A method and device for diagnosing vibration faults of ship equipment based on compressed sensing
By applying compressed sensing technology in ship equipment vibration and noise monitoring, sparse transformation and coding generate observation vectors and reconstruct vibration signals, solving the problem of huge data transmission volume caused by high-frequency vibration signals, achieving a significant reduction in data transmission volume and system stable operation.
Patent Information
- Application Number
- CN202211526258.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-12-01
- Publication Date
- 2025-01-21
- Estimated Expiration
- 2042-12-01
AI Technical Summary
In the monitoring of vibration noise of ship equipment, high-frequency vibration signals lead to huge data transmission, which is difficult to effectively transmit in the local area network of ship systems, making it difficult to realize real-time monitoring of vibration of real ship equipment.
Using a method based on compressed sensing, through sparse transformation and coding, an observation vector that is much smaller than the original signal dimension is generated, and the host computer is used for reconstruction processing to obtain the vibration reconstructed signal and reduce the amount of data transmission.
It greatly reduces the transmission volume of the original vibration signal. The amount of transmitted data is only about 10% of the original data volume. The reconstructed signal is highly similar to the original signal, which relieves the pressure of data transmission and realizes the long-term transmission and operation of large-capacity data. stability.
Smart Images
Figure CN115791241B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of ship vibration and noise monitoring, and in particular to a ship equipment vibration fault diagnosis method and device based on compressed sensing. Background Art
[0002] Ship vibration and noise control runs through the entire life cycle of scheme design, technical design, construction, sea trial delivery and even use. Due to the influence of many unavoidable factors, the equipment may have various faults, which may reduce or lose its intended function, resulting in abnormal vibration and noise. The research on vibration and noise monitoring and equipment fault diagnosis technology at home and abroad has a long history. With the development of computer technology and sensor technology, the vibration and noise testing technology of real ships has been very mature, but the long-term monitoring technology of vibration and noise applied to real ships is rarely reported. There are many kinds of failures in marine equipment. A lot of research has been conducted on the failure mechanism and fault diagnosis methods of different equipment at home and abroad, but mature cases applied to real ships are also relatively rare. One of the important reasons is the problem of data transmission. The characteristic frequency of equipment vibration and noise is usually high, so an extremely high frequency sampling rate is required to fully cover it. Long-term monitoring will generate extremely large data capacity, and the local area network of the ship system is difficult to withstand such a large amount of data transmission. Therefore, solving the transmission of large-capacity data is the primary condition for realizing real-time monitoring of vibration of real ship equipment. Summary of the invention
[0003] In view of the above problems and technical needs, the inventors have proposed a method and device for diagnosing vibration faults of ship equipment based on compressed sensing. The method utilizes compressed sensing technology. For sparse signals, linear projection can be used to obtain an observation vector that is much smaller than the dimension of the original signal. The original signal can be recovered using the observation vector and reconstruction algorithm, thus solving the problem of large-capacity data transmission during real-time monitoring of vibration of actual ship equipment.
[0004] The technical solution of the present invention is as follows:
[0005] In a first aspect, the present application provides a method for diagnosing vibration faults of ship equipment based on compressed sensing, comprising the following steps:
[0006] The lower computer determines the compressed sensing data packet corresponding to the original vibration signal with a dimension of N and sends it to the upper computer; the compressed sensing data packet includes the observation vector dimension M, the observation vector y and the position information p of its vector elements, where M<<N;
[0007] The host computer reconstructs the observation vector y according to the dimension N and the vector element position information p to obtain a vibration reconstructed signal;
[0008] The host computer determines whether the vibration reconstruction signal is an abnormal vibration signal. If it is an abnormal vibration signal, the name of the device generating the abnormal vibration and the fault type are determined.
