Abnormal vibration signal detection method and device, equipment and storage medium
By extracting the sub-sequence signal from the device vibration signal and calculating the similarity coefficient of its frequency domain sequence, it directly judges whether the vibration signal is an abnormal signal, which solves the problem of poor model training effect in the prior art, and improves the accuracy and reliability of fault analysis.
Patent Information
- Application Number
- CN202411819913.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-11
- Publication Date
- 2025-03-14
AI Technical Summary
In the analysis of equipment vibration signal, pre-trained anomaly detection model is required to distinguish normal and abnormal vibration signals. However, due to the stability of the equipment, there is less abnormal data and poor model training effect, which affects the detection accuracy and accuracy of fault analysis.
By extracting multiple subsequence signals from the vibration signal, performing discrete Fourier transform, the similarity coefficient between the frequency domain sequence of the subsequence signal and the frequency domain sequence of the first subsequence signal is calculated, and whether the vibration signal is an abnormal signal is not required, without using an abnormality detection model.
It realizes accurate detection of abnormal vibration signals that do not rely on an abnormal detection model, avoids the problem of low detection accuracy caused by poor model training effect, and improves the accuracy and reliability of fault analysis.
Smart Images
Figure CN119622361A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of equipment monitoring, and in particular to an abnormal vibration signal detection method, device, equipment and storage medium. Background Art
[0002] For rotating equipment, most common faults are closely related to vibration. A healthy and well-designed mechanical equipment has a low vibration amplitude, but when looseness, wear, imbalance and other changes occur inside it, its vibration signal will usually change accordingly. Therefore, the vibration signal is one of the key parameters for judging early equipment failures, and the analysis of vibration signals is also a key measure for equipment fault diagnosis.
[0003] Discrete Fourier Transform (DFT) is a key method for vibration signal analysis. DFT can decompose discrete time domain signals into their frequency components. Through DFT, the various frequency components in the vibration signal can be analyzed (i.e., spectrum analysis), thereby determining the source of the vibration and realizing fault diagnosis. According to the requirements of fault diagnosis by spectrum analysis and the characteristics of DFT methods, vibration signals usually need to meet certain requirements, such as (1) statistical characteristics do not change with time; (2) equipment operating conditions (such as rotation speed) do not change with time. When the vibration signal does not meet the above requirements, the DFT result will be wrong or the fault frequency component cannot be obtained through the spectrum diagram. Therefore, after collecting the equipment vibration signal and before DFT, it is generally necessary to clean and process the vibration signal.
[0004] In the prior art, abnormal vibration signals can be screened out based on a pre-trained anomaly detection model to eliminate abnormal vibration signals. However, it is necessary to use labeled data for training in advance to establish a model that distinguishes normal and abnormal vibration signals. However, due to the stability of the equipment, it generates fewer abnormal vibration signals during use, so there is less abnormal data for training the model, resulting in poor training results for the model, affecting the model's detection accuracy for abnormal vibration signals, thereby reducing the accuracy and reliability of subsequent fault analysis. Summary of the invention
[0005] The present application provides an abnormal vibration signal detection method, apparatus, device and storage medium to extract multiple subsequence signals from a vibration signal, accurately detect whether the vibration signal is an abnormal signal based on the similarity coefficient between the frequency domain sequence of the subsequence signal and the frequency domain sequence of the first subsequence signal, and accurately detect abnormal vibration signals without using an abnormal detection model, thereby avoiding the problem of low accuracy in abnormal vibration signal detection caused by poor model training effect, thereby improving the accuracy and reliability of subsequent fault analysis.
[0006] In a first aspect, the present application provides a method for detecting an abnormal vibration signal, comprising:
[0007] Acquire a vibration signal to be detected, and extract a plurality of subsequence signals from the vibration signal based on a preset sliding window and a preset sliding step size;
[0008] Performing discrete Fourier transform on each of the subsequence signals to obtain a corresponding frequency domain sequence, and determining the rank of each element value according to the order of magnitude of each element value in the frequency domain sequence;
[0009] Determine, according to the rank number of each element value in each of the frequency domain sequences and the rank number of each element value in the frequency domain sequence corresponding to the first subsequence signal, a similarity coefficient between each of the frequency domain sequences and the frequency domain sequence corresponding to the first subsequence signal;
[0010] When the similarity coefficient corresponding to any of the frequency domain sequences is less than a preset similarity threshold, the vibration signal is determined to be an abnormal vibration signal.
[0011] In a second aspect, the present application provides an abnormal vibration signal detection device, comprising:
[0012] A subsequence signal extraction module is configured to obtain a vibration signal to be detected, and extract a plurality of subsequence signals from the vibration signal based on a preset sliding window and a preset sliding step size;
[0013] a rank determination module, configured to perform discrete Fourier transform on each of the subsequence signals to obtain a corresponding frequency domain sequence, and determine the rank of each element value according to the order of magnitude of each element value in the frequency domain sequence;
[0014] a correlation determination module, configured to determine a similarity coefficient between each of the frequency domain sequences and the frequency domain sequence corresponding to the first subsequence signal according to the rank number of each element value in each of the frequency domain sequences and the rank number of each element value in the frequency domain sequence corresponding to the first subsequence signal;
[0015] The first abnormal signal detection module is configured to determine that the vibration signal is an abnormal vibration signal when a similarity coefficient corresponding to any of the frequency domain sequences is less than a preset similarity threshold.
[0016] In a third aspect, the present application provides an abnormal vibration signal detection device, comprising:
[0017] One or more processors; a storage device storing one or more programs, when the one or more programs are executed by the one or more processors, the one or more processors implement the abnormal vibration signal detection method as described in the first aspect.
[0018] In a fourth aspect, the present application provides a storage medium comprising computer executable instructions, which, when executed by a computer processor, are used to perform the abnormal vibration signal detection method as described in the first aspect.
