Non-steady-state signal processing method, signal identification method and related equipment

Abstract

The present application relates to the technical field of non-steady-state signal processing, and discloses a non-steady-state signal processing method, a signal identification method and related equipment, including: obtaining multiple original vibration signals corresponding to multiple moments in a target time period; determining first characteristic data according to the change trend of the original vibration signal, and determining second characteristic data according to the energy amplitude of the original vibration signal; clustering the first characteristic data and the second characteristic data of multiple original vibration signals in the target time period to obtain a first category corresponding to the first characteristic data and a second category corresponding to the second characteristic data; screening out a first characteristic threshold according to the first category, and screening out a second characteristic threshold according to the second category, the first characteristic threshold and the second characteristic threshold are used to identify whether the vibration signal is an interference signal. The present application improves the generalization ability of characteristic values ​​corresponding to interference signals.

Description

Technical Field

[0001] The present application relates to the technical field of non-steady-state signal processing, and in particular to a non-steady-state signal processing method, a signal recognition method, a device, an electronic device, and a computer-readable storage medium. Background Art

[0002] Fault diagnosis technology is a technology used to detect, identify and predict mechanical equipment failures. With the development of industrial automation and intelligence, fault diagnosis technology plays an increasingly important role in ensuring production efficiency, reducing maintenance costs and preventing equipment failures.

[0003] Current fault diagnosis technology has been able to achieve real-time monitoring and intelligent analysis, thereby providing early warning of potential faults. Among them, interference signal recognition technology is a key component in the field of fault diagnosis, which specifically involves distinguishing and excluding non-fault-related interference signals from equipment sensor data. In actual industrial environments, sensors may be affected by various noises and interferences, which may be mistakenly identified as fault signals, resulting in unnecessary maintenance work or false alarms. Interference signal recognition technology can improve the accuracy and reliability of fault diagnosis systems, reduce false alarms and missed alarms, and ensure that only real fault signals are used to trigger diagnosis. With the application of machine learning and artificial intelligence technologies, fault diagnosis systems have become more intelligent and can learn and identify faults from large amounts of complex data.

[0004] In mechanical equipment with non-steady-state operation, such as equipment that is frequently started and stopped, the identification of interference signals is particularly important. The working state of such equipment varies greatly, and it is easy to generate a large number of interference signals. However, most methods currently remain in the stage of traditional mathematical analysis and processing, and traditional mathematical analysis methods often rely on prior knowledge of a certain type of interference signal analysis. For the analysis of a large number of different interference signals, especially in mechanical equipment with non-steady-state operation, traditional mathematical analysis methods are often not universal and have weak generalization capabilities. When there are certain specific interference signals, if the prior knowledge of such interference signals is not analyzed, it is often impossible to accurately identify them from mechanical equipment with non-steady-state operation through traditional mathematical analysis methods, so that such signals are misjudged as fault signals, causing false fault alarms. Summary of the invention

[0005] The embodiments of the present application provide a non-steady-state signal processing method, a signal recognition method, an apparatus, an electronic device, and a computer-readable storage medium, which can solve the technical problem that the current eigenvalue generalization ability for identifying interference signals is weak.

[0006] The present application provides a non-steady-state signal processing method, including:

[0007] Acquire multiple original vibration signals corresponding to multiple moments in a target period;

[0008] Determine first characteristic data corresponding to the original vibration signal according to a change trend of the original vibration signal, and determine second characteristic data corresponding to the original vibration signal according to an energy amplitude of the original vibration signal;

[0009] Clustering the first characteristic data and the second characteristic data of the plurality of original vibration signals within the target time period to obtain a first category corresponding to the first characteristic data and a second category corresponding to the second characteristic data;

[0010] According to the first category, a plurality of first feature data corresponding to first feature thresholds are screened out, and according to the second category, a plurality of second feature data corresponding to second feature thresholds are screened out. The first feature threshold and the second feature threshold are used to identify whether the vibration signal is an interference signal.

[0011] Furthermore, the original vibration signal includes a plurality of vibration data, and the first characteristic data corresponding to the original vibration signal is determined according to the change trend of the original vibration signal, including:

[0012] Filtering out the first vibration data and the second vibration data from the original vibration signal according to the order of each vibration data in the original vibration signal;

[0013] According to a variation trend between the first vibration data and the second vibration data, first characteristic data corresponding to the original vibration signal is determined.

[0014] Furthermore, the above-mentioned determining the second characteristic data corresponding to the original vibration signal according to the energy amplitude of the original vibration signal includes:

[0015] Acquire a first energy value of the original vibration signal within a first frequency range, and a second energy value of the original vibration signal within a second frequency range;

[0016] Second characteristic data corresponding to the original vibration signal is determined according to the first energy value and the second energy value.

[0017] Furthermore, the first feature thresholds corresponding to the plurality of first feature data are screened out according to the first category, including:

[0018] Sort the plurality of first feature data according to the first category to obtain a target feature sequence;

[0019] A first feature threshold is screened out from the target feature sequence according to a position of the first feature data in the target feature sequence and a first category corresponding to the first feature data.

[0020] Furthermore, the first category includes a first subcategory and a second subcategory, and the target feature sequence includes a first feature sequence corresponding to the first subcategory and a second feature sequence corresponding to the second subcategory;

[0021] Filtering a first feature threshold from the target feature sequence according to a position of the first feature data in the target feature sequence and a first category corresponding to the first feature data includes:

[0022] Determine the boundary position between the first feature sequence and the second feature sequence in the target feature sequence;

[0023] Screening out at least two candidate feature data from the plurality of first feature data according to the boundary position and the position of each first feature data in the target feature sequence;

[0024] A first feature threshold is obtained iteratively according to at least two candidate feature data.

[0025] Furthermore, the iterative process of obtaining the first feature threshold according to at least two candidate feature data includes:

[0026] Obtaining the loss value corresponding to the candidate feature data, and determining whether the loss value meets the preset loss condition;

[0027] If the loss value satisfies the loss condition, the candidate feature data with the smallest loss value is selected as the first feature threshold;

[0028] If the loss value does not meet the loss condition, the candidate feature data is iterated according to the loss value, the iterated candidate feature data is used as the candidate feature data, and the execution is returned to obtain the loss value corresponding to the candidate feature data until the loss value meets the loss condition.

[0029] Furthermore, if the loss value does not meet the loss condition, the candidate feature data is iterated according to the loss value, the iterated candidate feature data is used as the candidate feature data, and the loss value corresponding to the candidate feature data is obtained by returning to execute until the loss value meets the loss condition, including:

[0030] If the loss value does not meet the loss condition, determining whether the number of iterations meets the preset number of iterations, and if the number of iterations meets the preset number of iterations, selecting a first feature threshold from the candidate feature data after the iteration;

[0031] If the loss value does not meet the loss condition and the number of iterations does not meet the preset number of iterations, the candidate feature data is iterated according to the loss value, and the candidate feature data after the iteration is used as the candidate feature data. Return to execute to obtain the loss value corresponding to the candidate feature data until the loss value meets the loss condition or the number of iterations meets the preset number of iterations.

[0032] The present application also provides a signal recognition method, including:

[0033] Acquire a vibration signal to be identified;

[0034] Determine first characteristic data corresponding to the vibration signal to be identified according to a change trend of the vibration signal to be identified, and determine second characteristic data corresponding to the vibration signal to be identified according to an energy amplitude of the vibration signal to be identified;

[0035] Obtaining a preset first feature threshold and a second feature threshold, comparing the first feature data corresponding to the vibration signal to be identified and the first feature threshold to obtain a first comparison result, and comparing the second feature data corresponding to the vibration signal to be identified and the second feature threshold to obtain a second comparison result;

[0036] If the first comparison result and / or the second comparison result meets the preset comparison result, it is determined that the vibration signal to be identified is an interference signal.