[0009] A further technical solution is that a method for determining a compressed sensing data packet corresponding to an original vibration signal having a dimension of N by a lower computer includes: in the lower computer:
[0010] Perform sparse transformation on the original vibration signal with dimension N to obtain the sparse expansion coefficient vector θ i ,θ i ∈R^(N×1);
[0011] According to the expansion coefficient vector θ i Encode the size and position of
[0012] Extract the required coding expansion coefficient vector θ′ according to the set error threshold ε i Composed of observation vector y, y∈R^(M×1), the encoding expansion coefficient vector θ′ is extracted i The number of observation vector dimensions M is used to encode the expansion coefficient vector θ′ i The corresponding encoding vector element position information p′ i The position information p of the observation vector elements constitutes the compressed sensing data packet (M, y, p).
[0013] A further technical solution is that the host computer reconstructs the observation vector y according to the dimension N and the vector element position information p, and the method for obtaining the vibration reconstruction signal includes: in the host computer:
[0014] Construct an N-dimensional zero vector according to a preset dimension N;
[0015] Read the observation vector y and its vector element position information p in the compressed sensing data packet, and replace the corresponding elements in the observation vector y with the corresponding positions of the N-dimensional zero vector in turn according to the position information of each element in the vector element position information p, and obtain the reconstructed expansion coefficient vector
[0016] Reconstruct the expansion coefficient vector Perform inverse sparse transform to obtain vibration reconstruction signal
[0017] A further technical solution is that the method for performing sparse transformation on the original vibration signal with dimension N includes:
[0018] The received original vibration signal is filtered and converted into digital form within the scope of the Nyquist sampling theorem according to the set sampling frequency to obtain a complete vibration acceleration digital signal x, where x∈R^(N×1), where N is the preset dimension of the original vibration signal;
[0019] Discrete cosine transform is selected as the sparse dictionary to perform sparse decomposition on the vibration acceleration digital signal x.
[0020] Its further technical solution is to use the expansion coefficient vector θ i Methods for encoding the size and position include:
[0021] The N-dimensional sparse expansion coefficient vector θ i Arrange them in order according to the numerical value to obtain the N-dimensional coding expansion coefficient vector θ i ′,θ i ′∈R^(N×1); and expand each coding coefficient vector θ i ′ in the sparse expansion coefficient vector θ i The position information corresponds to the position information p of the N-dimensional encoding vector element in the order of arrangement. i ′,p i ′∈N^(N×1); the order is either descending or ascending.
[0022] A further technical solution is to extract the required coding expansion coefficient vector θ according to the set error threshold ε i ′ The method of forming the observation vector y includes:
[0023] Calculate the N-dimensional encoding expansion coefficient vector θ i The total value of ′,
[0024] Initialize the number of iterations m = 1, the total iteration value T = 0;
[0025] Calculate Τ = Τ + (θ m ′) 2 ;
[0026] like Then let m=m+1, and re-execute the step of calculating the total iteration value T;
[0027] like Then the output M=m, y=[θ1′,θ2′,…,θ′ M ] T .
[0028] A further technical solution is that the host computer determines whether the vibration reconstruction signal is an abnormal vibration signal. If it is an abnormal vibration signal, the method for determining the name of the device generating the abnormal vibration and the fault type includes: in the host computer:
[0029] The vibration reconstruction signal is compared with the pre-established early warning model of the standard vibration signal. If it is determined that the vibration reconstruction signal is an abnormal vibration signal, the vibration reconstruction signal is input into the pre-established fault library for comparison and analysis to obtain the name of the equipment generating the abnormal vibration and the fault type; the pre-established fault library contains the fault vibration characteristic data and fault types corresponding to various types of mechanical equipment.
[0030] In the second aspect, the present application also provides a ship equipment vibration fault diagnosis device based on compressed sensing, including multiple vibration sensors, multiple compressed sensing acquisition terminals, a switch and a server terminal; the vibration sensor is placed on the equipment to be monitored, and is used to collect original vibration signals; the output end of the vibration sensor is connected to the corresponding input end of the compressed sensing acquisition terminal, and the output end of the compressed sensing acquisition terminal is connected to the server terminal through the switch; the compressed sensing acquisition terminal serves as a lower computer and the server terminal serves as a higher computer, and when processing the original vibration signal, the steps of the method described in the first aspect are implemented.
