A voiceprint online monitoring method and system for equipment defect analysis

Abstract

The present invention belongs to the technical field of defect detection. A voiceprint online monitoring method and system for equipment defect analysis are provided. The method comprises: using a first microphone to pick up a first voiceprint feature emitted by a target device in a working state; determining a number of second voiceprint features and a number of defect template data according to the device type of the target device and the state attributes of the working state; smoothing the feature part of the first voiceprint feature involving each second voiceprint feature to obtain a fourth voiceprint feature; matching and calculating the fourth voiceprint feature with the third voiceprint feature in each defect template data, and determining the defect type corresponding to the successfully matched third voiceprint feature as the defect type of the target device. The present invention not only takes into account the device type of the target device itself and the state attributes of the working state, but also takes into account the interference voiceprint features of the environment, thereby improving the accuracy of online defect analysis of the target device.

Description

Technical Field

[0001] The present invention relates to the field of defect detection technology, and in particular to a voiceprint online monitoring method and system for equipment defect analysis. Background Art

[0002] Equipment defect detection based on voiceprint analysis is an emerging technology, which is mainly used to identify and diagnose potential equipment defects by analyzing the acoustic signals generated when the equipment is running. Existing equipment defect detection methods based on voiceprint analysis are all based on the voiceprint characteristics of the equipment itself, lacking the influence of interference factors in the environment where the equipment is located. These interference factors may cause the collected voiceprint features to contain interference voiceprints, resulting in the final equipment defect detection results being inaccurate. The solution of the present invention is intended to solve this technical problem. Summary of the invention

[0003] In this regard, the present invention provides a voiceprint online monitoring method, system, electronic device, computer storage medium and computer program product for equipment defect analysis to solve the above technical problems.

[0004] The present invention discloses a voiceprint online monitoring method for equipment defect analysis, the method comprising the following steps:

[0005] Pick up a first voiceprint feature emitted by the target device in a working state using a first microphone;

[0006] Determine and obtain a plurality of second voiceprint features according to the device type and the state attributes of the working state of the target device; wherein the second voiceprint features are voiceprint features of the environment in which the target device is located;

[0007] Determine and obtain a plurality of defect template data according to the device type and the state attribute of the working state of the target device, each of the defect template data includes a defect type and a corresponding third voiceprint feature;

[0008] performing smoothing processing on a feature portion of the first voiceprint feature related to each of the second voiceprint features to obtain a fourth voiceprint feature;

[0009] The fourth voiceprint feature is matched and calculated with the third voiceprint feature in each defect template data, and the defect type corresponding to the successfully matched third voiceprint feature is determined as the defect type of the target device.

[0010] In some embodiments, the step of using the first microphone to pick up a first voiceprint feature emitted by the target device in a working state includes:

[0011] Detecting whether there is a defect record of the target device;

[0012] If so, the first collection start time of the first microphone is determined according to the defect record; if not, the first microphone is controlled to enter the voiceprint collection state when the target device is turned on.

[0013] In some embodiments, the determining of the defect template data according to the device type and the state attribute of the working state of the target device includes:

[0014] Determine and obtain a plurality of first defect template data according to the device type of the target device;

[0015] If the status attributes of the working status of the target device can be determined, the second defect template data is obtained by screening from several first defect template data according to the status attributes of the working status of the target device, and the second defect template data is determined as the defect template data; otherwise, several first defect template data are determined as the defect template data.

[0016] In some embodiments, the determining of a plurality of second voiceprint features according to the device type and the state attributes of the working state of the target device includes:

[0017] Predicting the probability of a change in the state attribute of the target device according to the device type and the real-time state attribute of the working state of the target device;

[0018] If the probability is lower than the probability threshold, controlling the second microphone to instantly collect the second voiceprint feature of the environment where the target device is located;

[0019] If the probability is higher than the probability threshold, predict the predicted state attributes according to the device type and the real-time state attributes of the working state of the target device, and determine a number of auxiliary devices and corresponding predicted auxiliary state attributes according to the predicted state attributes and the comparison table;

[0020] The second voiceprint feature corresponding to each of the auxiliary devices and the predicted auxiliary state attribute is obtained.

