CN113670610B - Fault detection method, system and medium based on wavelet transformation and neural network - Google Patents
Fault detection method, system and medium based on wavelet transformation and neural network Download PDFInfo
- Publication number
- CN113670610B CN113670610B CN202110823449.8A CN202110823449A CN113670610B CN 113670610 B CN113670610 B CN 113670610B CN 202110823449 A CN202110823449 A CN 202110823449A CN 113670610 B CN113670610 B CN 113670610B
- Authority
- CN
- China
- Prior art keywords
- neural network
- time
- wavelet
- fault
- frequency
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M13/00—Testing of machine parts
- G01M13/04—Bearings
- G01M13/045—Acoustic or vibration analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/14—Fourier, Walsh or analogous domain transformations, e.g. Laplace, Hilbert, Karhunen-Loeve, transforms
- G06F17/148—Wavelet transforms
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/15—Correlation function computation including computation of convolution operations
- G06F17/153—Multidimensional correlation or convolution
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Abstract
The invention discloses a fault detection method, a system and a medium based on wavelet transformation and a neural network, wherein the method comprises the following steps: acquiring a first vibration signal of a bearing to be detected; determining first frequency domain information of the first vibration signal, and selecting a corresponding wavelet basis function according to the first frequency domain information; performing continuous wavelet transformation on the first vibration signal according to the wavelet basis function to obtain a first time-frequency diagram; inputting the first time-frequency diagram into a pre-trained fault diagnosis model, and outputting to obtain a fault type identification result; the fault diagnosis model is obtained through convolutional neural network training. The embodiment of the invention can keep weak fault vibration information in continuous wavelet transformation by selecting proper wavelet basis functions, thereby avoiding noise interference, improving the precision of a time-frequency diagram and further improving the bearing fault detection accuracy; the fault diagnosis model is trained through the convolutional neural network, so that the bearing fault detection efficiency is improved, and the method can be widely applied to the technical field of equipment bearing fault detection.
Description
Technical Field
The invention relates to the technical field of equipment bearing fault detection, in particular to a fault detection method, a system and a medium based on wavelet transformation and a neural network.
Background
The vibration characteristics of the bearing fault in different stages are different, for the earliest ultrasonic stage, the vibration energy is not high, the characteristics are not obvious, and when the bearing fails to approach the tail sound in the later stage of the fault, the fault characteristic frequency and the natural frequency of the bearing can be submerged by random broadband high-frequency vibration noise. Therefore, the existing vibration processing method for the rolling bearing failure focuses more on the second and third stages, i.e., the natural frequency stage and the failure characteristic frequency stage. However, for the early failure of the bearing, an efficient and accurate bearing failure detection method is still lacked due to the weak failure vibration signal.
Disclosure of Invention
The present invention aims to solve at least to some extent one of the technical problems existing in the prior art.
Therefore, an object of an embodiment of the present invention is to provide a method for detecting a fault based on wavelet transform and a neural network, where the method selects an appropriate wavelet basis function according to frequency domain information of a first vibration signal of a bearing to be detected, and performs continuous wavelet transform on the first vibration signal according to the wavelet basis function to obtain a high-precision first time-frequency diagram, and further inputs the first time-frequency diagram into a fault diagnosis model obtained in advance through convolutional neural network training, so as to obtain a fault type identification result of the bearing to be detected.
Another objective of the embodiments of the present invention is to provide a fault detection system based on wavelet transform and neural network.
In order to achieve the technical purpose, the technical scheme adopted by the embodiment of the invention comprises the following steps:
in a first aspect, an embodiment of the present invention provides a fault detection method based on wavelet transform and a neural network, including the following steps:
acquiring a first vibration signal of a bearing to be detected;
determining first frequency domain information of the first vibration signal, and selecting a corresponding wavelet basis function according to the first frequency domain information;
performing continuous wavelet transformation on the first vibration signal according to the wavelet basis function to obtain a first time-frequency diagram;
inputting the first time-frequency graph into a pre-trained fault diagnosis model, and outputting to obtain a fault type identification result;
and the fault diagnosis model is obtained by training a convolutional neural network.
Further, in an embodiment of the present invention, the step of performing continuous wavelet transform on the first vibration signal according to the wavelet basis function to obtain a first time-frequency diagram specifically includes:
and under different translation factors and scale factors, performing inner product operation on the wavelet basis function and the first vibration signal to obtain a plurality of wavelet functions, and further obtaining a first time-frequency graph according to the plurality of wavelet functions.
Further, in an embodiment of the present invention, the first time-frequency diagram is:
wherein CWTf (a, b) represents a first time-frequency diagram, f (t) represents a first vibration signal, a represents a scale factor, b represents a translation factor,
representing the wavelet basis functions at a translation factor b and a scale factor a.
Further, in one embodiment of the present invention, the wavelet basis functions are:
where ψ (t) denotes a wavelet basis function, and C denotes a normalization constant.
Further, in an embodiment of the present invention, the method for detecting a fault based on a wavelet transform and a neural network further includes a step of training a fault diagnosis model, which specifically includes:
acquiring a time-frequency diagram data set, wherein the time-frequency diagram data set comprises second time-frequency diagrams of a plurality of vibration signals in each fault state of an equipment bearing and corresponding fault type label values;
denoising and thinning the time-frequency image data set to obtain a training image set;
and inputting the training picture set into a pre-constructed convolutional neural network for training to obtain a trained fault diagnosis model.