[0037] Accordingly, an embodiment of the present application provides a non-steady-state signal processing device, comprising:

[0038] A first acquisition module is used to acquire multiple original vibration signals corresponding to multiple moments in a target time period;

[0039] A first confirmation module, used to determine first characteristic data corresponding to the original vibration signal according to a change trend of the original vibration signal, and to determine second characteristic data corresponding to the original vibration signal according to an energy amplitude of the original vibration signal;

[0040] A clustering module, used for clustering the first characteristic data and the second characteristic data of the plurality of original vibration signals within the target time period to obtain a first category corresponding to the first characteristic data and a second category corresponding to the second characteristic data;

[0041] The screening module is used to screen out first feature thresholds corresponding to multiple first feature data according to the first category, and to screen out second feature thresholds corresponding to multiple second feature data according to the second category. The first feature threshold and the second feature threshold are used to identify whether the vibration signal is an interference signal.

[0042] Accordingly, an embodiment of the present application provides a signal recognition device, including:

[0043] A second acquisition module is used to acquire a vibration signal to be identified;

[0044] A second confirmation module is used to determine the first characteristic data corresponding to the vibration signal to be identified according to the change trend of the vibration signal to be identified, and to determine the second characteristic data corresponding to the vibration signal to be identified according to the energy amplitude of the vibration signal to be identified;

[0045] A comparison module, used to obtain a preset first feature threshold and a second feature threshold, compare the first feature data corresponding to the vibration signal to be identified with the first feature threshold to obtain a first comparison result, and compare the second feature data corresponding to the vibration signal to be identified with the second feature threshold to obtain a second comparison result;

[0046] The third confirmation module is used to determine that the vibration signal to be identified is an interference signal if the first comparison result and / or the second comparison result meets the preset comparison result.

[0047] In addition, an embodiment of the present application also provides an electronic device, including a processor and a memory, wherein the memory stores a computer program, and the processor is used to run the computer program in the memory to implement the non-steady-state signal processing method provided by the embodiment of the present application and / or the signal recognition method provided by the embodiment of the present application.

[0048] In addition, an embodiment of the present application also provides a computer-readable storage medium, which stores a computer program. The computer program is suitable for loading by a processor to execute any non-steady-state signal processing method provided by the embodiment of the present application and / or the signal identification method provided by the embodiment of the present application.

[0049] In addition, an embodiment of the present application also provides a computer program product, including a computer program, which, when executed by a processor, implements any non-steady-state signal processing method and / or signal recognition method provided by an embodiment of the present application.

[0050] In an embodiment of the present application, by selecting multiple original vibration signals corresponding to multiple moments in a target time period, the stability of the data and the reliability of the signal processing results are improved. According to the change trend of the original vibration signal, the first feature data corresponding to the original vibration signal is determined, and according to the energy amplitude of the original vibration signal, the second feature data corresponding to the original vibration signal is determined. Since the first feature data and the second feature data are strongly correlated with the interference signal, the first feature data and the second feature data of the multiple original vibration signals in the target time period are clustered to obtain the first category corresponding to the first feature data and the second category corresponding to the second feature data, so that the feature threshold can be accurately determined, so that the interference signal can be accurately identified through the feature threshold. According to the first category, the first feature threshold corresponding to the multiple first feature data is screened out, and according to the second category, the second feature threshold corresponding to the multiple second feature data is screened out. Since the change trend and energy of the signal are the characteristics generally possessed by the signal, the feature threshold determined in the present application can be applied to different interference signals, which improves the generalization of the feature threshold corresponding to the interference signal. The first feature threshold and the second feature threshold can be used to identify whether the vibration signal is an interference signal, so that different interference signals can be accurately identified through the feature threshold. When interference signals are identified based on the characteristic threshold, accurate identification of interference signals of different categories can be achieved, further improving the accuracy of mechanical fault diagnosis, avoiding false alarms, reducing maintenance costs and avoiding production interruptions. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0052] Figure 1 is a flowchart of a non-steady-state signal processing method provided in an embodiment of the present application;

[0053] Figure 2 is a flow chart of a signal recognition method provided in an embodiment of the present application;

[0054] Figure 3 It is a schematic diagram of vibration acceleration and velocity data of a reduction box measuring point of a port equipment provided in an embodiment of the present application;

[0055] Figure 4 It is a vibration acceleration and velocity spectrum diagram of the port equipment reduction box provided in an embodiment of the present application;

[0056] Figure 5It is a schematic diagram of the absolute mean of the first-order difference of the vibration data of the reduction box of the port equipment provided in the embodiment of the present application;

[0057] Figure 6 It is a distribution diagram of the clustering results of the vibration data of the reducer of the port equipment provided by the embodiment of the present application in the first-order difference absolute mean, low-frequency energy proportion and effective value dimensions;

[0058] Figure 7 It is a distribution diagram of the clustering results of the vibration data of the reducer of the port equipment provided by the embodiment of the present application in the first-order difference absolute mean, low-frequency energy proportion and equipment speed dimension;

[0059] Figure 8 It is a distribution diagram of the absolute mean of the first-order difference and the equipment speed dimension after the vibration data of the reducer of the port equipment provided in the embodiment of the present application are clustered;

[0060] Fig. 9 is a structural schematic diagram of a non-steady-state signal processing device provided in an embodiment of the present application;

[0061] Fig.10 is a schematic diagram of the structure of a signal recognition device provided in an embodiment of the present application;

[0062] Fig.11 It is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0063] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of this application.

[0064] The embodiments of the present application provide a non-steady-state signal processing method, device, electronic device and computer-readable storage medium. The non-steady-state signal processing device can be integrated in an electronic device, which can be a server or a terminal.

[0065] Among them, the server can be an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, network acceleration services (Content Delivery Network, CDN), as well as big data and artificial intelligence platforms.

[0066] The terminal may be a smart phone, a tablet computer, a laptop computer, a desktop computer, a smart speaker, a smart watch, etc., but is not limited thereto. The terminal and the server may be directly or indirectly connected via wired or wireless communication, and this application does not limit this.

[0067] In addition, the term "plurality" in the embodiments of the present application refers to two or more than two. The terms "first" and "second" in the embodiments of the present application are used to distinguish descriptions and should not be understood as implying relative importance.

[0068] Currently, there is a method for identifying outliers in gearbox vibration signals using spatiotemporal correlation, which specifically uses the first channel to input the shaft sensor signal, the second channel to output the shaft sensor signal, and establishes a corresponding relationship between the two. The real-time data sequence of the two channel signals is obtained, and the data feature data in each window, such as the effective value, peak-to-peak value, mean and variance, is calculated using a sliding window. The difference distance between the data feature data in the front and back windows of the two channel signals is calculated. If the difference distance is greater than the set threshold, it is used as a candidate outlier data sequence. For the candidate outlier data sequence, the correlation coefficient between it and the feature data in the corresponding window of the associated channel is calculated. If the correlation coefficient is lower than the preset threshold, it is confirmed as the outlier data of the real fault, otherwise it is regarded as normal data.

[0069] The threshold of the correlation coefficient in this method is a threshold set manually based on experience, which will make it difficult for the analysis algorithm to adapt to data from different devices and different working conditions. In terms of parameter selection, the length of the sliding window, the step size, and the selection of the correlation threshold require a lot of experiments and tuning to adapt to different application scenarios. In addition, this type of method relies heavily on prior knowledge, and there is a large room for optimization in the generalization and adaptability of the algorithm. By observing the actual data, it can be found that the interference signal is mistakenly identified as a pulse signal. However, when the operating frequency of the equipment is accelerating or decelerating, it will cause the data to exceed the limit when it passes through resonance. This is a common phenomenon in the mechanical working process. This type of signal is often mistakenly identified as a fault signal, resulting in false alarms during fault detection.

