CN113155422B - Fault detection method, device and storage medium - Google Patents
Fault detection method, device and storage medium Download PDFInfo
- Publication number
- CN113155422B CN113155422B CN202110339376.5A CN202110339376A CN113155422B CN 113155422 B CN113155422 B CN 113155422B CN 202110339376 A CN202110339376 A CN 202110339376A CN 113155422 B CN113155422 B CN 113155422B
- Authority
- CN
- China
- Prior art keywords
- vibration signal
- value
- determining
- difference
- edge
- 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
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01H—MEASUREMENT OF MECHANICAL VIBRATIONS OR ULTRASONIC, SONIC OR INFRASONIC WAVES
- G01H17/00—Measuring mechanical vibrations or ultrasonic, sonic or infrasonic waves, not provided for in the preceding groups
-
- 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
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- 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
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Biophysics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Acoustics & Sound (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Evolutionary Biology (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
Abstract
The invention discloses a fault detection method, a fault detection device and a storage medium, wherein the method comprises the following steps: acquiring a first vibration signal; determining an intrinsic mode function component set according to the first vibration signal and a preset threshold; determining a characteristic vector according to the eigenmode function component set; the feature vector characterizes the total energy of the eigenmode function component set; determining a fault detection result according to the feature vector and the classification decision function; according to the method, the first vibration signal is obtained, the intrinsic mode function component set is determined according to the first vibration signal and the preset threshold, then the characteristic vector representing the total energy of the intrinsic mode function component set is determined according to the intrinsic mode function component set, the first vibration signal is decomposed and subjected to stabilization treatment, and the processing result, namely the characteristic vector is combined with the classification decision function, so that the fault detection result is determined, the accuracy of fault detection can be effectively improved, and the method can be widely applied to the technical field of fault detection.
Description
Technical Field
The present invention relates to the field of fault detection, and in particular, to a fault detection method, apparatus, and storage medium.
Background
With the development of science and technology, the application field of the machine equipment is wider and wider, the development speed of the machine equipment is faster and faster, and the requirements on the machine equipment are higher and higher, such as the requirement that the machine equipment is suitable for different environments, the function and the precision of the machine equipment are higher, and the fault detection precision of the machine equipment is higher and higher. For example, when a machine device is detected by using a vibration signal, the vibration signal needs to be collected and analyzed, and the machine device is influenced by various factors of an external environment in the operation process, so that the collected vibration signal is actually changed and is usually a nonlinear signal.
Disclosure of Invention
In view of the above, in order to solve the above technical problems, an object of the present invention is to provide a fault detection method, device and storage medium, which improve the accuracy of fault detection.
The technical scheme adopted by the invention is as follows:
a fault detection method, comprising:
acquiring a first vibration signal;
determining an intrinsic mode function component set according to the first vibration signal and a preset threshold;
determining a characteristic vector according to the eigenmode function component set; the feature vector characterizes a total energy of the set of eigenmode function components;
and determining a fault detection result according to the feature vector and the classification decision function.
Further, the determining an eigenmode function component set according to the first vibration signal and a preset threshold includes:
selecting a first preset number of maximum value points from the plurality of maximum value points, calculating to obtain an edge maximum value according to the selected maximum value points, and taking the edge maximum value as the maximum value point at the leftmost end of the first vibration signal; the sum of the amplitudes between the edge maxima and the selected maxima point is minimal;
selecting a second preset number of minimum value points from the minimum value points, calculating to obtain an edge minimum value according to the selected minimum value points, and taking the edge minimum value as the minimum value point at the rightmost end of the first vibration signal; the sum of the amplitudes between the edge minimum and the selected minimum point is minimum;
connecting all the maximum values between the edge maximum value and the edge minimum value with the edge maximum value through a preset first sample line to obtain an upper envelope line, and connecting all the minimum values between the edge maximum value and the edge minimum value with the edge minimum value through a preset second sample line to obtain a lower envelope line;
calculating to obtain an average value according to the upper envelope line and the lower envelope line;
and determining the eigenmode function component set according to a first difference value between the first vibration signal and the average value and the preset threshold.
Further, the determining the set of eigenmode function components according to the first difference between the first vibration signal and the average value and the preset threshold includes:
when the first difference value is smaller than or equal to the preset threshold, taking the first difference value as a first component;
calculating a second difference value between the first vibration signal and the first component as a first intrinsic mode function component, and recalculating an average value of the first intrinsic mode function component as a new first vibration signal;
determining a second eigenmode function component according to the new first vibration signal, the recalculated average value and the preset threshold; the set of eigenmode function components comprises at least a first eigenmode function component and a second eigenmode function component.
