A full-cycle acoustic emission monitoring method for bolt loosening based on variational mode decomposition

Abstract

The present invention discloses a full-cycle acoustic emission monitoring method for bolt loosening based on variational mode decomposition, which relates to the technical field of nondestructive testing and includes: collecting acoustic emission signals at the bolt connection interface under several levels of pre-tightening force; performing variational mode pre-decomposition on the acoustic emission signals, and determining the optimal decomposition layer number by using the decomposition mode energy difference rate; based on the determined optimal decomposition layer number, performing variational mode decomposition on all acoustic emission signals to obtain several intrinsic mode functions; by calculating the energy values of each intrinsic mode function, establishing the variation relationship between the energy values of each intrinsic mode function and different levels of pre-tightening force, and screening out the characteristic mode that is most sensitive to the change of pre-tightening force; establishing the mapping relationship between the energy of the screened characteristic mode and the pre-tightening force, obtaining the loosening change curve, and dividing the bolt loosening process into different stages according to the loosening change curve to realize the quantitative monitoring of the full cycle of bolt loosening. The present invention improves the safety and reliability of the bolt connection structure.

Description

Technical Field

[0001] The present invention belongs to the technical field of nondestructive testing, and particularly relates to a full-cycle acoustic emission monitoring method for bolt loosening based on variational mode decomposition. Background Art

[0002] Bolt connection is a widely used fastening method in mechanical structures, and is widely applied in fields such as aerospace, mechanical manufacturing, bridge construction, etc. The loosening of bolts will not only lead to a decrease in the strength of the connection structure, but may also cause serious safety hazards and even lead to catastrophic accidents. Therefore, it is particularly important to effectively monitor and warn of bolt loosening. Traditional bolt loosening detection methods mainly include torque detection, ultrasonic detection, stress wave detection, etc. However, these methods have certain limitations in practical applications, such as low detection sensitivity, complex operation, and difficulty in realizing online monitoring. In recent years, acoustic emission (AE) technology has become an important means for bolt loosening detection due to its high sensitivity and ability to monitor in real time. The acoustic emission technology can capture the acoustic emission signals generated during the processes of microcrack propagation and frictional slip inside the material, and can timely reflect the state of bolt loosening. However, acoustic emission signals usually have non-linear and non-stationary characteristics, and are greatly affected by noise and other external interferences. Traditional signal processing methods are difficult to accurately extract the characteristic information related to bolt loosening. Therefore, there is an urgent need to propose a full-cycle acoustic emission monitoring method for bolt loosening based on variational mode decomposition. Summary of the Invention

[0003] To solve the above technical problems, the present invention proposes a full-cycle acoustic emission monitoring method for bolt loosening based on variational mode decomposition. By performing variational mode decomposition on the acoustic emission signals, the characteristic modes reflecting the bolt loosening state are extracted, and real-time monitoring and warning of the whole process of bolt loosening are realized, thereby improving the safety and reliability of the bolt connection structure.

[0004] To achieve the above object, the present invention provides a full-cycle acoustic emission monitoring method for bolt loosening based on variational mode decomposition, including:

[0005] Collect acoustic emission signals at the bolt connection interface under several levels of pre-tightening force;

[0006] Perform variational mode pre-decomposition on the acoustic emission signals. Perform variational mode pre-decomposition according to different decomposition layers, calculate the decomposition mode energies corresponding to different decomposition layers, calculate the decomposition mode energy difference rate based on the decomposition mode energies corresponding to different decomposition layers, and use the decomposition mode energy difference rate to determine the optimal decomposition layer;

[0007] Based on the optimal decomposition layer, perform variational mode decomposition on all the collected acoustic emission signals to obtain a number of intrinsic mode functions;

[0008] By calculating the energy values of each intrinsic mode function, establishing the variation relationship between the energy values of each intrinsic mode function and different levels of pre-tightening force, and screening out the characteristic modes that are most sensitive to the change of pre-tightening force;

[0009] Establish the mapping relationship between the energy of the screened characteristic mode and the pre-tightening force, obtain the loosening change curve, and divide the bolt loosening process into different stages according to the loosening change curve to realize the quantitative monitoring of the full cycle of bolt loosening.

