Intelligent fault diagnosis method, system and storage medium for mechanical equipment

Abstract

The present invention provides a mechanical equipment intelligent fault diagnosis method, system and storage medium, belonging to the technical field of intelligent mechanical equipment, by collecting vibration data of the mechanical equipment to be detected; according to the collected vibration data of the mechanical equipment to be detected, determine the sample to be detected and the spectrum characteristics of the sample to be detected; according to the spectrum characteristics of the sample to be detected and the pre-acquired spectrum integration complete dictionary, determine the optimized sparse coding of the spectrum characteristics of the sample to be detected for the spectrum integration complete dictionary; and decompose the optimized sparse coding into sub-sparse coding; according to the sub-sparse coding and the spectrum integration complete dictionary, determine the spectrum sparse reconstruction error; through the health state intelligent identification strategy based on the minimum discrimination criterion of the spectrum sparse reconstruction error, determine the health state of the mechanical equipment to be detected according to the spectrum sparse reconstruction error. The present invention has the characteristics of superior accuracy, noise robustness, hyperparameter selection robustness and high computational efficiency.

Description

Technical Field

[0001] The present invention belongs to the field of intelligent mechanical equipment, and specifically relates to a mechanical equipment intelligent fault diagnosis method, system and storage medium. Background Art

[0002] The condition monitoring and fault diagnosis technology of mechanical equipment is of great significance in reducing major safety accidents and greatly saving the operation and maintenance costs of major high-end mechanical equipment such as aircraft engines, heavy gas turbines, wind turbines, helicopters, etc.

[0003] The cutting-edge intelligent fault diagnosis methods for mechanical equipment mainly include intelligent fault diagnosis methods for mechanical equipment based on deep learning models and intelligent fault diagnosis methods for mechanical equipment based on discriminant sparse representation. Among them, the intelligent fault diagnosis method for mechanical equipment based on deep learning models realizes intelligent fault diagnosis through the stacking of neural networks and the design of deep network architecture. Although it realizes deep intelligent recognition of the health status of complex mechanical equipment through its powerful abstract feature extraction and learning functions, it has the disadvantages of complex deep network architecture design and cumbersome learning parameter tuning. For the intelligent fault diagnosis method for mechanical equipment based on discriminant sparse representation, the intelligent classification and recognition strategy is constructed through the sparse representation of signals, which can not only realize the intelligent diagnosis of different health states of mechanical equipment, but also overcome the disadvantages of complex model structure design and cumbersome procedures.

[0004] Representative methods of intelligent fault diagnosis methods for mechanical equipment based on discriminative sparse representation include a sparse representation classification method based on dictionary learning (Zhao et al., 2019). This method learns the sparse representation of vibration data signals through a dictionary learning algorithm and uses the sparse approximation error minimum criterion of time domain signals to realize the recognition and classification of vibration data signals under different health states of planetary bearings; an intelligent diagnosis method based on enhanced sparse representation (Kong et al., 2021). This method integrates the prior knowledge of rotating machinery vibration data signals to enhance the signal reconstruction and discrimination function of sparse representation, and further uses the sparse approximation error minimum criterion of vibration time domain signals to realize the intelligent recognition and classification of different health states of planetary bearings.

[0005] However, the existing intelligent fault diagnosis methods for mechanical equipment based on discriminant sparse representation still have the following drawbacks:

[0006] 1) The selection of algorithm hyperparameters has a significant impact on the performance of intelligent fault diagnosis;

[0007] 2) The computational efficiency is low and the noise robustness is insufficient.

[0008] Therefore, there is an urgent need for an ultra-robust and efficient intelligent fault diagnosis method for mechanical equipment. Summary of the invention

[0009] The present invention provides a mechanical equipment intelligent fault diagnosis method, system, electronic equipment and storage medium, which are used to overcome at least one technical problem existing in the prior art.

[0010] To achieve the above object, the present invention provides a method for intelligent fault diagnosis of mechanical equipment, the method comprising:

[0011] Collect vibration data of mechanical equipment to be tested;

[0012] Determine the sample to be detected and the frequency spectrum characteristics of the sample to be detected according to the collected vibration data of the mechanical equipment to be detected;

[0013] According to the spectral characteristics of the sample to be detected and the pre-acquired spectral integration complete dictionary, the optimized sparse coding of the spectral characteristics of the sample to be detected for the spectral integration complete dictionary is determined; and the optimized sparse coding is decomposed into sub-sparse coding; according to the sub-sparse coding and the spectral integration complete dictionary, the spectral sparse reconstruction error is determined;

[0014] Through the intelligent identification strategy of health status based on the minimum discrimination criterion of spectral sparse reconstruction error, the health status of the mechanical equipment to be detected is determined according to the spectral sparse reconstruction error.

[0015] Further, preferably, the step of determining the health state of the mechanical equipment to be detected according to the spectral sparse reconstruction error through the health state intelligent identification strategy based on the minimum discrimination criterion of the spectral sparse reconstruction error is implemented by the following formula:

[0016]

[0017] Among them, label(z n ) is the sample z to be tested n Health status class label; RE(Z n ,D whole ,α n (l) ) is the spectral sparse reconstruction error; Z n is the spectrum feature of the sample to be detected; D whole Complete dictionary for spectrum collection; α n (l) is the sub-sparse coding of the health state l, where l = 1, 2, ..., L, and L is the type of health state.

[0018] Further, preferably, the method for obtaining a complete dictionary of spectral collection includes:

[0019] Obtain vibration data of mechanical equipment under different operating health conditions;

[0020] Using a moving sliding window, the vibration data under different operating health states are constructed into training signal samples under different operating health states;

[0021] Performing Fourier transform on the training signal samples to obtain the frequency spectrum characteristics of the training signal samples;

[0022] According to the spectrum characteristics of the training signal samples, a spectrum integration sub-dictionary of each running health state is obtained;

[0023] According to the spectrum integration sub-dictionary of each running health state, a spectrum integration complete dictionary for multi-health state spectrum feature fusion is determined.