[0019] In the present application, by acquiring a vibration signal to be detected, multiple subsequence signals are extracted from the vibration signal based on a preset sliding window and a preset sliding step size; a discrete Fourier transform is performed on each subsequence signal to obtain a corresponding frequency domain sequence, and the rank number of each element value in the frequency domain sequence is determined according to the order of the magnitude of each element value in the frequency domain sequence; based on the rank number of each element value in each frequency domain sequence and the rank number of each element value in the frequency domain sequence corresponding to the first subsequence signal, a similarity coefficient between each frequency domain sequence and the frequency domain sequence corresponding to the first subsequence signal is determined; when the similarity coefficient corresponding to any frequency domain sequence is less than a preset similarity threshold, the vibration signal is determined to be an abnormal vibration signal. Through the above technical means, discrete Fourier transform is performed on each subsequence signal extracted corresponding to the vibration signal to obtain the frequency domain sequence corresponding to each subsequence signal. The rank of the element value in the frequency domain sequence represents the order of the amplitude corresponding to each frequency component. Therefore, the similarity of each frequency component can be calculated based on the rank of the element value at the same position in the frequency domain sequence and the frequency domain sequence of the first subsequence signal, so as to obtain the similarity coefficient between the frequency domain sequence and the frequency domain sequence of the first subsequence signal. When the similarity coefficient is greater than or equal to the similarity threshold, it can be considered that the working conditions of the frequency domain sequence and the frequency domain sequence of the first subsequence signal are similar. Therefore, when the similarity coefficient between any frequency domain sequence and the frequency domain sequence of the first subsequence signal is less than the similarity threshold, it can be determined that the working condition of the vibration signal changes with time, and then the vibration signal is determined to be an abnormal vibration signal, so that the abnormal vibration signal can be accurately detected without using an abnormal detection model, and the problem of low detection accuracy of abnormal vibration signals caused by poor model training effect is avoided, thereby improving the accuracy and reliability of subsequent fault analysis. BRIEF DESCRIPTION OF THE DRAWINGS
[0020] Figure 1 is a flow chart of an abnormal vibration signal detection method provided in an embodiment of the present application;
[0021] Figure 2 This is a flowchart of determining the rank number of an element value in a frequency domain sequence provided by an embodiment of the present application;
[0022] Figure 3 It is a schematic diagram of the sequence number and rank number of each element value in the frequency domain sequence provided by an embodiment of the present application;
[0023] Figure 4 is a flow chart of judging whether a vibration signal is an abnormal vibration signal based on a signal mean value provided by an embodiment of the present application;
[0024] Figure 5 is a flow chart of determining whether a vibration signal is an abnormal vibration signal based on signal variance provided by an embodiment of the present application;
[0025] Figure 6 is a schematic diagram of a process of retaining or deleting a vibration signal provided in an embodiment of the present application;
[0026] Figure 7 It is a structural schematic diagram of an abnormal vibration signal detection device provided in an embodiment of the present application;
[0027] Figure 8 It is a structural schematic diagram of an abnormal vibration signal detection device provided in an embodiment of the present application. DETAILED DESCRIPTION
[0028] In order to make the purpose, technical scheme and advantages of the present application clearer, the specific embodiments of the present application are further described in detail below in conjunction with the accompanying drawings. It is understood that the specific embodiments described herein are only used to explain the present application, rather than to limit the present application. It should also be noted that, for the convenience of description, only the part related to the present application but not all the contents are shown in the accompanying drawings. Before discussing the exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flow charts. Although the flow chart describes each operation (or step) as a sequential process, many of the operations therein can be implemented in parallel, concurrently or simultaneously. In addition, the order of each operation can be rearranged. The process can be terminated when its operation is completed, but it can also have additional steps not included in the accompanying drawings. The process can correspond to a method, a function, a procedure, a subroutine, a subprogram, etc.
[0029] The terms "first", "second", etc. in the specification and claims of the present application are used to distinguish similar objects, and are not used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable under appropriate circumstances, so that the embodiments of the present application can be implemented in an order other than those illustrated or described here, and the objects distinguished by "first", "second", etc. are generally of one type, and the number of objects is not limited. For example, the first object can be one or more. In addition, "and / or" in the specification and claims represents at least one of the connected objects, and the character " / " generally indicates that the objects associated with each other are in an "or" relationship.
[0030] In a related implementation, abnormal vibration signals can be screened out based on a pre-trained anomaly detection model to eliminate abnormal vibration signals. However, it is necessary to use labeled data for training in advance to establish a model that distinguishes normal and abnormal vibration signals. However, due to the stability of the equipment, it generates fewer abnormal vibration signals during use, so there is less abnormal data for training the model, resulting in poor training results for the model, affecting the model's detection accuracy for abnormal vibration signals, thereby reducing the accuracy and reliability of subsequent fault analysis.
[0031] To solve the above problems, the present embodiment provides an abnormal vibration signal detection method, which extracts multiple subsequence signals from the vibration signal, and accurately detects whether the vibration signal is an abnormal signal based on the similarity coefficient between the frequency domain sequence corresponding to the subsequence signal and the frequency domain sequence of the vibration signal. Abnormal vibration signals can be accurately detected without using an abnormal detection model, thereby improving the accuracy and reliability of subsequent fault analysis.
[0032] The abnormal vibration signal detection method provided in this embodiment can be performed by an abnormal vibration signal detection device, which can be implemented by software and / or hardware. The abnormal vibration signal detection device can be composed of two or more physical entities, or can be composed of one physical entity. For example, the abnormal vibration signal detection device can be an intelligent terminal with strong processing capabilities such as a computer and a server. Among them, the server can be implemented by an independent server or a server cluster composed of multiple servers.
[0033] The abnormal vibration signal detection device is installed with at least one type of operating system, and the abnormal vibration signal detection device can install at least one application based on the operating system, and the application can be an application that comes with the operating system, or an application downloaded from a third-party device or server. In this embodiment, the abnormal vibration signal detection device is installed with at least an application that can execute the abnormal vibration signal detection method.
[0034] For ease of understanding, this embodiment is described by taking a server as an example of a subject that executes the abnormal vibration signal detection method.
[0035] Figure 1 A flow chart of an abnormal vibration signal detection method provided by an embodiment of the present application is given. Figure 1 , the abnormal vibration signal detection method specifically includes:
[0036] S110, obtaining a vibration signal to be detected, and extracting a plurality of subsequence signals from the vibration signal based on a preset sliding window and a preset sliding step size.
[0037] Exemplarily, the sensor for collecting detection signals is installed at the key part of the equipment, and the sensor collects the vibration signal of the equipment in real time when the equipment is running. The sensor samples at a fixed sampling frequency f, and collects vibration signals of a fixed number of points n each time. The physical quantity of the vibration signal can be displacement, velocity or acceleration, which is specifically related to the sensor type and whether calculus operations are performed. However, vibration signals of different physical quantities have the same subsequent abnormal vibration signal detection process. In this regard, the vibration signal X is a signal sequence containing n signal values, which can be expressed as X = {x1,...,x i ,...,x n}, x i is the i-th signal value in the signal sequence corresponding to the vibration signal.
[0038] After the sensor collects the vibration signal in real time, the vibration signal is used as the vibration signal to be detected to determine whether the vibration signal is an abnormal vibration signal. If it is, the vibration signal is discarded to avoid using the abnormal vibration signal to erroneously analyze the operating status of the equipment. If not, the vibration signal is retained to accurately analyze the operating status of the equipment based on the vibration signal, thereby realizing real-time operating status monitoring of the equipment.