Further, the determining the set of eigenmode function components according to the first difference between the first vibration signal and the average value and the preset threshold includes:
when the first difference value is larger than the preset threshold, taking the first difference value as a second vibration signal;
calculating a second average value of the second vibration signal, and calculating a third difference value between the second vibration signal and the second average value;
calculating a fourth difference value between the first difference value and the third difference value, determining a component parameter according to the fourth difference value and the first difference value, when the component parameter is less than or equal to the preset threshold, taking the third difference value as a second component, calculating a fifth difference value between the second vibration signal and the second component as a third eigenmode function component, when the component parameter is greater than the preset threshold, taking the third difference value as a new second vibration signal, returning to the step of calculating a second average value of the second vibration signal until the component parameter is less than or equal to the preset threshold, and obtaining a third eigenmode function component; the set of eigenmode function components comprises at least a third eigenmode function component.
Further, the determining feature vectors according to the set of eigenmode function components includes:
performing integral processing on each intrinsic mode function component to obtain energy corresponding to each intrinsic mode function component;
and carrying out normalization processing on each energy to obtain a characteristic vector.
Further, the determining a fault detection result according to the feature vector and the classification decision function includes:
inputting the feature vector into the classification decision function to obtain an output result;
calculating an absolute value of a difference between the output result and a first threshold as a first difference, and calculating an absolute value of a difference between the output result and a second threshold as a second difference;
when the first difference is larger than the second difference, determining that the fault detection result is normal;
wherein the formula of the classification decision function is:
wherein f (x) is a classification decision function, x is a feature vector, x i Is a support vector, N is a support vector x i The number of the (c) is (c),
as Lagrangian parameter, y i As a class label, K mix (x,x i ) As a kernel function of a support vector machine, b * Is a classification threshold.
Further, the classification decision function is obtained by training through the following steps:
acquiring a training set; the training set comprises training feature vectors and classification labels corresponding to each element in the training feature vectors;
inputting the training feature vectors and the classification labels into a preset classification function to adjust parameters;
when the output result of the preset classification function meets the training condition, determining the classification decision function according to the adjusted parameter and the preset classification function; wherein the adjusted parameters comprise the support vector, the Lagrangian parameter, the class label, a hyperplane weight coefficient vector, a weight factor, a scale, an attenuation parameter and a kernel function parameter; the classification threshold value is determined according to the hyperplane weight coefficient vector, and the kernel function of the support vector machine is determined according to the weight factor, the scale, the attenuation parameter and the kernel function parameter.
The present invention also provides a fault detection device, comprising:
the acquisition module is used for acquiring a first vibration signal;
the first determining module is used for determining an intrinsic mode function component set according to the first vibration signal and a preset threshold;
a second determining module, configured to determine a feature vector according to the eigenmode function component set; the feature vector characterizes a total energy of the set of eigenmode function components;
and the third determining module is used for determining a fault detection result according to the feature vector and the classification decision function.
The invention also provides a fault detection device, which comprises a processor and a memory;
the memory stores a program;
the processor executes the program to implement the fault detection method.
The present invention also provides a computer-readable storage medium storing a program which, when executed by a processor, implements the fault detection method.
The invention has the beneficial effects that: the method comprises the steps of obtaining a first vibration signal, determining an intrinsic mode function component set according to the first vibration signal and a preset threshold, then determining a feature vector representing the total energy of the intrinsic mode function component set according to the intrinsic mode function component set, decomposing and stabilizing the first vibration signal, and combining a processing result, namely the feature vector, with a classification decision function, so as to determine a fault detection result, and effectively improve the accuracy of fault detection.
Drawings
FIG. 1 is a flow chart illustrating the steps of a method for detecting faults according to the present invention;
FIG. 2 is a schematic diagram illustrating the steps for determining a set of eigenmode function components according to an embodiment of the present invention;
FIG. 3 is a schematic diagram illustrating the steps for determining feature vectors according to an embodiment of the present invention;
FIG. 4 is a schematic diagram illustrating a step of determining a detection result according to an embodiment of the present invention.
Detailed Description
In order to make the technical solutions of the present application better understood by those skilled in the art, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only some embodiments of the present application, and not all embodiments. All other embodiments obtained by a person of ordinary skill in the art based on the embodiments in the present application without making any creative effort shall fall within the protection scope of the present application.
The terms "first," "second," "third," and "fourth," etc. in the description and claims of this application and in the accompanying drawings are used for distinguishing between different elements and not for describing a particular sequential order. Furthermore, the terms "include" and "have," as well as any variations thereof, are intended to cover a non-exclusive inclusion. For example, a process, method, system, article, or apparatus that comprises a list of steps or elements is not limited to only those steps or elements but may alternatively include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
Reference herein to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the application. The appearances of the phrase in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. It is explicitly and implicitly understood by one skilled in the art that the embodiments described herein may be combined with other embodiments.
As shown in fig. 1, the present embodiment provides a fault detection method, which includes steps S100-S400:
s100, acquiring a first vibration signal.