[0024] Further, preferably, the vibration data under different operating health states are constructed into training signal samples under different operating health states by using a moving sliding window, which is achieved by the following formula:

[0025] y n (l) =Γ n (y (l) )=y (l) (i n,start :i n,end ),

[0026]

[0027]

[0028] Among them, {y (l)}(l=1,2,...,L) is a training data set formed by vibration data under different operating health conditions; is the floor operator; Γ n is the sliding window operator; W is the window length parameter of the sliding window, and δ is the overlapping rate parameter of the sliding window.

[0029] Further, preferably, according to the spectrum characteristics of the training signal samples, the spectrum integration sub-dictionary of each running health state is obtained, which is achieved by the following formula:

[0030] D (l) =normalize{[Y1 (l) ,…,Y n (l) ,…,Y N (l) ]},

[0031] Among them, D (l) is a sub-dictionary of spectrum collection; {Y n (l)}(l=1,2,...,L;n=1,2,…,N) are the spectral features of the training signal samples; normalize{·} represents the column-by-column L2 norm normalization operation of the matrix.

[0032] Further, preferably, according to the collected vibration data of the mechanical equipment to be detected, the sample to be detected and the frequency spectrum characteristics of the sample to be detected are determined, which is achieved by the following formula:

[0033] Z n =Nyq{ABS{FFT{z n}}},

[0034] Among them, z n is the nth sample to be tested whose health status is unknown; FFT is the Fourier transform operator, ABS is the complex modulus operator, and Nyq means that only the spectrum amplitude characteristics within the Nyquist frequency range are extracted.

[0035] Further, preferably, in the step of determining the optimized sparse coding of the spectral features of the sample to be detected for the spectral integrated complete dictionary according to the spectral features of the sample to be detected and the pre-acquired spectral integrated complete dictionary, the optimization objective function of the optimized sparse coding is:

[0036]

[0037] Among them, D whole is the complete dictionary of spectral integration; T is the sparse threshold; Z n is the frequency spectrum feature of the nth sample to be detected whose health status is unknown.

[0038] In order to solve the above problems, the present invention also provides a mechanical equipment intelligent fault diagnosis system, comprising:

[0039] A collection unit, used for collecting vibration data of the mechanical equipment to be detected;

[0040] The data processing unit is used to determine the sample to be detected and the spectrum characteristics of the sample to be detected according to the collected vibration data of the mechanical equipment to be detected; determine the optimized sparse coding of the spectrum characteristics of the sample to be detected for the spectrum integration complete dictionary according to the spectrum characteristics of the sample to be detected and the pre-acquired spectrum integration complete dictionary; and decompose the optimized sparse coding into sub-sparse coding; determine the spectrum sparse reconstruction error according to the sub-sparse coding and the spectrum integration complete dictionary;

[0041] The mechanical equipment health status determination unit is used to determine the health status of the mechanical equipment to be detected according to the spectral sparse reconstruction error through a health status intelligent identification strategy based on the minimum spectral sparse reconstruction error judgment criterion.

[0042] In order to solve the above problem, the present invention further provides an electronic device, the electronic device comprising:

[0043] a memory storing at least one instruction; and

[0044] The processor executes the instructions stored in the memory to implement the steps in the above-mentioned mechanical equipment intelligent fault diagnosis method.

[0045] In order to solve the above problems, the present invention also provides a computer-readable storage medium, in which at least one instruction is stored, and the at least one instruction is executed by a processor in an electronic device to implement the above-mentioned intelligent fault diagnosis method for mechanical equipment.

[0046] The intelligent fault diagnosis method, system, electronic device and storage medium of mechanical equipment of the present invention avoid the tedious steps of feature engineering design and selection in deep learning models and do not rely on any explicit classifier. It is an intelligent fault diagnosis method for mechanical equipment with the characteristics of superior accuracy, robustness against noise, robustness of hyperparameter selection and high computational efficiency, and has achieved the technical effect of providing better technical support for fault prediction, health management and intelligent operation and maintenance of mechanical equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 A schematic diagram of a flow chart of a method for intelligent fault diagnosis of mechanical equipment according to an embodiment of the present invention;

[0048] Figure 2 A schematic diagram of the principle of a mechanical equipment intelligent fault diagnosis method according to an embodiment of the present invention;

[0049] Figure 3 A diagram showing the health status identification result of a planetary gearbox according to an intelligent fault diagnosis method for mechanical equipment according to an embodiment of the present invention;

[0050] Figure 4 A comparison chart of the health status diagnosis accuracy of a planetary gearbox according to an intelligent fault diagnosis method for mechanical equipment according to an embodiment of the present invention;

[0051] Figure 5 A comparison diagram of the anti-noise robustness effect of the intelligent fault diagnosis method for mechanical equipment according to an embodiment of the present invention;

[0052] Figure 6 A comparison diagram of the sensitivity effects of hyperparameter selection of the mechanical equipment intelligent fault diagnosis method according to an embodiment of the present invention;

[0053] Figure 7 A comparison diagram of the computational efficiency of the intelligent fault diagnosis method for mechanical equipment according to an embodiment of the present invention;

[0054] Figure 8A logical structure diagram of a mechanical equipment intelligent fault diagnosis system according to an embodiment of the present invention;

[0055] Fig. 9 Schematic diagram of the internal structure of an electronic device for implementing an intelligent fault diagnosis method for mechanical equipment according to an embodiment of the present invention.

[0056] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0057] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0058] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Among them, artificial intelligence (AI) is the theory, method, technology and application system of simulating, extending and expanding human intelligence, perceiving the environment, acquiring knowledge and using knowledge to obtain the best results using digital computers or machines controlled by digital computers. Basic artificial intelligence technology generally includes technologies such as sensors, special artificial intelligence chips, cloud computing, distributed storage, big data processing technology, operation / interaction system, mechatronics, etc. The artificial intelligence software technology in the present invention is spectrum integrated sparse recognition technology.

[0059] Glossary:

[0060] Sparse representation is intended to use as few non-zero coefficients as possible to represent the main information of the signal, thereby simplifying the solution process of signal processing problems.

[0061] A complete dictionary is complete if the atoms in the dictionary can just span an n-dimensional Euclidean space; the atom is the column vector of the dictionary. If the number of atoms m in the dictionary is much greater than n, and it can also span an n-dimensional Euclidean space, the dictionary is overcomplete. In the specific implementation process, the dictionary is overcomplete, because in the overcomplete dictionary scenario, the decomposition coefficients are not unique, which provides the possibility for adaptive processing, and then the most appropriate and sparse coefficients can be selected according to actual needs.