[0010] Optionally, the several levels of pre-tightening force satisfy that the number of levels is greater than or equal to 8.

[0011] Optionally, the energy in the decomposition mode energy difference rate is expressed as:

[0012]

[0013] Among them, E represents the energy of the signal, x(i) represents the collected signal sequence, n represents the number of sampling points, and i is the i-th sampling point in the signal sequence.

[0014] Optionally, the decomposition mode energy difference rate is expressed as:

[0015]

[0016] Among them, η represents the decomposition mode energy difference rate, E K represents the sum of the energies of all K components obtained by the decomposition layer number K, and E K-1 represents the sum of the energies of all K-1 components obtained by the decomposition layer number K-1.

[0017] Optionally, using the decomposition mode energy difference rate to determine the optimal decomposition layer number includes:

[0018] Step S1: Initialize the decomposition layer number, set the decomposition layer value, determine the search range of the decomposition layer number, set the step size, perform variational mode decomposition on the acoustic emission signal, and for the set decomposition layer value, calculate the energy of each intrinsic mode function obtained by the decomposition and calculate the sum of the energies of all intrinsic mode functions;

[0019] Step S2: Increase the decomposition layer number by 1 and continue with Step S1 until the maximum value of the search range, and then perform Step S3;

[0020] Step S3: Compare the decomposition mode energy difference rates corresponding to the decomposition layer values, and determine the optimal decomposition layer number according to the decomposition layer value corresponding to the maximum value of the decomposition mode energy difference rate.

[0021] Optionally, performing variational mode decomposition on all the collected acoustic emission signals to obtain several intrinsic mode functions includes:

[0022] Step 1. Parameter initialization: Before performing variational mode decomposition, set the required parameters, including the optimal decomposition level K, the penalty parameter α, and the convergence condition ε.

[0023] Step 2. Initialize the intrinsic mode function u k (t) and the frequency center ω k ;

[0024] Step 3. Iterative optimization: Use the Lagrange multiplier method to construct the objective function, and calculate each intrinsic mode function u k (t) by performing band-pass filtering on each intrinsic mode function; update its corresponding frequency center ω k for each intrinsic mode function to ensure that each mode is concentrated within its specific frequency band; update the Lagrange multiplier to ensure the reconstruction accuracy of the signal and reduce the decomposition error.

[0025] Step 4. Convergence determination: Repeat the iterative process in Step 3 until the convergence condition is reached. Optionally, the method for calculating the energy value of each intrinsic mode function includes:

[0026] For each intrinsic mode function IMF k (t), square it to calculate the instantaneous energy E k (t) = IMF k (t) 2 ; perform time integration on the instantaneous energy E k (t) of each intrinsic mode function to obtain the total energy value E k of this mode function over the entire time period. The formula is:

[0027]

[0028] where T is the total duration of the signal.

[0029] Optionally, establish the relationship between the energy value of each intrinsic mode function and the change of different levels of pre-tightening force, and screen out the characteristic modes that are most sensitive to the change of pre-tightening force, including:

[0030] For each intrinsic mode function, establish the relationship curve of its energy value E k varying with the pre-tightening force F, and the formula is as follows:

[0031] E k = f(F)

[0032] Plot the relationship diagram of the energy of each intrinsic mode function varying with the pre-tightening force; compare the energy curves of different intrinsic mode functions, analyze the amplitude of the energy value varying with the pre-tightening force, and select the intrinsic mode function with a larger amplitude of energy varying with the pre-tightening force and capable of monotonically indicating the change of pre-tightening force as the characteristic mode.

[0033] Optionally, establish the mapping relationship between the screened characteristic modal energy and the pre-tightening force, and obtain the loosening change curve including:

[0034] Use polynomial fitting or exponential, logarithmic or linear regression curve fitting methods to fit the relationship between the characteristic modal energy E 特征 and the pre-tightening force F.

[0035] Optionally, divide the bolt loosening process into different stages according to the loosening change curve, including:

[0036] Divide the bolt loosening process into different stages according to the slope characteristics of the loosening change curve.