[0070] Based on this, the embodiment of the present application proposes a non-steady-state signal processing method. Taking into account the drift of data, multiple original vibration signals corresponding to multiple moments in the target time period are selected to improve the stability of the data and the reliability of the signal processing results. According to the change trend of the original vibration signal, the first feature data corresponding to the original vibration signal is determined, and according to the energy amplitude of the original vibration signal, the second feature data corresponding to the original vibration signal is determined. Since the first feature data and the second feature data are strongly correlated with the interference signal, the first feature data and the second feature data of the multiple original vibration signals in the target time period are clustered to obtain the first category corresponding to the first feature data and the second category corresponding to the second feature data, so that the feature threshold can be accurately determined, so that the interference signal can be accurately identified by the feature threshold. According to the first category, the first feature threshold corresponding to the multiple first feature data is screened out, and according to the second category, the second feature threshold corresponding to the multiple second feature data is screened out. Since the change trend and energy of the signal are common characteristics of the signal, the feature threshold determined in the present application can be applied to different interference signals, which improves the generalization of the feature threshold corresponding to the interference signal. The first characteristic threshold and the second characteristic threshold can be used to identify whether the vibration signal is an interference signal, so that different interference signals can be accurately identified by the characteristic threshold. When the interference signal is identified based on the characteristic threshold, accurate identification of interference signals of different categories can be achieved, further improving the accuracy of mechanical fault diagnosis, avoiding false alarms, reducing maintenance costs and avoiding production interruptions.

[0071] It should be noted that the description order of the following embodiments is not intended to limit the preferred order of the embodiments.

[0072] See also Figure 1 , Figure 1 : is a flowchart of a non-steady-state signal processing method provided by an embodiment of the present application. The non-steady-state signal processing method may include:

[0073] S101, obtaining multiple original vibration signals corresponding to multiple moments in a target period;

[0074] S102, determining first characteristic data corresponding to the original vibration signal according to a change trend of the original vibration signal, and determining second characteristic data corresponding to the original vibration signal according to an energy amplitude of the original vibration signal;

[0075] S103, clustering the first characteristic data and the second characteristic data of the plurality of original vibration signals within the target time period to obtain a first category corresponding to the first characteristic data and a second category corresponding to the second characteristic data;

[0076] S104. Filter out first feature thresholds corresponding to a plurality of first feature data according to the first category, and filter out second feature thresholds corresponding to a plurality of second feature data according to the second category, wherein the first feature threshold and the second feature threshold are used to identify whether the vibration signal is an interference signal.

[0077] In this embodiment, the original vibration signal is a non-steady-state signal. The vibration signal of the target device is collected to obtain the original vibration signal. The target device may be a port device or other device that generates a non-steady-state signal. For example, in a port, since the equipment needs to be started and stopped frequently, a large number of non-steady-state signals are often generated. Multiple original vibration signals corresponding to multiple moments in the target time period are obtained.

[0078] For example, Figure 3 As shown, Figure 3 It is a schematic diagram of the vibration acceleration and velocity data of the measuring point of the reducer of the port equipment provided in the embodiment of the present application. The vibration acceleration and velocity data are the vibration data of the same measuring point at different times collected by the reducer of the port equipment in the working state. In order to intuitively and conveniently reveal the performance of data in different states, the vibration acceleration and vibration velocity data are displayed in pairs in the figure. The waveforms of the first five pairs of vibration acceleration and vibration velocity data from top to bottom in the figure are the performance under normal operating conditions, and the following are the data performance under acceleration or deceleration. Because in the acceleration or deceleration state, the total value of acceleration or velocity exceeds the limit specified by the standard, it will be mistakenly identified as fault signal data in some scenarios, causing a false fault alarm. This type of data is interference data.

[0079] In some embodiments, the target time period may be a recent historical time period. By acquiring multiple original vibration signals corresponding to multiple moments of the target device in the recent historical time period, the acquired original vibration signals can more accurately reflect the operating status of the target device, thereby improving the accuracy of the characteristic threshold determined based on the original vibration signal. In some embodiments, after acquiring the original vibration signal, the original vibration signal may be preprocessed, such as eliminating the mean trend of the original vibration signal to obtain a preprocessed vibration signal, thereby improving the signal processing efficiency and the accuracy of the characteristic threshold. The preprocessed vibration signal may be used as the original vibration signal in the embodiment of the present application. After preprocessing, the original vibration signal corresponding to each moment in the target time period may be expressed as: . Where t represents the time, X represents the original vibration signal, represents vibration data, and m represents the total number of vibration data in the original vibration signal at time t. Multiple vibration data at the same time constitute the original vibration signal.

[0080] According to the changing trend of the original vibration signal, the first characteristic data corresponding to the original vibration signal can be determined. The first characteristic data refers to data that can reflect the changing trend of the time series. For example, the absolute mean of the first-order difference (MAVFD), standard deviation, root mean square deviation, coefficient of variation or autocorrelation function, etc. Take the first characteristic data as the absolute mean of the first-order difference as an example. According to the changing trend of the original vibration signal, the absolute mean of the first-order difference of the original vibration signal is calculated, and the absolute mean of the first-order difference of the original vibration signal is the first characteristic data of the original vibration signal.

[0081] According to the energy amplitude of the original vibration signal, the second characteristic data corresponding to the original vibration signal can be determined. The second characteristic data refers to data that can reflect the proportion of signal energy within a preset frequency range. In some embodiments, the energy proportion of the original vibration signal within the preset frequency range can be extracted by performing Fourier transform on the original vibration signal, that is, the second characteristic data can be obtained.

[0082] like Figure 4 As shown, Figure 4 : is the vibration acceleration and velocity spectrum diagram of the port equipment reduction box provided by the embodiment of the present application. Figure 4 It can be found that the spectrum of the interference signal is concentrated in the low frequency segment, while the spectrum of the normal operation data is more reasonably distributed. Therefore, in some embodiments, the frequency range below 100 Hz can be determined as the preset frequency range, and the ratio of the sum of all energy amplitudes below 100 Hz to the sum of the total frequency domain energy amplitudes is used as the second characteristic data of the signal, that is, the low-frequency energy ratio.

[0083] like Figure 5 As shown, Figure 5 is the absolute mean of the first-order difference of the vibration data of the port equipment reducer provided in the embodiment of the present application. Figure 5 It can be found that after the interference signal is processed by the first-order difference absolute mean, its result is significantly smaller than the normal data. Therefore, in the time domain, the first-order difference absolute mean of the signal is used as the first feature data, and then the feature threshold of the first feature data can be used to accurately identify the interference signal.

[0084] Clustering the first feature data and the second feature data of multiple original vibration signals within the target time period can obtain the categories corresponding to each feature data. The category corresponding to the first feature data is the first category, and the category corresponding to the second feature data is the second category. Among them, the first category can be divided into at least two subcategories, and the second category can also be divided into at least two subcategories. The at least two subcategories can be used to distinguish normal data from abnormal data. For example, the first category includes category A (normal category) and category B (abnormal category), and the second category includes category C (normal category) and category D (abnormal category). Then, clustering the first feature data and the second feature data can determine that the category of each first feature data is A (normal category) or B (abnormal category), and determine that the category of each second feature data is C (normal category) or D (abnormal category). In some embodiments, the categories included in the first category and the second category can be the same category, that is, the first category is the same as the second category. For example, the first category includes category A and category B, and the second category also includes category A and category B.

[0085] Take the case where the first feature data is the absolute mean of the first-order difference and the second feature data is the low-frequency energy proportion of the original vibration signal as an example. Cluster analysis is performed on all data of the absolute mean of the first-order difference and the low-frequency energy proportion in the target period to obtain .in, Indicates cluster analysis of a given data set. Indicated in Time to The collection of all first-order difference absolute mean and low-frequency energy proportion data obtained by analyzing the original vibration signal within time. Time to The time is the target period. is the absolute mean of the first-order difference corresponding to time t And the proportion of low-frequency energy The cluster label after clustering. The cluster label is the category corresponding to each feature data. Figure 6 As shown, Figure 6 It is the distribution of the clustering result of the vibration data of the port equipment reducer provided in the embodiment of the present application in the first-order difference absolute mean, low-frequency energy proportion and effective value dimension. Figure 6 It can be seen that through clustering, two distinct clusters can be obtained, and the data points of the two clusters have basically distinguished normal vibration data from abnormal vibration data. Among them, the absolute mean of the first-order difference of normal vibration data is generally large, and the proportion of low-frequency energy is small.