Further, in an embodiment of the present invention, the step of inputting the training picture set into a pre-constructed convolutional neural network for training specifically includes:
inputting the training picture set into the convolutional neural network to obtain a fault type prediction result;
determining a training loss value according to the fault type prediction result and the fault type label value;
and updating the parameters of the convolutional neural network according to the loss value.
Further, in one embodiment of the present invention, the convolutional neural network comprises an input layer, a low hidden layer, a fully-connected layer and an output layer, wherein the low hidden layer is composed of a plurality of convolutional layers and a plurality of pooling layers alternately.
In a second aspect, an embodiment of the present invention provides a fault detection system based on wavelet transform and a neural network, including:
the vibration signal acquisition module is used for acquiring a first vibration signal of the equipment bearing;
the wavelet basis function selection module is used for determining first frequency domain information of the first vibration signal and selecting a corresponding wavelet basis function according to the first frequency domain information;
the first time-frequency graph determining module is used for performing continuous wavelet transformation on the first vibration signal according to the wavelet basis function to obtain a first time-frequency graph;
the fault type identification module is used for inputting the first time-frequency graph into a pre-trained fault diagnosis model and outputting to obtain a fault type identification result;
and the fault diagnosis model is obtained by training a convolutional neural network.
In a third aspect, an embodiment of the present invention provides a fault detection apparatus based on wavelet transform and a neural network, including:
at least one processor;
at least one memory for storing at least one program;
when executed by the at least one processor, the at least one program causes the at least one processor to implement a method of wavelet transform and neural network based failure detection as described above.
In a fourth aspect, the embodiment of the present invention further provides a computer-readable storage medium, in which a program executable by a processor is stored, where the program executable by the processor is used to execute the above-mentioned method for detecting a fault based on a wavelet transform and a neural network.
Advantages and benefits of the present invention will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the invention:
according to the embodiment of the invention, a proper wavelet basis function is selected according to the frequency domain information of the first vibration signal of the bearing to be detected, the first vibration signal is subjected to continuous wavelet transformation according to the wavelet basis function to obtain a high-precision first time-frequency diagram, and the first time-frequency diagram is input into a fault diagnosis model obtained through convolutional neural network training in advance to obtain a fault type identification result of the bearing to be detected. The embodiment of the invention can keep weak fault vibration information in continuous wavelet transformation by selecting proper wavelet basis functions, thereby avoiding noise interference, improving the precision of a time-frequency diagram and further improving the bearing fault detection accuracy; the fault diagnosis model is trained through the convolutional neural network, and the bearing fault detection efficiency is improved.
Drawings
In order to more clearly illustrate the technical solution in the embodiment of the present invention, the following description is made on the drawings required to be used in the embodiment of the present invention, and it should be understood that the drawings in the following description are only for convenience and clarity of describing some embodiments in the technical solution of the present invention, and it is obvious for those skilled in the art that other drawings can be obtained according to these drawings without creative efforts.
Fig. 1 is a flowchart illustrating steps of a method for detecting a fault based on wavelet transform and a neural network according to an embodiment of the present invention;
fig. 2 is a block diagram of a fault detection system based on wavelet transform and a neural network according to an embodiment of the present invention;
fig. 3 is a block diagram of a structure of a fault detection apparatus based on wavelet transform and a neural network according to an embodiment of the present invention.
Detailed Description
Reference will now be made in detail to embodiments of the present invention, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to the same or similar elements or elements having the same or similar function throughout. The embodiments described below with reference to the accompanying drawings are illustrative only for the purpose of explaining the present invention and are not to be construed as limiting the present invention. The step numbers in the following embodiments are provided only for convenience of illustration, the order between the steps is not limited at all, and the execution order of each step in the embodiments can be adapted according to the understanding of those skilled in the art.
In the description of the present invention, the meaning of a plurality is two or more, if there is a description to the first and the second for the purpose of distinguishing technical features, it is not understood as indicating or implying relative importance or implicitly indicating the number of indicated technical features or implicitly indicating the precedence of the indicated technical features. Furthermore, unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art.
Referring to fig. 1, an embodiment of the present invention provides a method for detecting a fault based on wavelet transform and a neural network, which specifically includes the following steps:
s101, acquiring a first vibration signal of a bearing to be detected;
specifically, a first vibration signal of the bearing to be detected can be directly acquired through the vibration sensor, and the infrared displacement sensor can also sense the change of infrared light flux to obtain the first vibration signal of the bearing to be detected. The first vibration signal is an amplitude-based function over time, and in the description of the embodiment of the present invention, the first vibration signal is denoted by f (t).
S102, determining first frequency domain information of the first vibration signal, and selecting a corresponding wavelet basis function according to the first frequency domain information;
in particular, good time-frequency conversion may be achieved by using fourier transform to obtain frequency domain information of the first vibration signal. The fourier transform is as follows:
wherein f (t) represents a first vibration signal, being a periodic function with respect to t; f (w) is a frequency domain function with respect to t, called the image function of F (t).
As can be seen from the result of the fourier transform, the transformed signal does not combine the time domain and the frequency domain, i.e., only one type of information in the time domain or the frequency domain can be read. Based on this, the embodiment of the invention considers that the wavelet transform is used for drawing the time-frequency diagram of the signal, and the information contained in the original vibration signal is stored as much as possible.
The wavelet is a function that must satisfy an "allowance" condition, which may be a constraint that defines the reversibility of the wavelet transform process, and is specifically expressed as follows:
wherein the content of the first and second substances,
representing wavelet basis functions, phi (w) representing
The frequency domain function after the Fourier transform,
indicating a tolerance condition for limiting the inverse wavelet transform, the value of which must be bounded. The wavelet function has only a small local non-zero domain, the function is 0 outside the window, and the wavelet itself is oscillatory and completely free of dc trend components, i.e. satisfies:
in the embodiment of the invention, after the frequency domain information of the first vibration signal is determined through Fourier transform, a proper wavelet basis function is selected according to the frequency domain information and the allowable conditions, so that subsequent continuous wavelet transform is facilitated.