[0037] Technical effects of the present invention: The present invention discloses a bolt loosening full-cycle acoustic emission monitoring method based on variational mode decomposition. By combining acoustic emission technology with variational mode decomposition method, real-time monitoring of the bolt loosening full cycle is realized. Using variational mode decomposition to decompose and extract features from complex acoustic emission signals effectively reduces the influence of noise and external interference, thereby improving the accuracy and stability of signal processing. This method can monitor the state of bolts in different loosening stages, provides comprehensive health monitoring for bolt connection structures, greatly improves the safety and reliability of bolt connections, reduces the risk of equipment failures and accidents, and has important application value for ensuring the stable operation of machinery and structures. Description of the Drawings

[0038] The drawings constituting a part of this application are used to provide a further understanding of this application. The schematic embodiments of this application and their descriptions are used to explain this application and do not constitute an improper limitation of this application. In the drawings:

[0039] Figure 1 is a schematic flow chart of a bolt loosening full-cycle acoustic emission monitoring method based on variational mode decomposition according to an embodiment of the present invention;

[0040] Figure 2 is a schematic diagram of a bolt loosening acoustic emission monitoring test bench according to an embodiment of the present invention;

[0041] Figure 3 is a schematic diagram of the relationship between the decomposition layer number K and the decomposition modal energy difference rate η according to an embodiment of the present invention;

[0042] Figure 4 is a schematic diagram of the energy change trend of IMF1 according to an embodiment of the present invention;

[0043] Figure 5 is a schematic diagram of the energy change trend of IMF2 according to an embodiment of the present invention;

[0044] Figure 6Schematic diagram of the energy change trend of IMF3 in the embodiment of the present invention;

[0045] Figure 7 Schematic diagram of the energy change trend of IMF4 in the embodiment of the present invention;

[0046] Figure 8 Schematic diagram of the classification monitoring of the full cycle stage of bolt loosening in the embodiment of the present invention;

[0047] Among them, 1 is a signal generator, 2 is a power amplifier, 3 is a vibrator, 4 is a push rod, 5 is a 5-mm steel plate, 6 is a digital display, 7 is an M8 bolt, 8 is a pressure sensor, 9 is an acoustic emission sensor, 10 is a preamplifier, 11 is an acoustic emission acquisition instrument, 12 is a PC, 13 is a vise, 14 is a foundation, and 15 is a 5-mm steel plate. Detailed implementation manners

[0048] It should be noted that, without conflict, the embodiments in the present application and the features in the embodiments may be combined with each other. The present application will be described in detail below with reference to the drawings and in combination with the embodiments.

[0049] It should be noted that the steps shown in the flowchart of the drawings can be executed in a computer system such as a set of computer-executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than here.

[0050] Introduction of related terms:

[0051] Variational mode decomposition (VMD): The VMD method decomposes a complex signal into several intrinsic mode functions (IMFs), and each mode function represents different characteristic components of the signal. As an emerging signal processing method, variational mode decomposition (VMD) has gradually attracted the attention of researchers. Compared with traditional signal decomposition methods, VMD has better adaptability and resolution, can effectively retain the local characteristics of the signal, and significantly improve the accuracy of signal processing. Therefore, VMD has broad application prospects in acoustic emission signal analysis, especially in bolt loosening monitoring under complex working conditions, and can provide more accurate feature extraction and analysis means.

[0052] As Figure 1 shown, in this embodiment, a method for acoustic emission monitoring of bolt loosening in the full cycle based on variational mode decomposition is provided, including:

[0053] Collect acoustic emission signals at several levels of pre-tightening force for the bolt connection interface;

[0054] Perform variational mode pre-decomposition on the acoustic emission signals according to different decomposition levels, calculate the decomposition mode energy corresponding to different decomposition levels, then calculate the decomposition mode energy difference rate based on the decomposition mode energy corresponding to different decomposition levels, and use the decomposition mode energy difference rate to determine the optimal decomposition level;

[0055] Based on the determined optimal decomposition level, perform variational mode decomposition on all the collected acoustic emission signals to obtain several intrinsic mode functions;

[0056] By calculating the energy value of each intrinsic mode function, establish the variation relationship between the energy value of each intrinsic mode function and different levels of pre-tightening force, and screen out the characteristic mode that is most sensitive to the change of pre-tightening force;

[0057] Establish the mapping relationship between the energy of the screened characteristic mode and the pre-tightening force, obtain the loosening change curve, and divide the bolt loosening process into different stages according to the loosening change curve to realize the quantitative monitoring of the entire cycle of bolt loosening.