Specifically, the first vibration signal may be operation data of the machine device collected by various types of sensors, or may also be operation data of the machine device received through a transmission interface or the like. Optionally, the machine device includes, but is not limited to, a robot, and the operation data includes, but is not limited to, data related to the rotation speed, the rotation shaft position, and the like of the rotation shaft. It should be noted that the first vibration signal includes a plurality of maximum value points and a plurality of minimum value points, and the first vibration signal may be a certain section of signal selected from the collected signals, and the abscissa is time and the ordinate is amplitude.
S200, determining an intrinsic mode function component set according to the first vibration signal and a preset threshold.
As shown in fig. 2, optionally, step S200 may include steps S211-S215:
s211, selecting a first preset number of maximum value points from the plurality of maximum value points, calculating to obtain an edge maximum value according to the selected maximum value points, and taking the edge maximum value as the maximum value point at the leftmost end of the first vibration signal.
Optionally, the first preset number may be adjusted as needed, for example, three maximum points are selected from the plurality of maximum points, an edge maximum is determined through calculation, so that the sum of the edge maximum and the amplitudes of the three maximum points is minimum, and then the edge maximum is used as the maximum point at the leftmost end of the first vibration signal. For example, the sum of the amplitudes is
Is the edge maximum, x (t) i ) Is the amplitude of the maximum point, t i For time, i is the number. It should be noted that, the amplitude of the leftmost maximum point is the edge maximum, and the time may be the time of the leftmost end of the current first vibration signal, or the time may also be the time of the leftmost maximum point. In the embodiment of the present invention, the three maximum value points selected are the three maximum value points closest to the starting point, which are selected by using the end point of the leftmost end of the first vibration signal as the starting point.
S212, selecting a second preset number of minimum value points from the minimum value points, calculating to obtain an edge minimum value according to the selected minimum value points, and taking the edge minimum value as the minimum value point at the rightmost end of the first vibration signal; the sum of the amplitudes between the edge minima and the selected minima point is minimal.
Optionally, the second preset number may be adjusted as needed, and may be the same as or different from the first preset number, and in this embodiment, the second preset number is also three, and then three minimum value points are selected from the multiple minimum value points, and an edge minimum value is determined through calculation, so that the sum of the edge minimum value and the amplitudes between the three minimum value points is minimum, and then the edge minimum value is used as the minimum value point at the rightmost end of the first vibration signal; the calculation method can refer to the principle of step S211. The amplitude of the minimum value point at the rightmost end is the edge minimum value, and the time may be the time of the end point at the rightmost end of the current first vibration signal, or the time may also be the time of the minimum value point at the rightmost end. In the embodiment of the present invention, the three minimum value points selected are three minimum value points which are closest to the starting point and are selected by using the end point of the rightmost end of the first vibration signal as the starting point.
And S213, connecting all the maximum values between the edge maximum value and the edge minimum value with the edge maximum value through a preset first sample line to obtain an upper envelope line, and connecting all the minimum values between the edge maximum value and the edge minimum value with the edge minimum value through a preset second sample line to obtain a lower envelope line.
Specifically, after determining an edge maximum value and an edge minimum value, a signal range is formed between the edge maximum value and the edge minimum value, all maximum value points (including the edge maximum value) in the signal range are connected through a preset first sample line to obtain an upper envelope line, and all minimum value points (including the edge minimum value) in the signal range are connected through a preset second sample line to obtain a lower envelope line. It should be noted that the form of the preset first spline line and the preset second spline line may be set according to actual needs, in this embodiment of the present application, both the preset first spline line and the preset second spline line are cubic spline curves, and the upper envelope line and the lower envelope line may be formed by a cubic spline interpolation method.
And S214, calculating to obtain an average value according to the upper envelope line and the lower envelope line.
Specifically, the amplitudes of the points on the upper envelope and the lower envelope are added (time is kept coincident), and then the average m is obtained 1 。
S215, determining an intrinsic mode function component set according to a first difference value between the first vibration signal and the average value and a preset threshold.
Optionally step S215 may be implemented by steps S221-S223:
and S221, when the first difference value is smaller than or equal to a preset threshold, taking the first difference value as a first component.
Optionally, a preset threshold epsilon may be set according to actual needs, and when the first difference is smaller than the preset threshold, the first difference is used as the first component.
S222, calculating a second difference value between the first vibration signal and the first component to serve as a first eigenmode function component, and calculating an average value again by taking the first eigenmode function component as a new first vibration signal.
Specifically, a second difference of the first vibration signal and the first component is calculated as a first eigenmode function component, and then the first eigenmode function component is recalculated as a new first vibration signal. It should be noted that, the recalculation of the average value can be realized through steps S211 to S214, and the principle is the same and is not described again.
And S223, determining a second eigenmode function component according to the new first vibration signal, the recalculated average value and a preset threshold.
Specifically, determining the second eigenmode function component according to the new first vibration signal, the recalculated average value and the preset threshold may be implemented through steps S221 to S222, and the principle is the same, so as to obtain the second eigenmode function component. It should be noted that the set of eigenmode function components at least includes a first eigenmode function component and a second eigenmode function component, and a plurality of eigenmode function components included in the set of eigenmode function components can be determined by performing multiple cycles according to the principle of the above steps.
