Method for detecting pre-tightening force of GIS flange bolt

Abstract

The present application relates to the technical field of bolt fastener, and specifically provides a GIS flange bolt pre-tightening force detection method, comprising: acquiring ultrasonic echo signals of both ends of a flange bolt of a gas insulated switchgear (GIS); decomposing the ultrasonic echo signals of each end into intrinsic mode components based on a characteristic mode decomposition algorithm; optimizing characteristic parameters of the characteristic mode decomposition algorithm based on a grey wolf optimization algorithm and the intrinsic mode components to obtain optimized characteristic parameters; decomposing the ultrasonic echo signals of each end into reconstruction components based on the characteristic mode decomposition algorithm with the optimized characteristic parameters; and performing time-frequency correlation analysis on the reconstruction components corresponding to the ultrasonic echo signals of both ends respectively to obtain a pre-tightening force detection result of the flange bolt. The present application solves the problem of low accuracy and stability of related art flange bolt pre-tightening force detection.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of bolt fasteners, in particular to a GIS flange bolt pre-tightening force detection method. BACKGROUND

[0002] As a core component for connecting a pot-type insulator and a gas insulated switchgear (GIS) cylinder, a flange bolt is prone to pre-tightening force attenuation or interface failure under the combined action of long-term mechanical vibration, temperature difference deformation and environmental corrosion, which leads to unbalanced sealing pressure distribution and may cause SF6 gas leakage or insulator cracking accidents.

[0003] The conventional maintenance means relying on a passive artificial detection mode can stage by stage alleviate the influence of the above problems, but it is difficult to accurately capture dynamic stress changes, and the data analysis efficiency is low, so it is urgent to develop an in-situ monitoring technology suitable for the flange structure. In this regard, related pre-tightening force detection methods include: a torque method, which is significantly affected by the friction coefficient and has a large error in the detection result; a strain gauge and a conventional ultrasonic method, which are sensitive to installation conditions and lack anti-interference performance; and an emerging vibration acoustic method, which has deployment convenience, but the low-frequency characteristics used are easily coupled with the body vibration of the flange bolt, resulting in low detection sensitivity.

[0004] Currently, no effective solutions have been proposed to solve the problem of low accuracy and stability of related art flange bolt pre-tightening force detection methods. SUMMARY

[0005] The GIS flange bolt pre-tightening force detection method provided by the embodiments of the application at least solves the problem of low accuracy and stability of related art flange bolt pre-tightening force detection methods.

[0006] The GIS flange bolt pre-tightening force detection method provided by the embodiments of the application comprises: acquiring ultrasonic echo signals at both ends of a flange bolt of a gas insulated switchgear (GIS); decomposing the ultrasonic echo signals at each end into intrinsic modal components based on a characteristic modal decomposition algorithm; optimizing characteristic parameters of the characteristic modal decomposition algorithm based on a grey wolf optimization algorithm and the intrinsic modal components to obtain optimized characteristic parameters; decomposing the ultrasonic echo signals at each end into reconstruction components based on the characteristic modal decomposition algorithm using the optimized characteristic parameters; and performing time-frequency correlation analysis on the reconstruction components corresponding to the ultrasonic echo signals at both ends to obtain a pre-tightening force detection result of the flange bolt.

[0007] The present invention provides a method for detecting the preload force of GIS flange bolts, which decomposes the ultrasonic echo signal at each end into eigenmode components based on an eigenmode decomposition algorithm. The method includes: inputting the ultrasonic echo signal into a constrained variational model of the eigenmode decomposition algorithm to obtain the signal eigenmode; and performing modal aliasing suppression processing on the signal eigenmode based on the constraints of the eigenmode decomposition algorithm to obtain the eigenmode components.

[0008] The present invention provides a method for detecting the preload force of GIS flange bolts. Based on the constraints of the characteristic mode decomposition algorithm, the method performs modal aliasing suppression processing on the signal eigenmode to obtain the eigenmode components. The method includes: based on the constraints of the characteristic mode decomposition algorithm, performing modal aliasing suppression processing on the signal eigenmode to obtain the constrained mode components; based on the filter length and the number of frequency bands, evenly dividing the signal spectrum of the constrained mode components into multiple frequency bands; and constraining each frequency band to be occupied by a constrained mode component through frequency domain truncation optimization to obtain the eigenmode components. The formula for frequency domain truncation optimization is expressed as follows:

[0009] ;

[0010] ;

[0011] ;

[0012] Where, Indicates the The signal spectrum of the eigenmode components, Indicates the The spectral components of the eigenmode components, for frequency, Indicates the The selection weights of the eigenmode components, , represents the filter length, Indicates the number of frequency bands, In other cases.