[0021] In some embodiments, the acquiring the second voiceprint feature corresponding to each of the auxiliary devices and the predicted auxiliary state attribute includes:

[0022] Inputting the device type of each auxiliary device and the corresponding predicted auxiliary state attribute into a switching duration prediction model based on artificial intelligence, wherein the switching duration prediction model predicts the switching duration;

[0023] A second collection start time is determined according to the switching duration, and at the second collection start time, the second microphone is controlled to instantly collect the second voiceprint features of each of the auxiliary devices.

[0024] In some embodiments, the step of smoothing the feature portion of the first voiceprint feature related to each of the second voiceprint features to obtain the fourth voiceprint feature includes:

[0025] The significant feature parts corresponding to the corresponding auxiliary devices in each of the second voiceprint features are identified, and the voiceprint curves of each significant feature part existing in the first voiceprint feature are smoothed by using a smoothing algorithm to obtain a fourth voiceprint feature.

[0026] The present invention also discloses a voiceprint online monitoring system for equipment defect analysis, the system comprising a microphone, a processing module, and a storage module, the processing module being connected to the storage module and the microphone respectively;

[0027] The microphone is at least used to collect voiceprint feature data of the target device;

[0028] The storage module stores a computer program;

[0029] The processing module is used to call the computer program to analyze and process the voiceprint feature data to implement any of the methods described above.

[0030] The present invention also discloses an electronic device, comprising: at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, wherein the processor executes the computer program to implement any of the above methods.

[0031] The present invention also discloses a computer storage medium, wherein the computer-readable storage medium stores a computer program, and the computer program is executed by a processor to implement any of the above methods.

[0032] The present invention also discloses a computer program product, which includes a computer program for implementing any of the above methods.

[0033] When detecting the defect type of the target device based on the voiceprint, the present invention not only takes into account the device type and state attributes of the working state of the target device itself, but also takes into account the interference voiceprint characteristics of the environment, thereby further improving the accuracy of online defect analysis of the target device. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0035] Figure 1 It is a flow chart of a voiceprint online monitoring method for equipment defect analysis disclosed in an embodiment of the present invention;

[0036] Figure 2 It is a structural schematic diagram of a voiceprint online monitoring system for equipment defect analysis disclosed in an embodiment of the present invention. DETAILED DESCRIPTION

[0037] The following is a description of the implementation of the present application by specific specific embodiments. People familiar with the technology can easily understand other advantages and effects of the present application from the contents disclosed in this specification. Obviously, the described embodiments are 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 ordinary technicians in the field without creative work are within the scope of protection of the present application.

[0038] In addition, the technical features involved in the different embodiments of the present application described below can be combined with each other as long as they do not conflict with each other.

[0039] like Figure 1 As shown, an embodiment of the present invention discloses a voiceprint online monitoring method for equipment defect analysis, the method comprising the following steps:

[0040] Pick up a first voiceprint feature emitted by the target device in a working state using a first microphone;

[0041] Determine and obtain a plurality of second voiceprint features according to the device type and the state attributes of the working state of the target device; wherein the second voiceprint features are voiceprint features of the environment in which the target device is located;

[0042] Determine and obtain a plurality of defect template data according to the device type and the state attribute of the working state of the target device, each of the defect template data includes a defect type and a corresponding third voiceprint feature;

[0043] performing smoothing processing on a feature portion of the first voiceprint feature related to each of the second voiceprint features to obtain a fourth voiceprint feature;

[0044] The fourth voiceprint feature is matched and calculated with the third voiceprint feature in each defect template data, and the defect type corresponding to the successfully matched third voiceprint feature is determined as the defect type of the target device.

[0045] The present invention adopts a conventional method to pick up the first voiceprint feature emitted by the target device when it is working, and then, according to the device type and the state attribute of the working state of the target device, several second voiceprint features and several defect template data are determined. The second voiceprint feature is the environmental feature of the area where the target device is located, and the defect template data is a pre-established standard defect template. Each defect template data contains a set of defect types and corresponding third voiceprint features. The feature part of the first voiceprint feature involving each second voiceprint feature is smoothed, that is, the fourth voiceprint feature with the environmental interference voiceprint removed is obtained, and then the fourth voiceprint feature is matched and calculated with each defect template data. The defect type corresponding to the successfully matched defect template data is the defect type of the target device.

[0046] Therefore, when detecting the defect type of the target device based on the voiceprint, the present invention not only takes into account the device type and state attributes of the working state of the target device itself, but also takes into account the interference voiceprint characteristics of the environment, thereby further improving the accuracy of online defect analysis of the target device.

[0047] In some embodiments, the step of using the first microphone to pick up a first voiceprint feature emitted by the target device in a working state includes:

[0048] Detecting whether there is a defect record of the target device;

[0049] If so, the first collection start time of the first microphone is determined according to the defect record; if not, the first microphone is controlled to enter the voiceprint collection state when the target device is turned on.

[0050] In this embodiment, the target device may be a device involved in a routine defect inspection task, or may be a device for repair. For the latter, there is a corresponding defect record, which may be automatically generated by the management system of the target device or manually generated by relevant management personnel.

[0051] Therefore, the present invention first detects whether there is a defect record of the target device. If so, it means that the target device has a corresponding defect. This time, the type of the defect needs to be analyzed specifically. Correspondingly, the microphone collection start time can be determined based on the corresponding defect location in the defect record, the working stage of the device when the defect occurs, and other information; if not, it is necessary to collect the voiceprint of the target device in all stages to achieve subsequent full-stage defect inspections. Such a setting can effectively reduce the amount of voiceprint data collected from the target device, which is conducive to faster online analysis of device defects, and is particularly suitable for the situation where a single set of defect analysis equipment analyzes multiple target devices at the same time.

[0052] In some embodiments, the determining of the defect template data according to the device type and the state attribute of the working state of the target device includes:

[0053] Determine and obtain a plurality of first defect template data according to the device type of the target device;

[0054] If the status attributes of the working status of the target device can be determined, the second defect template data is obtained by screening from several first defect template data according to the status attributes of the working status of the target device, and the second defect template data is determined as the defect template data; otherwise, several first defect template data are determined as the defect template data.

[0055] In this embodiment, a defect template database for multiple types of equipment is pre-established, and multiple defect template data stored in the defect template database are respectively stored according to the equipment type. Each equipment type corresponds to multiple defect template data, and the equipment types include cutters, transformers, mixers, etc. in cutting equipment; each equipment also includes state attributes of different working states, and the state attributes include the startup phase, normal working phase (low load, medium load, high load), shutdown phase, etc. The above-mentioned multiple defect template data are respectively associated with the corresponding state attributes.

[0056] During specific use, firstly, a number of first defect template data are determined based on the device type of the target device. At the same time, since the state attributes of different types of target devices may not be determined, for example, the state attributes of electrical equipment are relatively easy to obtain, while the working attributes of mechanical equipment (especially semi-automatic equipment and manually operated equipment) are generally difficult to obtain. In this regard, the present invention is arranged to filter out accurate second defect template data from a number of first defect template data based on the state attributes when the state attributes of the target device can be determined, otherwise all the first defect template data are determined as defect template data. In this way, the defect template data can be determined regardless of the type of the target device.

[0057] In some embodiments, the determining of a plurality of second voiceprint features according to the device type and the state attributes of the working state of the target device includes:

[0058] Predicting the probability of a change in the state attribute of the target device according to the device type and the real-time state attribute of the working state of the target device;

[0059] If the probability is lower than the probability threshold, controlling the second microphone to instantly collect the second voiceprint feature of the environment where the target device is located;

[0060] If the probability is higher than the probability threshold, predict the predicted state attributes according to the device type and the real-time state attributes of the working state of the target device, and determine a number of auxiliary devices and corresponding predicted auxiliary state attributes according to the predicted state attributes and the comparison table;

[0061] The second voiceprint feature corresponding to each of the auxiliary devices and the predicted auxiliary state attribute is obtained.