[0039] For the vibration signal to be detected, the preset sliding window is slid according to the preset sliding step length, and the corresponding signal value is intercepted from the vibration signal based on the preset sliding window as a subsequence signal. The length of each subsequence signal is equal to the length of the preset sliding window, and there are non-overlapping signal values between two adjacent subsequence signals, and the number of non-overlapping signal values is equal to the preset sliding step length.
[0040] Optionally, the implementation process of extracting the subsequence signal is as follows: determine the starting point of the vibration signal as the starting point of the preset sliding window, slide the preset sliding window based on the preset sliding step; extract the signal value of the vibration signal in the preset sliding window as a subsequence signal each time the sliding is performed until the preset sliding window reaches the end of the vibration signal. Exemplarily, the starting point of the vibration signal is used as the starting point of the starting end of the preset sliding window, and multiple signal values of the vibration signal in the preset sliding window are extracted as the first subsequence signal. After the preset sliding window is slid by the preset sliding step, multiple signal values of the vibration signal in the preset sliding window after sliding are extracted as the second subsequence signal. Then, the preset sliding window is slid by the preset sliding step, and then the third subsequence signal is extracted. Similarly, when the end of the preset sliding window reaches the end of the vibration signal, the last subsequence signal is extracted. For example, assuming that the length of the preset sliding window is w and the preset sliding step is 1, the subsequence signals extracted from the vibration signal include {X1,...,X i ,...,X n-w+1}, X iis the i-th subsequence signal extracted from the vibration signal, X i ={x i ,x i+1 ,...,x i+w-1}.
[0041] It should be noted that the length of the preset sliding window is generally an integer between one quarter and three quarters of the length of the vibration signal. The larger the length of the preset sliding window, the greater the number of signal values in the extracted subsequence signal. In addition to being set to 1, the preset sliding step can also be set to other integers, but the subsequence signal extracted when the preset sliding step is greater than 1 is a subset of the subsequence signal extracted when the preset sliding step is equal to 1. Obviously, the preset sliding step is set to 1 to extract a more comprehensive and detailed subsequence signal, so the preset sliding step is generally set to 1. This embodiment is described exemplarily with the length of the preset sliding window being w and the preset sliding step being 1.
[0042] S120, performing discrete Fourier transform on each subsequence signal to obtain a corresponding frequency domain sequence, and determining the rank of each element value according to the order of magnitude of each element value in the frequency domain sequence.
[0043] Exemplarily, the discrete Fourier transform formula is used to convert the subsequence signal from the time domain to the frequency domain to obtain the frequency domain sequence corresponding to the subsequence signal. For example, the subsequence signal extracted from the vibration signal includes {X1,...,X i ,...,X n-w+1}, correspondingly, the subsequence signal X i The corresponding transformed frequency domain sequence is F i , then the multiple subsequence signals are converted into multiple frequency domain sequences including {D1,...,F i ,...,F n-w+1}.
[0044] Frequency domain sequence F i It can be expressed as is the frequency domain sequence F i The kth element value in is also the subsequence signal X i The kth sampling point in the frequency domain, m is the frequency domain sequence F i The total number of element values in is the length of the frequency domain sequence. It should be noted that any two subsequence signals are subsequence signals of a vibration signal that is continuous in time, so any two subsequence signals have the same sampling frequency and number of sampling points, so the length m of any two subsequence signals after Fourier transform is also the same.
[0045] After performing discrete Fourier transform on the subsequence signal to obtain the corresponding frequency domain sequence, the rank number of each element value is determined according to the sequence number when each element value in the frequency domain sequence is sorted from large to small or from small to large, so as to subsequently calculate the correlation between the frequency domain sequence and the frequency domain sequence of the first subsequence signal based on the rank number of the element value. Figure 2 1 is a flowchart of determining the rank of an element value in a frequency domain sequence provided by an embodiment of the present application. Figure 2 As shown, the step of determining the rank of the element value in the frequency domain sequence specifically includes S1201-S1203:
[0046] S1201. Sort each element value in the frequency domain sequence in ascending order to obtain a sequence number for each element value.
[0047] Exemplarily, the frequency domain sequence F i The values of each element in Sort in ascending order, and determine the corresponding sequence number according to the order of each element value in the sort, such as When it ranks fifth, its sequence number is determined to be 5. It should be noted that for multiple identical element values, their sequence numbers are different. For example, and Same, if is the second smallest element value, then, and The serial numbers are 2, 3 and 4 respectively.
[0048] S1202: When the element value is different from other element values of the corresponding frequency domain sequence, determine the sequence number of the element value as the rank number of the element value.
[0049] For example, if If the value is different from the other elements, The sequence number of is determined as the corresponding rank number. For example, The sequence number is 5 and the other element values are the same as Different, then confirm The rank number is 5.
[0050] S1203: When the element value is identical to some element values of the corresponding frequency domain sequence, determine the element value and the mean of the sequence numbers of the identical element values as the rank number of the element value.
[0051] For example, if If the value is the same as some of the elements, The average value of the sequence numbers corresponding to the same element value is determined as The rank number of the corresponding element value. For example, and same, and The serial numbers are 2, 3 and 4 respectively. and The rank numbers of all elements are 3. In this embodiment, the rank number of each element value is determined by the sequence number of each element value in the frequency domain sequence, so as to accurately represent the size order of each element value in the frequency domain sequence by the rank number.
[0052] In order to more clearly understand the process of determining the rank of each element value in the frequency domain sequence, this embodiment uses Figure 3 The sequence number and rank number of each element value in the frequency domain sequence shown in FIG. Figure 3 As shown, the original sequence of element values in the frequency domain sequence is {3.1,,5.8,0,1.2,1.2,1.2,4.0}. After sorting the element values from small to large, the sequence is {0,1.2,1.2,1.2,3.1,4.0,5.8}. The sequence numbers of the sequence {0,1.2,1.2,1.2,3.1,4.0,5.8} are {1,2,3,4,5,6,7}. Based on the element values and the corresponding sequence numbers, it can be determined that the rank numbers of the sequence {0,1.2,1.2,1.2,3.1,4.0,5.8} are {1,3,3,3,5,6,7}. Then, the rank numbers of the original sequence {3.1,5.8,0,1.2,1.2,1.2,4.0} are {5,7,1,3,3,3,6}. It can be understood that the rank of the element value can accurately represent its size order in the frequency domain sequence, and the correlation between any frequency domain sequence and the frequency domain sequence corresponding to the first subsequence signal in the frequency component can be calculated based on the rank.