S103, performing continuous wavelet transformation on the first vibration signal according to the wavelet basis function to obtain a first time-frequency diagram;
as a further optional implementation, the step of performing continuous wavelet transform on the first vibration signal according to the wavelet basis function to obtain the first time-frequency diagram specifically includes:
and performing inner product operation on the wavelet basis function and the first vibration signal under different translation factors and scale factors to obtain a plurality of wavelet functions, and further obtaining a first time-frequency graph according to the plurality of wavelet functions.
In particular, wavelet basis functions
After a displacement b, at a different scale a, withThe original first vibration signal f (t) is subjected to inner product to complete the wavelet transform process, which can be expressed as follows:
wherein a represents a scale factor and a>0, the effect is on the wavelet basis function
Making telescopic change; b represents a shift factor reflecting the wavelet basis function
The displacement of (a) may be positive or negative; in the embodiment of the present invention, a and b are both continuous variables, so that the method is called Continuous Wavelet Transform (CWT).
The implementation process of the continuous wavelet transform in the embodiment of the invention is as follows:
a1, selecting a proper wavelet basis function, fixing a scale factor, and comparing the scale factor with an initial segment of a signal;
a2, calculating a wavelet coefficient, wherein the wavelet coefficient reflects the similarity of a wavelet under the current scale and the corresponding signal segment;
a3, changing the translation factor to enable the wavelet to generate a certain displacement along a time axis, and repeating the steps to complete one-time transformation;
a4, increasing scale factors, and repeating the three steps to finish the second analysis;
and A5, circularly executing the steps until the analysis requirement is met.
As a further optional implementation, the first time-frequency diagram is:
wherein CWTf (a, b) represents a first time-frequency diagram, f (t) represents a first vibration signal, a represents a scale factor, b represents a translation factor,
representing the wavelet basis functions at a translation factor b and a scale factor a.
Specifically, the first time-frequency diagram obtained through the continuous wavelet transform is as follows:
as can be seen from the above equation, the result of the continuous wavelet transform can be expressed as a function of the translation factor b and the scale factor a.
As a further optional implementation, the wavelet basis function is:
where ψ (t) denotes a wavelet basis function, and C denotes a normalization constant.
In particular, among the numerous wavelet basis function types, the classical continuous wavelet basis Morlet, also known as Morl wavelet, is chosen.
The Morl wavelet is a single frequency sine function under a Gaussian envelope and is expressed as follows:
where C represents the normalization constant during reconstruction, the Morl wavelet basis has no scale function phi (t), and is a non-orthogonal decomposition.
When the computer actually processes, the sampling frequency and the total data quantity of the acquired signal data f (t) are used as parameters, a pywt function in python is called, continuous wavelet transformation of the data is completed, the output value of the continuous wavelet transformation is recorded as cwtmatr, the output frequency is frequency, finally, the signal is read in a one-dimensional time sequence in a segmentation mode according to the time interval of 3s, after the wavelet transformation, the cwtmatr value is stored in a local file, namely a first time-frequency diagram, in a two-dimensional graph format, and the files are named according to fault types and then are respectively stored.
The time-frequency graph obtained by wavelet basis function processing is stored in a local file in a picture form, so that the method is beneficial to manual viewing on one hand and online compiling on the other hand.
It can be understood that in the embodiment of the invention, data acquisition and signal processing can be performed simultaneously, so that the time for reading data and analyzing data is greatly shortened; the adopted wavelet base stores the time-frequency information of the signals, reduces the signal loss amount in the signal transmission process, and retains weak fault information to a considerable extent, so that the subsequent signal analysis is more convincing; the collection and the processing of the signals are basically automated, and much manual intervention is not needed, so that the labor capacity of workers is reduced, the working efficiency is improved, and the cost is reduced.
S104, inputting the first time-frequency graph into a pre-trained fault diagnosis model, and outputting to obtain a fault type identification result;
the fault diagnosis model is obtained through convolutional neural network training.
Specifically, the continuous wavelet transform process completes the processing of the original vibration signal f (t), and the first time-frequency graph in the picture format retains the time-frequency information of the signal and is completely stored in a local file, so that the direct input of the convolutional neural network is facilitated. In the python platform, after being read in, the pictures can be converted into a matrix form, so that subsequent calculation, extraction and identification are facilitated.
Further as an optional embodiment, the convolutional neural network comprises an input layer, a low hidden layer, a full connection layer and an output layer, wherein the low hidden layer is composed of a plurality of convolutional layers and a plurality of pooling layers alternately.
In the embodiment of the invention, the convolutional neural network is adopted to finish the training of the fault diagnosis model. The convolutional neural network is a multi-layer supervised learning neural network, and the convolutional layer and the pooling layer of the hidden layer are core modules for realizing the function of extracting the characteristics of the convolutional neural network. The low hidden layer of the convolutional neural network is composed of convolutional layers and pooling layers alternately, and the high hidden layer is a full-connection layer and corresponds to the hidden layer and the logistic regression classifier of the traditional multilayer perceptron.
The basic convolutional neural network comprises five parts, an input layer, a convolutional layer, a pooling layer, a full-connection layer and an output layer. Each layer has a plurality of feature maps, each feature map extracting a feature of the input through a convolution filter, each feature map having a plurality of neurons. The input layer inputs each pixel representing a feature node. The convolution layer is composed of a plurality of filters, and the input data is subjected to characteristic extraction, and each element composing the convolution kernel corresponds to a weight coefficient and a deviation value.