[0062] In this embodiment, the second voiceprint feature refers to the voiceprint feature of the environment in which the target device is located, but the voiceprint feature of the environment is generally relatively fixed and can be measured in advance or in real time. However, the change of the state attribute of the target device itself may cause the operating state of some associated devices in the environment to change. For example, when the target device switches from low load to high load, it will cause the associated heat dissipation device to switch to startup or switch to high-efficiency heat dissipation mode, so that the first voiceprint feature collected by the first microphone contains interference voiceprints of non-target devices, which need to be excluded.

[0063] Therefore, before the first microphone collects data, the present invention first predicts the probability of changes in the state attributes of the target device based on the device type and the real-time state attributes of the working state of the target device. The real-time state attributes include state attributes of multiple different moments (historical moments and real-time moments) that have been recorded. The state attributes may also include specific working parameters. Based on the characteristics of the target device itself and each state attribute, the state trend of the target device may be analyzed to obtain the probability of changes in the state attributes of the target device.

[0064] If the probability is lower than the probability threshold, it means that the device is running smoothly and there is no sign of state switching at the current stage. At this time, the second microphone is directly used to collect the second voiceprint characteristics of the environment. The second microphone should be kept at a certain distance from the first microphone, for example, 5m, which can be determined according to the actual situation on site.

[0065] If the probability is higher than the probability threshold, it means that the device is about to switch the operating state. At this time, several associated auxiliary devices are determined based on the predicted state attributes and the preset comparison table, and the corresponding predicted auxiliary state attributes can be further predicted. Auxiliary devices refer to devices that will be forced to change their own operating states when the target device switches to the predicted state attributes, such as the aforementioned heat dissipation device. The preset comparison table includes the auxiliary devices corresponding to the target device when it switches between different state attributes. The predicted auxiliary state attributes refer to the predicted state attributes of the auxiliary devices corresponding to the predicted state attributes of the target device, that is, the working state to which the auxiliary device should switch to adapt to the new working state of the target device. In specific implementation, it can be predicted by the constructed dedicated prediction model, and the present invention does not limit this.

[0066] After determining each auxiliary device and its corresponding predicted auxiliary state attribute, the corresponding second voiceprint feature can be determined. The second voiceprint feature can be non-measured, for example, it is predicted in advance by looking up a table or a dedicated prediction model according to the type of each auxiliary device and the corresponding predicted auxiliary state attribute, and it represents the voiceprint feature outputted by at least one auxiliary device in an ideal state when entering the predicted auxiliary state attribute. Of course, the second voiceprint feature can also be measured, which will be described in detail in the following embodiments.

[0067] In some embodiments, the acquiring the second voiceprint feature corresponding to each of the auxiliary devices and the predicted auxiliary state attribute includes:

[0068] Inputting the device type of each auxiliary device and the corresponding predicted auxiliary state attribute into a switching duration prediction model based on artificial intelligence, wherein the switching duration prediction model predicts the switching duration;

[0069] A second collection start time is determined according to the switching duration, and at the second collection start time, the second microphone is controlled to instantly collect the second voiceprint features of each of the auxiliary devices.

[0070] In this embodiment, a switching duration prediction model is constructed based on artificial intelligence technology in advance, and the device type of each auxiliary device and the corresponding predicted auxiliary state attribute are input into the switching duration prediction model to obtain the switching duration, which refers to the time required for each auxiliary device to switch from the state attribute when the target device starts to switch the predicted state attribute to the predicted auxiliary state attribute. Considering the moment when the target device starts to switch the predicted state attribute, the second collection start time can be determined, that is, the second microphone starts to collect the second voiceprint features of each auxiliary device that has completed the switch at the second collection start time. Based on each second voiceprint feature, the interfering voiceprint features (under the predicted auxiliary state attribute) belonging to the auxiliary device in the first voiceprint feature collected by the first microphone can be removed, especially those prominent voiceprint feature areas unique to the auxiliary device.

[0071] It should be noted that the second collection start time may be later than or earlier than the first collection start time. However, when the first voiceprint feature is smoothed, only the part where the second voiceprint feature overlaps with the first voiceprint feature is smoothed.

[0072] In some embodiments, the step of smoothing the feature portion of the first voiceprint feature related to each of the second voiceprint features to obtain the fourth voiceprint feature includes:

[0073] The significant feature parts corresponding to the corresponding auxiliary devices in each of the second voiceprint features are identified, and the voiceprint curves of each significant feature part existing in the first voiceprint feature are smoothed by using a smoothing algorithm to obtain a fourth voiceprint feature.

[0074] In this embodiment, firstly, the significant characteristic parts in the second voiceprint feature are screened out based on the specific type of the auxiliary device, such as the high-frequency part, the regularly occurring high-frequency sound (such as the regular stirring sound emitted by the stirring device when working), etc. If the second voiceprint feature contains these significant characteristic parts, the voiceprint curves of these significant characteristic parts can be smoothed by using a smoothing algorithm such as Savitzky-Golay filter and sliding average filter method, so as to obtain the fourth voiceprint feature, which basically contains the voiceprint feature of the target device itself.

[0075] At this time, based on the existing defect analysis method, the fourth voiceprint feature can be subjected to defect analysis. For example, if the fourth voiceprint feature contains abnormal high-frequency sound, it can be determined that foreign matter has entered the transmission component. Preferably, a dedicated defect type analysis model can be used to process the fourth voiceprint feature, for example, the dedicated defect type analysis model can be constructed based on Transformer.

[0076] like Figure 2 As shown, an embodiment of the present invention further discloses a voiceprint online monitoring system for equipment defect analysis, the system comprising a microphone, a processing module, and a storage module, the processing module being connected to the storage module and the microphone respectively;

[0077] The microphone is at least used to collect voiceprint feature data of the target device;

[0078] The storage module stores a computer program;

[0079] The processing module is used to call the computer program to analyze and process the voiceprint feature data to implement any of the methods described above.

[0080] An embodiment of the present invention further discloses an electronic device, comprising: at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, wherein the processor executes the computer program to implement the method described in the above embodiment.

[0081] An embodiment of the present invention further discloses a computer storage medium, wherein the computer storage medium stores a computer program, and the computer program is executed by a processor to implement the method described in the above embodiment.

[0082] The embodiment of the present invention further discloses a computer program product, which includes a computer program for implementing the method described in the above embodiment.

[0083] The various component embodiments of the present invention can be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof. It should be understood by those skilled in the art that a microprocessor or digital signal processor (DSP) can be used in practice to implement some or all functions of some or all components in the device according to an embodiment of the present invention. The present invention can also be implemented as a device or apparatus program (e.g., computer program and computer program product) for executing part or all of the methods described herein. Such a program implementing the present invention can be stored on a computer-readable medium, or can have the form of one or more signals. Such a signal can be downloaded from an Internet website, or provided on a carrier signal, or provided in any other form.

[0084] In the description provided herein, a large number of specific details are described. However, it is understood that embodiments of the present invention can be practiced without these specific details. In some instances, well-known methods, structures and techniques are not shown in detail so as not to obscure the understanding of this description.

[0085] It should be noted that the above embodiments illustrate the present invention rather than limit it, and that those skilled in the art may devise alternative embodiments without departing from the scope of the appended claims. In the claims, any reference symbol between brackets shall not be construed as a limitation on the claims. The word "comprising" does not exclude the presence of elements or steps not listed in the claims. The word "one" or "an" preceding an element does not exclude the presence of a plurality of such elements. The present invention may be implemented by means of hardware comprising a number of different elements and by means of a suitably programmed computer. In a unit claim enumerating a number of devices, several of these devices may be embodied by the same hardware item. The use of the words first, second, and third, etc., does not indicate any order. These words may be interpreted as names.

[0086] In addition, it should be noted that the language used in this specification is selected primarily for readability and teaching purposes, rather than for explaining or limiting the subject matter of the present invention. Therefore, many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the appended claims. The disclosure of the present invention is illustrative rather than restrictive with respect to the scope of the present invention, and the scope of the present invention is defined by the appended claims.

Claims

1. A voiceprint online monitoring method for equipment defect analysis, characterized in that: The method comprises the following steps: Pick up a first voiceprint feature emitted by the target device in a working state using a first microphone; Determine and obtain a plurality of second voiceprint features according to the device type and the state attributes of the working state of the target device; wherein the second voiceprint features are voiceprint features of the environment in which the target device is located; Determine and obtain a plurality of defect template data according to the device type and the state attribute of the working state of the target device, each of the defect template data includes a defect type and a corresponding third voiceprint feature; performing smoothing processing on a feature portion of the first voiceprint feature related to each of the second voiceprint features to obtain a fourth voiceprint feature; Matching and calculating the fourth voiceprint feature with the third voiceprint feature in each defect template data, and determining the defect type corresponding to the successfully matched third voiceprint feature as the defect type of the target device; According to the device type and the state attributes of the working state of the target device, several second voiceprint features are determined, including: Predicting the probability of a change in the state attribute of the target device according to the device type and the real-time state attribute of the working state of the target device; If the probability is lower than the probability threshold, controlling the second microphone to instantly collect the second voiceprint feature of the environment where the target device is located; If the probability is higher than the probability threshold, predict the predicted state attributes according to the device type and the real-time state attributes of the working state of the target device, and determine a number of auxiliary devices and corresponding predicted auxiliary state attributes according to the predicted state attributes and the comparison table; The second voiceprint feature corresponding to each of the auxiliary devices and the predicted auxiliary state attribute is obtained.

2. The method for online voiceprint monitoring for equipment defect analysis according to claim 1, characterized in that: Using the first microphone to pick up a first voiceprint feature emitted by the target device in a working state, including: Detecting whether there is a defect record of the target device; If so, the first collection start time of the first microphone is determined according to the defect record; if not, the first microphone is controlled to enter the voiceprint collection state when the target device is turned on.

3. The method for online voiceprint monitoring for equipment defect analysis according to claim 1, characterized in that: According to the device type and the state attributes of the working state of the target device, several defect template data are determined, including: Determine and obtain a plurality of first defect template data according to the device type of the target device; If the status attributes of the working status of the target device can be determined, the second defect template data is obtained by screening from several first defect template data according to the status attributes of the working status of the target device, and the second defect template data is determined as the defect template data; otherwise, several first defect template data are determined as the defect template data.

4. The method for online voiceprint monitoring for equipment defect analysis according to claim 3 is characterized by: Acquiring the second voiceprint feature corresponding to each of the auxiliary devices and the predicted auxiliary state attribute includes: Inputting the device type of each auxiliary device and the corresponding predicted auxiliary state attribute into a switching duration prediction model based on artificial intelligence, wherein the switching duration prediction model predicts the switching duration; A second collection start time is determined according to the switching duration, and at the second collection start time, the second microphone is controlled to instantly collect the second voiceprint features of each of the auxiliary devices.

5. The method for online voiceprint monitoring for equipment defect analysis according to claim 4, characterized in that: Smoothing the feature parts of the first voiceprint features related to the second voiceprint features to obtain a fourth voiceprint feature, including: The significant feature parts corresponding to the corresponding auxiliary devices in each of the second voiceprint features are identified, and the voiceprint curves of each significant feature part existing in the first voiceprint feature are smoothed by using a smoothing algorithm to obtain a fourth voiceprint feature.

6. A voiceprint online monitoring system for equipment defect analysis, the system comprising a microphone, a processing module, and a storage module, the processing module being connected to the storage module and the microphone respectively; The microphone is at least used to collect voiceprint feature data of the target device; The storage module stores a computer program; Features: The processing module is used to call the computer program to analyze and process the voiceprint feature data to implement the method according to any one of claims 1-5.

7. An electronic device comprising: At least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, wherein the processor executes the computer program to implement the method according to any one of claims 1 to 5.

8. A computer storage medium storing a computer program, characterized in that: The computer program is executed by a processor to implement the method according to any one of claims 1 to 5.

9. A computer program product, characterized in that: The computer program product comprises a computer program for implementing the method according to any one of claims 1 to 5.

Images