[0031] Its further technical solution is that the server terminal includes a lower computer interaction module, a database interaction module and a human-computer interaction module;
[0032] The lower computer interaction module includes a lower computer communication interface and a data reconstruction and analysis unit; the lower computer communication interface is used to provide an Ethernet connection interface for receiving a compressed sensing data packet uploaded by a compressed sensing acquisition terminal; the data reconstruction and analysis unit is used to reconstruct the observation vector y in the compressed sensing data packet to obtain a vibration reconstruction signal;
[0033] The database interaction module includes a standard library and a fault library. The standard library is used to store early warning models containing threshold information of various types of equipment and standard vibration signals of equipment under normal conditions. The fault library is used to store fault vibration characteristic data and fault types corresponding to various types of mechanical equipment.
[0034] The human-computer interaction module includes a fault analysis and display unit and a hardware control unit. The fault analysis and display unit is used to compare the vibration reconstruction signal output by the data reconstruction and analysis unit with the early warning model in the standard library. If the vibration reconstruction signal is determined to be an abnormal vibration signal, the vibration reconstruction signal is input into the fault library for comparison and analysis to obtain the name of the device generating the abnormal vibration and the type of fault, which are displayed on the server terminal interface. The hardware control unit is used to adjust the sampling rate and sampling method of the compressed sensing acquisition terminal.
[0035] Its further technical solution is that the compressed sensing acquisition terminal includes an FPGA chip and an encoder chip. The FPGA chip is used for conditioning and Nyquist sampling of the original vibration signal to obtain a complete vibration acceleration digital signal x and transmit it to the encoder chip; the encoder chip is used to perform sparse transformation, encoding, and compression processing on the vibration acceleration digital signal x to obtain a vibration reconstruction signal and transmit it to the server terminal.
[0036] The beneficial technical effects of the present invention are:
[0037] The method proposed in this application uses compressed sensing technology to greatly reduce the transmission volume of the original vibration signal. The data volume of the compressed sensing data packet transmitted between the lower computer and the upper computer is only about 10% of the data volume of the original vibration signal, and the vibration reconstruction signal has a high similarity with the original vibration signal. On the basis of meeting the needs of vibration monitoring and fault diagnosis, the data transmission pressure between the lower computer and the upper computer is greatly alleviated, thereby achieving the stability of long-term transmission and operation of large-capacity data; through the pre-established early warning model and fault library, when an abnormal vibration signal appears, the equipment and cause of the possible failure are judged according to different characteristics. The device proposed in this application adopts a distributed compressed sensing acquisition terminal and a centralized server terminal, which can be connected to a whole through the Ethernet on the ship, or used through an independent network system, and is simple and convenient to install and use; the database interaction module proposed in this application can upload and update data at any time to continuously improve the fault diagnosis device; the server terminal proposed in this application can remotely complete system configuration through the Internet, and can also remotely analyze and download monitoring data. BRIEF DESCRIPTION OF THE DRAWINGS
[0038] Figure 1 This is a flow chart of the ship equipment vibration fault diagnosis method based on compressed sensing provided in this application.
[0039] Figure 2 It is a principle block diagram of signal compression acquisition and reconstruction provided by this application.
[0040] Figure 3 It is a signal compression flow chart provided by this application.
[0041] Figure 4 It is a comparison diagram of the original vibration signal and the vibration reconstruction signal provided by this application.
[0042] Figure 5 It is a schematic diagram of a ship equipment vibration fault diagnosis device based on compressed sensing provided in this application.
[0043] Figure 6 It is a schematic diagram of the software logic relationship of the server terminal provided by this application. DETAILED DESCRIPTION
[0044] The specific implementation of the present invention will be further described below in conjunction with the accompanying drawings.
[0045] In order to break through the bottleneck of massive data transmission in the process of vibration signal monitoring and fault diagnosis, this application provides a ship equipment vibration fault diagnosis method based on compressed sensing, targeting the sparse characteristics of vibration signals in the frequency space. Figure 1 As shown, the following steps are included:
[0046] Step 1: The lower computer determines the compressed sensing data packet corresponding to the original vibration signal with dimension N and sends it to the upper computer.