[0086] like Figure 7 As shown, Figure 7The clustering result of the vibration data of the port equipment reducer provided in the embodiment of the present application is the distribution of the absolute mean of the first-order difference, the proportion of low-frequency energy, and the equipment speed dimension. Figure 7 It can be seen that the clustering results are strongly correlated with the rotational speed, that is, the data under high-speed operation conditions are all normal, which is consistent with the actual working conditions. Figure 7 There is a small amount of data in the low-speed working condition with a large absolute mean of the first-order difference and a small proportion of low-frequency energy. This part of the data belongs to the data of low-speed steady-state operation, and the overall vibration total value does not exceed the limit, which is normal data. Through the thresholds corresponding to the absolute mean of the first-order difference and the proportion of low-frequency energy, this type of data can be correctly identified as normal data.

[0087] Therefore, in some embodiments, the effective value and pulse characteristics of the vibration signal can also be obtained, and the effective value and pulse are used as the third characteristic data and the fourth characteristic data. The interference signal in the multiple vibration signals is identified by the characteristic threshold corresponding to at least one characteristic data among the first characteristic data, the second characteristic data, the third characteristic data and the fourth characteristic data. The method for determining the characteristic threshold corresponding to the third characteristic data and the fourth characteristic data is the same as the method for determining the characteristic threshold corresponding to the first characteristic data and the second characteristic data in the embodiment of the present application, and will not be repeated here.

[0088] A plurality of first feature data are screened according to the first category to obtain a first feature threshold, and a plurality of second feature data are screened according to the second category to obtain a second feature threshold. Whether the vibration signal is an interference signal can be determined by the first feature threshold and the second feature threshold. In some implementations, a vibration signal to be identified can be obtained, and the first feature data corresponding to the vibration signal to be identified can be determined according to the change trend of the vibration signal to be identified, and the second feature data corresponding to the vibration signal to be identified can be determined according to the energy amplitude of the vibration signal to be identified. The first feature data corresponding to the vibration signal to be identified is compared with the first feature threshold to obtain a first comparison result, and the second feature data corresponding to the vibration signal to be identified is compared with the second feature threshold to obtain a second comparison result. If the first comparison result and / or the second comparison result meet the preset comparison result, it is determined that the vibration signal to be identified is an interference signal.

[0089] like Figure 8 As shown, Figure 8 It is the distribution of the absolute mean of the first-order difference and the equipment speed dimension after the vibration data of the port equipment reducer provided in the embodiment of the present application is clustered. Figure 8 It can be seen that some data points are assigned to the normal cluster at low speed. Through further analysis, it can be found that this part is relatively stable under low speed conditions, and the proportion of low-frequency energy is not high. This shows that the classification of this type of cluster is correct and the effect is obvious.

[0090] Furthermore, it should be emphasized that the present application can achieve accurate acquisition of the threshold only through single-channel data (i.e., the vibration data of the vibration channel), and then realize adaptive recognition of interference signals through the threshold, and the hardware conditions for the on-site equipment are relatively loose.

[0091] In some implementations, multi-channel data may be added for fusion analysis to determine a corresponding feature threshold, and the interference signal may be further identified based on the feature threshold to improve the accuracy of interference signal identification.

[0092] This embodiment improves the stability of data and the reliability of signal processing results by selecting multiple original vibration signals corresponding to multiple moments in the target time period. According to the change trend of the original vibration signal, the first feature data corresponding to the original vibration signal is determined, and according to the energy amplitude of the original vibration signal, the second feature data corresponding to the original vibration signal is determined. Since the first feature data and the second feature data are strongly correlated with the interference signal, the first feature data and the second feature data of the multiple original vibration signals in the target time period are clustered to obtain the first category corresponding to the first feature data and the second category corresponding to the second feature data, so that the feature threshold can be accurately determined, so that the interference signal can be accurately identified through the feature threshold. According to the first category, the first feature threshold corresponding to the multiple first feature data is screened out, and according to the second category, the second feature threshold corresponding to the multiple second feature data is screened out. Since the change trend and energy of the signal are common characteristics of the signal, the feature threshold determined in this application can be applied to different interference signals, which improves the generalization of the feature threshold corresponding to the interference signal. The first feature threshold and the second feature threshold can be used to identify whether the vibration signal is an interference signal, so that different interference signals can be accurately identified through the feature threshold. When interference signals are identified based on the characteristic threshold, accurate identification of interference signals of different categories can be achieved, further improving the accuracy of mechanical fault diagnosis, avoiding false alarms, reducing maintenance costs and avoiding production interruptions.

[0093] In some embodiments of the present application, the original vibration signal includes a plurality of vibration data, and the first characteristic data corresponding to the original vibration signal is determined according to the change trend of the original vibration signal, including:

[0094] Filtering out the first vibration data and the second vibration data from the original vibration signal according to the order of each vibration data in the original vibration signal;

[0095] According to a variation trend between the first vibration data and the second vibration data, first characteristic data corresponding to the original vibration signal is determined.

[0096] In this embodiment, multiple moments correspond to multiple original vibration signals, and one moment corresponds to one original vibration signal. The original vibration signal includes multiple vibration data. For example, the vibration signal is usually represented by a waveform, and the waveform corresponding to the original vibration signal is marked by multiple data points, so that multiple vibration data of the original vibration signal can be obtained. According to the order of each vibration data in the original vibration signal, the first vibration data and the second vibration data can be screened out from the original vibration signal. For example, two adjacent vibration data from multiple vibration data are selected, which are the first vibration data and the second vibration data, respectively, or two vibration data are selected within a preset range, which are the first vibration data and the second vibration data, respectively. According to the change trend between the first vibration data and the second vibration data, the first characteristic data corresponding to the original vibration signal can be determined.

[0097] For example, the first characteristic data is taken as the absolute mean of the first-order difference. The original vibration signal obtained after preprocessing is subjected to data standardization to obtain the value of the vibration data after standardization. ,in, is the mean value of the vibration data, is the standard deviation of the vibration data. According to the standardized values ​​of each vibration data, the absolute mean of the first-order difference corresponding to the original vibration signal is determined .in, and is the normalized value corresponding to two adjacent vibration data, and m represents the total number of vibration data in the original vibration signal at time t. The absolute mean of the first-order difference is the first characteristic data corresponding to the original vibration signal.

[0098] The first characteristic data corresponding to the original vibration signal is determined by the change trend between the first vibration data and the second vibration data in this embodiment, so that the characteristic threshold of the interference signal can be accurately determined by the first characteristic data.

[0099] In some embodiments of the present application, the above-mentioned determining the second characteristic data corresponding to the original vibration signal according to the energy amplitude of the original vibration signal includes:

[0100] Acquire a first energy value of the original vibration signal within a first frequency range, and a second energy value of the original vibration signal within a second frequency range;

[0101] Second characteristic data corresponding to the original vibration signal is determined according to the first energy value and the second energy value.

[0102] In this embodiment, the first frequency range and the second frequency range are two pre-set frequency ranges. The first frequency range and the second frequency range are different, but there may be an overlapping portion. According to the ratio of the first energy value to the second energy value, the energy data corresponding to the original vibration signal, i.e., the second characteristic data, can be obtained.

[0103] In some embodiments, the first frequency range is the low frequency range of the original vibration signal, and the second frequency range is the entire frequency range of the original vibration signal. The first energy value of the original vibration signal in the low frequency range and the total energy value of the original vibration signal in the entire frequency range, i.e., the second energy value, are obtained. The ratio of the first energy value to the second energy value is calculated to obtain the low frequency energy proportion corresponding to the original vibration signal, and the low frequency energy proportion is the second characteristic data. The calculation formula of the second characteristic data is as follows:

[0104]

[0105] in, represents the second characteristic data, t represents the time corresponding to the original vibration signal, Indicates the energy amplitude corresponding to the low-frequency dividing line frequency, j to n represents the low-frequency range, and i to m represents the entire frequency range.

[0106] By using the energy amplitude of the original vibration signal in this embodiment, the second characteristic data corresponding to the original vibration signal is determined, so that the characteristic threshold of the interference signal can be accurately determined through the second characteristic data.