Optionally, step S200 may also include steps S231-S233:
and S231, when the first difference value is larger than a preset threshold, taking the first difference value as a second vibration signal.
Optionally, a preset threshold epsilon may be set according to actual needs, and when the first difference is greater than the preset threshold, the first difference is used as the second vibration signal.
S232, calculating a second average value of the second vibration signal, and calculating a third difference value between the second vibration signal and the second average value.
Specifically, the second average value of the second vibration signal refers to an average value obtained by acquiring a new upper envelope and a new lower envelope using the second vibration signal and then determined according to the new upper envelope and the new lower envelope, and a specific principle may be found by referring to steps S211 to S214. A third difference between the second vibration signal and the second average is then calculated.
S233, calculating a fourth difference value between the first difference value and the third difference value, determining a component parameter according to the fourth difference value and the first difference value, when the component parameter is less than or equal to a preset threshold, taking the third difference value as a second component, calculating a fifth difference value between the second vibration signal and the second component as a third eigenmode function component, when the component parameter is greater than the preset threshold, taking the third difference value as a new second vibration signal, returning to the step of calculating a second average value of the second vibration signal, until the component parameter is less than or equal to the preset threshold, and obtaining a third eigenmode function component; the set of eigenmode function components comprises at least a third eigenmode function component.
Specifically, a fourth difference between the first difference and the third difference is calculated, a component parameter is determined according to the fourth difference and the first difference, when the component parameter is less than or equal to a preset threshold, the third difference is used as a second component, and a fifth difference between the second vibration signal and the second component is calculated as a third eigenmode function component. If the component parameter is greater than the preset threshold, taking the third difference as a new second vibration signal, returning to the step of calculating the second average value of the second vibration signal until the component parameter is less than or equal to the preset threshold, and obtaining a third eigenmode function component; the set of eigenmode function components comprises at least a third eigenmode function component. Optionally, the component parameter is determined according to a ratio of a square of the fourth difference to a square of the first difference when the component parameter is determined for the first time, and the component parameter is determined by combining the new average value and the determined difference value of the new second vibration signal on the basis of the fourth difference and the first difference when the component parameter is determined for the subsequent time.
It should be noted that steps S221-S223 and S231-S235 may be performed alternately, and the finally determined set of eigenmode function components may include the eigenmode function components determined by steps S221-S22 and may also include the eigenmode function components determined by steps S231-S235.
For example: when the upper envelope line and the lower envelope line are obtained for the first time, the average value of the first determination is recorded as m 1 ，x(t)-m 1 ＝h 1 X (t) is a first vibration signal, h 1 Is a first difference value when h 1 If the threshold condition is satisfied (less than or equal to a predetermined threshold epsilon, e.g., 0.3), the first component is determined. When h is 1 Not satisfying the threshold condition, h 1 The average m of the upper and lower envelope lines is obtained as the primary data (i.e., the first vibration signal) by the second calculation 1i Then, a new difference value h is determined 11 ＝h 1 -m 11 Whether a threshold condition is met, if so, a new difference value h is obtained 11 If not, recycling k times to obtain h 1(k-1) -m 1k ＝h 1k Wherein h is 1(0) I.e. h 1 So that h is 1k Satisfying the threshold condition, determining the first component, h 1(k-1) Is the difference between the k-1 th original data and the average value calculated from the original data in the k-th cycle, h 1k Is h 1(k-1) Difference from the mean value calculated from the original data of the k-th cycle, m 1k The average of the upper and lower envelope at the k-th cycle. Wherein the first component in the above step is denoted as c 1 (t) then use r 1 (t)＝x(t)-c 1 (t)，r 1 (t) removing the first eigenmode function component of the high frequency component, and then dividing r 1 And (t) as original data, namely x (t), circulating to obtain a plurality of components meeting the threshold condition so as to obtain a plurality of intrinsic mode function components, and stopping circulating when the finally obtained intrinsic mode function components are constants which are monotonous quantities, and determining a final intrinsic mode function component set. It should be noted that, when the k cycles are passed, the threshold parameter SD needs to be calculated and compared with the preset threshold epsilon,judging whether a threshold condition is reached, wherein the specific formula is as follows:
wherein h is 1(k-1) Is the difference between the k-1 th original data and the average value calculated from the original data in the k-th cycle, h 1k Is h 1(k-1) And the difference value of the average value obtained by calculation with the original data of the kth cycle, wherein the 0 th original data is the first vibration signal. After obtaining the components meeting the threshold condition, the intrinsic mode function components are calculated. For example: r is n-1 (t)-c n (t)＝r n (t)，n≥1，c n (t) is the nth component satisfying the threshold condition, r n (t) is the nth eigenmode function component, r n-1 (t) is the (n-1) th eigenmode function component.
And S300, determining a characteristic vector according to the eigenmode function component set.