[0047] Step 1.1: In the lower computer, perform sparse transformation on the original vibration signal with dimension N to obtain the sparse expansion coefficient vector θ i ,θ i ∈R^(N×1), including:
[0048] like Figure 2 As shown, the original vibration signal output by the vibration sensor is an analog signal. The lower computer filters the received original vibration signal through an anti-aliasing filter, and performs analog-to-digital conversion within the scope of the Nyquist sampling theorem according to the set sampling frequency to obtain a complete vibration acceleration digital signal x, x∈R^(N×1), N is the preset dimension of the original vibration signal, that is, the set length of the original vibration signal. Since the vibration signal is not sparse in the time domain, in order to achieve sparse representation of the signal, a sparse transformation must first be performed in a specific space. This embodiment selects a discrete cosine transform (DCT) with good universality and easy implementation as a sparse dictionary, performs sparse decomposition on the vibration acceleration digital signal x, and obtains a sparse expansion coefficient vector θ i ,θ i ∈R^(N×1).
[0049] Step 1.2: According to the expansion coefficient vector θ i Encodes the size and position of the
[0050] The N-dimensional sparse expansion coefficient vector θ i Arrange them in order according to the numerical value to obtain the N-dimensional coding expansion coefficient vector θ i ′,θ i ′∈R^(N×1). And each coding expansion coefficient vector θ i ′ in the sparse expansion coefficient vector θ i The position information corresponds to the position information p of the N-dimensional encoding vector element in the order of arrangement. i ′,p i ′∈N^(N×1).
[0051] The sequential arrangement may be descending arrangement or ascending arrangement. In this embodiment, descending arrangement is preferred, which is also convenient for subsequent compression processing.
[0052] Step 1.3: Extract the required coding expansion coefficient vector θ according to the set error threshold ε i ′ forms the observation vector y, y∈R^(M×1).
[0053] like Figure 3 As shown, calculate the N-dimensional encoding expansion coefficient vector θ i The total value of ′,
[0054] Initialize the number of iterations m = 1, the total iteration value T = 0;
[0055] Calculate Τ = Τ + (θ m ′) 2 ;
[0056] like Then let m=m+1, and re-execute the step of calculating the total iteration value T;
[0057] like Then the output M=m, y=[θ′1,θ′2,…,θ′ M [ T , p=[p′1,p′2,…,p′ M ] T .
[0058] Step 1.4: Finally, the encoding expansion coefficient vector θ′ is extracted i The number of observation vector dimensions M is used to encode the expansion coefficient vector θ′ i The corresponding encoding vector element position information p i ′ constitutes the observation vector element position information p, and together with the observation vector y constitutes the compressed sensing data packet (M, y, p), where M<<N.
[0059] Step 2: The host computer reconstructs the observation vector y according to the dimension N and the vector element position information p to obtain a vibration reconstructed signal.
[0060] Step 2.1: Figure 2 As shown, in the host computer, an N-dimensional zero vector is constructed according to a preset dimension N.
[0061] Step 2.2: Read the observation vector y and its vector element position information p in the compressed sensing data packet, and replace the corresponding elements in the observation vector y with the corresponding positions of the N-dimensional zero vector in turn according to the position information of each element in the vector element position information p, and obtain the reconstructed expansion coefficient vector
[0062] Step 2.3: Reconstruct the expansion coefficient vector Perform inverse sparse transform, that is, inverse discrete cosine transform (IDCT), to obtain the vibration reconstruction signal Vibration reconstruction signal The comparison chart with the vibration acceleration digital signal x is as follows Figure 4 As shown in the figure, it can be seen that the trend of the vibration signal reconstructed by compressed sensing technology is basically consistent with that of the original vibration signal.
[0063] Step 3: The host computer determines whether the vibration reconstruction signal is an abnormal vibration signal. If it is an abnormal vibration signal, the name of the device generating the abnormal vibration and the fault type are determined.