[0107] In some embodiments of the present application, the first feature threshold corresponding to the plurality of first feature data is screened out according to the first category, including:

[0108] Sort the plurality of first feature data according to the first category to obtain a target feature sequence;

[0109] A first feature threshold is screened out from the target feature sequence according to a position of the first feature data in the target feature sequence and a first category corresponding to the first feature data.

[0110] In this embodiment, the plurality of first characteristic data may be sorted according to the size and the first category of the first characteristic data to obtain a target characteristic sequence, wherein the plurality of first characteristic data refers to the first characteristic data corresponding to the plurality of original vibration signals within the target time period.

[0111] In some embodiments, by clustering the first feature data and the second feature data of multiple original vibration signals, a labeled cluster analysis result can be obtained. In the cluster analysis result, each first feature data and each second feature data has its corresponding cluster label, and the cluster label is the category corresponding to the first feature data and the second feature data. Therefore, after obtaining the cluster analysis result, the first feature data with labels in the cluster analysis result can be sorted on the dimension of the first feature data to obtain the target feature sequence corresponding to the first feature data; the second feature data with labels in the cluster analysis result can be sorted on the dimension of the second feature data to obtain the target feature sequence corresponding to the second feature data.

[0112] For example, cluster analysis is performed on all data of the absolute mean of the first-order difference and the proportion of low-frequency energy in the target period to obtain the cluster analysis results The cluster analysis results The absolute mean of the first-order difference And the proportion of low-frequency energy Dimension, sorted with labels. For example, the cluster analysis results are sorted in the first-order difference absolute mean dimension with labels, and we get , represents a data set Dimensions are sorted to obtain the target feature sequence corresponding to the first feature data The cluster analysis results are sorted with labels in the low-frequency energy proportion dimension to obtain , represents a data set exist The dimension is sorted to obtain the target feature sequence corresponding to the second feature data .

[0113] in, Indicates cluster analysis of a given data set. Indicated in Time to The collection of all first-order difference absolute mean and low-frequency energy proportion data obtained by analyzing the original vibration signal within time. Time to The time is the target period. is the absolute mean of the first-order differences, is the proportion of low-frequency energy, is the cluster label after clustering the absolute mean of the first-order difference and the proportion of low-frequency energy corresponding to time t.

[0114] In the embodiment of the present application, the process of sorting the second feature data and screening out the second feature threshold is the same as the process of sorting the first feature data and screening out the first feature threshold, and will not be repeated here.

[0115] A first feature threshold is screened out from the target feature sequence according to a position of the first feature data in the target feature sequence and a first category corresponding to the first feature data.

[0116] In some embodiments of the present application, the first category includes a first subcategory and a second subcategory, and the target feature sequence includes a first feature sequence corresponding to the first subcategory and a second feature sequence corresponding to the second subcategory;

[0117] Filtering a first feature threshold from the target feature sequence according to a position of the first feature data in the target feature sequence and a first category corresponding to the first feature data includes:

[0118] Determine the boundary position between the first feature sequence and the second feature sequence in the target feature sequence;

[0119] Screening out at least two candidate feature data from the plurality of first feature data according to the boundary position and the position of each first feature data in the target feature sequence;

[0120] A first feature threshold is obtained iteratively according to at least two candidate feature data.

[0121] In this embodiment, the first category includes the first subcategory and the second subcategory, and similarly, the second category also includes the third subcategory and the fourth subcategory. Therefore, the target feature sequence obtained by sorting the first feature data includes the first feature data of the first subcategory and the first feature data of the second subcategory. In the target feature sequence, the first feature data corresponding to the first subcategory constitutes the first feature sequence, and the first feature data corresponding to the second subcategory constitutes the second feature sequence.

[0122] Determine the boundary position between the first feature sequence and the second feature sequence in the target feature sequence. For example, we analyze the locations where cluster labels are converted from both ends of the data. and Thus, we can get the fuzzy boundary band of two clusters in the cluster analysis of the first-order difference absolute mean d dimension. Similarly, the cluster analysis of the low-frequency energy ratio R dimension can be divided into two clusters of fuzzy boundary bands. In the fuzzy boundary band, the position where the cluster label changes is the boundary position between the first feature sequence and the second feature sequence.

[0123] To determine the feature threshold from the fuzzy boundary band, further data processing is required. According to the data characteristics, it can be determined that there must be a threshold in the fuzzy boundary band, so that the data points in the fuzzy boundary band have the least number of misclassified data points on both sides of the threshold. Therefore, according to the boundary position and the position of each first feature data in the target feature sequence, at least two candidate feature data are selected from the multiple first feature data. For example, the first feature data at both ends of the fuzzy boundary band are selected. and , are selected as two candidate feature data. Among them, The position in the target feature sequence The corresponding first feature data, The position in the target feature sequence The first feature threshold is obtained by iteratively obtaining the corresponding first feature data according to the two candidate feature data.

[0124] In some embodiments of the present application, the iterative step of obtaining the first feature threshold according to at least two candidate feature data includes:

[0125] Obtaining the loss value corresponding to the candidate feature data, and determining whether the loss value meets the preset loss condition;

[0126] If the loss value satisfies the loss condition, the candidate feature data with the smallest loss value is selected as the first feature threshold;

[0127] If the loss value does not meet the loss condition, the candidate feature data is iterated according to the loss value, the iterated candidate feature data is used as the candidate feature data, and the execution is returned to obtain the loss value corresponding to the candidate feature data until the loss value meets the loss condition.

[0128] In this embodiment, in order to quickly determine the feature threshold, the two determined candidate feature data can be used as two data endpoints, and the third candidate feature data can be determined based on the two candidate feature data. For example, the median of the two candidate feature data is calculated, and the median is used as the third candidate feature data. Calculate the loss values ​​corresponding to the third candidate feature data and the two candidate feature data respectively. Specifically, obtain a preset loss function, which can be a loss function for the amount of data in which the feature data is misclassified into two clusters (i.e., two different subcategories) on both sides of the threshold. According to the loss function, the loss values ​​of the two candidate feature data and the third candidate feature data are calculated respectively to obtain the loss values ​​corresponding to each candidate feature data, that is, to obtain the loss values ​​corresponding to the three candidate feature data respectively. Select the current minimum loss value from the loss values ​​corresponding to the three candidate feature data. Compare the current minimum loss value with the historical minimum loss value to determine whether the comparison result meets the preset comparison result. If so, determine that the current loss value meets the preset loss condition. For example, the calculated minimum loss value The minimum loss value calculated from the previous iteration For comparison, if , it is determined that the loss condition is met, and the candidate feature data with the smallest loss value is used as the first feature threshold.

[0129] like , it is determined that the loss condition is not met, and the candidate feature data are iterated according to the loss values ​​corresponding to the third candidate feature data and the two candidate feature data. Specifically, two smaller loss values ​​are selected from the three loss values, and the candidate feature data corresponding to the two smaller loss values ​​are used as the candidate feature data after iteration. For example, the three loss values ​​are A, B, and C, respectively, where A is greater than B and B is greater than C, then the candidate feature data corresponding to B and C are determined as the candidate feature data after iteration. The candidate feature data after iteration is used as the new data endpoint, that is, as the candidate feature data, and the execution is returned to obtain the loss value corresponding to the candidate feature data, and it is determined whether the loss value meets the preset loss condition until the loss value meets the preset loss condition.

[0130] The loss condition in this embodiment can be used to quickly iterate the candidate feature data, thereby improving the efficiency of acquiring the feature threshold.

[0131] In some embodiments of the present application, if the loss value does not meet the loss condition, the candidate feature data is iterated according to the loss value, the iterated candidate feature data is used as the candidate feature data, and the loss value corresponding to the candidate feature data is obtained by returning to execute until the loss value meets the loss condition, including:

[0132] If the loss value does not meet the loss condition, determining whether the number of iterations meets the preset number of iterations, and if the number of iterations meets the preset number of iterations, selecting a first feature threshold from the candidate feature data after the iteration;

[0133] If the loss value does not meet the loss condition and the number of iterations does not meet the preset number of iterations, the candidate feature data is iterated according to the loss value, and the candidate feature data after the iteration is used as the candidate feature data. Return to execute to obtain the loss value corresponding to the candidate feature data until the loss value meets the loss condition or the number of iterations meets the preset number of iterations.