[0062] Sparse decomposition, the process of optimal sparse representation or sparse approximation under an overcomplete dictionary is the sparse decomposition of the signal.

[0063] Specifically, as an example, Figure 1 A flow chart of a method for intelligent fault diagnosis of mechanical equipment provided by an embodiment of the present invention. Figure 1As shown, the present invention provides a mechanical equipment intelligent fault diagnosis method, which can be executed by a device, and the device can be implemented by software and / or hardware. The mechanical equipment intelligent fault diagnosis method includes: steps S110 to S140.

[0064] Specifically, S110, collecting vibration data of the mechanical equipment to be detected; S120, determining the samples to be detected and the spectral characteristics of the samples to be detected based on the collected vibration data of the mechanical equipment to be detected; S130, determining the optimized sparse coding of the spectral characteristics of the samples to be detected for the spectral integrated complete dictionary based on the spectral characteristics of the samples to be detected and the pre-acquired spectral integrated complete dictionary; and decomposing the optimized sparse coding into sub-sparse coding; determining the spectral sparse reconstruction error based on the sub-sparse coding and the spectral integrated complete dictionary; S140, judging the health status of the mechanical equipment to be detected according to the spectral sparse reconstruction error through an intelligent health status recognition strategy based on the minimum discrimination criterion of the spectral sparse reconstruction error.

[0065] Figure 2 FIG. 1 is a schematic diagram showing the principle of an intelligent fault diagnosis method for mechanical equipment according to an embodiment of the present invention; Figure 2 As shown, in the prior art, in the scenario of complex rotating machinery structure, the selection of algorithm hyperparameters has a great influence on the intelligent fault diagnosis performance; the computational efficiency is low and the anti-noise robustness is insufficient. The intelligent fault diagnosis method of mechanical equipment of the present invention includes 4 parts: 1. Data acquisition; 2. Construction of training samples and test samples and acquisition of their respective spectral features; 3. Design of spectral integration dictionary considering spectral feature fusion; 4. Sparse coding and spectral sparse reconstruction error calculation; 5. Intelligent identification of health status.

[0066] Specifically, in the data collection process, training data of mechanical equipment in different operating health states and test data of unknown health states are obtained; in the specific execution process, an acceleration sensor is installed on the casing of the rotating mechanical equipment, and the acceleration sensor is used to collect data of the mechanical equipment in different operating health states to form a data set; and the data set is divided into a training data set and a test data set.

[0067] In the process of constructing training samples and test samples and obtaining their respective spectral features, the training data set is constructed into training signal samples {y n (l)}, and then use the Fourier transform operator to obtain the spectrum feature Y of the training signal sample n (l) ; L is the type of health status. The test data set is constructed into test signal samples {z n}, and then use the Fourier transform operator to obtain the spectrum characteristics Z of the test signal samplen .

[0068] In the process of designing the spectrum integration dictionary considering the fusion of spectrum features, a spectrum integration sub-dictionary specific to the running health state is constructed based on the spectrum features of the training signal, and then a spectrum integration complete dictionary considering the fusion of spectrum features of multiple health states is constructed. The spectrum integration sub-dictionary specific to the running health state is continuously constructed based on the spectrum features of the training signal until the spectrum integration sub-dictionary D (l) The health state index l = L. Specifically, set the health state index l = 1; construct the spectrum integration sub-dictionary D specific to the running health state (l) ; Use the health status index l and L to determine whether to continue to construct the health status specific spectrum integration sub-dictionary; if the spectrum integration sub-dictionary D (l) If the health status index l = L, then stop constructing the spectrum integration sub-dictionary D (l) , and output L spectral integration sub-dictionaries {D (l)}; If the spectrum is integrated into the dictionary D (l) If the health state index l<L, then update l=l+1 and return to step 1 to construct the spectrum integration sub-dictionary D specific to the running health state. (l) ; Using L spectrum ensemble sub-dictionaries {D (l)}, considering the spectral feature fusion of multiple health states, constructing the spectral integration complete dictionary D of spectral feature fusion whole .

[0069] In the process of sparse coding and spectral sparse reconstruction error calculation, the complete dictionary D is integrated by obtaining the spectrum characteristics of the test signal. whole Sparse coding And decomposition optimization sparse coding is a series of sub-sparse codes {α n (l)}; Set the health status index l = 1; Calculate the test signal sample spectrum feature Z n About the Complete Dictionary of Spectral Collection D whole With all sub-sparse codes α n (l) The spectral sparse reconstruction error RE(Z n ,D whole ,α n (l)); continuously calculate the spectral sparse reconstruction error of the test signal until the healthy state index l of the spectral sparse reconstruction error of the test signal is equal to L; that is, whether to continue calculating the spectral sparse reconstruction error of the test signal is determined by whether the healthy state index l is equal to L; if the healthy state index l of the spectral sparse reconstruction error is equal to L, then stop the calculation and output L spectral sparse reconstruction errors of the test signal; if the healthy state index l of the spectral sparse reconstruction error is less than L, then update l=l+1 and return to the step of calculating the spectral sparse reconstruction error of the test signal.

[0070] Finally, in the process of intelligent identification of health status, the health status class label of the test signal sample is obtained according to the health status intelligent identification strategy based on the minimum discrimination criterion of spectral sparse reconstruction error, and then the health status of the mechanical equipment test signal is identified by the health status class label of the test signal sample.

[0071] In a specific implementation process, the intelligent fault diagnosis method for mechanical equipment includes: steps S110 to S140.

[0072] S110, collecting vibration data of the mechanical equipment to be tested.

[0073] S120 . Determine the sample to be detected and the frequency spectrum characteristics of the sample to be detected according to the collected vibration data of the mechanical equipment to be detected.

[0074] According to the collected vibration data of the mechanical equipment to be tested, the samples to be tested and the spectrum characteristics of the samples to be tested are determined, which is achieved by the following formula:

[0075] Z n =Nyq{ABS{FFT{z n}}},

[0076] Among them, z n is the nth test signal sample of the mechanical equipment to be tested whose health status is unknown; FFT is the Fourier transform operator, ABS is the complex modulus operator, and Nyq means that only the spectrum amplitude characteristics within the Nyquist frequency range are extracted.