[0064] In the host computer, the vibration reconstructed signal The reconstructed vibration signal is compared with the pre-established early warning model of the standard vibration signal. If it is determined that the reconstructed vibration signal is an abnormal vibration signal, the reconstructed vibration signal is input into the pre-established fault library for comparison and analysis to obtain the name of the equipment that generates the abnormal vibration and the type of fault. The pre-established fault library contains the fault vibration characteristic data and fault types corresponding to various types of mechanical equipment.
[0065] In this embodiment, according to the sparseness of vibration signals in the frequency domain, under the compressed sensing framework, a technical route of first collecting and then compressing is adopted to reduce the amount of redundant data between the lower computer and the upper computer, alleviate the transmission pressure of massive information in the actual ship vibration monitoring and equipment fault diagnosis system, and provide support and guarantee for the realization of system functions and long-term stable operation.
[0066] Based on the same inventive concept, this embodiment also provides a ship equipment vibration fault diagnosis device based on compressed sensing, such as Figure 5 As shown, the device includes multiple vibration sensors, multiple compressed sensing acquisition terminals, switches and server terminals, wherein the compressed sensing acquisition terminals and corresponding vibration sensors can be increased or decreased according to monitoring needs. The vibration sensor is placed on the equipment to be monitored to collect the original vibration signal, and the compressed sensing acquisition terminal is arranged near the monitoring equipment; the output end of the vibration sensor is connected to the corresponding compressed sensing acquisition terminal input end, and the output end of the compressed sensing acquisition terminal is connected to the server terminal through the switch. The compressed sensing acquisition terminal serves as a lower computer and the server terminal serves as a higher computer. When processing the original vibration signal, the steps of the ship equipment vibration fault diagnosis method provided above are executed, which will not be repeated here.
[0067] The vibration sensor uses an IEPE type accelerometer with a measurement frequency range of 10Hz to 10kHz and an operating temperature range of -40℃ to +120℃. The shell material is 304 stainless steel, meeting the IP65 protection level.
[0068] The compressed sensing acquisition terminal includes an FPGA chip and an encoder chip. The FPGA chip is used for conditioning and Nyquist sampling of the original vibration signal, that is, the original vibration signal is converted into an A / D signal after passing through the signal conditioning circuit (including filtering processing), and a complete vibration acceleration digital signal x is obtained and transmitted to the encoder chip. The encoder chip can be connected to 4 sensor channels and has a power-on self-start function. It is used to perform sparse transformation, encoding, and compression processing on the vibration acceleration digital signal x, obtain a vibration reconstruction signal, and transmit it to the server terminal.
[0069] The switch is used for Ethernet connection between compression sensing acquisition terminals and server terminals, and the interface speed is generally 1000M.
[0070] The server terminal is a high-performance server with the capabilities of data analysis and reconstruction, monitoring signal display and fault diagnosis. It is generally located in the ship's main control center and can directly view monitoring data and perform system configuration. Figure 6 As shown, the software system of the server terminal includes a lower computer interaction module, a database interaction module and a human-computer interaction module. <1> The lower computer interaction module includes a lower computer communication interface, a data reconstruction and analysis unit and a data storage unit. The lower computer communication interface is used to provide an Ethernet connection interface for receiving compressed sensing data packets uploaded by the compressed sensing acquisition terminal. The data reconstruction and analysis unit is used to reconstruct the observation vector y in the compressed sensing data packet and process the obtained vibration reconstruction signal into vibration feature data according to the vibration characteristics of the equipment. <2> The database interaction module includes a standard library and a fault library. The standard library is used to store early warning models containing threshold information of various types of equipment and standard vibration signals of equipment in normal state, and also stores vibration characteristic data of mechanical equipment under normal operation under real-time monitoring. The fault library is used to store fault vibration characteristic data and fault types corresponding to various types of mechanical equipment, and also stores vibration characteristic data of equipment in fault state during real-time monitoring. <3> The human-computer interaction module is mainly used for hardware control, fault analysis and display of the compressed sensing acquisition terminal, including a fault analysis and display unit and a hardware control unit. The fault analysis and display unit is used to compare the vibration characteristic data output by the data reconstruction and analysis unit with the early warning model in the standard library. If the vibration characteristic data is determined to be an abnormal vibration signal, the vibration characteristic data is input into the fault library for comparison and analysis to obtain the name of the device that generates the abnormal vibration and the type of fault, and the diagnosis results and real-time monitoring signals are displayed on the server terminal interface, where the monitoring signal display is a graphical interface to display the vibration information of all measuring points. The hardware control unit is used to adjust the sampling rate, sampling mode and other parameter settings of the compressed sensing acquisition terminal.