[0053] S130. Determine a similarity coefficient between each frequency domain sequence and the frequency domain sequence corresponding to the first subsequence signal according to the rank of each element value in each frequency domain sequence and the rank of each element value in the frequency domain sequence corresponding to the first subsequence signal.
[0054] Exemplarily, the first subsequence signal is the subsequence signal extracted at the starting point of the vibration signal through a preset sliding window, that is, the first signal value of the first subsequence signal is the first signal value of the vibration signal. The first subsequence signal can be represented by X1, and accordingly, the frequency domain sequence of the first subsequence signal can be represented by F1.
[0055] It can be understood that after the subsequence signal undergoes discrete Fourier transform, the size order of the element values in its frequency domain sequence represents the size order of the corresponding amplitudes of each frequency component. If the size order of two frequency domain sequences is similar, then the frequency components of the two frequency domain sequences are also similar, indicating that the working conditions represented by the two frequency domain sequences are similar. When the frequency domain sequence of each subsequence signal is similar to the frequency domain sequence of the first subsequence signal, the working conditions represented by each subsequence signal of the vibration signal are similar, and the working conditions of the vibration signal do not change with time; when the frequency domain sequence of any subsequence signal is not similar to the frequency domain sequence of the first subsequence signal, the working conditions represented by some subsequence signals in the vibration signal are not similar, and the working conditions of the vibration signal change with time, indicating that the vibration signal will have an error in the DFT result or the fault frequency component cannot be obtained through the spectrum diagram, thereby determining that the vibration signal is an abnormal vibration signal.
[0056] From the above content, we can know that the rank of each element value in the frequency domain sequence can be used to represent the order of the corresponding element values, so the frequency domain sequence F i The rank of each element value in and the rank of each element value in the frequency domain sequence F1 are used to calculate the frequency domain sequence F i The similarity coefficient with the frequency domain sequence F1. Among them, the similarity coefficient can be used to characterize the frequency domain sequence F i The similarity with the frequency domain sequence F1, the larger the similarity coefficient, the higher the frequency domain sequence F i The greater the correlation with the frequency domain sequence F1. Exemplarily, the similarity coefficient between the frequency domain sequence and the frequency domain sequence of the first subsequence signal can be determined based on the rank difference of the element value at the same position in each frequency domain sequence and the frequency domain sequence of the first subsequence signal. i Expressed as The frequency domain sequence F1 is expressed as Then the rank difference of the element value at the kth position is and They are and It can be understood that the element values of the two frequency domain sequences at the same position represent the amplitudes of the two frequency domain sequences on the same frequency component. Then the rank difference of the element values of the two frequency domain sequences at the same position can represent the order difference of the amplitudes of the two frequency domain sequences on the same frequency component. Therefore, the larger the rank difference, the smaller the similarity of the frequency components of the two frequency domain sequences. The rank difference of the element values of the frequency domain sequence and the frequency domain sequence of the first subsequence signal at the same position can be used to accurately calculate the similarity coefficient between the frequency domain sequence and the frequency domain sequence of the first subsequence signal.
[0057] Optionally, the similarity coefficient between the frequency domain sequence and the frequency domain sequence corresponding to the first subsequence signal may be determined based on a similarity coefficient calculation formula and the rank number of each element value in the frequency domain sequence and the rank number of each element value in the frequency domain sequence corresponding to the first subsequence signal; wherein the similarity coefficient calculation formula is:
[0058]
[0059] Among them, C is the similarity coefficient, p k is the difference between the rank of the kth element value in the frequency domain sequence and the rank of the kth element value in the frequency domain sequence corresponding to the first subsequence signal; m is the number of elements in the frequency domain sequence. For example, when calculating the frequency domain sequence F i When the similarity coefficient C between the frequency domain sequence F1 is calculated, the frequency domain sequence F i The rank difference of the element value at the same position as the frequency domain sequence F1 is substituted into the above similarity coefficient calculation formula to obtain the frequency domain sequence F i The similarity coefficient C between the frequency domain sequence F1 and the frequency domain sequence F2 is shown in the formula for calculating the similarity coefficient. The larger the rank difference is, the smaller the similarity coefficient is. i The greater the difference between the working conditions of the frequency domain sequence F1, the greater the change in the working conditions of the vibration signal. When the rank difference is smaller, the similarity coefficient is larger, then the frequency domain sequence F i The smaller the difference between the operating conditions and the frequency domain sequence F1, the smaller the operating condition change of the vibration signal.
[0060] S140: When the similarity coefficient corresponding to any frequency domain sequence is less than a preset similarity threshold, determine that the vibration signal is an abnormal vibration signal.
[0061] The similarity threshold is the minimum similarity coefficient between the frequency domain sequence of the corresponding extracted subsequence signal and the frequency domain sequence of the first subsequence signal when the equipment working condition of the vibration signal does not change with time. It can be understood that when the frequency domain sequence F i When the similarity coefficient with the frequency domain sequence F1 is less than the similarity threshold, it indicates that the equipment operating condition of the vibration signal changes over time, thereby determining that the vibration signal is an abnormal vibration signal.
[0062] Optionally, the similarity threshold can be generally set to 0.9, that is, in the frequency domain sequence F i When the similarity coefficient with the frequency domain sequence F1 is less than 0.9, the vibration signal is determined to be an abnormal vibration signal.
[0063] Furthermore, when the frequency domain sequence F iWhen the similarity coefficient with the frequency domain sequence F1 is greater than the similarity threshold, it only proves that the equipment condition of the vibration signal does not change with time, but the vibration signal must also meet the statistical characteristics that do not change with time to avoid DFT result errors or the problem of being unable to obtain the fault frequency component through the spectrum diagram. In this regard, it is also possible to determine whether it meets the condition that the statistical characteristics do not change with time based on each subsequence signal of the vibration signal, so that when it is determined that the vibration signal meets the condition that the statistical characteristics do not change with time and the condition that the equipment condition does not change with time, the vibration signal is determined to be a normal vibration signal, so that the vibration signal can be used to perform fault analysis on the equipment.
[0064] In one embodiment, whether the vibration signal satisfies the condition that the statistical characteristics do not change over time may be determined based on the signal mean of each subsequence signal and the signal mean of the vibration signal. Figure 4 1 is a flow chart of judging whether a vibration signal is an abnormal vibration signal based on the signal mean value provided by an embodiment of the present application. Figure 4 As shown, the step of judging whether the vibration signal is an abnormal vibration signal based on the signal mean value specifically includes S1501-S1503:
[0065] S1501. Calculate a first signal mean value of each subsequence signal and a second signal mean value of the vibration signal.