[0077] S130. Determine the optimized sparse coding of the spectral characteristics of the sample to be detected for the spectral integrated complete dictionary according to the spectral characteristics of the sample to be detected and the pre-acquired spectral integrated complete dictionary; and decompose the optimized sparse coding into sub-sparse coding; determine the spectral sparse reconstruction error according to the sub-sparse coding and the spectral integrated complete dictionary.

[0078] The methods for obtaining the complete dictionary of the spectrum collection include:

[0079] S1311. Obtain vibration data of mechanical equipment under different operating health conditions;

[0080] S1312. Using a moving sliding window, construct the vibration data under different operating health states into training signal samples under different operating health states; perform Fourier transform on the training signal samples to obtain frequency spectrum features of the training signal samples.

[0081] S13121. Select appropriate window length parameter W and overlap rate parameter δ to design vibration data signal y∈R 1×m The moving sliding window operator Γ n :R 1×m →R W×1 , realize the vibration data signal y∈R 1×m Extraction of the nth specific data segment,

[0082] Γ n (y) = y(i n,start :i n,end ),

[0083] Among them, index i n,start with i n,end is determined as follows:

[0084]

[0085]

[0086] in, is the floor operator, Γ n is the sliding window operator; W is the window length parameter of the sliding window, and δ is the overlapping rate parameter of the sliding window.

[0087] S13122, the training data set {y (l)}(l=1,2,...,L), and construct training signal samples {y n (l)}(l=1,2,...,L;n=1,2,…,N), which is realized by the following formula:

[0088] y n (l) =Γ n (y (l) )=y (l) (i n,start :i n,end ),

[0089] Where, l = 1, 2, ..., L; n = 1, 2, ..., N; y n (l) is the nth training signal sample of health state l.

[0090] S13123, using the Fourier transform operator to obtain the frequency spectrum characteristics of the training signal sample, which is achieved by the following formula:

[0091] Y n (l) =Nyq{ABS{FFT{y n (l)}}},

[0092] Wherein, l = 1, 2, ..., L; n = 1, 2, ..., N; FFT is the Fourier transform operator, ABS is the complex modulus operator, and Nyq means that only the spectrum amplitude characteristics within the Nyquist frequency range are extracted.

[0093] S1313. Obtain a spectrum integration sub-dictionary of each operating health state according to the spectrum characteristics of the training signal samples.

[0094] Among them, according to the spectrum characteristics of the training signal samples {Y n (l)}(l=1,2,...,L;n=1,2,...,N), obtain the spectrum integration sub-dictionary D of each running health state (l) , achieved through the following formula:

[0095] D (l) =normalize{[Y1 (l) ,…,Y n (l) ,…,Y N (l) ]},

[0096] Among them, D (l) is a sub-dictionary of spectrum collection; {Y n (l)}(l=1,2,...,L;n=1,2,…,N) are the spectral features of the training signal samples; normalize{·} represents the column-by-column L2 norm normalization operation of the matrix.

[0097] S1314. Determine a spectrum integration complete dictionary for fusion of multi-health state spectrum features based on the spectrum integration sub-dictionaries of each operational health state.

[0098] According to the spectrum of each running health state, the sub-dictionary D (l) , determine the spectral integration complete dictionary D for multi-health state spectral feature fusion whole , achieved through the following formula:

[0099] D whole =[D (1) ,…,D (l) ,…,D (L) ].

[0100] S140 , determining the health status of the mechanical equipment to be detected according to the spectral sparse reconstruction error through a health status intelligent identification strategy based on the minimum spectral sparse reconstruction error judgment criterion.

[0101] The step of determining the health status of the mechanical equipment to be detected according to the spectral sparse reconstruction error is implemented by the following formula through the healthy state intelligent identification strategy based on the minimum discrimination criterion of the spectral sparse reconstruction error:

[0102]

[0103] Among them, label(z n ) is the nth sample z to be tested n Health status class label; RE(Z n ,D whole ,α n (l) ) is the spectral sparse reconstruction error; Z n is the spectrum feature of the nth sample to be detected; D whole Complete dictionary for spectrum collection; α n (l) is the sub-sparse coding of the health state l, where l = 1, 2, ..., L, and L is the type of health state.

[0104] Taking the planetary gearbox model NGW11-10 as an example, by installing a vibration acceleration sensor on the housing of the planetary gearbox, the vibration data acquisition system is used to obtain the training vibration data set and the test vibration data set with unknown health state of the planetary gearbox under L types (L=8) of different health states (including normal NS, input shaft bearing outer ring fault ORF, input shaft bearing inner ring fault IRF, input shaft bearing rolling element fault REF, ring gear fault RGF, sun gear fault SGF, planetary gear fault PGF, planetary bearing inner ring fault PBF). The input shaft speed of the planetary gearbox is 1500 rpm, and the sampling frequency and sampling time of the training and test vibration data signals are 25600 Hz and 30 seconds respectively.

[0105] In a specific implementation process, the intelligent fault diagnosis method for mechanical equipment includes: steps S110 to S140.

[0106] S110, collecting vibration data of the mechanical equipment to be tested.

[0107] S120 . Determine the sample to be detected and the frequency spectrum characteristics of the sample to be detected according to the collected vibration data of the mechanical equipment to be detected.

[0108] Step S120 is a process of determining the sample to be detected and the frequency spectrum characteristics of the sample to be detected according to the collected vibration data of the mechanical equipment to be detected, including:

[0109] S121: Select appropriate window length parameter W and overlap rate parameter δ to design vibration data signal y∈R 1×m The moving sliding window operator Γ n :R 1×m →R W×1 , realize the vibration data signal y∈R 1×m Extraction of the nth specific data segment,

[0110] Γ n (y) = y(i n,start :i n,end ),

[0111] Among them, index i n,start with i n,end is determined as follows:

[0112]

[0113]

[0114] in, is the floor operator.

[0115] In the specific implementation process, the window length parameter W and the overlap rate parameter δ are 900 and 0.98 respectively; the number of training and test signal samples N corresponding to each health state is 4218.