[0058] Furthermore, the number of levels of pre-tightening force satisfies that the number of levels is greater than or equal to 8.

[0059] By increasing the number of pre-tightening force levels, the influence of pre-tightening force changes on acoustic emission signals can be captured more precisely. Collecting data at more than or equal to 8 different pre-tightening force levels can ensure that the relationship between energy changes and pre-tightening force is clearer, forming a more accurate mapping model. This helps to more accurately evaluate the state of the bolt at different loosening stages and improve the monitoring accuracy.

[0060] Furthermore, the energy in the decomposition mode energy difference rate is expressed as:

[0061]

[0062] where E represents the energy of the signal, x(i) represents the collected signal sequence, n represents the number of sampling points, and i is the i-th sampling point in the signal sequence.

[0063] By calculating the total energy of the signal using this formula, the intensity of the acoustic emission signal can be effectively quantified, which is convenient for evaluating the energy contribution of each decomposition mode during the bolt loosening process.

[0064] Furthermore, the decomposition mode energy difference rate is expressed as:

[0065]

[0066] where η represents the decomposition mode energy difference rate, E KDenotes the sum of the energies of all K components obtained by decomposing the number of layers K, E K-1 Denotes the sum of the energies of all K - 1 components obtained by decomposing the number of layers K - 1.

[0067] Calculating the energy difference rate of the decomposition mode through this formula can reflect the change trend between the modal energies of different decomposition levels, providing a basis for subsequent screening of characteristic modes.

[0068] Furthermore, determining the decomposition number of layers using the energy difference rate of the decomposition mode includes:

[0069] Step S1: Initialize the decomposition number of layers, set the decomposition layer value, determine the search range of the decomposition number of layers, set the step size, perform variational mode decomposition on the acoustic emission signal. For the set decomposition layer value, calculate the energy of each intrinsic mode function obtained by decomposition, and calculate the sum of the energies of all intrinsic mode functions;

[0070] Step S2: Increase the decomposition number of layers by 1, and continue with Step S1 until the maximum value of the search range, then perform Step S3;

[0071] Step S3: Compare the energy difference rates of the decomposition modes corresponding to the decomposition layer values, and determine the decomposition number of layers based on the decomposition layer value corresponding to the maximum energy difference rate of the decomposition mode.

[0072] By presetting the search range and step size of the decomposition number of layers, it is possible to effectively avoid excessive redundant decomposition, reduce unnecessary computational work, and improve the efficiency of signal processing. The finally selected decomposition number of layers is determined by the principle of maximizing the energy difference rate, which means that the system will not over - decompose the signal, nor will it lose key information due to insufficient decomposition, achieving a balance between computational efficiency and accuracy.

[0073] Furthermore, performing variational mode decomposition on all the collected acoustic emission signals to obtain several intrinsic mode functions includes:

[0074] Step 1: Parameter initialization. Before performing variational mode decomposition, set the required parameters, including the optimal decomposition number of layers K, the penalty parameter α, and the convergence condition ε.

[0075] Step 2: Initialize the intrinsic mode function u k (t) and the frequency center ω k .

[0076] Step 3: Iterative optimization. Use the Lagrange multiplier method to construct the objective function, and calculate each intrinsic mode function u k (t) by performing band - pass filtering on each intrinsic mode function; update its corresponding frequency center ω for each intrinsic mode function k, ensure that each mode is concentrated within its specific frequency band; update the Lagrange multipliers to ensure the reconstruction accuracy of the signal and reduce the decomposition error.

[0077] Step 4: Convergence determination. Repeat the iterative process in Step 3 until the convergence condition is reached.

[0078] Through variational mode decomposition, complex acoustic emission signals can be decomposed into several intrinsic mode functions with different frequency bands. Since each intrinsic mode function is concentrated within a specific frequency band, this precise decomposition ensures the mutual independence between different mode signals, can better capture the characteristic frequencies at different stages during the bolt loosening process, and greatly improves the analysis and monitoring accuracy of the loosening process.