The convolutional layer parameters are made up of a series of learnable filters, each filter being small in width and height, the input and data being consistent. When the filter slides along the width and height of the image, a two-dimensional activation map is generated.
Each filter will have an entire set of filters that will form multiple activation maps.
In the above formula, b is the deviation amount, Z l And Z l+1 Represents the convolutional input and output of layer L +1, L l+1 Is Z l+1 Z (i, j) corresponds to the pixels of the feature map, K is the number of channels of the feature map, f corresponds to the convolution kernel size, s 0 Convolution step size, p number of padding layers.
The distribution of data is mostly nonlinear, and the introduction of an excitation function is to introduce a nonlinear relation in a neural network so as to strengthen the learning capability of the network. The stimulus function subjects the input data of the network to a specific distribution:
a. zero-averaging of data distribution, namely: the mean value obtained by the distribution calculation is approximately equal to 0. A non-zero averaged distribution may result in a gradient vanishing or training jitter.
b. The data distribution is normal. A non-normal distribution may result in an algorithm overfitting.
c. In the training process, when different data scales batch are faced, the input data distribution of each layer of the neural network should be kept consistent all the time, and the phenomenon that the input data distribution cannot be kept consistent is called Internal Covaraite Shift, so that the training process is seriously influenced.
The convolutional layer contains an excitation function to assist in expressing complex features, and the expression form is as follows:
in the above formula, f (z) is a characterizing activation function, z l L is the output of the L-th layer, and the output value after the excitation function processing is A l 。
Although the current ReLU function has some defects, the ReLU function can also achieve good effect. ReLU has the lowest computational cost and simplest code implementation compared to other activation functions. Activation functions that have the ability to generate a zero mean distribution are preferred over other activation functions. It should be noted that neural network training and reasoning using ReLU will be slower because more complex exponential operations are required to obtain the function activation values.
The pooling layer is used for carrying out feature selection and information filtering on the convolution result, the size of the parameter matrix can be effectively reduced, the number of parameters in the last connection layer is reduced, and the effects of accelerating the calculation speed and preventing over-fitting are achieved. The pooling layer contains a pre-set pooling function whose function is to replace the result of a single point in the feature map with the feature map statistics of its neighboring regions. The step of selecting the pooling area by the pooling layer is the same as the step of scanning the characteristic diagram by the convolution kernel, and the pooling size, the step length and the filling are controlled. The general representation of the pooling model is:
in the formula, step length s 0 Meaning and volume of pixel (i, j)The build-up layers are the same, and p is a predetermined parameter. Analogous to the vector norm, when p =1, the pooling process is averaged over the pooling area, referred to as average pooling (averaging); when p =2, the pooling process takes a maximum within the region, referred to as maximal pooling (max pooling).
The fully connected layers (FC) function as "classifiers" throughout the convolutional neural network. Operations such as convolution, pooling and activation map raw data to hidden layer feature controls, and the fully-connected layer plays a role in mapping learned 'distributed features' to sample labels. The last layer of the network serves as the input to the fully connected layer. The correlation formula is as follows:
Z j ＝W j X+b j ＝ω j1 x 1 +ω j2 x 2 +…+ω jn x n +b j
w is to be j And (4) regarding the weight of the features under the jth class, namely the importance degree of the features in each dimension, obtaining the score of each class by weighting and summing the features, and mapping the score into the probability through a Softmax function. Through the full connection layer, the fraction Z in the range of K categories (- ∞, + ∞) j To obtain the probability of belonging to each class, first pass e Zj The score is mapped to (0, + ∞) and then normalized to (0, 1). The following formula:
all Z of e j And calculating the power, summing, calculating the ratio of each value, and ensuring that the sum is 1, wherein the probability of classification is obtained by softmax.
dropout refers to temporarily discarding a neural network unit from a network according to a certain probability in the training process of a deep learning network. After Dropout is used, the transfer process changes, in the following specific form:
1) Randomly (temporarily) deleting hidden neurons in the network, and keeping input and output neurons unchanged;
2) The input x is propagated forward through the modified network and backward in the modified network using the obtained loss result. After a small batch of training samples finishes the process, updating corresponding parameters (w, b) on the non-deleted neurons according to a random gradient descent method;
3) This process then continues to repeat:
a. recovering deleted neurons (deleted neurons remain intact, non-deleted neurons have been updated);
b. randomly selecting a subset from hidden layer neurons to delete temporarily (backup parameters of deleted neurons);
c. for a small batch of training samples, the propagation is forward and then backward lost, and the parameters are updated according to a stochastic gradient descent method (the part of the parameters that are not deleted is updated, and the deleted neuron parameters keep the result before deletion).
4) This process is repeated until the number of iterations is complete.
The corresponding calculation formula is as follows:
y ～(l) ＝r (l) y (l)
the Bernoulli function in the above formula is to generate a probability r vector, i.e. a vector of 0 and 1 is randomly generated.
The weight of each neuron is multiplied by a probability p, which in the "population" makes the test data and the training data approximately the same. The standard model has no Dropout layer, and the same training data to train 5 different neural networks generally results in 5 different results, and the final result is determined by using "5 result averaging" or "majority winning voting strategy". Different networks produce different overfitting, and some of the "inverse" fitting mutually offset to achieve overall reduction of overfitting. The updating of the weight value is not dependent on the joint action of the implicit nodes with fixed relations, and the situation that some characteristics are effective only under other specific characteristics is prevented.