[0116] S122: For the test vibration data z of the planetary gearbox whose health status is unknown, the moving sliding window operator is used to construct the test signal samples {z n}:

[0117] For n=1,2,…,N,z n =Γ n (z)=z(i n,start :i n,end ),

[0118] Among them, z n is the nth sample to be tested whose health status is unknown.

[0119] S123: Use the Fourier transform operator to obtain the frequency spectrum characteristics of the sample to be detected, which is achieved through the following formula:

[0120] For n = 1, 2, ..., N, Z n =Nyq{ABS{FFT{z n}}},

[0121] Among them, FFT is the Fourier transform operator, ABS is the complex modulus operator, and Nyq means that only the spectrum amplitude characteristics within the Nyquist frequency range are extracted.

[0122] S130. Determine the optimized sparse coding of the spectral characteristics of the sample to be detected for the spectral integrated complete dictionary according to the spectral characteristics of the sample to be detected and the pre-acquired spectral integrated complete dictionary; and decompose the optimized sparse coding into sub-sparse coding; determine the spectral sparse reconstruction error according to the sub-sparse coding and the spectral integrated complete dictionary.

[0123] S1301, using the orthogonal matching pursuit algorithm, calculate the spectral feature Z of the sample to be detected n About the Complete Dictionary of Spectral Collection D whole Optimized sparse coding The optimization objective function of sparse coding is as follows:

[0124]

[0125] Wherein, T is a sparse threshold. In this embodiment, the sparse threshold T is 10.

[0126] S1302: Optimize sparse coding of the sample to be detected Decomposed into a series of spectrum integration sub-dictionaries corresponding to different health states {D (l)}(l=1,2,...,L) of sub-sparse coding {α n (l)}(l=1,2,...,L), which is realized by the following formula:

[0127]

[0128] Among them, the sub-sparse code α corresponding to the health state l n (l) , which can be expressed as:

[0129]

[0130] S1303, calculate the frequency spectrum feature Z of the sample to be detected n About the Complete Dictionary of Spectral Collection D whole With all sub-sparse codes α n (l) The spectral sparse reconstruction error RE(Z n ,D whole ,α n (l) ), which is achieved by the following formula:

[0131] For l = 1, 2, ..., L;

[0132] S140, determine the health status of the mechanical equipment to be tested according to the spectral sparse reconstruction error through the health status intelligent identification strategy based on the spectral sparse reconstruction error minimum discrimination criterion. That is, according to the health status intelligent identification strategy based on the spectral sparse reconstruction error minimum discrimination criterion, obtain the health status class label of the test signal sample, and then identify the health status of the mechanical equipment test signal through the health status class label of the test signal sample.

[0133] According to the intelligent identification strategy of health status based on the minimum discrimination criterion of spectral sparse reconstruction error, the health status of the planetary gearbox test signal sample is identified, which is achieved by the following formula:

[0134]

[0135] Among them, label(z n ) is the sample z to be tested n Health status category label.

[0136] Figure 3 : is a diagram showing the health status identification result of a planetary gearbox according to an intelligent fault diagnosis method for mechanical equipment according to an embodiment of the present invention; Figure 3 As shown, the intelligent fault diagnosis method for mechanical equipment in an embodiment of the present invention can perfectly identify the true health status of 8 different planetary gearbox health states, totaling 33,744 test samples, that is, the diagnostic accuracy for 8 different planetary gearbox health states can reach 100%, which fully demonstrates the superior diagnostic performance of the provided invention in realizing planetary gearbox fault diagnosis.

[0137] Figures 4 to 6 The diagnostic effect of the mechanical equipment intelligent fault diagnosis method (S-ESRC) of the embodiment of the present invention is compared with the diagnostic effect of the mechanical equipment intelligent fault diagnosis method in the prior art. The prior art adopts the enhanced sparse representation intelligent recognition method ESRIR, the sparse representation classification method based on dictionary learning DLSRC, and the deep neural network method DCNN. Specifically, Figure 4 A comparison chart of the health status diagnosis accuracy of a planetary gearbox according to an intelligent fault diagnosis method for mechanical equipment according to an embodiment of the present invention; Figure 5 A comparison diagram of the anti-noise robustness effect of the intelligent fault diagnosis method for mechanical equipment according to an embodiment of the present invention; Figure 6 A comparison chart of the sensitivity effects of hyperparameter selection for the intelligent fault diagnosis method for mechanical equipment according to an embodiment of the present invention.

[0138] like Figure 4As shown, the planetary gearbox health status diagnosis accuracy of the mechanical equipment intelligent fault diagnosis method according to the embodiment of the present invention is compared with the planetary gearbox health status diagnosis accuracy of the three methods of ESRIR, DLSRC and DCNN in the prior art. Through the comparison, it can be seen that the mechanical equipment intelligent fault diagnosis method (S-ESRC) according to the present invention can achieve the highest diagnostic accuracy for the identification of each planetary gearbox health status, and the overall average diagnostic accuracy is the highest, which reflects the superior diagnostic performance of the provided invention in realizing planetary gearbox fault diagnosis.

[0139] like Figure 5 As shown, by adding random Gaussian white noise to the training data set and the test signal data set at the same time, the noise robustness of the intelligent fault diagnosis method for mechanical equipment in the embodiment of the present invention is compared with the noise robustness of the three methods of ESRIR, DLSRC and DCNN in the prior art. Through the comparison, it can be seen that when the signal-to-noise ratio SNR is -5dB, 0dB, 5dB, 10dB, 15dB and 20dB respectively, the intelligent fault diagnosis method for mechanical equipment in the embodiment of the present invention, that is, the super-robust intelligent fault diagnosis method (S-ESRC) driven by the spectral integrated sparse recognition model, can achieve the highest overall average diagnostic accuracy compared with the three cutting-edge comparison methods (ESRIR, DLSRC and DCNN), which reflects the strong noise robustness of the provided invention in realizing planetary gearbox fault diagnosis.