[0066] Exemplarily, each signal value in the subsequence signal is accumulated and then the accumulated value is divided by the length of the subsequence signal to obtain the first signal mean of the subsequence signal. Each signal value in the vibration signal is accumulated and then the accumulated value is divided by the length of the vibration signal to obtain the second signal mean of the vibration signal.
[0067] S1502: Determine a mean coefficient corresponding to each subsequence signal according to the first signal mean and the second signal mean.
[0068] Exemplarily, the cumulative value of the first signal mean and the second signal mean is divided by twice the second signal mean to obtain the mean coefficient of the subsequence signal corresponding to the first signal mean. For example, the calculation formula of the mean coefficient is:
[0069]
[0070] in, is the ith subsequence signal X i The mean coefficient of is the second signal mean, is the ith subsequence signal X i The first signal mean.
[0071] S1503: When the mean coefficient corresponding to any subsequence signal is not within the first preset coefficient range, determine that the vibration signal is an abnormal vibration signal.
[0072] It can be understood that the statistical characteristics of the vibration signal do not change with time, that is, the vibration signal has a certain stability, and the subsequence signal is a signal subset of the vibration signal in the corresponding time sliding window. When the vibration signal has a certain stability, the signal mean of the vibration signal is slightly different from the signal mean of each subsequence signal. The mean coefficient of the subsequence signal can be used to characterize the similarity between the signal mean of the vibration signal and the corresponding subsequence signal. The closer the mean coefficient is to 1, the higher the similarity between the signal mean of the vibration signal and the corresponding subsequence signal. The first preset coefficient range is a coefficient range with a certain deviation floating up and down around 1, such as the first preset coefficient range can be set to (0.95, 1.05). Correspondingly, when the mean coefficient corresponding to the subsequence signal is not in the first preset coefficient range, it indicates that the similarity between the signal mean of the subsequence signal and the vibration signal is low, thereby determining that the statistical characteristics of the vibration signal change with time, and then determining that the vibration signal is an abnormal vibration signal.
[0073] Furthermore, when the corresponding mean coefficients of each subsequence signal are within the first preset coefficient range, it is not possible to directly determine that the vibration signal meets the condition that the statistical characteristics do not change over time, because the signal value fluctuation of the vibration signal cannot be fully presented by the mean. In this regard, it is possible to determine whether the vibration signal meets the condition that the statistical characteristics do not change over time based on the signal variances of each subsequence signal and the signal variance of the vibration signal. Optionally, Figure 5 1 is a flow chart of judging whether a vibration signal is an abnormal vibration signal based on signal variance provided by an embodiment of the present application. Figure 5 As shown, the step of judging whether the vibration signal is an abnormal vibration signal based on the signal variance specifically includes S1601-S1603:
[0074] 1601. Calculate a first signal variance of each subsequence signal and a second signal variance of the vibration signal.
[0075] Exemplarily, the difference between each signal value in the subsequence signal and the corresponding first signal mean is calculated, the squares of the differences are accumulated, and the accumulated value is divided by the length of the subsequence signal to obtain the first signal variance of the subsequence signal. The difference between each signal value in the vibration signal and the second signal mean is calculated, the squares of the differences are accumulated, and the accumulated value is divided by the length of the vibration signal to obtain the second signal variance of the vibration signal.
[0076] 1602. Determine a variance coefficient corresponding to each subsequence signal according to the first signal variance and the second signal variance.
[0077] Exemplarily, the cumulative value of the first signal variance and the second signal variance is divided by twice the second signal variance to obtain the variance coefficient of the subsequence signal corresponding to the first signal variance. For example, the calculation formula of the variance coefficient is:
[0078]
[0079] in, is the ith subsequence signal X i The coefficient of variance, is the second signal variance, is the ith subsequence signal X i The first signal variance is .
[0080] 1603. When the variance coefficient corresponding to any subsequence signal is not within a second preset coefficient range, determine that the vibration signal is an abnormal vibration signal.
[0081] It can be understood that the statistical characteristics of the vibration signal do not change with time, that is, the vibration signal has a certain stability, and the subsequence signal is a signal subset of the vibration signal in the corresponding time sliding window. When the vibration signal has a certain stability, the signal variance of the vibration signal is slightly different from the signal variance of each subsequence signal. The variance coefficient of the subsequence signal can be used to characterize the similarity of the signal variance of the vibration signal and the corresponding subsequence signal. The closer the variance coefficient is to 1, the higher the similarity of the signal variance of the vibration signal and the corresponding subsequence signal. The second preset coefficient range is a coefficient range with a certain deviation floating up and down around 1, such as the second preset coefficient range can be set to (0.95, 1.05). Correspondingly, when the variance coefficient corresponding to the subsequence signal is not in the second preset coefficient range, it indicates that the similarity of the signal variance of the subsequence signal and the vibration signal is low, thereby determining that the statistical characteristics of the vibration signal change over time, and then determining that the vibration signal is an abnormal vibration signal.
[0082] Further, when the mean coefficient corresponding to each subsequence signal is within the first preset coefficient range, the variance coefficient corresponding to each subsequence signal is within the second preset coefficient range, and the similarity coefficient corresponding to each frequency domain sequence is greater than the preset similarity threshold, the vibration signal is determined to be a normal vibration signal. Exemplarily, when the similarity coefficient corresponding to each frequency domain sequence is greater than the preset similarity threshold, it is confirmed that the equipment operating condition of the vibration signal does not change with time, and when the mean coefficient corresponding to each subsequence signal is within the first preset coefficient range and the variance coefficient corresponding to each subsequence signal is within the second preset coefficient range, it is confirmed that the statistical characteristics of the vibration signal do not change with time. At this time, the vibration signal can be determined to be a normal vibration signal, so that the vibration signal pair is retained for subsequent use in fault analysis of the equipment.