The calculation of the maximum likelihood loss function softmax _ loss comprises 2 steps:
1) Calculating the normalized probability of softmax, wherein the formula is as follows:
x i ＝x i -max(x 1 ,...,x n )
2) Calculating a loss function, the formula is as follows:
wherein N is the number of samples and K is the number of tags.
The Adam algorithm dynamically adjusts the learning rate for each parameter using first and second moment estimates of the gradient. The tf.train.AdamaOptizer provided by TensorFlow can control the learning speed, and after offset correction, the learning rate of each iteration has a certain range, so that the parameters are relatively stable. In the default initial background of TensorFlow, the step size α =0.001 and the magnitude of the kinetic value β 1 ＝0.9，β 2 =0.999,epsilon is a very small number to avoid divisor 0, let epsilon =10 -8 . In determining the parameters alpha and beta 1 、β 2 And the random objective function f (w), it is necessary to initialize the parameter vector, the first moment vector, the second moment vector and the time step.
Direct incorporation of momentum into the first moment m of the gradient in Adam 0 (exponentially weighted) estimation. Phase(s)Adam includes bias corrections, correcting first order moments (momentum terms) and (non-central) second order moment estimates initialized from the origin, rather than RMSProp where lack of correction factors results in second order moment estimates that may have very high bias during the initial stages of training.
Wherein m is t And v t First order momentum terms and second order momentum terms, respectively.
The respective correction values are obtained. w is a t Representing the parameters of the t-th iterative model at time t,
the gradient magnitude of the loss function with respect to w at the t-th iteration is shown. When the parameter w does not converge, the loop iteratively updates the various parts. I.e. the time step t plus 1, updating the gradient of the target function over the parameter w at this time step, updating the first moment estimate and the second original moment estimate of the bias, calculating the first moment estimate of the bias correction and the second moment estimate of the bias correction, and then updating the parameter w of the model with the values calculated above t 。
And selecting functions of each layer of the network according to actual needs, and connecting each layer to complete the construction of the integral structure of the convolutional neural network. Adding dimensionality of input data in a first-layer network, and selecting an activation function as a relu function; selecting padding to ensure the consistency of data dimensions; the maximum pooling is selected with a step size of 2. The latter network is gracefully modified with reference to the first tier network settings. The Dropout layer setting parameter is 0.5, half of neurons are shielded in operation, and the operation speed of the network model is increased. The full connection layer selects the activation functions as a relu function and a softmax function, and is suitable for multi-classification problems. The optimizer selects Adam, the loss function selects catagorical _ crosssentryp, and the output matrix selects the parameter accuracy. And at this point, the initial convolutional neural network model is built.
The convolutional neural network used in the embodiment of the present invention is described above. According to the embodiment of the invention, a data training set (used for training a model and determining parameter values) and a test set (predicting model accuracy) are constructed on the second time-frequency graph after wavelet transformation according to the fault types, the set file name is used as a label value (used for expressing the fault type), and finally the data is led into the convolutional neural network model to finish the training of the fault diagnosis module. In the subsequent working process, the vibration signals acquired in real time are directly input into the trained fault diagnosis model to carry out corresponding fault judgment.
The method has the advantages that the convolutional neural network model directly reads the local folder, so that the information processing process is reduced, the operation speed of the computer is greatly improved, and the computer identification time is reduced; on the other hand, under the condition of experience permission, for particularly obvious picture signals, a worker can directly judge a diagnosis result according to a mode of visual inspection; the number of parameters used by the convolutional layers can be automatically adjusted, the number of layers is moderately increased for a complex model, the number of layers is moderately reduced for a simple model, the total parameter quantity is reasonably controlled, and the time cost of network identification can be reduced; under the support of big data, the recognition rate of the convolutional neural network on fault signals can reach more than 95%, and the convolutional neural network has considerable guarantee on the normal operation of equipment in industrial production; the migratability is that the fault signal identified by the convolutional neural network in a certain device can be used for identifying other devices, so that multiple purposes are achieved, and the labor time of workers is greatly reduced; the network carries out automatic identification on the fault signal, can monitor the running state of the equipment in real time, reduces the risk of equipment fault, has lower requirements on maintenance and monitoring personnel, and basically realizes the visualization of the equipment fault.
As a further optional implementation, the method for detecting a fault based on wavelet transform and neural network further includes a step of training a fault diagnosis model, which specifically includes:
b1, acquiring a time-frequency diagram data set, wherein the time-frequency diagram data set comprises second time-frequency diagrams of a plurality of vibration signals in each fault state of an equipment bearing and corresponding fault type label values;
b2, denoising and thinning the time-frequency image data set to obtain a training image set;
and B3, inputting the training picture set into a pre-constructed convolutional neural network for training to obtain a trained fault diagnosis model.
As a further optional implementation, the step of inputting the training picture set into a pre-constructed convolutional neural network for training specifically includes:
c1, inputting the training picture set into a convolutional neural network to obtain a fault type prediction result;
c2, determining a loss value of training according to the fault type prediction result and the fault type label value;
and C3, updating the parameters of the convolutional neural network according to the loss value.