[0140] like Figure 6 As shown in FIG. 1 , the selection sensitivity of the algorithm hyperparameters of the mechanical equipment intelligent fault diagnosis method according to the embodiment of the present invention is compared with the selection sensitivity of the algorithm hyperparameters of the two methods ESRIR and DLSRC in the prior art. Figure 6 In the embodiments shown in (a)-(c), the sensitivity analysis comparison results of the mechanical equipment intelligent fault diagnosis method (S-ESRC) provided by the present invention and two comparative methods in the prior art (ESRIR and DLSRC) on the selection of algorithm hyperparameters (window length W and overlap rate δ) show that the mechanical equipment intelligent fault diagnosis method (S-ESRC) of the embodiment of the present invention has superior diagnostic performance, which is least sensitive to the selection of algorithm hyperparameters window length W and overlap rate δ, and can obtain superior diagnostic performance within a wide range of hyperparameters (W, δ), which reflects that the invention provided has the characteristic of being insensitive to the selection of algorithm hyperparameters when realizing planetary gearbox fault diagnosis. Figure 6(d) shows an embodiment in which the sensitivity analysis comparison results of the intelligent fault diagnosis method for mechanical equipment according to the embodiment of the present invention and two cutting-edge comparison methods (ESRIR and DLSRC) on the selection of algorithm hyperparameters (sparse threshold T) show that the superior diagnostic performance of the intelligent fault diagnosis method for mechanical equipment (S-ESRC) according to the embodiment of the present invention is the least sensitive to the selection of the algorithm hyperparameter sparse threshold T, and the superior diagnostic performance can be obtained within a very wide selection range of the sparse threshold T, which also reflects that the provided invention has the characteristic of being insensitive to the selection of algorithm hyperparameters when realizing planetary gearbox fault diagnosis.

[0141] like Figure 7 As shown, by comparing the computational efficiency of the intelligent fault diagnosis method for mechanical equipment in the embodiment of the present invention with the three methods of ESRIR, DLSRC and DCNN in the prior art, it is found that the intelligent fault diagnosis method for mechanical equipment (S-ESRC) in the embodiment of the present invention has a very low training time cost, and the cumulative calculation time cost for the 33744 training signal samples and 33744 test signal samples in the embodiment is the shortest, which is only 166.14 seconds, which reflects that the intelligent fault diagnosis method for mechanical equipment in the embodiment of the present invention has high computational efficiency in realizing planetary gearbox fault diagnosis.

[0142] In summary, the intelligent fault diagnosis method for mechanical equipment of the present invention is a super-robust intelligent fault diagnosis method driven by a spectral integrated sparse recognition model. First, it can realize the intelligent identification of the health status of mechanical equipment without relying on complex deep network architecture design and cumbersome learning parameter tuning. Secondly, by adopting a spectral feature fusion strategy considering multiple health states, the difference and discriminability of spectral features of mechanical equipment under different health states are fully explored, and a health state intelligent identification strategy based on the minimum discrimination criterion of spectral sparse reconstruction error is constructed, which can have superior diagnostic performance, algorithm hyperparameter selection insensitivity and good noise robustness. Third, by adopting the spectral features of the vibration signal of mechanical equipment under different health states, the spectral integrated sub-dictionary design and the spectral integrated complete dictionary construction considering the fusion of spectral features of multiple health states are carried out. The model establishment process is very simple and fast, and has the advantage of high computational efficiency. In summary, the intelligent fault diagnosis method for mechanical equipment of the present invention is particularly suitable for the intelligent identification of the health status of rotating mechanical equipment under constant operating conditions, and can provide important technical support for fault prediction, health management and intelligent operation and maintenance of rotating mechanical equipment.

[0143] Corresponding to the above-mentioned intelligent fault diagnosis method for mechanical equipment, the present invention also provides an intelligent fault diagnosis system for mechanical equipment. Figure 8 The functional modules of the intelligent fault diagnosis system for mechanical equipment according to an embodiment of the present invention are shown.

[0144] like Figure 8As shown, the mechanical equipment intelligent fault diagnosis system 800 provided by the present invention can be installed in an electronic device. According to the functions to be implemented, the mechanical equipment intelligent fault diagnosis system 800 may include a collection unit 810, a data processing unit 820 and a mechanical equipment health status determination unit 830. The unit described in the present invention can also be called a module, which refers to a series of computer program segments that can be executed by a processor of an electronic device and can complete a fixed function, which is stored in the memory of the electronic device.

[0145] In this embodiment, the functions of each module / unit are as follows:

[0146] The collection unit 810 is used to collect vibration data of the mechanical equipment to be detected;

[0147] The data processing unit 820 is used to determine the sample to be detected and the spectrum characteristics of the sample to be detected according to the collected vibration data of the mechanical equipment to be detected; determine the optimized sparse coding of the spectrum characteristics of the sample to be detected for the spectrum integration complete dictionary according to the spectrum characteristics of the sample to be detected and the pre-acquired spectrum integration complete dictionary; and decompose the optimized sparse coding into sub-sparse coding; determine the spectrum sparse reconstruction error according to the sub-sparse coding and the spectrum integration complete dictionary;

[0148] The mechanical equipment health status determination unit 830 is used to determine the health status of the mechanical equipment to be detected according to the spectral sparse reconstruction error through a health status intelligent identification strategy based on the minimum spectral sparse reconstruction error judgment criterion.

[0149] More specific implementations of the intelligent fault diagnosis system for mechanical equipment provided by the present invention can all be described with reference to the above-mentioned embodiments of the intelligent fault diagnosis method for mechanical equipment, and will not be listed one by one here.

[0150] like Fig. 9 As shown, the present invention provides an electronic device 9 for an intelligent fault diagnosis method for mechanical equipment.

[0151] The electronic device 9 may include a processor 90 , a memory 91 and a bus, and may also include a computer program stored in the memory 91 and executable on the processor 90 , such as a mechanical equipment intelligent fault diagnosis program 92 .

[0152] The memory 91 includes at least one type of readable storage medium, and the readable storage medium includes flash memory, mobile hard disk, multimedia card, card-type memory (for example: SD or DX memory, etc.), magnetic memory, disk, optical disk, etc. The memory 91 may be an internal storage unit of the electronic device 9 in some embodiments, such as a mobile hard disk of the electronic device 9. The memory 91 may also be an external storage device of the electronic device 9 in other embodiments, such as a plug-in mobile hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (SecureDigital, SD) card, a flash card (Flash Card), etc. equipped on the electronic device 9. Further, the memory 91 may also include both an internal storage unit of the electronic device 9 and an external storage device. The memory 91 may not only be used to store application software and various types of data installed in the electronic device 9, such as the code of the mechanical equipment intelligent fault diagnosis program, etc., but also be used to temporarily store data that has been output or is to be output.