[0134] In this embodiment, in order to save processor computing power, the feature threshold can also be constrained by the number of iterations. Specifically, if the loss value does not meet the loss condition, it is determined whether the number of iterations of the candidate feature data meets the preset number of iterations. If the number of iterations meets the preset number of iterations, the first feature threshold is selected from the candidate feature data after the iteration. For example, the preset number of iterations is 100 times. If the number of iterations is greater than or equal to 100 times, it is determined that the number of iterations meets the preset number of iterations, and the first feature threshold is selected from the candidate feature data after the iteration. In some embodiments, the candidate feature data corresponding to the minimum loss value can be selected as the first feature threshold from the candidate feature data obtained in the last iteration.

[0135] If the loss value does not meet the loss condition and the number of iterations does not meet the preset number of iterations, the candidate feature data will continue to be iterated according to the loss value, and the candidate feature data after the iteration will be used as the candidate feature data. Return to execute and obtain the loss value corresponding to the candidate feature data until the loss value meets the loss condition or the number of iterations meets the preset number of iterations.

[0136] By constraining the candidate feature data by presetting the number of iterations in this embodiment, processor computing power can be saved and waste of computing resources can be avoided.

[0137] See also Figure 2 , Figure 2 : is a flow chart of a signal recognition method provided by an embodiment of the present application. The signal recognition method may include:

[0138] S201, obtaining a vibration signal to be identified;

[0139] S202, determining first characteristic data corresponding to the vibration signal to be identified according to a change trend of the vibration signal to be identified, and determining second characteristic data corresponding to the vibration signal to be identified according to an energy amplitude of the vibration signal to be identified;

[0140] S203, obtaining a preset first feature threshold and a second feature threshold, comparing the first feature data corresponding to the vibration signal to be identified and the first feature threshold to obtain a first comparison result, and comparing the second feature data corresponding to the vibration signal to be identified and the second feature threshold to obtain a second comparison result;

[0141] S204: If the first comparison result and / or the second comparison result meets the preset comparison result, it is determined that the vibration signal to be identified is an interference signal.

[0142] In this embodiment, after the first characteristic threshold and the second characteristic threshold are obtained, if the vibration signal to be identified is obtained, the interference signal can be identified from the multiple vibration signals to be identified according to the first characteristic threshold and the second characteristic threshold. Specifically, according to the change trend of each vibration signal to be identified, the first characteristic data corresponding to each vibration signal to be identified is determined, such as the absolute mean of the first-order difference of the vibration signal to be identified, and according to the energy amplitude of the vibration signal to be identified, the second characteristic data corresponding to the vibration signal to be identified is determined, such as the low-frequency energy proportion of the vibration signal to be identified. A preset first characteristic threshold, such as the characteristic threshold corresponding to the absolute mean of the first-order difference is obtained, and the first-order difference absolute mean of the vibration signal to be identified is compared with the first characteristic threshold to obtain a first comparison result. A preset second characteristic threshold, such as the characteristic threshold corresponding to the low-frequency energy proportion is obtained, and the low-frequency energy proportion of the vibration signal to be identified and the second characteristic threshold are compared to obtain a second comparison result. If the first comparison result and / or the second comparison result meet the preset comparison result, it is determined that the vibration signal to be identified is an interference signal.

[0143] In this embodiment, the vibration signal to be identified may be a vibration signal at a certain moment. If multiple vibration signals to be identified within a period of time are obtained, the interference signal can be identified from the multiple vibration signals to be identified by the signal identification method in the embodiment of the present application.

[0144] The signal recognition method proposed in this embodiment compares the first characteristic data corresponding to the vibration signal to be identified with the first characteristic threshold to obtain a first comparison result, and compares the second characteristic data corresponding to the vibration signal to be identified with the second characteristic threshold to obtain a second comparison result, thereby achieving an accurate judgment on whether the vibration signal to be identified is an interference signal. In addition, since the first characteristic threshold and the second characteristic threshold are universal, interference signals generated by different noises can be accurately identified by the first characteristic threshold and / or the second characteristic threshold. In addition, false alarms of the equipment are avoided, and the efficiency of equipment fault detection and the accuracy of fault identification are improved.

[0145] In order to better implement the non-steady-state signal processing method provided in the embodiment of the present application, the embodiment of the present application also provides a device based on the non-steady-state signal processing method. The meanings of the terms are the same as those in the non-steady-state signal processing method, and the specific implementation details can refer to the description in the method embodiment.

[0146] For example, Fig. 9 As shown, the non-steady-state signal processing device may include: a first acquisition module 301, a first confirmation module 302, a clustering module 303 and a screening module 304.

[0147] The first acquisition module 301 is used to acquire multiple original vibration signals corresponding to multiple moments in a target period;

[0148] A first confirmation module 302 is used to determine first characteristic data corresponding to the original vibration signal according to a change trend of the original vibration signal, and to determine second characteristic data corresponding to the original vibration signal according to an energy amplitude of the original vibration signal;

[0149] A clustering module 303, used to cluster the first characteristic data and the second characteristic data of the plurality of original vibration signals within the target time period to obtain a first category corresponding to the first characteristic data and a second category corresponding to the second characteristic data;

[0150] The screening module 304 is used to screen out first feature thresholds corresponding to multiple first feature data according to the first category, and to screen out second feature thresholds corresponding to multiple second feature data according to the second category. The first feature threshold and the second feature threshold are used to identify whether the vibration signal is an interference signal.

[0151] In an embodiment of the present application, the original vibration signal includes a plurality of vibration data, and the first confirmation module 302 includes:

[0152] A first screening unit, configured to screen out first vibration data and second vibration data from the original vibration signal according to the order of each vibration data in the original vibration signal;

[0153] The first confirmation unit is used to determine first characteristic data corresponding to the original vibration signal according to a change trend between the first vibration data and the second vibration data.

[0154] In an embodiment of the present application, the first confirmation module 302 further includes:

[0155] A first acquisition unit, used to acquire a first energy value of the original vibration signal within a first frequency range, and a second energy value of the original vibration signal within a second frequency range;

[0156] The second confirmation unit is used to determine second characteristic data corresponding to the original vibration signal according to the first energy value and the second energy value.

[0157] In one embodiment of the present application, the screening module 304 includes:

[0158] A sorting unit, used for sorting the plurality of first feature data according to the first category to obtain a target feature sequence;

[0159] The second screening unit is used to screen out a first feature threshold from the target feature sequence according to a position of the first feature data in the target feature sequence and a first category corresponding to the first feature data.

[0160] In an embodiment of the present application, the first category includes a first subcategory and a second subcategory, and the target feature sequence includes a first feature sequence corresponding to the first subcategory and a second feature sequence corresponding to the second subcategory; the second screening unit includes:

[0161] A third confirmation unit is used to determine the boundary position between the first feature sequence and the second feature sequence in the target feature sequence;

[0162] A third screening unit, configured to screen out at least two candidate feature data from the plurality of first feature data according to the boundary position and the position of each first feature data in the target feature sequence;

[0163] A processing unit is used to iteratively obtain a first feature threshold according to at least two candidate feature data.

[0164] In one embodiment of the present application, the processing unit includes:

[0165] A second acquisition unit, used to acquire a loss value corresponding to the candidate feature data, and determine whether the loss value meets a preset loss condition;

[0166] A first processing subunit is configured to select candidate feature data with a minimum loss value as a first feature threshold if the loss value satisfies the loss condition;

[0167] The second processing sub-unit is used to iterate the candidate feature data according to the loss value if the loss value does not meet the loss condition, use the iterated candidate feature data as the candidate feature data, and return to execute to obtain the loss value corresponding to the candidate feature data until the loss value meets the loss condition.