[0083] In one embodiment, it is possible to determine whether each subsequence signal satisfies the corresponding first preset coefficient range, second preset coefficient range and similarity threshold based on the mean coefficient, variance coefficient and similarity coefficient of the subsequence signal and the corresponding frequency domain sequence, so that when it is determined that a subsequence signal does not meet one of the conditions, the vibration signal is determined to be an abnormal vibration signal and the vibration signal is deleted. Exemplary, Figure 6 Schematic diagram of the process of retaining or deleting a vibration signal provided by an embodiment of the present application. Figure 6 As shown, it is determined whether the next subsequence signal can be extracted from the vibration signal. If it can, the extracted next subsequence signal is used as the subsequence signal to be detected. The mean coefficient of the subsequence signal to be detected is determined according to the first signal mean of the subsequence signal to be detected and the second signal mean of the vibration signal, and it is determined whether the mean coefficient of the subsequence signal to be detected is within the first preset coefficient range. If the mean coefficient of the subsequence signal to be detected is not within the first preset coefficient range, the vibration signal is deleted. If the mean coefficient of the subsequence signal to be detected is within the first preset coefficient range, the variance coefficient of the subsequence signal to be detected is determined according to the first signal variance of the subsequence signal to be detected and the second signal variance of the vibration signal. If the variance coefficient of the subsequence signal to be detected is not within the second preset coefficient range, the vibration signal is deleted. If the variance coefficient of the subsequence signal to be detected is within the second preset coefficient range, the subsequence signal to be detected is subjected to discrete Fourier transform to obtain the frequency domain sequence to be detected, and it is determined whether the similarity coefficient of the frequency domain sequence to be detected is less than the preset similarity threshold. If the similarity coefficient of the frequency domain sequence to be detected is less than the preset similarity threshold, the vibration signal is deleted. If the similarity coefficient of the frequency domain sequence to be detected is greater than or equal to the preset similarity threshold, the vibration signal is retained and it is determined whether the vibration signal can be extracted into the next subsequence signal. If the vibration signal cannot be extracted into the next subsequence signal, the vibration signal is retained so that the equipment can be subsequently analyzed for faults based on the vibration signal.
[0084] In summary, the abnormal vibration signal detection method provided in the embodiment of the present application obtains a vibration signal to be detected, and extracts multiple subsequence signals from the vibration signal based on a preset sliding window and a preset sliding step size; performs a discrete Fourier transform on each subsequence signal to obtain a corresponding frequency domain sequence, and determines the rank number of each element value according to the order of the size of each element value in the frequency domain sequence; determines the similarity coefficient between each frequency domain sequence and the frequency domain sequence corresponding to the first subsequence signal according to the rank number of each element value in each frequency domain sequence and the rank number of each element value in the frequency domain sequence corresponding to the first subsequence signal; when the similarity coefficient corresponding to any frequency domain sequence is less than a preset similarity threshold, the vibration signal is determined to be an abnormal vibration signal. Through the above technical means, the frequency domain sequence corresponding to each subsequence signal extracted from the vibration signal is obtained by performing discrete Fourier transform. The rank of the element value in the frequency domain sequence represents the order of the amplitude corresponding to each frequency component. Therefore, the similarity of each frequency component can be calculated based on the rank of the element value at the same position in the frequency domain sequence and the frequency domain sequence of the first subsequence signal, so as to obtain the similarity coefficient between the frequency domain sequence and the frequency domain sequence of the first subsequence signal. When the similarity coefficient is greater than or equal to the similarity threshold, the working conditions of the frequency domain sequence and the frequency domain sequence of the first subsequence signal are similar. Therefore, when the similarity coefficient between any frequency domain sequence and the frequency domain sequence of the first subsequence signal is less than the similarity threshold, it can be determined that the working condition of the vibration signal changes with time, and then the vibration signal is determined to be an abnormal vibration signal, so that the abnormal vibration signal can be accurately detected without using an abnormal detection model, avoiding the problem of low detection accuracy of abnormal vibration signals caused by poor model training effect, thereby improving the accuracy and reliability of subsequent fault analysis. .
[0085] Based on the above embodiments, Figure 7 This is a schematic diagram of the structure of an abnormal vibration signal detection device provided in an embodiment of the present application. Figure 7 The abnormal vibration signal detection device provided in this embodiment specifically includes: a subsequence signal extraction module 21, a rank determination module 22, a correlation determination module 23 and a first abnormal signal detection module 24.
[0086] The subsequence signal extraction module 21 is configured to obtain a vibration signal to be detected, and extract a plurality of subsequence signals from the vibration signal based on a preset sliding window and a preset sliding step size;
[0087] The rank determination module 22 is configured to perform a discrete Fourier transform on each subsequence signal to obtain a corresponding frequency domain sequence, and determine the rank of each element value according to the order of magnitude of each element value in the frequency domain sequence;
[0088] A correlation determination module 23 is configured to determine a similarity coefficient between each frequency domain sequence and the frequency domain sequence corresponding to the first subsequence signal according to the rank number of each element value in each frequency domain sequence and the rank number of each element value in the frequency domain sequence corresponding to the first subsequence signal;
[0089] The first abnormal signal detection module 24 is configured to determine that the vibration signal is an abnormal vibration signal when the similarity coefficient corresponding to any frequency domain sequence is less than a preset similarity threshold.
[0090] On the basis of the above embodiment, the subsequence signal extraction module 21 includes: a window sliding unit, configured to determine the starting point of the vibration signal as the starting point of a preset sliding window, and slide the preset sliding window based on a preset sliding step; a subsequence signal extraction unit, configured to extract the signal value of the vibration signal within the preset sliding window as a subsequence signal each time the sliding is performed until the preset sliding window reaches the end point of the vibration signal.
[0091] On the basis of the above embodiment, the rank number determination module 22 includes: a sequence number determination unit, configured to sort the element values in the frequency domain sequence in ascending order to obtain the sequence number of each element value; a first rank number determination unit, configured to determine the sequence number of the element value as the rank number of the element value when the element value is different from the remaining element values of the corresponding frequency domain sequence; a second rank number determination unit, configured to determine the element value and the average of the sequence numbers of the corresponding identical element values as the rank number of the element value when the element value is the same as some element values of the corresponding frequency domain sequence.
[0092] On the basis of the above embodiment, the correlation determination module 23 includes: a correlation determination unit, configured to determine the similarity coefficient between the frequency domain sequence and the frequency domain sequence corresponding to the first subsequence signal based on the similarity coefficient calculation formula and the rank number of each element value in the frequency domain sequence and the rank number of each element value in the frequency domain sequence corresponding to the first subsequence signal; wherein the similarity coefficient calculation formula is:
[0093]
[0094] Among them, C is the similarity coefficient, p k is the difference between the rank number of the kth element value in the frequency domain sequence and the rank number of the kth element value in the frequency domain sequence corresponding to the first subsequence signal; m is the number of elements in the frequency domain sequence.
[0095] On the basis of the above embodiment, the abnormal vibration signal detection device includes a second abnormal signal detection module, and the second abnormal signal detection module includes: a signal mean determination unit, configured to calculate the first signal mean of each subsequence signal and the second signal mean of the vibration signal after extracting multiple subsequence signals from the vibration signal based on a preset sliding window and a sliding step; a mean coefficient determination unit, configured to determine the mean coefficient corresponding to each subsequence signal according to the first signal mean and the second signal mean; the second abnormal signal detection unit, configured to determine that the vibration signal is an abnormal vibration signal when the mean coefficient corresponding to any subsequence signal is not within a first preset coefficient range.