[0153] The processor 90 may be composed of an integrated circuit in some embodiments, for example, a single packaged integrated circuit, or a plurality of packaged integrated circuits with the same or different functions, including one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and combinations of various control chips, etc. The processor 90 is the control core (Control Unit) of the electronic device, and uses various interfaces and lines to connect various components of the entire electronic device, and executes or executes programs or modules (such as mechanical equipment intelligent fault diagnosis programs, etc.) stored in the memory 91, and calls data stored in the memory 91 to execute various functions of the electronic device 9 and process data.

[0154] The bus may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc. The bus is configured to realize connection and communication between the memory 91 and at least one processor 90, etc.

[0155] Fig. 9 Only an electronic device with components is shown, and those skilled in the art will understand that Fig. 9The structure shown does not constitute a limitation on the electronic device 9, and may include fewer or more components than shown in the figure, or combine certain components, or arrange the components differently.

[0156] For example, although not shown, the electronic device 9 may also include a power source (such as a battery) for supplying power to each component. Preferably, the power source may be logically connected to the at least one processor 90 through a power management device, so that the power management device can realize functions such as charging management, discharging management, and power consumption management. The power source may also include any components such as one or more DC or AC power sources, recharging devices, power failure detection circuits, power converters or inverters, and power status indicators. The electronic device 9 may also include a variety of sensors, Bluetooth modules, Wi-Fi modules, etc., which will not be repeated here.

[0157] Furthermore, the electronic device 9 may also include a network interface. Optionally, the network interface may include a wired interface and / or a wireless interface (such as a WI-FI interface, a Bluetooth interface, etc.), which is generally used to establish a communication connection between the electronic device 9 and other electronic devices.

[0158] Optionally, the electronic device 9 may further include a user interface, which may be a display, an input unit (such as a keyboard), or a standard wired interface or a wireless interface. Optionally, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, and an OLED (Organic Light-Emitting Diode) touch device. The display may also be appropriately referred to as a display screen or a display unit, which is used to display information processed in the electronic device 9 and to display a visual user interface.

[0159] It should be understood that the embodiment is for illustration only and the scope of the patent application is not limited to this structure.

[0160] The mechanical equipment intelligent fault diagnosis program 92 stored in the memory 91 in the electronic device 9 is a combination of multiple instructions. When running in the processor 90, it can achieve: S110, collecting vibration data of the mechanical equipment to be detected; S120, determining the sample to be detected and the spectral characteristics of the sample to be detected based on the collected vibration data of the mechanical equipment to be detected; S130, determining the optimized sparse coding of the spectral characteristics of the sample to be detected for the spectral integration complete dictionary based on the spectral characteristics of the sample to be detected and the pre-acquired spectral integration complete dictionary; and decomposing the optimized sparse coding into sub-sparse coding; determining the spectral sparse reconstruction error based on the sub-sparse coding and the spectral integration complete dictionary; S140, judging the health status of the mechanical equipment to be detected based on the spectral sparse reconstruction error through an intelligent health status identification strategy based on the minimum discrimination criterion of the spectral sparse reconstruction error.

[0161] Specifically, the specific implementation method of the processor 90 for the above instructions can refer to Figure 1 The description of the relevant steps in the corresponding embodiment is not repeated here. It should be emphasized that in order to further ensure the privacy and security of the above-mentioned mechanical equipment intelligent fault diagnosis program, the above-mentioned mechanical equipment intelligent fault diagnosis program is stored in the node of the blockchain where the server cluster is located.

[0162] Furthermore, if the module / unit integrated in the electronic device 9 is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. The computer-readable medium may include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a mobile hard disk, a magnetic disk, an optical disk, a computer memory, and a read-only memory (ROM).

[0163] An embodiment of the present invention also provides a computer-readable storage medium, which may be non-volatile or volatile, and stores a computer program, which, when executed by a processor, implements: S110, collecting vibration data of a mechanical device to be detected; S120, determining a sample to be detected and spectral characteristics of the sample to be detected based on the collected vibration data of the mechanical device to be detected; S130, determining an optimized sparse coding of the spectral characteristics of the sample to be detected for the spectral integrated complete dictionary based on the spectral characteristics of the sample to be detected and a pre-acquired spectral integrated complete dictionary; and decomposing the optimized sparse coding into sub-sparse coding; determining a spectral sparse reconstruction error based on the sub-sparse coding and the spectral integrated complete dictionary; S140, determining a health status of the mechanical device to be detected based on the spectral sparse reconstruction error through an intelligent health status recognition strategy based on a minimum discrimination criterion for the spectral sparse reconstruction error.

[0164] Specifically, the specific implementation method when the computer program is executed by the processor can refer to the description of the relevant steps in the intelligent fault diagnosis method for mechanical equipment in the embodiment, and will not be repeated here.

[0165] In the several embodiments provided by the present invention, it should be understood that the disclosed devices, apparatuses and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of the modules is only a logical function division, and there may be other division methods in actual implementation.

[0166] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0167] In addition, each functional module in each embodiment of the present invention may be integrated into one processing unit, each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of hardware plus software functional modules.

[0168] It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0169] Therefore, no matter from which point of view, the embodiments should be regarded as illustrative and non-restrictive, and the scope of the present invention is limited by the appended claims rather than the above description, so it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims are included in the present invention. Any attached figure mark in the claims should not be regarded as limiting the claims involved.

[0170] The blockchain referred to in the present invention is a new application mode of computer technologies such as distributed data storage, peer-to-peer transmission, consensus mechanism, encryption algorithm, etc. Blockchain is essentially a decentralized database, a string of data blocks generated by cryptographic methods. Each data block contains a batch of network transaction information, which is used to verify the validity of its information (anti-counterfeiting) and generate the next block. Blockchain can include the underlying blockchain platform, platform product service layer, and application service layer. Blockchain can store medical data, such as personal health records, kitchens, examination reports, etc.