Specifically, after the data in the training data set is input into the initialized convolutional neural network model, the recognition result output by the model, that is, the fault type prediction result, can be obtained, and the accuracy of the recognition model prediction can be evaluated according to the fault type prediction result and the label value, so that the parameters of the model are updated. For the fault diagnosis model, the accuracy of the model prediction result can be measured by a Loss Function (Loss Function), the Loss Function is defined on a single training data and is used for measuring the prediction error of the training data, and specifically, the Loss value of the training data is determined by the label of the single training data and the prediction result of the model on the training data. In actual training, a training data set has many training data, so a Cost Function (Cost Function) is generally adopted to measure the overall error of the training data set, and the Cost Function is defined on the whole training data set and is used for calculating the average value of prediction errors of all the training data, so that the prediction effect of the model can be measured better. For a general machine learning model, based on the cost function, and a regularization term for measuring the complexity of the model, the regularization term can be used as a training objective function, and based on the objective function, the loss value of the whole training data set can be obtained. There are many kinds of commonly used loss functions, such as 0-1 loss function, square loss function, absolute loss function, logarithmic loss function, cross entropy loss function, etc. all can be used as the loss function of the machine learning model, and are not described one by one here. In the embodiment of the invention, a loss function can be selected from the loss functions to determine the loss value of the training. And updating the parameters of the model by adopting a back propagation algorithm based on the trained loss value, and iterating for several rounds to obtain the trained fault diagnosis model. Specifically, the number of iteration rounds may be preset, or training may be considered complete when the test set meets the accuracy requirement.
The steps of the method are explained above, and the embodiment of the invention can retain weak fault vibration information in continuous wavelet transform by selecting proper wavelet basis functions, thereby avoiding noise interference, improving the precision of a time-frequency diagram and further improving the accuracy of bearing fault detection; the fault diagnosis model is trained through the convolutional neural network, and the bearing fault detection efficiency is improved.
Referring to fig. 2, an embodiment of the present invention provides a fault detection system based on wavelet transform and a neural network, including:
the vibration signal acquisition module is used for acquiring a first vibration signal of the equipment bearing;
the wavelet basis function selection module is used for determining first frequency domain information of the first vibration signal and selecting a corresponding wavelet basis function according to the first frequency domain information;
the first time-frequency graph determining module is used for carrying out continuous wavelet transformation on the first vibration signal according to the wavelet basis function to obtain a first time-frequency graph;
the fault type identification module is used for inputting the first time-frequency diagram into a pre-trained fault diagnosis model and outputting to obtain a fault type identification result;
the fault diagnosis model is obtained through convolutional neural network training.
The contents in the above method embodiments are all applicable to the present system embodiment, the functions specifically implemented by the present system embodiment are the same as those in the above method embodiment, and the beneficial effects achieved by the present system embodiment are also the same as those achieved by the above method embodiment.
Referring to fig. 3, an embodiment of the present invention provides a fault detection apparatus based on wavelet transform and a neural network, including:
at least one processor;
at least one memory for storing at least one program;
when the at least one program is executed by the at least one processor, the at least one processor is enabled to implement the method for detecting a fault based on a wavelet transform and a neural network.
The contents in the method embodiments are all applicable to the device embodiments, the functions specifically implemented by the device embodiments are the same as those in the method embodiments, and the beneficial effects achieved by the device embodiments are also the same as those achieved by the method embodiments.
An embodiment of the present invention further provides a computer-readable storage medium, in which a program executable by a processor is stored, and the program executable by the processor is used for executing the above-mentioned fault detection method based on the wavelet transform and the neural network.
The computer-readable storage medium of the embodiment of the invention can execute the fault detection method based on wavelet transformation and neural network provided by the embodiment of the method of the invention, can execute any combination of the implementation steps of the embodiment of the method, and has corresponding functions and beneficial effects of the method.
The embodiment of the invention also discloses a computer program product or a computer program, which comprises computer instructions, and the computer instructions are stored in a computer readable storage medium. The computer instructions may be read by a processor of a computer device from a computer-readable storage medium, and executed by the processor to cause the computer device to perform the method illustrated in fig. 1.
In alternative embodiments, the functions/acts noted in the block diagrams may occur out of the order noted in the operational illustrations. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality/acts involved. Furthermore, the embodiments presented and described in the flow charts of the present invention are provided by way of example in order to provide a more thorough understanding of the technology. The disclosed methods are not limited to the operations and logic flows presented herein. Alternative embodiments are contemplated in which the order of various operations is changed, and in which sub-operations described as part of larger operations are performed independently.
Furthermore, although the present invention is described in the context of functional modules, it should be understood that, unless otherwise stated to the contrary, one or more of the above-described functions and/or features may be integrated in a single physical device and/or software module, or one or more of the functions and/or features may be implemented in a separate physical device or software module. It will also be appreciated that a detailed discussion of the actual implementation of each module is not necessary for an understanding of the present invention. Rather, the actual implementation of the various functional modules in the apparatus disclosed herein will be understood within the ordinary skill of an engineer given the nature, function, and interrelationships of the modules. Accordingly, those of ordinary skill in the art will be able to practice the invention as set forth in the claims without undue experimentation. It is also to be understood that the specific concepts disclosed are merely illustrative of and not intended to limit the scope of the invention, which is to be determined from the appended claims along with their full scope of equivalents.
The above functions, if implemented in the form of software functional units and sold or used as a separate product, may be stored in a computer-readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the above method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
The logic and/or steps represented in the flowcharts or otherwise described herein, e.g., an ordered listing of executable instructions that can be considered to implement logical functions, can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, processor-containing system, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions. For the purposes of this description, a "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device.
More specific examples (a non-exhaustive list) of the computer-readable medium would include the following: an electrical connection (electronic device) having one or more wires, a portable computer diskette (magnetic device), a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Further, the computer readable medium could even be paper or another suitable medium upon which the above described program is printed, as the program can be electronically captured, via for instance optical scanning of the paper or other medium, then compiled, interpreted or otherwise processed in a suitable manner if necessary, and then stored in a computer memory.