[0168] In one embodiment of the present application, the second processing subunit includes:

[0169] A fourth screening unit, configured to determine whether the number of iterations satisfies a preset number of iterations if the loss value does not satisfy the loss condition, and select a first feature threshold from the candidate feature data after the iteration if the number of iterations satisfies the preset number of iterations;

[0170] The third processing sub-unit is used to iterate the candidate feature data according to the loss value if the loss value does not meet the loss condition and the number of iterations does not meet the preset number of iterations, use the iterated candidate feature data as the candidate feature data, and return to execute to obtain the loss value corresponding to the candidate feature data until the loss value meets the loss condition or the number of iterations meets the preset number of iterations.

[0171] The present application also provides a device based on the above signal recognition method. The meanings of the terms are the same as those in the above signal recognition method, and the specific implementation details can refer to the description in the method embodiment.

[0172] For example, Fig.10 As shown, the signal recognition device may include: a second acquisition module 401, a second confirmation module 402, a comparison module 403, and a third confirmation module 404.

[0173] The second acquisition module 401 is used to acquire the vibration signal to be identified;

[0174] A second confirmation module 402 is used to determine the first characteristic data corresponding to the vibration signal to be identified according to the change trend of the vibration signal to be identified, and to determine the second characteristic data corresponding to the vibration signal to be identified according to the energy amplitude of the vibration signal to be identified;

[0175] The comparison module 403 is used to obtain a preset first feature threshold and a second feature threshold, compare the first feature data corresponding to the vibration signal to be identified with the first feature threshold to obtain a first comparison result, and compare the second feature data corresponding to the vibration signal to be identified with the second feature threshold to obtain a second comparison result;

[0176] The third confirmation module 404 is used to determine that the vibration signal to be identified is an interference signal if the first comparison result and / or the second comparison result meets the preset comparison result. In specific implementation, the above modules can be implemented as independent entities, or can be arbitrarily combined and implemented as the same or several entities. The specific implementation methods and corresponding beneficial effects of the above modules can be referred to the previous method embodiments, which will not be repeated here.

[0177] The present application also provides an electronic device, which may be a server or a terminal, etc. Fig.11 As shown, it shows a schematic diagram of the structure of the electronic device involved in the embodiment of the present application, specifically:

[0178] The electronic device may include components such as a processor 601 with one or more processing cores, a memory 602 with one or more computer-readable storage media, a power supply 603, and an input unit 604. Those skilled in the art will appreciate that Fig.11 The electronic device structure shown in the figure does not constitute a limitation on the electronic device, and may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently.

[0179] The processor 601 is the control center of the electronic device, which uses various interfaces and lines to connect various parts of the entire electronic device, and executes various functions of the electronic device and processes data by running or executing computer programs and / or modules stored in the memory 602, and calling data stored in the memory 602. Optionally, the processor 601 may include one or more processing cores; preferably, the processor 601 may integrate an application processor and a modem processor, wherein the application processor mainly processes the operating system, user interface, and application programs, and the modem processor mainly processes wireless communications. It is understandable that the above-mentioned modem processor may not be integrated into the processor 601.

[0180] The memory 602 can be used to store computer programs and modules. The processor 601 executes various functional applications and data processing by running the computer programs and modules stored in the memory 602. The memory 602 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, a computer program required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area may store data created according to the use of the electronic device, etc. In addition, the memory 602 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 volatile solid-state storage devices. Accordingly, the memory 602 may also include a memory controller to provide the processor 601 with access to the memory 602.

[0181] The electronic device also includes a power supply 603 for supplying power to each component. Preferably, the power supply 603 can be logically connected to the processor 601 through a power management system, so as to manage charging, discharging, power consumption and other functions through the power management system. The power supply 603 can also include one or more DC or AC power supplies, recharging systems, power failure detection circuits, power converters or inverters, power status indicators and other arbitrary components.

[0182] The electronic device may further include an input unit 604, which may be used to receive input digital or character information and generate keyboard, mouse, joystick, optical or trackball signal input related to user settings and function control.

[0183] Although not shown, the electronic device may further include a display unit, etc., which will not be described in detail herein. Specifically in this embodiment, the processor 601 in the electronic device will load the executable files corresponding to the processes of one or more computer programs into the memory 602 according to the following instructions, and the processor 601 will run the computer programs stored in the memory 602, thereby realizing various functions, such as:

[0184] Acquire multiple original vibration signals corresponding to multiple moments in a target period;

[0185] Determine first characteristic data corresponding to the original vibration signal according to a change trend of the original vibration signal, and determine second characteristic data corresponding to the original vibration signal according to an energy amplitude of the original vibration signal;

[0186] Clustering the first characteristic data and the second characteristic data of the plurality of original vibration signals within the target time period to obtain a first category corresponding to the first characteristic data and a second category corresponding to the second characteristic data;

[0187] According to the first category, a plurality of first feature data corresponding to first feature thresholds are screened out, and according to the second category, a plurality of second feature data corresponding to second feature thresholds are screened out, and the first feature threshold and the second feature threshold are used to identify whether the vibration signal is an interference signal. And / or,

[0188] Acquire a vibration signal to be identified;

[0189] Determine first characteristic data corresponding to the vibration signal to be identified according to a change trend of the vibration signal to be identified, and determine second characteristic data corresponding to the vibration signal to be identified according to an energy amplitude of the vibration signal to be identified;

[0190] Obtaining a preset first feature threshold and a second feature threshold, comparing the first feature data corresponding to the vibration signal to be identified and the first feature threshold to obtain a first comparison result, and comparing the second feature data corresponding to the vibration signal to be identified and the second feature threshold to obtain a second comparison result;

[0191] If the first comparison result and / or the second comparison result meets the preset comparison result, it is determined that the vibration signal to be identified is an interference signal.

[0192] The specific implementation methods and corresponding beneficial effects of the above operations can be found in the above detailed description of the non-steady-state signal processing method, which will not be elaborated here.

[0193] A person of ordinary skill in the art will appreciate that all or part of the steps in the various methods of the above embodiments may be completed by a computer program, or by controlling related hardware through a computer program. The computer program may be stored in a computer-readable storage medium and loaded and executed by a processor.

[0194] To this end, an embodiment of the present application provides a computer-readable storage medium, in which a computer program is stored, and the computer program can be loaded by a processor to execute the steps in any of the non-steady-state signal processing methods provided in the embodiments of the present application. For example, the computer program can execute the following steps:

[0195] Acquire multiple original vibration signals corresponding to multiple moments in a target period;

[0196] Determine first characteristic data corresponding to the original vibration signal according to a change trend of the original vibration signal, and determine second characteristic data corresponding to the original vibration signal according to an energy amplitude of the original vibration signal;

[0197] Clustering the first characteristic data and the second characteristic data of the plurality of original vibration signals within the target time period to obtain a first category corresponding to the first characteristic data and a second category corresponding to the second characteristic data;

[0198] According to the first category, a plurality of first feature data corresponding to first feature thresholds are screened out, and according to the second category, a plurality of second feature data corresponding to second feature thresholds are screened out, and the first feature threshold and the second feature threshold are used to identify whether the vibration signal is an interference signal. And / or,

[0199] Acquire a vibration signal to be identified;

[0200] Determine first characteristic data corresponding to the vibration signal to be identified according to a change trend of the vibration signal to be identified, and determine second characteristic data corresponding to the vibration signal to be identified according to an energy amplitude of the vibration signal to be identified;

[0201] Obtaining a preset first feature threshold and a second feature threshold, comparing the first feature data corresponding to the vibration signal to be identified and the first feature threshold to obtain a first comparison result, and comparing the second feature data corresponding to the vibration signal to be identified and the second feature threshold to obtain a second comparison result;

[0202] If the first comparison result and / or the second comparison result meets the preset comparison result, it is determined that the vibration signal to be identified is an interference signal.

[0203] The specific implementation methods and corresponding beneficial effects of the above operations can be found in the previous embodiments and will not be described in detail here.