[0171] In addition, it is clear that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units or devices stated in the system claim can also be implemented by one unit or device through software or hardware. The second and other words are used to indicate names, but not to indicate any particular order.

[0172] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention.

Claims

1. A method for intelligent fault diagnosis of mechanical equipment, characterized in that: include: Collect vibration data of mechanical equipment to be tested; Determine the sample to be detected and the frequency spectrum characteristics of the sample to be detected according to the collected vibration data of the mechanical equipment to be detected; According to the spectrum characteristics of the sample to be detected and the pre-acquired spectrum integration complete dictionary, determining the optimized sparse coding of the spectrum characteristics of the sample to be detected for the spectrum integration complete dictionary; and decomposing the optimized sparse coding into sub-sparse coding; Determining a spectral sparse reconstruction error based on the sub-sparse coding and the spectral integrated complete dictionary; By using a health status intelligent identification strategy based on a minimum spectral sparse reconstruction error discrimination criterion, the health status of the mechanical equipment to be detected is determined according to the spectral sparse reconstruction error; The method for obtaining the complete dictionary of the spectral collection includes: Obtain vibration data of mechanical equipment under different operating health conditions; Using a moving sliding window, constructing the vibration data under the different running health states into training signal samples under the different running health states; Performing Fourier transform on the training signal samples to obtain frequency spectrum characteristics of the training signal samples; According to the spectrum characteristics of the training signal samples, obtaining the spectrum integration sub-dictionary of each running health state; According to the spectrum integration sub-dictionaries of the various operating health states, a spectrum integration complete dictionary for fusion of multi-health state spectrum features is determined.

2. The intelligent fault diagnosis method for mechanical equipment as claimed in claim 1, characterized in that: The step of determining the health state of the mechanical equipment to be detected according to the spectral sparse reconstruction error by using the health state intelligent identification strategy based on the minimum discrimination criterion of the spectral sparse reconstruction error is implemented by the following formula: Among them, label(z n ) is the sample z to be tested n Health status class label; RE(Z n ,D whole ,α n (l) ) is the spectral sparse reconstruction error; Z n is the frequency spectrum feature of the sample to be detected; D whole Complete dictionary for spectrum collection; α n (l) is the sub-sparse coding of the health state l, where l = 1, 2, ..., L, and L is the type of health state.

3. The intelligent fault diagnosis method for mechanical equipment according to claim 1, characterized in that: The use of a moving sliding window to construct the vibration data under different operating health states into training signal samples under different operating health states is achieved by the following formula: y n (l) =Γ n (y (l) )=y (l) (i n,start :i n,end ), Among them, {y (l) }(l=1,2,...,L) is a training data set formed by vibration data under different operating health conditions; is the floor operator; Γ n is the sliding window operator; W is the window length parameter of the sliding window, and δ is the overlapping rate parameter of the sliding window.

4. The intelligent fault diagnosis method for mechanical equipment according to claim 1, characterized in that: The spectral integration sub-dictionary of each running health state is obtained according to the spectral characteristics of the training signal sample, which is implemented by the following formula: D (l) =normalize{[Y1 (l) ,…,Y n (l) ,…,Y N (l) ]}, Among them, D (l) is a sub-dictionary of spectrum collection; {Y n (l) }(l=1,2,...,L;n=1,2,…,N) are the frequency spectrum characteristics of the training signal samples; normalize{·} represents the column-by-column L2 norm normalization operation of the matrix.

5. The intelligent fault diagnosis method for mechanical equipment according to claim 1, characterized in that: The method of determining the sample to be detected and the frequency spectrum characteristics of the sample to be detected based on the collected vibration data of the mechanical equipment to be detected is implemented by the following formula: Z n =Nyq{ABS{FFT{z n }}}, Among them, z n is the nth sample to be tested whose health status is unknown; FFT is the Fourier transform operator, ABS is the complex modulus operator, and Nyq means that only the spectrum amplitude characteristics within the Nyquist frequency range are extracted.

6. The intelligent fault diagnosis method for mechanical equipment according to claim 1, characterized in that: In the step of determining the optimized sparse coding of the spectrum characteristics of the sample to be detected for the spectrum integration complete dictionary according to the spectrum characteristics of the sample to be detected and the pre-acquired spectrum integration complete dictionary, the optimization objective function of the optimized sparse coding is: Among them, D whole is the complete dictionary of spectral integration; T is the sparse threshold; Z n is the spectrum feature of the nth sample to be tested whose health status is unknown.

7. A mechanical equipment intelligent fault diagnosis system, characterized in that: include: A collection unit, used for collecting vibration data of the mechanical equipment to be detected; A data processing unit is used to determine the sample to be detected and the spectrum characteristics of the sample to be detected according to the collected vibration data of the mechanical equipment to be detected; according to the spectrum characteristics of the sample to be detected and the pre-acquired spectrum integration complete dictionary, determine the optimized sparse coding of the spectrum characteristics of the sample to be detected for the spectrum integration complete dictionary; and decomposing the optimized sparse coding into sub-sparse coding; Determining a spectral sparse reconstruction error based on the sub-sparse coding and the spectral integrated complete dictionary; A mechanical equipment health status determination unit, configured to determine the health status of the mechanical equipment to be detected according to the spectral sparse reconstruction error through a health status intelligent identification strategy based on a minimum spectral sparse reconstruction error determination criterion; The method for obtaining a complete dictionary of spectral collection includes: Obtain vibration data of mechanical equipment under different operating health conditions; Using a moving sliding window, constructing the vibration data under the different running health states into training signal samples under the different running health states; Performing Fourier transform on the training signal samples to obtain frequency spectrum characteristics of the training signal samples; According to the spectrum characteristics of the training signal samples, obtaining the spectrum integration sub-dictionary of each running health state; According to the spectrum integration sub-dictionaries of the various operating health states, a spectrum integration complete dictionary for fusion of multi-health state spectrum features is determined.

8. An electronic device, characterized in that: The electronic device comprises: at least one processor; and, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the steps in the intelligent fault diagnosis method for mechanical equipment as described in any one of claims 1 to 6.

9. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the intelligent fault diagnosis method for mechanical equipment as described in any one of claims 1 to 6 is implemented.

Images