[0079] Furthermore, the method for calculating the energy value of each intrinsic mode function includes:

[0080] For each intrinsic mode function IMF k (t), perform a square operation on it to calculate the instantaneous energy E k (t) = IMF k (t) 2 ; perform a time integration on the instantaneous energy E k (t) of each intrinsic mode function to obtain the total energy value E k of this mode function over the entire time period, and the formula is:

[0081]

[0082] where T is the total duration of the signal.

[0083] By performing a square operation and integration on each intrinsic mode function, the modal energy can be accurately calculated. It can clearly reflect the energy distribution of each intrinsic mode over the entire signal time period, making the energy characteristics of the signal more explicit.

[0084] Furthermore, establish the variation relationship between the energy value of each intrinsic mode function and different levels of pre-tightening force. The characteristic modes that are most sensitive to the change in pre-tightening force include:

[0085] For each intrinsic mode function, establish the relationship curve of its energy value E k varying with the pre-tightening force F, and the formula is as follows:

[0086] E k = f(F)

[0087] Plot the relationship diagram of the energy of each intrinsic mode function varying with the pre-tightening force. Compare the energy curves of different intrinsic mode functions, analyze the amplitude of the energy value varying with the pre-tightening force, and select the intrinsic mode function with a larger amplitude of energy varying with the pre-tightening force and capable of monotonically indicating the change in pre-tightening force as the characteristic mode.

[0088] By establishing the relationship curve between the energy value of the intrinsic mode function and the change in pre-tightening force, the influence of the pre-tightening force change on the energy of each mode can be intuitively reflected. Such a relationship diagram can reveal the response degree of different modes to the pre-tightening force change, which helps to screen out the characteristic modes that are most sensitive to the pre-tightening force change, thus improving the accurate monitoring of the bolt loosening state.

[0089] Furthermore, establish the mapping relationship between the energy of the screened characteristic modes and the pre-tightening force, and obtain the loosening change curve including:

[0090] Use polynomial fitting or other suitable curve fitting methods (such as exponential, logarithmic or linear regression) to fit the relationship between the characteristic mode energy E 特征 and the pre-tightening force F.

[0091] By using suitable curve fitting methods such as polynomial fitting, exponential, logarithmic or linear regression, an accurate mapping relationship between the characteristic mode energy and the pre-tightening force can be established. This mapping relationship not only provides a quantitative evaluation basis for the bolt loosening process, but also enables the visualization of the pre-tightening force change, providing solid data support for the accurate monitoring and early warning of bolt loosening.

[0092] Furthermore, divide the bolt loosening process into different stages according to the loosening change curve, including:

[0093] Divide the bolt loosening process into different stages according to the slope characteristics of the loosening change curve.

[0094] Through the analysis of the loosening change curve, an effective division of the bolt loosening stage can be achieved. The loosening characteristics of different stages are significantly different, and dividing the stages can provide support for the system to formulate differential early warning strategies. In the early loosening stage, more sensitive early warning signals can be triggered, while in the late loosening stage, stronger alarm measures can be taken to reduce the impact of bolt loosening on the structural safety.

[0095] The implementation process of a specific application example of the present invention includes:

[0096] Step 1: As Figure 2As shown in the figure, a bolt loosening acoustic emission monitoring test bench is built to collect the acoustic emission signals at the bolt connection interface. The signal generator 1 outputs a 5 Hz sine signal and transmits it to the power amplifier 2. The power amplifier 2 outputs a voltage of 1 V to amplify the signal and provide sufficient power for the exciter 3. The exciter 3 converts the amplified electrical signal into mechanical vibration and transmits it to the steel plate 5 through the ejector rod 4. The steel plates 5 and 15 serve as the media for vibration propagation and are fixed by the vise 13 in the experiment to ensure stability. The digital display 6 shows the pre-tightening force applied to the bolt 7 in real time. The M8 bolt 7 monitors the applied pre-tightening force through the pressure sensor 8, and the pressure sensor transmits the measured data to the digital display 6. The acoustic emission sensor 9 is used to detect the acoustic emission signals generated during the loosening of the bolt. The sensor is pasted on the steel plate 15 to capture the acoustic emission signals. The preamplifier 10 amplifies the weak signals from the acoustic emission sensor to facilitate the processing by the acoustic emission collector 11. The acoustic emission collector converts the amplified signals into digital data and transmits them to the PC 12 for real-time monitoring, analysis, and storage. The vise 13 fixes the steel plate, and the base 14 provides stable support for the entire device to ensure that no displacement or interference occurs among the components during the experiment. Among them, the M8 bolt used has a strength grade of 8.8, and according to the German Engineers' Association standard VDI 2230: Part 1, its allowable pre-tightening force is 16.5 kN. In the experiment, the pre-tightening force is incremented by 2 kN from 2 kN to 16 kN. The sampling frequency is set to 10 MHz. Six groups of data are collected for each pre-tightening force level.

[0097] Step 2: Use VMD to process the original AE signals. First, use the decomposition mode energy difference rate to screen the parameter K so that the VMD algorithm can perform mode decomposition with the optimal decomposition layer number. Randomly select a set of AE signals for analysis, and the signal length is taken as one vibration period of 0.2 seconds. Figure 3 The relationship between the decomposition layer number and the decomposition mode energy difference rate is given. It can be seen that the decomposition mode energy difference rate of the AE signals after VMD decomposition reaches the maximum value when K = 4. Therefore, the optimal K value for VMD decomposition is taken as 4.

[0098] Step 3: According to the determined optimal decomposition layer number of 4 layers, perform VMD decomposition on all the acoustic emission signals of 8 levels of pre-tightening force collected. Each acoustic emission signal is decomposed into 4 IMFs, namely IMF1, IMF2, IMF3, and IMF4.

[0099] Step 4: As Figures 4 to 7 shown, calculate the energy values of each IMF obtained, and establish the variation relationship between the energy values from IMF1 to IMF4 and the 8 levels of pre-tightening force. Among them, Figure 5 the energy of IMF2 is the most sensitive to the change of the pre-tightening force and can monotonically indicate the change of the pre-tightening force. Therefore, the IMF2 component is selected as the characteristic mode.

[0100] Step Five: As Figure 8 shown, establish the mapping relationship between the characteristic modal energy and the pre-tightening force to obtain the loosening change curve. According to the different slopes of the loosening change curve, the loosening process is divided into three stages, namely the early loosening stage, the mid-term loosening stage, and the late loosening stage. By using the fitting equations of each stage, quantitative monitoring of the entire cycle of bolt loosening can be achieved.

[0101] The present invention discloses a method for acoustic emission monitoring of the entire cycle of bolt loosening based on variational mode decomposition. By combining the acoustic emission technology with the variational mode decomposition method, real-time monitoring of the entire cycle of bolt loosening is realized. Using variational mode decomposition to decompose and extract features from complex acoustic emission signals effectively reduces the influence of noise and external interference, thereby improving the accuracy and stability of signal processing. This method can monitor the state of bolts at different loosening stages, provides comprehensive health monitoring for bolt connection structures, greatly improves the safety and reliability of bolt connections, reduces the risk of equipment failures and accidents, and has important application value for ensuring the stable operation of machinery and structures.

[0102] The above is only a preferred specific embodiment of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by those skilled in the art within the technical scope disclosed in the present application should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A full-cycle acoustic emission monitoring method for bolt loosening based on variational mode decomposition, characterized in that: include: Collect acoustic emission signals of bolt connection interfaces under several levels of preload; Performing variational modal pre-decomposition on the acoustic emission signal, performing variational modal pre-decomposition according to different decomposition levels, calculating decomposition modal energies corresponding to different decomposition levels, calculating decomposition modal energy difference rates according to the decomposition modal energies corresponding to different decomposition levels, and determining an optimal decomposition level using the decomposition modal energy difference rates; The decomposed modal energy difference rate is expressed as: Among them, η represents the energy difference rate of the decomposed mode, E K represents the sum of the energies of all K components obtained by decomposing the Kth level, E K-1 Represents the sum of the energies of all K-1 components obtained by decomposing the K-1 layer; Determining the optimal number of decomposition layers using the decomposition modal energy difference rate includes: Step S1, initialize the number of decomposition layers, set the decomposition layer value, determine the search range of the decomposition layer number, set the step size, perform variational mode decomposition on the acoustic emission signal, calculate the energy of each decomposed intrinsic mode function for the set decomposition layer value, and calculate the sum of the energies of all intrinsic mode functions; Step S2, increase the number of decomposition levels by 1, and continue step S1 until the maximum value of the search range is reached, and then proceed to step S3; Step S3, comparing the decomposition modal energy difference rate corresponding to the decomposition layer value and the decomposition layer value corresponding to the maximum value of the decomposition modal energy difference rate, and determining the optimal decomposition layer number; Based on the optimal decomposition layer number, performing variational modal decomposition on all collected acoustic emission signals to obtain a number of intrinsic mode functions; By calculating the energy value of each eigenmode function, the relationship between the energy value of each eigenmode function and the change of preload force at different levels is established, and the characteristic mode that is most sensitive to the change of preload force is screened out; A mapping relationship between the selected characteristic modal energy and the preload force is established to obtain a loosening change curve. The bolt loosening process is divided into different stages according to the loosening change curve to achieve quantitative monitoring of the entire bolt loosening cycle.

2. The full-cycle acoustic emission monitoring method for bolt loosening based on variational modal decomposition according to claim 1 is characterized in that: The plurality of levels of preload force satisfy a number that is greater than or equal to 8.

3. The full-cycle acoustic emission monitoring method for bolt loosening based on variational modal decomposition according to claim 1 is characterized in that: The energy in the decomposed modal energy difference rate is expressed as: Where E represents the energy of the signal, x(i) represents the acquired signal sequence, n represents the number of sampling points, and i is the i-th sampling point in the signal sequence.

4. The method for monitoring bolt loosening full-cycle acoustic emission based on variational modal decomposition according to claim 1 is characterized in that: All collected acoustic emission signals are subjected to variational mode decomposition to obtain several eigenmode functions including: Step 1: Parameter initialization: Before performing variational mode decomposition, set the required parameters, including the optimal decomposition layer number K, penalty parameter α, and convergence condition ε; Step 2: Initialize the intrinsic mode function u k (t) and frequency center ω k ; Step 3: Iterative optimization: construct the objective function using the Lagrange multiplier method, and calculate each eigenmode function u by performing bandpass filtering on each eigenmode function. k (t); Update the corresponding frequency center ω for each eigenmode function k , ensuring that each mode is concentrated within its frequency band; updating the Lagrange multiplier to ensure the reconstruction accuracy of the signal and reduce the decomposition error; Step 4: Convergence determination: repeat the iterative process in step 3 until the convergence condition is reached.

5. The method for monitoring bolt loosening in full cycle based on variational modal decomposition according to claim 1 is characterized in that: The method for calculating the energy value of each eigenmode function includes: For each intrinsic mode function IMF k (t), square it and calculate the instantaneous energy E k (t) = IMF k (t) 2 , the instantaneous energy E of each eigenmode function k (t) Perform time integration to obtain the total energy value E of the modal function in the entire time period k , the formula is: Where T is the total duration of the signal.

6. The full-cycle acoustic emission monitoring method for bolt loosening based on variational modal decomposition according to claim 1 is characterized in that: The relationship between the energy value of each eigenmode function and the change of preload at different levels is established, and the eigenmodes most sensitive to the change of preload are screened out, including: For each eigenmode function, establish its energy value E k The relationship curve with the change of preload force F is as follows: E k =f(F) The energy of each intrinsic mode function is plotted as a function of preload. The energy curves of different intrinsic mode functions are compared, the amplitude of the energy value changing with preload is analyzed, and the intrinsic mode function with a large amplitude of energy changing with preload and which can monotonically indicate the change of preload is selected as the characteristic mode.

7. The method for monitoring bolt loosening in full cycle based on variational modal decomposition according to claim 1, characterized in that: The mapping relationship between the selected characteristic modal energy and the preload force is established, and the looseness change curve is obtained, including: Fit the eigenmode energy E using polynomial fitting or exponential, logarithmic or linear regression curve fitting methods 特征 The relationship between and preload force F.

8. The full-cycle acoustic emission monitoring method for bolt loosening based on variational modal decomposition according to claim 1 is characterized in that: Dividing the bolt loosening process into different stages according to the loosening change curve includes: dividing the bolt loosening process into different stages according to the slope characteristics of the loosening change curve.

Images