[0071] Optionally, the software system of the server terminal also has a system configuration function for performing corresponding configuration on the compressed sensing acquisition terminal, the standard library and the fault library, and the server terminal.
[0072] In this embodiment, the device is based on a distributed compressed sensing acquisition terminal and a centralized server terminal, which are connected through Ethernet to form a whole, and can automatically establish a vibration signal history database for monitoring the vibration of ship machinery and equipment. It can grasp the vibration status of various types of ship machinery and equipment at any time, analyze abnormal vibration signals in real time, and combine with the database interaction module to locate faulty equipment in time and determine the fault type to ensure the safety of the ship.
[0073] The above is only a preferred embodiment of the present application, and the present invention is not limited to the above embodiments. It is understood that other improvements and changes directly derived or associated by those skilled in the art without departing from the spirit and concept of the present invention should be considered to be included in the protection scope of the present invention.
Claims
1. A ship equipment vibration fault diagnosis method based on compressed sensing, characterized in that: The method comprises: The lower computer determines a compressed sensing data packet corresponding to the original vibration signal of dimension N and sends it to the upper computer; the compressed sensing data packet includes the observation vector dimension M, the observation vector y and the observation vector element position information p, where M<<N; The host computer reconstructs the observation vector y according to the dimension N and the position information p of the observation vector element to obtain a vibration reconstructed signal; The host computer determines whether the vibration reconstruction signal is an abnormal vibration signal, and if it is an abnormal vibration signal, determines the name of the device generating the abnormal vibration and the fault type; The method for determining the compressed sensing data packet corresponding to the original vibration signal with a dimension of N by the lower computer includes: in the lower computer: Perform sparse transformation on the original vibration signal with dimension N to obtain the sparse expansion coefficient vector θ i ,θ i ∈R^(N×1); According to the sparse expansion coefficient vector θ i Encode the size and position of Extract the required coding expansion coefficient vector θ according to the set error threshold ε i ′ forms the observation vector y, y∈R^(M×1), and extracts the coding expansion coefficient vector θ i ′ as the observation vector dimension M, and the encoding expansion coefficient vector θ i ′ corresponds to the encoding vector element position information p i ′ constitutes the position information p of the observation vector element, forming a compressed sensing data packet (M, y, p); The required coding expansion coefficient vector θ is extracted according to the set error threshold ε. i ′ The method of forming the observation vector y includes: Calculate the N-dimensional encoding expansion coefficient vector θ i The total value of ′ is T0, Initialize the number of iterations m = 1, the total iteration value T = 0; Calculate T=T+(θ m ′) 2 ; like Then let m=m+1, and re-execute the step of calculating the total iteration value T; like Then the output M=m, y=[θ1′,θ2′,...,θ′ M ] T .
2. The ship equipment vibration fault diagnosis method based on compressed sensing according to claim 1 is characterized in that: The method of obtaining a vibration reconstruction signal by reconstructing the observation vector y according to the dimension N and the observation vector element position information p by the host computer includes: Construct an N-dimensional zero vector according to a preset dimension N; Read the observation vector y and the observation vector element position information p in the compressed sensing data packet, and replace the corresponding elements in the observation vector y with the corresponding positions of the N-dimensional zero vector in turn according to the position information of each element in the observation vector element position information p, and obtain the reconstructed expansion coefficient vector Reconstruct the expansion coefficient vector Perform inverse sparse transform to obtain vibration reconstruction signal 3. The ship equipment vibration fault diagnosis method based on compressed sensing according to claim 1 is characterized in that: The method for performing sparse transformation on an original vibration signal having a dimension of N comprises: The received original vibration signal is filtered and converted into digital form within the scope of the Nyquist sampling theorem according to the set sampling frequency to obtain a complete vibration acceleration digital signal x, where x∈R^(N×1), where N is the preset dimension of the original vibration signal; Discrete cosine transform is selected as the sparse dictionary to perform sparse decomposition on the vibration acceleration digital signal x.