It should be understood that portions of the present invention may be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, various steps or methods may be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any one or combination of the following technologies, which are well known in the art, may be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application specific integrated circuit having an appropriate combinational logic gate circuit, a Programmable Gate Array (PGA), a Field Programmable Gate Array (FPGA), or the like.
In the foregoing description of the specification, reference to the description of "one embodiment/example," "another embodiment/example," or "certain embodiments/examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
While embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that: various changes, modifications, substitutions and alterations can be made to the embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.
While the preferred embodiments of the present invention have been illustrated and described, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention as defined by the appended claims.
Claims (9)
1. A fault detection method based on wavelet transformation and neural network is characterized by comprising the following steps:
acquiring a first vibration signal of a bearing to be detected;
determining first frequency domain information of the first vibration signal, and selecting a corresponding wavelet basis function according to the first frequency domain information;
performing continuous wavelet transformation on the first vibration signal according to the wavelet basis function to obtain a first time-frequency diagram;
inputting the first time-frequency diagram into a pre-trained fault diagnosis model, and outputting to obtain a fault type identification result;
wherein the fault diagnosis model is obtained by training through the following steps:
acquiring a time-frequency diagram data set, wherein the time-frequency diagram data set comprises second time-frequency diagrams of a plurality of vibration signals in each fault state of an equipment bearing and corresponding fault type label values;
denoising and thinning the time-frequency image data set to obtain a training image set;
and inputting the training picture set into a pre-constructed convolutional neural network for training to obtain a trained fault diagnosis model.
2. The method according to claim 1, wherein the step of performing continuous wavelet transform on the first vibration signal according to the wavelet basis function to obtain the first time-frequency diagram specifically comprises:
and under different translation factors and scale factors, performing inner product operation on the wavelet basis functions and the first vibration signals to obtain a plurality of wavelet functions, and further obtaining a first time-frequency diagram according to the wavelet functions.
3. The wavelet transform and neural network based fault detection method of claim 2, wherein the first time-frequency diagram is:
wherein CWTf (a, b) represents a first time-frequency diagram, f (t) represents a first vibration signal, a represents a scale factor, b represents a translation factor,
representing the wavelet basis functions at a translation factor b and a scale factor a.
4. The wavelet transform and neural network based fault detection method according to claim 2, wherein the wavelet basis functions are:
where ψ (t) denotes a wavelet basis function, and C denotes a normalization constant.
5. The method according to claim 1, wherein the step of inputting the training picture set to a pre-constructed convolutional neural network for training specifically comprises:
inputting the training picture set into the convolutional neural network to obtain a fault type prediction result;
determining a training loss value according to the fault type prediction result and the fault type label value;
and updating the parameters of the convolutional neural network according to the loss value.
6. The wavelet transform and neural network based failure detection method of any one of claims 1 to 5, wherein: the convolutional neural network comprises an input layer, a low hidden layer, a full connection layer and an output layer, wherein the low hidden layer is formed by alternately forming a plurality of convolutional layers and a plurality of pooling layers.
7. A fault detection system based on wavelet transform and neural network, comprising:
the vibration signal acquisition module is used for acquiring a first vibration signal of the equipment bearing;
the wavelet basis function selection module is used for determining first frequency domain information of the first vibration signal and selecting a corresponding wavelet basis function according to the first frequency domain information;
the first time-frequency diagram determining module is used for carrying out continuous wavelet transformation on the first vibration signal according to the wavelet basis function to obtain a first time-frequency diagram;
the fault type identification module is used for inputting the first time-frequency diagram into a pre-trained fault diagnosis model and outputting to obtain a fault type identification result;
wherein the fault diagnosis model is obtained by training through the following steps:
acquiring a time-frequency diagram data set, wherein the time-frequency diagram data set comprises second time-frequency diagrams of a plurality of vibration signals in each fault state of an equipment bearing and corresponding fault type label values;
denoising and thinning the time-frequency image data set to obtain a training image set;
and inputting the training picture set into a pre-constructed convolutional neural network for training to obtain a trained fault diagnosis model.
8. A fault detection device based on wavelet transformation and neural network is characterized by comprising:
at least one processor;
at least one memory for storing at least one program;
when the at least one program is executed by the at least one processor, the at least one processor may implement a wavelet transform and neural network-based failure detection method according to any one of claims 1 to 6.