[0013] The present invention provides a method for detecting the preload force of GIS flange bolts. The method optimizes the characteristic parameters of the characteristic mode decomposition algorithm based on the Gray Wolf optimization algorithm and the intrinsic mode component to obtain the optimized characteristic parameters. The method includes: encoding the selection weights and the number of frequency bands of the intrinsic mode component to obtain the corresponding Gray Wolf position vector; iteratively updating the Gray Wolf position vector based on the fitness function to obtain the final position vector, wherein the optimization is stopped when the iterative change rate of the fitness function is less than a preset threshold; and determining the optimized characteristic parameters based on the final position vector. The formula for determining the fitness function according to the energy entropy minimization principle is:

[0014] ;

[0015] ;

[0016] wherein, represents energy entropy, and the characteristic parameters of the empirical mode decomposition algorithm include filter length and frequency band number , represents maximum frequency band number, represents maximum sampling point number, represents time, and represent the intrinsic mode components corresponding to the ultrasonic echo signals received at the two ends of the flange bolt respectively, and and come from the same excitation signal.

[0017] The GIS flange bolt pre-tightening force detection method provided by the embodiments of the application performs time-frequency correlation analysis on the reconstructed components corresponding to the ultrasonic echo signals at the two ends, and obtains the pre-tightening force detection result of the flange bolt, including: calculating the wave energy ratio and the double-end time delay distortion degree corresponding to the reconstructed components; inputting the wave energy ratio and the double-end time delay distortion degree into a mapping model of support vector regression to obtain the pre-tightening force value, and determining the pre-tightening force detection result, wherein the kernel function of the mapping model is determined in a cross-validation manner.

[0018] The GIS flange bolt pre-tightening force detection method provided by the embodiments of the application calculates the double-end time delay distortion degree corresponding to the reconstructed components, including: performing zero-phase filtering on the reconstructed components to obtain double-end received signals; calculating a normalized cross-correlation function based on the double-end received signals to determine a time delay value corresponding to a maximum correlation coefficient point; and calculating the double-end time delay distortion degree based on the time delay value and a preset nonlinear term.

[0019] The GIS flange bolt pre-tightening force detection method provided by the embodiments of the application is used for a flange bolt in a single screw rod structure without a screw head; before obtaining the ultrasonic echo signals at the two ends of the flange bolt of the gas insulated switchgear (GIS), the above method further includes: passing a hollow column piezoelectric ceramic through a screw rod of the flange bolt, and installing the hollow column piezoelectric ceramic on an outer surface of any nut of the flange bolt as an excitation source, and installing piezoelectric ultrasonic transducers at the two ends of the screw rod as receiving ends; controlling a pulse driver to generate a high-frequency narrow pulse drive excitation signal according to a preset center frequency to drive the excitation source to generate an excitation signal, so as to obtain ultrasonic echo signals at the receiving ends, wherein the pulse driver is connected with the hollow column piezoelectric ceramic.

[0020] The method for detecting the pre-tightening force of the GIS flange bolt provided by the embodiment of the application obtains the ultrasonic echo signals at both ends of the flange bolt of the gas insulated switchgear (GIS), and comprises the following steps: repeatedly obtaining a plurality of ultrasonic echo signals at both ends of the flange bolt according to a preset number of times under different levels of measured pre-tightening force; and constructing a sample library based on the plurality of ultrasonic echo signals, so as to obtain a plurality of pre-tightening force detection results corresponding to different levels of measured pre-tightening force and a pre-tightening force estimation curve of the flange bolt.

[0021] The electronic device provided by the embodiment of the application comprises a processor and a memory storing programs, the programs comprising instructions which, when executed by the processor, cause the processor to perform any of the above methods.

[0022] The non-transitory machine-readable medium storing computer instructions for causing a computer to perform the method according to any of the above methods is provided by the embodiment of the application.

[0023] The method for detecting the pre-tightening force of the GIS flange bolt provided by the embodiment of the application obtains the ultrasonic echo signals at both ends of the GIS flange bolt, decomposes the ultrasonic echo signals at each end into intrinsic modal components based on a characteristic modal decomposition algorithm, optimizes the characteristic parameters of the characteristic modal decomposition algorithm based on a grey wolf optimization algorithm and the intrinsic modal components to obtain optimized characteristic parameters, decomposes the ultrasonic echo signals at each end into reconstructed components based on the characteristic modal decomposition algorithm using the optimized characteristic parameters, and performs time-frequency correlation analysis on the reconstructed components corresponding to the ultrasonic echo signals at both ends to obtain the pre-tightening force detection result of the flange bolt. The method can simultaneously perceive the nonlinear attenuation characteristics of micron-level pre-tightening deformation and interface contact state, does not rely on a multi-sensor array, has strong adaptive analysis capability in complex working conditions, and has high detection accuracy and stability. The method solves the problem of low accuracy and stability of the related art in detecting the pre-tightening force of the flange bolt. BRIEF DESCRIPTION OF DRAWINGS

[0024] In order to more clearly illustrate the embodiments of the application or the technical solutions in the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description are only some embodiments of the application, and for those skilled in the art, other embodiments can also be obtained from these drawings without creative labor.

[0025] Figure 1 is a step flow chart of a method for detecting the pre-tightening force of a GIS flange bolt in the embodiment of the application.

[0026] Figure 2 is a schematic diagram of an experimental platform built for applying Figure 1 the method.

[0027] Figure 3 is a schematic diagram of a near-end original ultrasonic echo signal and a far-end original ultrasonic echo signal in an embodiment of the present application.