[0096] On the basis of the above embodiment, the abnormal vibration signal detection device includes a third abnormal signal detection module, and the third abnormal signal detection module includes: a signal variance determination unit, configured to calculate the first signal variance of each subsequence signal and the second signal variance of the vibration signal after extracting multiple subsequence signals from the vibration signal based on a preset sliding window and a sliding step; a variance coefficient determination unit, configured to determine the variance coefficient corresponding to each subsequence signal according to the first signal variance and the second signal variance; the third abnormal signal detection unit, configured to determine that the vibration signal is an abnormal vibration signal when the variance coefficient corresponding to any subsequence signal is not within a second preset coefficient range.
[0097] On the basis of the above embodiment, the abnormal vibration signal detection device also includes: a normal vibration signal determination module, which is configured to determine that the vibration signal is a normal vibration signal when the mean coefficient corresponding to each subsequence signal is within a first preset coefficient range, the variance coefficient corresponding to each subsequence signal is within a second preset coefficient range, and the similarity coefficient corresponding to each frequency domain sequence is greater than a preset similarity threshold.
[0098] As described above, the abnormal vibration signal detection device provided in the embodiment of the present application obtains a vibration signal to be detected, and extracts multiple subsequence signals from the vibration signal based on a preset sliding window and a preset sliding step size; performs a discrete Fourier transform on each subsequence signal to obtain a corresponding frequency domain sequence, and determines the rank number of each element value according to the order of the size of each element value in the frequency domain sequence; determines the similarity coefficient between each frequency domain sequence and the frequency domain sequence corresponding to the first subsequence signal according to the rank number of each element value in each frequency domain sequence and the rank number of each element value in the frequency domain sequence corresponding to the first subsequence signal; when the similarity coefficient corresponding to any frequency domain sequence is less than a preset similarity threshold, the vibration signal is determined to be an abnormal vibration signal. Through the above technical means, discrete Fourier transform is performed on each subsequence signal extracted corresponding to the vibration signal to obtain the frequency domain sequence corresponding to each subsequence signal. The rank of the element value in the frequency domain sequence represents the order of the magnitude of the amplitude corresponding to each frequency component. Therefore, the similarity of each frequency component can be calculated based on the rank of the element value at the same position in the frequency domain sequence and the frequency domain sequence of the first subsequence signal, so as to obtain the similarity coefficient between the frequency domain sequence and the frequency domain sequence of the first subsequence signal. When the similarity coefficient is greater than or equal to the similarity threshold, the operating conditions of the frequency domain sequence and the frequency domain sequence of the first subsequence signal are similar. Therefore, when the similarity coefficient between any frequency domain sequence and the frequency domain sequence of the first subsequence signal is less than the similarity threshold, it can be determined that the operating condition of the vibration signal changes with time, and then the vibration signal is determined to be an abnormal vibration signal, so that the abnormal vibration signal can be accurately detected without using an abnormal detection model, and the problem of low detection accuracy of abnormal vibration signals caused by poor model training effect is avoided, thereby improving the accuracy and reliability of subsequent fault analysis.
[0099] The abnormal vibration signal detection device provided in the embodiment of the present application can be used to execute the abnormal vibration signal detection method provided in the above embodiment, and has corresponding functions and beneficial effects.
[0100] Figure 8 is a schematic diagram of the structure of an abnormal vibration signal detection device provided in an embodiment of the present application, with reference to Figure 8 The abnormal vibration signal detection device includes: a processor 31, a memory 32, a communication device 33, an input device 34 and an output device 35. The number of processors 31 in the abnormal vibration signal detection device can be one or more, and the number of memories 32 in the abnormal vibration signal detection device can be one or more. The processor 31, memory 32, communication device 33, input device 34 and output device 35 of the abnormal vibration signal detection device can be connected through a bus or other methods.
[0101] The memory 32, as a computer-readable storage medium, can be used to store software programs, computer executable programs and modules, such as program instructions / modules corresponding to the abnormal vibration signal detection method of any embodiment of the present application (for example, the subsequence signal extraction module 21, the rank determination module 22, the correlation determination module 23 and the first abnormal signal detection module 24 included in the abnormal vibration signal detection device). The memory 32 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created according to the use of the device, etc. In addition, the memory 32 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some instances, the memory may further include a memory remotely arranged relative to the processor, and these remote memories may be connected to the device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network and a combination thereof.
[0102] The communication device 33 is used for data transmission.
[0103] The processor 31 executes various functional applications and data processing of the device by running the software programs, instructions and modules stored in the memory 32, that is, implements the above-mentioned abnormal vibration signal detection method.
[0104] The input device 34 may be used to receive input digital or character information and generate key signal input related to user settings and function control of the device. The output device 35 may include a display device such as a display screen.
[0105] The abnormal vibration signal detection device provided above can be used to execute the abnormal vibration signal detection method provided in the above embodiment, and has corresponding functions and beneficial effects.
[0106] An embodiment of the present application also provides a storage medium containing computer executable instructions. When the computer executable instructions are executed by a computer processor, they are used to execute an abnormal vibration signal detection method. The abnormal vibration signal detection method includes: obtaining a vibration signal to be detected, and extracting multiple subsequence signals from the vibration signal based on a preset sliding window and a preset sliding step size; performing a discrete Fourier transform on each subsequence signal to obtain a corresponding frequency domain sequence, and determining the rank number of each element value according to the order of the size of each element value in the frequency domain sequence; determining a similarity coefficient between each frequency domain sequence and a frequency domain sequence corresponding to the first subsequence signal according to the rank number of each element value in each frequency domain sequence and the rank number of each element value in the frequency domain sequence corresponding to the first subsequence signal; when the similarity coefficient corresponding to any frequency domain sequence is less than a preset similarity threshold, determining that the vibration signal is an abnormal vibration signal.
[0107] Storage medium - any of various types of memory devices or storage devices. The term "storage medium" is intended to include: installation media, such as CD-ROM, floppy disk or tape device; computer system memory or random access memory, such as DRAM, DDR RAM, SRAM, EDO RAM, Rambus RAM, etc.; non-volatile memory, such as flash memory, magnetic media (such as hard disk or optical storage); registers or other similar types of memory elements, etc. Storage media may also include other types of memory or combinations thereof. In addition, the storage medium may be located in the first computer system in which the program is executed, or may be located in a different second computer system, which is connected to the first computer system via a network (such as the Internet). The second computer system can provide program instructions to the first computer for execution. The term "storage medium" may include two or more storage media residing in different locations (for example, in different computer systems connected by a network). The storage medium may store program instructions (for example, embodied as a computer program) that can be executed by one or more processors.
[0108] Of course, the storage medium containing computer executable instructions provided in an embodiment of the present application, whose computer executable instructions are not limited to the above abnormal vibration signal detection method, can also execute related operations in the abnormal vibration signal detection method provided in any embodiment of the present application.
[0109] The abnormal vibration signal detection device, abnormal vibration signal detection system, storage medium and abnormal vibration signal detection equipment provided in the above embodiments can execute the abnormal vibration signal detection method provided in any embodiment of the present application. For technical details not described in detail in the above embodiments, please refer to the abnormal vibration signal detection method provided in any embodiment of the present application.
[0110] The above are only preferred embodiments of the present application and the technical principles used. The present application is not limited to the specific embodiments herein, and various obvious changes, readjustments and substitutions that can be made by those skilled in the art will not deviate from the protection scope of the present application. Therefore, although the present application is described in more detail through the above embodiments, the present application is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present application, and the scope of the present application is determined by the scope of the claims.
Claims
1. A method for detecting abnormal vibration signals, characterized in that: include: Acquire a vibration signal to be detected, and extract a plurality of subsequence signals from the vibration signal based on a preset sliding window and a preset sliding step size; Performing discrete Fourier transform on each of the subsequence signals to obtain a corresponding frequency domain sequence, and determining the rank of each element value according to the order of magnitude of each element value in the frequency domain sequence; Determine, according to the rank number of each element value in each of the frequency domain sequences and the rank number of each element value in the frequency domain sequence corresponding to the first subsequence signal, a similarity coefficient between each of the frequency domain sequences and the frequency domain sequence corresponding to the first subsequence signal; When the similarity coefficient corresponding to any of the frequency domain sequences is less than a preset similarity threshold, the vibration signal is determined to be an abnormal vibration signal.
2. The abnormal vibration signal detection method according to claim 1, characterized in that: The step of extracting a plurality of subsequence signals from the vibration signal based on a preset sliding window and a preset sliding step size comprises: Determine the starting point of the vibration signal as the starting point of a preset sliding window, and slide the preset sliding window based on a preset sliding step length; Each time the sliding is performed, a signal value of the vibration signal within the preset sliding window is extracted as a subsequence signal until the preset sliding window reaches the end point of the vibration signal.
3. The abnormal vibration signal detection method according to claim 1, characterized in that: The determining the rank of each element value according to the order of magnitude of each element value in the frequency domain sequence includes: Sort the element values in the frequency domain sequence in ascending order to obtain a sequence number of each element value; In a case where the element value is different from other element values of the corresponding frequency domain sequence, determining the sequence number of the element value as the rank number of the element value; In the case that the element value is identical to some element values of the corresponding frequency domain sequence, the element value and the average of the sequence numbers of the corresponding identical element values are determined as the rank number of the element value.
4. The abnormal vibration signal detection method according to claim 1, characterized in that: The determining, according to the rank number of each element value in each of the frequency domain sequences and the rank number of each element value in the frequency domain sequence corresponding to the first subsequence signal, a similarity coefficient between each of the frequency domain sequences and the frequency domain sequence corresponding to the first subsequence signal comprises: Based on the similarity coefficient calculation formula and the rank number of each element value in the frequency domain sequence and the rank number of each element value in the frequency domain sequence corresponding to the first subsequence signal, the similarity coefficient between the frequency domain sequence and the frequency domain sequence corresponding to the first subsequence signal is determined; wherein the similarity coefficient calculation formula is: Among them, C is the similarity coefficient, p k is the difference between the rank number of the kth element value in the frequency domain sequence and the rank number of the kth element value in the frequency domain sequence corresponding to the first subsequence signal; m is the number of elements in the frequency domain sequence.
5. The abnormal vibration signal detection method according to claim 1, characterized in that: After extracting a plurality of subsequence signals from the vibration signal based on the preset sliding window and the sliding step size, the method further includes: Calculating a first signal mean value of each of the subsequence signals and a second signal mean value of the vibration signal; Determine a mean coefficient corresponding to each of the subsequence signals according to the first signal mean and the second signal mean; When the mean coefficient corresponding to any of the subsequence signals is not within the first preset coefficient range, it is determined that the vibration signal is an abnormal vibration signal.
6. The abnormal vibration signal detection method according to claim 5, characterized in that: After extracting a plurality of subsequence signals from the vibration signal based on the preset sliding window and the sliding step size, the method further includes: Calculating a first signal variance of each of the subsequence signals and a second signal variance of the vibration signal; Determine a variance coefficient corresponding to each of the subsequence signals according to the first signal variance and the second signal variance; When the variance coefficient corresponding to any of the subsequence signals is not within the second preset coefficient range, it is determined that the vibration signal is an abnormal vibration signal.
7. The abnormal vibration signal detection method according to claim 6, further comprising: When the mean coefficient corresponding to each of the subsequence signals is within a first preset coefficient range, the variance coefficient corresponding to each of the subsequence signals is within a second preset coefficient range, and the similarity coefficient corresponding to each of the frequency domain sequences is greater than a preset similarity threshold, the vibration signal is determined to be a normal vibration signal.
8. An abnormal vibration signal detection device, characterized in that: include: A subsequence signal extraction module is configured to obtain a vibration signal to be detected, and extract a plurality of subsequence signals from the vibration signal based on a preset sliding window and a preset sliding step size; a rank determination module, configured to perform a discrete Fourier transform on each of the subsequence signals to obtain a corresponding frequency domain sequence, and determine the rank of each element value according to the order of magnitude of each element value in the frequency domain sequence; a correlation determination module, configured to determine a similarity coefficient between each of the frequency domain sequences and the frequency domain sequence corresponding to the first subsequence signal according to the rank number of each element value in each of the frequency domain sequences and the rank number of each element value in the frequency domain sequence corresponding to the first subsequence signal; The first abnormal signal detection module is configured to determine that the vibration signal is an abnormal vibration signal when a similarity coefficient corresponding to any of the frequency domain sequences is less than a preset similarity threshold.
9. An abnormal vibration signal detection device, characterized in that: include: one or more processors; A storage device stores one or more programs. When the one or more programs are executed by the one or more processors, the one or more processors implement the abnormal vibration signal detection method as described in any one of claims 1-7.
10. A storage medium containing computer executable instructions, characterized in that: The computer executable instructions are used to execute the abnormal vibration signal detection method as described in any one of claims 1-7 when executed by a computer processor.