9. A computer-readable storage medium in which a processor-executable program is stored, wherein the processor-executable program, when executed by a processor, is configured to perform a wavelet transform and neural network based failure detection method according to any one of claims 1 to 6.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110644123 | 2021-06-09 | ||
| CN2021106441239 | 2021-06-09 |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113670610A CN113670610A (en) | 2021-11-19 |
| CN113670610B true CN113670610B (en) | 2022-10-04 |
Family
ID=78539729
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110823449.8A Active CN113670610B (en) | 2021-06-09 | 2021-07-21 | Fault detection method, system and medium based on wavelet transformation and neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113670610B (en) |
Families Citing this family (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114445027B (en) * | 2021-12-27 | 2023-04-07 | 江西省交通工程集团有限公司 | Attendance frequency control method and system for spraying vehicle |
| CN114330439B (en) * | 2021-12-28 | 2023-04-07 | 盐城工学院 | Bearing diagnosis method based on convolutional neural network |
| CN114638149B (en) * | 2022-01-20 | 2023-08-08 | 佳都科技集团股份有限公司 | Convolution-based fault diagnosis method and device |
| CN114742108B (en) * | 2022-04-20 | 2022-12-20 | 中科航迈数控软件(深圳)有限公司 | Method and system for detecting fault of bearing of numerical control machine tool |
| CN114548190B (en) * | 2022-04-27 | 2022-08-30 | 西安易诺敬业电子科技有限责任公司 | Wind turbine fault diagnosis method and device based on self-adaptive residual error neural network |
| CN114818831B (en) * | 2022-05-27 | 2023-04-07 | 深圳先进技术研究院 | Bidirectional lithium ion battery fault detection method and system based on multi-source perception |
| CN114993710B (en) * | 2022-06-02 | 2023-01-31 | 浙江加力仓储设备股份有限公司 | Remote control type electric-drive folding forklift hub rapid maintenance device and method thereof |
| CN115510925B (en) * | 2022-11-17 | 2023-04-07 | 中国人民解放军国防科技大学 | Mechanical fault diagnosis method, device and medium |
| CN115546613A (en) * | 2022-12-01 | 2022-12-30 | 成都理工大学 | Debris flow identification method and device based on infrasonic waves |
| CN116523146B (en) * | 2023-06-30 | 2023-09-26 | 天津蓝天环科环境科技股份有限公司 | Neural network-based hospital sewage prediction method and device |
| CN117520926B (en) * | 2024-01-04 | 2024-04-16 | 南京灿能电力自动化股份有限公司 | Electric energy quality analysis method and device |
Family Cites Families (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| FR2900745B1 (en) * | 2006-05-05 | 2008-10-10 | Eurocopter France | METHOD AND DEVICE FOR DIAGNOSING A MECHANISM |
| CN103512765A (en) * | 2013-09-13 | 2014-01-15 | 中国科学院苏州生物医学工程技术研究所 | Fault detection method for variable learning rate wavelet BP neural network of blood type centrifugal machine |
| CN104198184A (en) * | 2014-08-11 | 2014-12-10 | 中国人民解放军空军工程大学 | Bearing fault diagnostic method based on second generation wavelet transform and BP neural network |
| CN107122802A (en) * | 2017-05-02 | 2017-09-01 | 哈尔滨理工大学 | A kind of fault detection method based on the rolling bearing for improving wavelet neural network |
| CN108426713B (en) * | 2018-02-26 | 2020-05-01 | 成都昊铭科技有限公司 | Rolling bearing weak fault diagnosis method based on wavelet transformation and deep learning |
| CN108830127B (en) * | 2018-03-22 | 2021-08-10 | 南京航空航天大学 | Rotary machine fault feature intelligent diagnosis method based on deep convolutional neural network structure |
| CN108896296A (en) * | 2018-04-18 | 2018-11-27 | 北京信息科技大学 | A kind of wind turbine gearbox method for diagnosing faults based on convolutional neural networks |
| CN110553839A (en) * | 2019-08-27 | 2019-12-10 | 华中科技大学 | Single and composite fault diagnosis method, equipment and system for gearbox |
| CN112257528B (en) * | 2020-10-12 | 2023-07-18 | 南京工业大学 | Wind power gear box fault diagnosis method based on wavelet transformation and dense connection expansion convolutional neural network |
| CN112926644A (en) * | 2021-02-22 | 2021-06-08 | 山东大学 | Method and system for predicting residual service life of rolling bearing |
-
2021
- 2021-07-21 CN CN202110823449.8A patent/CN113670610B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN113670610A (en) | 2021-11-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113670610B (en) | Fault detection method, system and medium based on wavelet transformation and neural network | |
| Maya et al. | dLSTM: a new approach for anomaly detection using deep learning with delayed prediction | |
| CN108334936B (en) | Fault prediction method based on migration convolutional neural network | |
| CN112990280B (en) | Class increment classification method, system, device and medium for image big data | |
| CN111340238A (en) | Fault diagnosis method, device, equipment and storage medium of industrial system | |
| Guo et al. | Deep convolution feature learning for health indicator construction of bearings | |
| CN108898238B (en) | Medical equipment fault prediction system and related method, device and equipment | |
| CN111428874A (en) | Wind control method, electronic device and computer readable storage medium | |
| JP2015087903A (en) | Apparatus and method for information processing | |
| CN112668164A (en) | Transformer fault diagnosis method and system for inducing ordered weighted evidence reasoning | |
| CN113344295B (en) | Method, system and medium for predicting residual life of equipment based on industrial big data | |
| CN107977748B (en) | Multivariable distorted time sequence prediction method | |
| EP3996035A1 (en) | Methods and systems for training convolutional neural networks | |
| CN112116957A (en) | Disease subtype prediction method, system, device and medium based on small sample | |
| CN113592071B (en) | Equipment fault recognition model training and recognition method, system, device and medium | |
| CN116740728A (en) | Dynamic acquisition method and system for wafer code reader | |
| US20210319269A1 (en) | Apparatus for determining a classifier for identifying objects in an image, an apparatus for identifying objects in an image and corresponding methods | |
| CN113989838A (en) | Pedestrian re-recognition model training method, recognition method, system, device and medium | |
| CN116522070A (en) | Non-supervision intelligent fault diagnosis method and system for mechanical parts | |
| US20220138573A1 (en) | Methods and systems for training convolutional neural networks | |
| CN115240843A (en) | Fairness prediction system based on structure causal model | |
| Saranya et al. | FBCNN-TSA: An optimal deep learning model for banana ripening stages classification | |
| CN113609933B (en) | Fault detection method, system, device and storage medium based on suffix tree | |
| CN112508080A (en) | Vehicle model identification method, device, equipment and medium based on experience playback | |
| CN112884093B (en) | Rotary machine fault diagnosis method and equipment based on DSCRN model and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |