A new intelligent vibration voiceprint sensor perception system

Abstract

The present disclosure provides a new intelligent vibration voiceprint sensor sensing system, belonging to the field of data optimization, which is used in mobile or fixed devices to detect the vibration, displacement and angle change of the measured object, equipment, etc. through cloud server, extracts vibration voiceprint characteristic value, converts it into digital audio signal, and sends it to the background for analysis and processing. Alternatively, local characteristic value analysis is performed, and abnormal conditions exceeding the threshold value are output for early warning, achieving the effect of predicting abnormal operating conditions, and being suitable for various interfaces, small in size and high in cost performance.

Description

TECHNICAL FIELD

[0001] The present disclosure belongs to the field of data optimization and data communication, and particularly relates to a new intelligent vibration voiceprint sensor perception system. BACKGROUND

[0002] The vibration voiceprint sensor can be widely applied in factory and mine production, logistics transportation, machine equipment, building, bridge, pipeline, elevator, even mobile equipment such as automobile, and can realize 24-hour detection of vibration, displacement and angle change of measured objects and equipment by cooperating with a cloud server, extracting vibration voiceprint characteristic values through a specific algorithm, converting the vibration voiceprint characteristic values into digital audio signals, and sending the digital audio signals to a background for analysis and processing. However, the existing system device is not sensitive to abnormal conditions in the detection time. In the patent document with the publication number CN113721167B, a transformer voiceprint vibration detection system based on an Internet of Things architecture is provided, which can be used to collect vibration signals, transformer temperature and audio signals at each monitoring position on the surface of the transformer, but the sensitivity to time sequence needs to be improved. SUMMARY

[0003] The present disclosure aims to provide a new intelligent vibration voiceprint sensor perception system to solve one or more technical problems in the prior art and at least provide a beneficial choice or create conditions.

[0004] The present disclosure provides a new intelligent vibration voiceprint sensor perception system, which is used in mobile or fixed equipment, cooperates with a cloud server to detect vibration, displacement and angle change of measured objects and equipment, extracts vibration voiceprint characteristic values, converts the vibration voiceprint characteristic values into digital audio signals, and sends the digital audio signals to a background for analysis and processing, or performs local characteristic value analysis, and outputs a warning for abnormal conditions exceeding a threshold value.

[0005] In order to achieve the above-mentioned purpose, according to an aspect of the present disclosure, a new intelligent vibration voiceprint sensor perception system is provided, which is used in mobile or fixed equipment, cooperates with a cloud server to detect vibration, displacement and angle change of measured objects and equipment, extracts vibration voiceprint characteristic values, converts the vibration voiceprint characteristic values into digital audio signals, and sends the digital audio signals to a background for analysis and processing, or performs local characteristic value analysis, and outputs a warning for abnormal conditions exceeding a threshold value.

[0006] The system is composed of, but not limited to, a vibration monitoring sensor, a remote acquisition terminal and the like, and can be used as a perception sensor of an intelligent AI cloud platform and a server.

[0007] The vibration monitoring sensor adopts a high-speed communication circuit, a high-speed acceleration and angular velocity acquisition circuit, and a MCU / FPGA included in the high-speed communication circuit. The MCU / FPGA obtains real-time 3D vibration characteristic values of the equipment in operation by calculating the acceleration and angular velocity values returned by the high-speed acquisition sensor, and the acceleration returned by the high-speed acquisition sensor is u i , and the angular velocity value is g i , and the 3D vibration characteristic value is Fi, and the calculation formula of Fi is:

[0008] ;

[0009] Further, the system uses a 24-bit or 16-bit digital audio signal with a speed of not less than 48k / s, which is transmitted to a remote monitoring terminal through an interface. The remote monitoring terminal uses an industrial computer platform based on multiple operating systems, on which a core algorithm is run to convert vibration data into a sound spectrum graph, and then 3D vibration voiceprint characteristics are obtained through MFFC and DFT algorithm analysis. Then, through DL deep learning algorithm, abnormal vibration in any direction of the equipment is judged and an alarm is given.

[0010] Further, the interface of the vibration monitoring sensor adopts a standard USB audio transmission protocol, and does not need to install special driver software.

[0011] Further, the system is transplanted from the remote monitoring terminal to the vibration monitoring sensor, or to a server background with higher computing power than the system. The vibration monitoring sensor analyzes voiceprint characteristic values locally, and directly alarms the server background for abnormal conditions exceeding the threshold. In application scenarios requiring real-time monitoring and recording of operating status, the remote monitoring terminal interface obtains vibration real-time curves of multiple vibration monitoring sensors and performs voiceprint characteristic value algorithm analysis. The obtained data is saved locally and transmitted to the server background database through industrial Ethernet.

[0012] Further, in the vibration monitoring sensor, the MCU uses an interpolation algorithm to fit 48k~192k / s audio signal output based on the sampling values of the low-speed single-axis or multi-axis MEMS sensor.

[0013] Further, if only one-axis vibration change needs to be monitored, the sampling value uses an interpolation algorithm to simplify it to one axis.

[0014] Further, the sampling value uses an interpolation algorithm to extract voiceprint characteristics, analyze background noise, filter and self-calibrate.

[0015] Further, the digital audio signal output format is 48k~192k / s.

[0016] Further, the system also supports interface signal expansion including Ethernet / WIFI / LoRa / Bluetooth / LVDS / CAN or RS-485.

[0017] Further, the system provides installation methods including magnetic attraction, screws and cable ties.

[0018] The beneficial effects of the present disclosure are that the present application proposes a new intelligent vibration voiceprint sensor, which adopts innovative new technology, can not only perceive the vibration, movement and other information of the monitored object and equipment, but also can extract voiceprint features, model and achieve the effect of predicting abnormal operating conditions, can adapt to various interfaces, has small size and high cost performance. BRIEF DESCRIPTION OF DRAWINGS

[0019] The above and other features of the present disclosure will become more apparent with the following detailed description taken in conjunction with the accompanying drawings, in which like reference numerals refer to like elements in the several views. As will be obvious to those of ordinary skill in the art, the drawings included in the present disclosure are diagrammatic or schematic and are not to precise scale, and therefore should not be interpreted in a way defining or limiting the scope of the present disclosure, and appropriately reflect the features of the present disclosure. In the drawings:

[0020] Figure 1 Fig. 1 shows a schematic diagram of embodiment 1 of a new intelligent vibration voiceprint sensor perception system;

[0021] Figure 2 Fig. 2 shows a schematic diagram of embodiment 2 of a new intelligent vibration voiceprint sensor perception system;

[0022] Figure 3 Fig. 3 shows a schematic diagram of embodiment 3 of a new intelligent vibration voiceprint sensor perception system;

[0023] Figure 4 Fig. 4 shows a schematic diagram of embodiment 4 of a new intelligent vibration voiceprint sensor perception system. DETAILED DESCRIPTION

[0024] The concept, specific structure and generated technical effects of the present disclosure will be described clearly and completely in the following combined with embodiments and drawings, so as to fully understand the purpose, scheme and effect of the present disclosure. It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other without conflict.

[0025] In the description of the present application, the meaning of one or more is one or more, the meaning of multiple is more than two, greater than, less than, more than, etc. are understood as not including the number, above, below, within, etc. are understood as including the number. If it is described that the first, second is only used for the purpose of distinguishing technical features, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of indicated technical features or the sequence of indicated technical features.

[0026] As Figures 1-4 The flow chart of a new intelligent vibration voiceprint sensor sensing system according to the present application is shown, and the new intelligent vibration voiceprint sensor sensing system according to the embodiment of the present application is described below in combination with Figures 1-4

[0027] The present disclosure proposes a new intelligent vibration voiceprint sensor sensing system, which is used in mobile or fixed equipment, cooperates with a cloud server to detect the vibration, displacement and angle change of the measured object, equipment, etc., extracts the vibration voiceprint characteristic value, converts it into a digital audio signal, and sends it to the background for analysis and processing; or, performs local characteristic value analysis, and outputs a warning for abnormal conditions exceeding the threshold value;

[0028] The system is composed of a vibration monitoring sensor and a remote acquisition terminal, and can be used as a sensing sensor of an intelligent AI cloud platform and server.

[0029] The vibration monitoring sensor uses a high-speed communication circuit, a high-speed acceleration and angular velocity acquisition circuit, and a high-performance MCU / FPGA to acquire the acceleration and angular velocity values returned by the high-speed acquisition sensor, and calculate the real-time 3D vibration characteristic value of the equipment during operation.

[0030] In some embodiments, the acceleration and angular velocity values returned by the high-speed acquisition sensor are sampled multiple times in a continuous sampling time period, and multiple different time points are used as sampling time points. The number of sampling time points is denoted as n, which can theoretically approach positive infinity. In practice, the continuous sampling time period can be preferably a preset time period, and n should be no less than 12. The serial number of each sampling time point is denoted as i, i∈[1,n].

[0031] For the sampling time point with serial number i, the acceleration value returned by the high-speed acquisition sensor is denoted as u i , and the angular velocity value is denoted as g i . The subscript i of u i and g i indicates the serial number of the corresponding sampling time point, and the 3D vibration characteristic value is denoted as Fi.

[0032] The calculation method of Fi is as follows: ​

[0033] First calculate the special speed initiator rGu:

[0034] ;

[0035] The special speed initiator is a value that effectively guides the calculation of the 3D vibration speed value, which is beneficial to the collection of data characteristics of the acceleration and angular velocity values returned by the high-speed acquisition sensor in the sampling time distribution, and this collection is fast and harmless to the equipment;

[0036] Then calculate the sampling time corresponding to each serial number i,

[0037] ;

[0038] Here, to calculate the Fi corresponding to the sampling time of serial number i, it is necessary to record the data of each time in the n sampling time, and then calculate it, which corresponds to In rGu, ∫ and di represent the integral of The integral is performed at the n sampling time in the continuous sampling time period, and when it is not integrable, it represents The result value of the cumulative sum at the sampling time with serial number belonging to [1, n]; when i is 1, F(i-1) is 1 / π, and π represents the value of the circular constant; As can be seen, the 3D vibration speed value has sensitivity in reflecting numerical fluctuations, and can obviously transmit signals exceeding the frequency, which is used to identify abnormal conditions exceeding the threshold, and is better used for probability prediction and judgment of early warning output;

[0039] Calculate the arithmetic mean of the 3D vibration speed values Fi corresponding to each sampling time as the vibration sub, obtain the maximum value of the 3D vibration speed values Fi corresponding to each sampling time as the vibration base, calculate the ratio of the vibration sub to the vibration base as the vibration base sub, calculate the number of Fi in the 3D vibration speed values Fi corresponding to each sampling time that is greater than or equal to the vibration sub as trop, and calculate the quotient of trop divided by n as the exceeding frequency. Determine whether the exceeding frequency is greater than the vibration base sub in value, if so, it is the abnormal condition, and perform early warning output to the database.

[0040] Further, the system uses a 24-bit or 16-bit digital audio signal of no less than 48k / s, which is transmitted to the remote monitoring terminal through the interface. The remote monitoring terminal uses an industrial computer platform based on multiple operating systems, runs the core algorithm thereon, converts the vibration data into a sound spectrum diagram, and then analyzes the 3D vibration voiceprint characteristics through MFFC and DFT algorithms. Then, through the DL deep learning algorithm, it judges the occurrence of abnormal vibration in any direction of the equipment and gives an alarm prompt.

[0041] Further, the interface of the vibration monitoring sensor adopts a standard USB audio transmission protocol, and does not need to install special driver software.

[0042] Further, the system is transplanted from a remote monitoring terminal to a vibration monitoring sensor, or to a server background with higher computing power than the system; the vibration monitoring sensor locally performs voiceprint feature value analysis, and directly alarms an abnormal situation exceeding a threshold value to the server background; in an application scenario requiring real-time monitoring and recording of a running state, a remote monitoring terminal interface acquires vibration real-time curves of multiple vibration monitoring sensors, and performs voiceprint feature value algorithm analysis, saves obtained data locally, and transmits the data to a server background database through an industrial Ethernet.

[0043] Further, in the vibration monitoring sensor, an MCU uses an interpolation algorithm on a sampling value of a low-speed single-axis or multi-axis MEMS sensor to fit a 48k-192k / s audio signal output.

[0044] Further, if only one-axis vibration change needs to be monitored, the sampling value uses an interpolation algorithm to simplify to one axis.

[0045] Further, the sampling value uses an interpolation algorithm to extract a voiceprint feature, analyze background noise, and perform filtering and self-calibration.

[0046] Further, a digital audio signal output format is 48k-192k / s.

[0047] Further, the system also supports interface signal expansion including Ethernet / WIFI / LoRa / Bluetooth / LVDS / CAN or RS-485.

[0048] Further, the system provides installation methods including magnetic attraction, screws, and cable ties.

[0049] The new intelligent vibration voiceprint sensor perception system can also include a processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps in the above-mentioned new intelligent vibration voiceprint sensor perception system embodiments. The new intelligent vibration voiceprint sensor perception system can run in desktop computers, notebook computers, palm computers, and cloud data centers, and the executable system can include, but is not limited to, a processor, a memory, a server cluster.

[0050] Application scenarios of system architecture

[0051] According to different field application conditions, the application proposes, including but not limited to, the following four system construction modes of the new intelligent vibration soundprint sensor perception system, namely four embodiments:

[0052] Embodiment 1. Low-cost vibration sensor small-scale installation

[0053] Embodiment 2. Low-cost vibration sensor medium-scale installation

[0054] Embodiment 3. Large-scale long-distance installation of vibration sensors

[0055] Embodiment 4. Large-scale super-long distance, low-power consumption installation of vibration sensors

[0056] Therefore, the four embodiments are as follows:

[0057] 1. Low-cost vibration sensor small-scale installation

[0058] As shown in Figure 1 , the system architecture is suitable for 24-hour real-time detection of the operation of the equipment, and the detected soundprint curve is also stored in real time. The core soundprint extraction algorithm is transplanted and run on the remote monitoring terminal at this time, and the vibration monitoring sensor is only responsible for real-time collection of the vibration amplitude of the measured equipment, and is transmitted to the remote monitoring terminal at high speed through the USB audio protocol. The remote monitoring terminal extracts the real-time vibration soundprint and performs comparison. At the same time, the real-time collected soundprint data is transmitted to the server background storage analysis through the Ethernet industrial Internet protocol. Each remote monitoring terminal can hang 6-8 vibration monitoring sensors.

[0059] 2. Low-cost vibration sensor medium-scale installation

[0060] As shown in Figure 2 , the system architecture is suitable for 24-hour real-time detection of the operation of the equipment, and the detected soundprint curve is also stored in real time. However, it is not suitable for wired installation or transmission distance exceeding 10 meters or more. The core soundprint extraction algorithm is transplanted and run on the remote monitoring terminal at this time, and the vibration monitoring sensor is only responsible for real-time collection of the vibration amplitude of the measured equipment, and is transmitted to the remote monitoring terminal at high speed through the WIFI Mesh audio protocol. The remote monitoring terminal extracts the real-time vibration soundprint and performs comparison. At the same time, the real-time collected soundprint data is transmitted to the server background storage analysis through the Ethernet industrial Internet protocol. Each remote monitoring terminal can hang up to 32 vibration monitoring sensors.

[0061] 3. Large-scale long-distance installation of vibration sensors

[0062] As shown in Figure 3As shown, this system architecture is suitable for applications that do not require real-time monitoring of device operation. The core acoustic signature extraction algorithm is ported and runs on a vibration monitoring sensor. This sensor is responsible for real-time acquisition of the vibration amplitude of the device under test, calculating and extracting the real-time vibration acoustic signature, and comparing it. If the threshold is exceeded, the alarm information is transmitted at high speed to the PC server backend via Wi-Fi Mesh for recording and analysis. The vibration sensor can transmit over a distance exceeding 30,000 meters, and a single Wi-Fi Mesh router can connect up to 1,000 vibration monitoring sensors. Adding more Wi-Fi Mesh routers allows for the addition of even more vibration monitoring sensors, the exact number limited by server processing power and Ethernet bandwidth.

[0063] 4. Large-scale, ultra-long-distance, low-power installation of vibration sensors

[0064] like Figure 4 As shown, this system architecture is similar to the third type mentioned above. The difference is that it uses LoRa Mesh to transmit alarm information to a PC server for recording and analysis. Vibration sensors not only have power consumption as low as microamps (µA), but also have transmission distances exceeding tens or even hundreds of kilometers. A single LoRa Mesh router can connect to over 1000 vibration monitoring sensors. Adding more LoRa Mesh routers allows for an even greater number of vibration monitoring sensors, the exact number being limited by server computing power and Ethernet bandwidth.

[0065] Based on the above four embodiments, the novel intelligent vibration and acoustic fingerprint sensor system can operate in computing devices such as desktop computers, laptops, handheld computers, and cloud data centers. The novel intelligent vibration and acoustic fingerprint sensor system includes, but is not limited to, a processor and a memory. Those skilled in the art will understand that the examples described are merely illustrations of a novel intelligent vibration and acoustic fingerprint sensor system and do not constitute a limitation on the system. It may include more or fewer components, or combine certain components, or different components. For example, the novel intelligent vibration and acoustic fingerprint sensor system may also include input / output devices, network access devices, buses, etc.

[0066] The processor can be a central processing unit (CPU), and can also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or can also be any conventional processor, etc. The processor is a control center of the new intelligent vibration voiceprint sensor perception system, and is connected to various sub-regions of the new intelligent vibration voiceprint sensor perception system through various interfaces and lines.

[0067] The memory can be used to store the computer program and / or modules, and the processor realizes various functions of the new intelligent vibration voiceprint sensor perception system by running or executing the computer program and / or modules stored in the memory, and calling the data stored in the memory. The memory can mainly include a program storage area and a data storage area. The program storage area can store an operating system, at least one application program required for a function (such as a sound playing function, an image playing function, etc.), etc.; and the data storage area can store data created according to the use of the mobile phone (such as audio data, a phone book, etc.), etc. In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, for example, a hard disk, a memory, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0068] The present disclosure provides a new intelligent vibration voiceprint sensor perception system, which is used in mobile or fixed devices, cooperates with a cloud server to detect the vibration, displacement and angle change of a measured object or device, converts vibration voiceprint characteristic values into digital audio signals through extraction, and sends the digital audio signals to a background for analysis and processing; or performs characteristic value analysis locally, and outputs a warning for abnormal conditions exceeding a threshold value, so as to achieve the effect of predicting abnormal operating conditions, and can be adapted to various interfaces, has a small size and a high cost performance.

[0069] While the description of the present disclosure has been quite extensive and particular in describing several described embodiments, it is intended that the scope of the present disclosure effectively encompass any such embodiments falling within the predetermined scope of the present disclosure. Moreover, the foregoing description has been presented for the purpose of providing a detailed description of the present disclosure, as well as for the purpose of enabling one skilled in the art to make and use the present disclosure, and it is intended that the description be taken as a whole, not as separate embodiments.

Claims

1. An intelligent vibratory voiceprint sensor perception system, characterized in that, The system is used in mobile or fixed equipment, cooperates with a cloud server to detect vibration, displacement and angle change of a measured object or equipment, extracts vibration voiceprint characteristic values, converts into digital audio signals, sends to a background for analysis and processing, or performs local characteristic value analysis, and outputs an early warning for an abnormal situation exceeding a threshold value. The system is composed of a vibration monitoring sensor and a remote acquisition terminal. The vibration monitoring sensor includes a high-speed communication circuit, a high-speed acceleration and angular velocity acquisition circuit, and the high-speed communication circuit includes an MCU / FPGA. Specifically, in a continuous sampling time period, the acceleration and angular velocity values returned by the high-speed acquisition sensor are sampled at multiple different time points as sampling time points, the number of sampling time points is denoted as n, and the serial number of each sampling time point is denoted as i, i∈[1,n]. Corresponding to the sampling time of the serial number i, the acceleration returned by the high-speed acquisition sensor is recorded as u i , and the angular velocity value is g i , so the 3D vibration characteristic value is Fi; The calculation method of Fi is as follows: First, calculate the speed primer rGu: ， The speed primer is an effective and guiding value generated in the calculation of the 3D vibration speed value, which is beneficial to collect the data characteristics of the acceleration and angular velocity values returned by the high-speed acquisition sensor in the distribution of the sampling time, and this collection is fast and non-destructive to the equipment. Then, calculate the 3D vibration speed value Fi corresponding to each sampling time point, ， Here, to calculate the Fi corresponding to the sampling time point with serial number i, it is necessary to calculate after recording the data of each time point in the n sampling time points, which corresponds to In rGu, ∫ and di represent the integral of the function f (t) in the sampling time interval [t i-1, t i], and di represents the integral of the function f (t) in the sampling time interval [t i-1, t i] when it is not integrable. In the n sampling time points in the continuous sampling time period, the integral is performed, and when it is not integrable, it represents The result value of the cumulative summation at the sampling time point with serial number i; when i is 1, F(i-1) is 1 / π, and π represents the value of the circular constant. Calculate the arithmetic mean of the 3D vibration speed values Fi corresponding to each sampling time point as the vibration sub, obtain the maximum value of the 3D vibration speed values Fi corresponding to each sampling time point as the vibration base, calculate the ratio of the vibration sub to the vibration base as the vibration base sub, and calculate the number of Fi values greater than or equal to the vibration sub in the 3D vibration speed values Fi corresponding to each sampling time point as trop, calculate the quotient of trop divided by n as the exceeding frequency, and judge whether the exceeding frequency is greater than the vibration base sub in value, if yes, it is the abnormal situation, and an early warning is output to the database.

2. The intelligent vibratory voiceprint sensor perception system of claim 1, wherein, The system includes a vibration monitoring sensor and a remote acquisition terminal.

3. The intelligent vibratory voiceprint sensor perception system of claim 2, wherein, The system uses a 24bit or 16bit digital audio signal with a speed of not less than 48k / s, transmits the signal to a remote monitoring terminal through an interface, the remote monitoring terminal runs on an industrial computer based on multiple operating systems, converts the vibration data into a sound spectrum graph, and then analyzes the 3D vibration voiceprint characteristics through MFFC and DFT algorithms; then, through DL deep learning algorithm, the system judges the occurrence of abnormal vibration in any direction of the equipment and gives an alarm prompt.

4. The intelligent vibratory acoustic print sensor perception system of claim 2, wherein, The interface of the vibration monitoring sensor adopts a standard USB audio transmission protocol.

5. The intelligent vibratory acoustic print sensor perception system of claim 2, wherein, The system is transplanted from a remote monitoring terminal to a vibration monitoring sensor, or to a server background with higher computing power than the system; the vibration monitoring sensor analyzes the voiceprint characteristic value locally, and directly alarms and outputs to the server background for abnormal conditions exceeding the threshold; in application scenarios requiring real-time monitoring and recording of operating status, the remote monitoring terminal interface obtains the vibration real-time curve of multiple vibration monitoring sensors and performs voiceprint characteristic value algorithm analysis, saves the obtained data locally, and transmits it to the server background database through industrial Ethernet.

6. The intelligent vibratory acoustic print sensor perception system of claim 2, wherein, In the vibration monitoring sensor, the MCU uses an interpolation algorithm on the sampling value of a low-speed single-axis or multi-axis MEMS sensor to fit a 48k-192k / s audio signal output.

7. The intelligent vibratory acoustic print sensor perception system of claim 6, wherein, If only one-axis vibration change needs to be monitored, the sampling value interpolation algorithm is simplified to one axis.

8. The intelligent vibratory acoustic print sensor perception system of claim 6, wherein, The sampling value interpolation algorithm can extract voiceprint features, analyze background noise, filter, and self-calibrate; The system also supports interface signal expansion including Ethernet / WIFI / LoRa / Bluetooth / LVDS / CAN or RS-485; The system provides installation methods including magnetic attraction, screws, and cable ties.

9. The intelligent vibratory acoustic print sensor perception system of claim 1 or 3, wherein, The digital audio signal output format is 48k-192k / s.

Images