It should be noted that the set of eigenmode function components includes a plurality of eigenmode function components, and the feature vector characterizes total energy of the set of eigenmode function components.
As shown in fig. 3, optionally, step S300 may include steps S310-S320:
and S310, performing integral processing on each intrinsic mode function component to obtain energy corresponding to each intrinsic mode function component.
Specifically, integral processing is performed on each eigenmode function component, and a specific formula is as follows:
wherein r is i (t) is the ith eigenmode function component, E i The energy corresponding to the ith eigenmode function component.
And S320, normalizing each energy to obtain a feature vector.
In the embodiment of the invention, energy is used as an element to construct a preliminary feature vector T:
T＝[E 1 ,E 2 ,…,E n ]
generally, the energy value is large, so that for subsequent analysis and processing, the preliminary feature vector T is normalized, specifically:
T′＝[E 1 /E,E 2 /E,…,E n /E]
wherein E is a normalization parameter, the value range of i is n, and T' is the finally obtained characteristic vector.
And S400, determining a fault detection result according to the feature vector and the classification decision function.
As shown in fig. 4, optionally, step S400 may include the steps of: S410-S430:
and S410, inputting the feature vector into a classification decision function to obtain an output result.
Specifically, the formula of the classification decision function is:
wherein f (x) is a classification decision function, x is a feature vector, x i Is a support vector, N is a support vector x i The number of the (c) component(s),
as Lagrangian parameter, y i As a class label, K mix (x,x i ) As a kernel function of a support vector machine, b * Is a classification threshold. It should be noted that the classification decision function f (x) is a trained classification decision function, and the support vectors, the number of the support vectors, the lagrangian parameters, the class labels, the kernel functions of the support vector machine, and the classification threshold are determined to continuously optimize the training conditions in the training processAnd the degree is the optimal parameter.
S420, calculating an absolute value of a difference between the output result and the first threshold as a first difference, and calculating an absolute value of a difference between the output result and the second threshold as a second difference.
In the embodiment of the present invention, the first threshold and the second threshold may be set as required, for example, the first threshold is +1, the second threshold is-1, when the output result is 0.7, the first difference is 0.3, and the second difference is 1.3; when the output result is-0.7, the first difference is 1.7 and the second difference is 0.3.
S430, when the first difference is larger than the second difference, determining that the fault detection result is normal.
In the embodiment of the present invention, the fault detection result is determined to be normal or abnormal by comparing the first difference with the second difference, specifically, when the first difference is greater than the second difference, the fault detection result is determined to be normal, otherwise, the fault detection result is determined to be a fault. In the above example, when the output result is 0.7 and the first difference 0.3 is smaller than the second difference 1.3, the fault detection result is considered as abnormal; and when the output result is-0.7, the first difference 1.7 is larger than the second difference 0.3, and the fault detection result is determined to be normal.
Optionally, the classification decision function in the embodiment of the present invention is obtained by training, and specifically the training process includes steps S501 to S503:
s501, acquiring a training set.
Specifically, the training set includes training feature vectors and class labels corresponding to each element in the training feature vectors. It should be noted that the training feature vector is obtained by collecting a large amount of vibration signal data in a manner similar to steps S200-S300, and then manually setting a corresponding classification label, such as normal or abnormal, for each element in the training feature vector.
And S502, inputting the training feature vectors and the classification labels into a preset classification function to adjust parameters.
Specifically, the form of the preset classification function is the same as that of the classification decision function, and the training feature vector and the classification label are input to the preset classification function to perform training and parameter adjustment.
And S503, when the output result of the preset classification function meets the training condition, determining a classification decision function according to the adjusted parameter and the preset classification function.
Specifically, the adjusted parameters include support vectors, lagrangian parameters, class labels, hyperplane weight coefficient vectors, weight factors, scales, attenuation parameters, and kernel function parameters. In the embodiment of the invention, the classification threshold b * The concrete formula is as follows:
omega is a hyperplane weight coefficient vector, x (1) is any support vector in a fault sample, and x (-1) is any support vector in a normal sample. Wherein, the hyperplane weight coefficient vector ω has a limiting condition in the training process, specifically:
wherein C is a penalty parameter, b is an offset, y i Is labeled as category (take +1 or-1), x i To support the vector, xi i Is a relaxation variable, m is a relaxation variable xi i The number of the cells.
In the embodiment of the invention, the kernel function K of the support vector machine mix (x,x i ) The concrete formula is as follows:
wherein, δ is a weight factor, and the value range is as follows: 0<δ<1, v is the scale, x is the training feature vector, x i The parameters in the kernel function of the support vector machine can be finally determined by training and adjusting. It will be appreciated that from the above formula, the support vector is knownAnd the kernel function of the machine is determined according to the weight factor, the scale, the attenuation parameter and the kernel function parameter.
In summary, the fault detection method of the present invention can decompose and stabilize the first vibration signal, and combine the processing result, i.e., the feature vector, with the classification decision function, thereby determining the fault detection result, and effectively improving the accuracy of fault detection. The invention utilizes the chemical support vector machine to detect and classify the faults, can well solve the problem of small samples, has higher detection and diagnosis precision than a neural network and lower structural dependence than the neural network, and has higher training speed and higher stability. In addition, the fault detection method can detect the fault in use more timely, reduce misjudgment, find problems timely, avoid delaying rescue of trapped personnel and timely retrieve property loss. The detection method can efficiently and accurately acquire the vibration signals and filter out noise, avoids the end effect of the detection method, and can quickly and accurately diagnose the faults of the machine equipment, such as the faults of the rolling bearing, by using the optimized support vector machine by adopting small sample training.
The present invention also provides a fault detection device, comprising:
the acquisition module is used for acquiring a first vibration signal;
the first determining module is used for determining an eigenmode function component set according to the first vibration signal and a preset threshold;
the second determining module is used for determining a characteristic vector according to the eigenmode function component set; the feature vector characterizes the total energy of the eigenmode function component set;
and the third determining module is used for determining a fault detection result according to the feature vector and the classification decision function.
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.
The embodiment of the invention also provides a device, and the equipment comprises a processor and a memory;
the memory is used for storing programs;
the processor is used for executing programs to realize the fault detection method of the embodiment of the invention. The device of the embodiment of the invention can realize the function of fault detection. The device can be any intelligent terminal such as a mobile phone, a tablet Personal computer, a Personal Digital Assistant (PDA for short), a Point of Sales (POS for short), a vehicle-mounted computer and the like.
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.
The embodiment of the present invention further provides a computer-readable storage medium, where a program is stored in the computer-readable storage medium, and the program is executed by a processor to implement the fault detection method according to the foregoing embodiment of the present invention.
Embodiments of the present invention also provide a computer program product including instructions, which when run on a computer, cause the computer to perform the fault detection method of the foregoing embodiments of the invention.
The terms "first," "second," "third," "fourth," and the like (if any) in the description of the present application and in the drawings described above are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used is interchangeable under appropriate circumstances such that the embodiments of the application described herein are capable of operation in sequences other than those illustrated or described herein. Moreover, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
It should be understood that, in this application, "at least one" means one or more, "a plurality" means two or more. "and/or" is used to describe the association relationship of the associated object, indicating that there may be three relationships, for example, "a and/or B" may indicate: only A, only B and both A and B are present, wherein A and B may be singular or plural. The character "/" generally indicates that the former and latter associated objects are in an "or" relationship. "at least one of the following" or similar expressions refer to any combination of these items, including any combination of single item(s) or plural items. For example, at least one (one) of a, b, or c, may represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", wherein a, b, c may be single or plural.
In the several embodiments provided in the present application, it should be understood that the disclosed apparatus and method may be implemented in other manners. For example, the above-described apparatus embodiments are merely illustrative, and for example, a division of a unit is only a logical division, and other divisions may be realized in practice, for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or units, and may be in an electrical, mechanical or other form. The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one position, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment. In addition, functional units in the embodiments of the present application may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, and can also be realized in a form of a software functional unit.
The integrated unit, if implemented in the form of a software functional unit 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 application may be substantially implemented or contributed to by the prior art, or all or part of the technical solution may be embodied in a software product, which is stored in a storage medium and includes multiple 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 method of the embodiments of the present application. And the aforementioned storage medium includes: various media capable of storing programs, such as a usb disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk.
The above embodiments are only used to illustrate the technical solutions of the present application, and not to limit the same; although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present application.
Claims (9)
1. A method of fault detection, comprising:
acquiring a first vibration signal;
determining an intrinsic mode function component set according to the first vibration signal and a preset threshold;
determining a characteristic vector according to the eigenmode function component set; the feature vector characterizes a total energy of the set of eigenmode function components;
determining a fault detection result according to the feature vector and a classification decision function;
the determining of the eigenmode function component set according to the first vibration signal and a preset threshold includes:
selecting a first preset number of maximum value points from the maximum value points, calculating to obtain an edge maximum value according to the selected maximum value points, and taking the edge maximum value as the maximum value point at the leftmost end of the first vibration signal; the sum of the amplitudes between the edge maxima and the selected maxima point is minimal;
selecting a second preset number of minimum value points from the minimum value points, calculating to obtain an edge minimum value according to the selected minimum value points, and taking the edge minimum value as the minimum value point at the rightmost end of the first vibration signal; the sum of the amplitudes between the edge minimum value and the selected minimum value point is minimum;
connecting all the maximum values between the edge maximum value and the edge minimum value with the edge maximum value through a preset first sample line to obtain an upper envelope line, and connecting all the minimum values between the edge maximum value and the edge minimum value with the edge minimum value through a preset second sample line to obtain a lower envelope line;
calculating to obtain an average value according to the upper envelope line and the lower envelope line;
and determining the intrinsic mode function component set according to a first difference value between the first vibration signal and the average value and the preset threshold.
2. The fault detection method of claim 1, wherein: determining the set of eigenmode function components according to a first difference between the first vibration signal and the average and the preset threshold, including:
when the first difference value is smaller than or equal to the preset threshold, taking the first difference value as a first component;
calculating a second difference value between the first vibration signal and the first component as a first eigenmode function component, and recalculating an average value of the first eigenmode function component as a new first vibration signal;
determining a second eigenmode function component according to the new first vibration signal, the recalculated average value and the preset threshold; the set of eigenmode function components comprises at least a first eigenmode function component and a second eigenmode function component.
3. The fault detection method of claim 1, wherein: determining the eigenmode function component set according to a first difference between the first vibration signal and the average value and the preset threshold, including:
when the first difference value is larger than the preset threshold, taking the first difference value as a second vibration signal;
calculating a second average value of the second vibration signal, and calculating a third difference value between the second vibration signal and the second average value;
calculating a fourth difference value between the first difference value and the third difference value, determining a component parameter according to the fourth difference value and the first difference value, when the component parameter is less than or equal to the preset threshold, taking the third difference value as a second component, calculating a fifth difference value between the second vibration signal and the second component as a third eigenmode function component, when the component parameter is greater than the preset threshold, taking the third difference value as a new second vibration signal, returning to the step of calculating a second average value of the second vibration signal until the component parameter is less than or equal to the preset threshold, and obtaining a third eigenmode function component; the set of eigenmode function components comprises at least a third eigenmode function component.
4. The fault detection method of claim 1, wherein: the set of eigenmode function components comprises a plurality of eigenmode function components, and determining a feature vector according to the set of eigenmode function components comprises:
performing integral processing on each intrinsic mode function component to obtain energy corresponding to each intrinsic mode function component;
and carrying out normalization processing on each energy to obtain a feature vector.
5. The fault detection method of claim 1, wherein: the determining a fault detection result according to the feature vector and the classification decision function includes:
inputting the feature vector into the classification decision function to obtain an output result;
calculating an absolute value of a difference between the output result and a first threshold as a first difference, and calculating an absolute value of a difference between the output result and a second threshold as a second difference;
when the first difference is larger than the second difference, determining that the fault detection result is normal;
wherein the formula of the classification decision function is:
wherein f (x) is a classification decision function, x is a feature vector, x i Is a support vector, N is a support vector x i The number of the (c) component(s),
as lagrange parameter, y i As a class label, K mix (x,x i ) As a kernel function of a support vector machine, b * Is a classification threshold.
6. The fault detection method of claim 5, wherein: the classification decision function is obtained by training through the following steps:
acquiring a training set; the training set comprises training feature vectors and classification labels corresponding to each element in the training feature vectors;
inputting the training feature vectors and the classification labels into a preset classification function to adjust parameters;
when the output result of the preset classification function meets the training condition, determining the classification decision function according to the adjusted parameters and the preset classification function; wherein the adjusted parameters comprise the support vector, the Lagrangian parameter, the class label, a hyperplane weight coefficient vector, a weight factor, a scale, an attenuation parameter and a kernel function parameter; the classification threshold is determined according to the hyperplane weight coefficient vector, and the kernel function of the support vector machine is determined according to the weight factor, the scale, the attenuation parameter and the kernel function parameter.
7. A fault detection device, comprising:
the acquisition module is used for acquiring a first vibration signal;
the first determining module is used for determining an intrinsic mode function component set according to the first vibration signal and a preset threshold;
a second determining module, configured to determine a feature vector according to the eigenmode function component set; the feature vector characterizes a total energy of the set of eigenmode function components;
the third determining module is used for determining a fault detection result according to the feature vector and the classification decision function;
the determining of the eigenmode function component set according to the first vibration signal and a preset threshold includes:
selecting a first preset number of maximum value points from the maximum value points, calculating to obtain an edge maximum value according to the selected maximum value points, and taking the edge maximum value as the maximum value point at the leftmost end of the first vibration signal; the sum of the amplitudes between the edge maxima and the selected maxima point is minimal;
selecting a second preset number of minimum value points from the minimum value points, calculating to obtain an edge minimum value according to the selected minimum value points, and taking the edge minimum value as the minimum value point at the rightmost end of the first vibration signal; the sum of the amplitudes between the edge minimum and the selected minimum point is minimum;
connecting all the maximum values between the edge maximum value and the edge minimum value with the edge maximum value through a preset first sample line to obtain an upper envelope line, and connecting all the minimum values between the edge maximum value and the edge minimum value with the edge minimum value through a preset second sample line to obtain a lower envelope line;
calculating to obtain an average value according to the upper envelope line and the lower envelope line;
and determining the intrinsic mode function component set according to a first difference value between the first vibration signal and the average value and the preset threshold.
8. A fault detection device is characterized by comprising a processor and a memory;
the memory stores a program;
the processor executes the program to implement the method of any one of claims 1-6.
