spectrum to eliminate the interference of other components and highlight the fault characteristic frequency and its multiples, so as to obtain the fault type of the bearing; The taste concentration value function of the fruit fly algorithm is the objective function S: S = H i + 1 / abs(qd i - 3) Among them, H i represents the information entropy increment of the i-th fruit fly, and qd i represents the kurtosis index of the i-th fruit fly, and abs represents the absolute value; The information entropy increment and kurtosis index are selected as the objective function, and the optimization objective is determined according to the objective function, including: When the $i$-th fruit fly in the population is at different positions, find all the taste concentration values at this time The component corresponding to the taste concentration value is the optimal component, and take as the optimization objective.

7. A computer-readable storage medium storing one or more programs, characterized in that, The one or more programs include instructions that, when executed by a computing device, cause the computing device to perform any of the methods described in claims 1 to 5.

8. A computing device, characterized in that, Including: One or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs include instructions for performing any of the methods described in claims 1 to 5.

Images