[0204] The computer-readable storage medium may include: a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0205] Since the computer program stored in the computer-readable storage medium can execute the steps in any one of the non-steady-state signal processing methods provided in the embodiments of the present application, the beneficial effects that can be achieved by any one of the non-steady-state signal processing methods provided in the embodiments of the present application can be achieved. Please refer to the previous embodiments for details and will not be repeated here.

[0206] According to one aspect of the present application, a computer program product or a computer program is provided, the computer program product or the computer program includes computer instructions, and the computer instructions are stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device performs the above-mentioned non-steady-state signal processing method.

[0207] The above is a detailed introduction to a non-steady-state signal processing method, device, electronic device and computer-readable storage medium provided in the embodiments of the present application. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea; at the same time, for technical personnel in this field, according to the ideas of the present application, there will be changes in the specific implementation methods and application scopes. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A non-steady-state signal processing method, characterized in that: include: Acquire multiple original vibration signals corresponding to multiple moments in a target period; Determining first characteristic data corresponding to the original vibration signal according to a change trend of the original vibration signal, and determining second characteristic data corresponding to the original vibration signal according to an energy amplitude of the original vibration signal; Clustering the first characteristic data and the second characteristic data of the plurality of original vibration signals within the target time period to obtain a first category corresponding to the first characteristic data and a second category corresponding to the second characteristic data; According to the first category, a plurality of first feature thresholds corresponding to the first feature data are screened out, and according to the second category, a plurality of second feature thresholds corresponding to the second feature data are screened out, wherein the first feature threshold and the second feature threshold are used to identify whether the vibration signal is an interference signal; The original vibration signal includes a plurality of vibration data, and determining the first characteristic data corresponding to the original vibration signal according to the change trend of the original vibration signal includes: Filtering out first vibration data and second vibration data from the original vibration signal according to the order of each vibration data in the original vibration signal; First characteristic data corresponding to the original vibration signal is determined according to a variation trend between the first vibration data and the second vibration data.

2. The non-steady-state signal processing method according to claim 1, characterized in that: The determining, according to the energy amplitude of the original vibration signal, second characteristic data corresponding to the original vibration signal comprises: Acquire a first energy value of the original vibration signal within a first frequency range, and a second energy value of the original vibration signal within a second frequency range; Second characteristic data corresponding to the original vibration signal is determined according to the first energy value and the second energy value.

3. The non-steady-state signal processing method according to claim 1, characterized in that: The step of screening out a plurality of first feature thresholds corresponding to the first feature data according to the first category includes: Sort the plurality of the first feature data according to the first category to obtain a target feature sequence; The first feature threshold is screened out from the target feature sequence according to the position of the first feature data in the target feature sequence and the first category corresponding to the first feature data.

4. The non-steady-state signal processing method according to claim 3, characterized in that: The first category includes a first subcategory and a second subcategory, and the target feature sequence includes a first feature sequence corresponding to the first subcategory and a second feature sequence corresponding to the second subcategory; The step of selecting the first feature threshold from the target feature sequence according to the position of the first feature data in the target feature sequence and the first category corresponding to the first feature data includes: Determining a boundary position between the first feature sequence and the second feature sequence in the target feature sequence; Screening out at least two candidate feature data from a plurality of the first feature data according to the boundary position and the position of each of the first feature data in the target feature sequence; The first feature threshold is obtained iteratively according to at least two of the candidate feature data.

5. The non-steady-state signal processing method according to claim 4, characterized in that: The iteratively obtaining the first feature threshold according to at least two of the candidate feature data includes: Obtaining a loss value corresponding to the candidate feature data, and determining whether the loss value satisfies a preset loss condition; If the loss value satisfies the loss condition, selecting the candidate feature data with the smallest loss value as the first feature threshold; If the loss value does not satisfy the loss condition, the candidate feature data is iterated according to the loss value, the iterated candidate feature data is used as the candidate feature data, and the process returns to obtain the loss value corresponding to the candidate feature data until the loss value satisfies the loss condition.

6. The non-steady-state signal processing method according to claim 5, characterized in that: If the loss value does not satisfy the loss condition, iterating the candidate feature data according to the loss value, taking the iterated candidate feature data as the candidate feature data, returning to execute to obtain the loss value corresponding to the candidate feature data until the loss value satisfies the loss condition, including: If the loss value does not satisfy the loss condition, determining whether the number of iterations satisfies a preset number of iterations, and if the number of iterations satisfies the preset number of iterations, selecting the first feature threshold from the candidate feature data after the iteration; If the loss value does not meet the loss condition and the number of iterations does not meet the preset number of iterations, the candidate feature data is iterated according to the loss value, the candidate feature data after the iteration is used as the candidate feature data, and the process returns to obtain the loss value corresponding to the candidate feature data until the loss value meets the loss condition or the number of iterations meets the preset number of iterations.

7. A signal recognition method, characterized in that: include: Acquire a vibration signal to be identified; Determining first characteristic data corresponding to the vibration signal to be identified according to a change trend of the vibration signal to be identified, and determining second characteristic data corresponding to the vibration signal to be identified according to an energy amplitude of the vibration signal to be identified; Obtaining a preset first feature threshold and a second feature threshold, comparing the first feature data corresponding to the vibration signal to be identified with the first feature threshold to obtain a first comparison result, and comparing the second feature data corresponding to the vibration signal to be identified with the second feature threshold to obtain a second comparison result; If the first comparison result and / or the second comparison result meets the preset comparison result, determining that the vibration signal to be identified is an interference signal; The first feature threshold and the second feature threshold are obtained by the following steps: Acquire multiple original vibration signals corresponding to multiple moments in a target period; Determining first characteristic data corresponding to the original vibration signal according to a change trend of the original vibration signal, and determining second characteristic data corresponding to the original vibration signal according to an energy amplitude of the original vibration signal; Clustering the first characteristic data and the second characteristic data of the plurality of original vibration signals within the target time period to obtain a first category corresponding to the first characteristic data and a second category corresponding to the second characteristic data; According to the first category, filter out a plurality of first feature thresholds corresponding to the first feature data, and according to the second category, filter out a plurality of second feature thresholds corresponding to the second feature data; The original vibration signal includes a plurality of vibration data, and determining the first characteristic data corresponding to the original vibration signal according to the change trend of the original vibration signal includes: Filtering out first vibration data and second vibration data from the original vibration signal according to the order of each vibration data in the original vibration signal; First characteristic data corresponding to the original vibration signal is determined according to a variation trend between the first vibration data and the second vibration data.

8. A non-steady-state signal processing device, characterized in that: include: A first acquisition module is used to acquire multiple original vibration signals corresponding to multiple moments in a target time period; A first confirmation module, used to determine first characteristic data corresponding to the original vibration signal according to a change trend of the original vibration signal, and to determine second characteristic data corresponding to the original vibration signal according to an energy amplitude of the original vibration signal; A clustering module, used for clustering the first characteristic data and the second characteristic data of the plurality of original vibration signals in the target time period to obtain a first category corresponding to the first characteristic data and a second category corresponding to the second characteristic data; a screening module, configured to screen out a plurality of first feature thresholds corresponding to the first feature data according to the first category, and screen out a plurality of second feature thresholds corresponding to the second feature data according to the second category, wherein the first feature threshold and the second feature threshold are used to identify whether the vibration signal is an interference signal; The original vibration signal includes a plurality of vibration data, and determining the first characteristic data corresponding to the original vibration signal according to the change trend of the original vibration signal includes: Filtering out first vibration data and second vibration data from the original vibration signal according to the order of each vibration data in the original vibration signal; First characteristic data corresponding to the original vibration signal is determined according to a variation trend between the first vibration data and the second vibration data.

9. An electronic device, characterized in that: It comprises a processor and a memory, the memory stores a computer program, and the processor is used to run the computer program in the memory to execute the non-steady-state signal processing method according to any one of claims 1 to 6 and / or the signal recognition method according to claim 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and the computer program is suitable for being loaded by a processor to execute the non-steady-state signal processing method according to any one of claims 1 to 6 and / or the signal identification method according to claim 7.

Images