9. A computer-readable storage medium, characterized in that the storage medium stores a program which, when executed by a processor, implements the method according to any one of claims 1-6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110339376.5A CN113155422B (en) | 2021-03-30 | 2021-03-30 | Fault detection method, device and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110339376.5A CN113155422B (en) | 2021-03-30 | 2021-03-30 | Fault detection method, device and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113155422A CN113155422A (en) | 2021-07-23 |
| CN113155422B true CN113155422B (en) | 2022-07-26 |
Family
ID=76885324
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110339376.5A Active CN113155422B (en) | 2021-03-30 | 2021-03-30 | Fault detection method, device and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113155422B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH07128134A (en) * | 1993-11-04 | 1995-05-19 | Toshiba Corp | Monitoring-diagnosing apparatus for rotary machine |
| CN106092574A (en) * | 2016-05-30 | 2016-11-09 | 西安工业大学 | The Method for Bearing Fault Diagnosis selected with sensitive features is decomposed based on improving EMD |
| CN110824281A (en) * | 2019-11-19 | 2020-02-21 | 国网江苏省电力有限公司检修分公司 | Method and system for on-line monitoring and fault diagnosis of synchronous phase modulator |
| CN111832353A (en) * | 2019-04-19 | 2020-10-27 | 中国科学院沈阳自动化研究所 | Steam turbine rotor fault diagnosis method based on EMD and BA optimization SVM |
| CN112461546A (en) * | 2020-10-27 | 2021-03-09 | 江苏大学 | Construction method and diagnosis method of pump bearing fault diagnosis model based on improved binary tree support vector machine |
-
2021
- 2021-03-30 CN CN202110339376.5A patent/CN113155422B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH07128134A (en) * | 1993-11-04 | 1995-05-19 | Toshiba Corp | Monitoring-diagnosing apparatus for rotary machine |
| CN106092574A (en) * | 2016-05-30 | 2016-11-09 | 西安工业大学 | The Method for Bearing Fault Diagnosis selected with sensitive features is decomposed based on improving EMD |
| CN111832353A (en) * | 2019-04-19 | 2020-10-27 | 中国科学院沈阳自动化研究所 | Steam turbine rotor fault diagnosis method based on EMD and BA optimization SVM |
| CN110824281A (en) * | 2019-11-19 | 2020-02-21 | 国网江苏省电力有限公司检修分公司 | Method and system for on-line monitoring and fault diagnosis of synchronous phase modulator |
| CN112461546A (en) * | 2020-10-27 | 2021-03-09 | 江苏大学 | Construction method and diagnosis method of pump bearing fault diagnosis model based on improved binary tree support vector machine |
Non-Patent Citations (1)
| Title |
|---|
| 汪学渊等.基于EMD的电机轴承故障识别研究.《煤矿机械》.2009,第30卷(第02期),第215-217页. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113155422A (en) | 2021-07-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102893294A (en) | Probability density function estimator | |
| US20130338965A1 (en) | Anomaly Detection Method, Program, and System | |
| CN113091872B (en) | Method and device for diagnosing fault sensor | |
| Chen et al. | An efficient ICA-DW-SVDD fault detection and diagnosis method for non-Gaussian processes | |
| CN108399434B (en) | Analysis and prediction method of high-dimensional time series data based on feature extraction | |
| CN116861798B (en) | Online real-time residual life prediction method for vacuum dry pump based on XGBoost algorithm | |
| CN111666757A (en) | Commodity comment emotional tendency analysis method, device and equipment and readable storage medium | |
| CN110955862B (en) | Evaluation method and device for equipment model trend similarity | |
| CN115359846A (en) | Batch correction method and device for group data, storage medium and electronic equipment | |
| CN111191671A (en) | Electrical appliance waveform detection method and system, electronic equipment and storage medium | |
| CN112202630A (en) | Network quality abnormity detection method and device based on unsupervised model | |
| CN113989519B (en) | Long-tail target detection method and system | |
| CN111461923A (en) | Electricity stealing monitoring system and method based on deep convolutional neural network | |
| CN113155422B (en) | Fault detection method, device and storage medium | |
| CN116564418B (en) | Cell group correlation network construction method, device, equipment and storage medium | |
| CN115270998B (en) | Method and device for orderly classifying time series, computer equipment and storage medium | |
| WO2019177130A1 (en) | Information processing device and information processing method | |
| CN115526882A (en) | Medical image classification method, device, equipment and storage medium | |
| Khalid et al. | Performance Comparison of Feature Selection Methods for Prediction in Medical Data | |
| CN115111182A (en) | Method and device for detecting running state of direct current fan and terminal equipment | |
| CN114692693A (en) | Distributed optical fiber signal identification method, device and storage medium based on fractal theory | |
| CN110046770B (en) | Grain mildew prediction method and device | |
| CN112217749A (en) | Blind signal separation method and device | |
| CN109436980B (en) | Method and system for detecting state of elevator component | |
| CN113256093A (en) | Dynamic risk pool monitoring method, system, equipment and readable 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 |