4. The ship equipment vibration fault diagnosis method based on compressed sensing according to claim 1 is characterized in that: According to the sparse expansion coefficient vector θ i Methods for encoding the size and position include: The N-dimensional sparse expansion coefficient vector θ i Arrange them in order according to the numerical value to obtain the N-dimensional coding expansion coefficient vector θ i ′,θ i ′∈R^(N×1); and expand each coding coefficient vector θ i ′ in the sparse expansion coefficient vector θ i The position information corresponds to the position information p of the N-dimensional encoding vector element in the order of arrangement. i ′,p i ′∈N^(N×1); wherein the order is descending or ascending.
5. The ship equipment vibration fault diagnosis method based on compressed sensing according to claim 1, characterized in that: The method of determining by the host computer whether the vibration reconstruction signal is an abnormal vibration signal, and if it is an abnormal vibration signal, determining the name of the device generating the abnormal vibration and the fault type includes, in the host computer: The vibration reconstruction signal is compared with a pre-established early warning model of a standard vibration signal. If it is determined that the vibration reconstruction signal is an abnormal vibration signal, the vibration reconstruction signal is input into a pre-established fault library for comparison and analysis to obtain the name of the equipment generating the abnormal vibration and the type of fault; the pre-established fault library contains fault vibration characteristic data and fault types corresponding to various types of mechanical equipment.
6. A ship equipment vibration fault diagnosis device based on compressed sensing, characterized in that: The device includes multiple vibration sensors, multiple compressed sensing acquisition terminals, a switch and a server terminal; the vibration sensor is placed on the equipment to be monitored and is used to collect original vibration signals; the output end of the vibration sensor is connected to the input end of the corresponding compressed sensing acquisition terminal, and the output end of the compressed sensing acquisition terminal is connected to the server terminal through the switch; the compressed sensing acquisition terminal serves as a lower computer and the server terminal serves as a higher computer, and when processing the original vibration signal, the steps of the method described in any one of claims 1 to 5 are implemented.
7. The ship equipment vibration fault diagnosis device based on compressed sensing according to claim 6, characterized in that: The server terminal includes a lower computer interaction module, a database interaction module and a human-computer interaction module; The lower computer interaction module includes a lower computer communication interface and a data reconstruction and analysis unit; the lower computer communication interface is used to provide an Ethernet connection interface for receiving a compressed sensing data packet uploaded by a compressed sensing acquisition terminal; the data reconstruction and analysis unit is used to reconstruct the observation vector y in the compressed sensing data packet to obtain a vibration reconstruction signal; The database interaction module includes a standard library and a fault library. The standard library is used to store early warning models containing threshold information of various types of equipment and standard vibration signals of equipment in normal state. The fault library is used to store fault vibration characteristic data and fault types corresponding to various types of mechanical equipment. The human-computer interaction module includes a fault analysis and display unit and a hardware control unit. The fault analysis and display unit is used to compare the vibration reconstruction signal output by the data reconstruction and analysis unit with the early warning model in the standard library. If the vibration reconstruction signal is determined to be an abnormal vibration signal, the vibration reconstruction signal is input into the fault library for comparison and analysis to obtain the name of the device generating the abnormal vibration and the fault type, and display them on the server terminal interface. The hardware control unit is used to adjust the sampling rate and sampling mode of the compressed sensing acquisition terminal.
8. The ship equipment vibration fault diagnosis device based on compressed sensing according to claim 6, characterized in that: The compressed sensing acquisition terminal includes an FPGA chip and an encoder chip. The FPGA chip is used for conditioning and Nyquist sampling of the original vibration signal to obtain a complete vibration acceleration digital signal x and transmit it to the encoder chip; the encoder chip is used to perform sparse transformation, encoding, and compression processing on the vibration acceleration digital signal x to obtain a vibration reconstruction signal and transmit it to the server terminal.