[0028] Figure 4 is a schematic diagram of a high-frequency intrinsic modal component obtained by decomposing the far-end original ultrasonic echo signal shown in the figure. Figure 3

[0029] Figure 5 is a schematic diagram of a high-frequency reconstructed component obtained by decomposing the far-end original ultrasonic echo signal shown in the figure. Figure 3

[0030] Figure 6 is a schematic diagram of a medium-frequency intrinsic modal component obtained by decomposing the far-end original ultrasonic echo signal shown in the figure. Figure 3

[0031] Figure 7 is a schematic diagram of a medium-frequency reconstructed component obtained by decomposing the far-end original ultrasonic echo signal shown in the figure. Figure 3

[0032] Figure 8 is a schematic diagram of a convergence curve of a grey wolf optimization algorithm GWO in an embodiment of the present application.

[0033] Figure 9 is a schematic diagram of SVR super parameter optimization by a mapping model in an embodiment of the present application.

[0034] Figure 10 is a schematic diagram of a pretightening force estimation curve in an embodiment of the present application.

[0035] Figure 11 is a schematic diagram of a structure of an electronic device in an embodiment of the present application.

[0036] Among the above figures, the following reference signs are included:

[0037] 21-GIS flange; 22-piezoelectric ultrasonic transducer; 23-screw rod; 24-nut; 25-nut gasket; 26-hollow column piezoelectric ceramic; 27-pulse driver; 28-dynamic signal acquisition system. DETAILED DESCRIPTION

[0038] Embodiments of the present application will be described in more detail by referring to the attached drawings. Although certain embodiments of the present application are shown in the drawings, it is understood that the present application can be implemented in various forms and should not be interpreted as being limited to the embodiments set forth herein, but rather these embodiments are provided so as to more completely and thoroughly understand the present application. It is understood that the drawings and embodiments of the present application are for exemplary purposes only and are not intended to limit the scope of protection of the present application.​​​​

[0039] The conventional maintenance means relying on passive artificial detection mode can alleviate the impact of the above problems in stages, but it is difficult to accurately capture dynamic stress changes, and the data analysis efficiency is low, and it is urgent to develop in-situ monitoring technology suitable for flange structure. In this regard, the relevant pre-tightening force detection methods are: torque method, which is significantly affected by the friction coefficient, and the error of the detection result is large; strain gauge and conventional ultrasonic method, which are sensitive to installation conditions and lack of anti-interference performance; emerging vibration acoustic method, which has deployment convenience, but the low frequency characteristics used are easily coupled with the body vibration of the flange bolt, resulting in low detection sensitivity.

[0040] Therefore, as shown in Figure 1 The embodiment of the present application provides a GIS flange bolt pre-tightening force detection method, which comprises steps S101 to S105.

[0041] Step S101, acquiring ultrasonic echo signals at both ends of the flange bolt of the gas insulated switchgear GIS.

[0042] Step S102, decomposing the ultrasonic echo signals at each end into intrinsic modal components based on a characteristic modal decomposition algorithm.

[0043] Step S103, optimizing the characteristic parameters of the characteristic modal decomposition algorithm based on the intrinsic modal components and the gray wolf optimization algorithm, to obtain optimized characteristic parameters.

[0044] Step S104, decomposing the ultrasonic echo signals at each end into reconstructed components based on the characteristic modal decomposition algorithm with the optimized characteristic parameters.

[0045] Step S105, performing time-frequency correlation analysis on the reconstructed components corresponding to the ultrasonic echo signals at both ends, to obtain the pre-tightening force detection result of the flange bolt.

[0046] The subsequent embodiment takes the GIS flange bolt with a single screw rod structure without a screw head as an example for specific introduction, which has a higher degree of adaptation to the above detection method provided by the embodiment, but is not limited thereto. Those skilled in the art can refer to the above method to detect the pre-tightening force of the GIS flange bolt with other structures.

[0047] In order to obtain the above-mentioned ultrasonic echo signals, an excitation source is needed to provide an excitation signal, which belongs to the prior art. The embodiment preferably uses a single-end excitation method to generate double-end ultrasonic echo signals, which can establish a longitudinal wave coupling propagation path in the axial direction of the above-mentioned flange bolt, which will be described in detail later.

[0048] The ultrasonic signal reflected by the double path can be decomposed into an intrinsic modal component with clear physical meaning by the characteristic modal decomposition algorithm. The double path refers to the path from the excitation source to the two ends of the flange bolt. The characteristic modal decomposition algorithm includes but is not limited to a constrained variation model, an objective function, and a constraint condition corresponding to the objective function, which will be described later. The characteristic modal decomposition algorithm provided by the preferred mode of the embodiment further includes a modal aliasing suppression processing by a frequency domain truncation constraint condition, which helps to improve the detection accuracy.

[0049] The grey wolf optimization algorithm simulates the hierarchical system and hunting behavior of a wolf pack, and guides the iterative optimization of the wolf pack through alpha wolves, beta wolves and delta wolves. Specifically, the alpha wolf is the leader of the pack, representing the optimal solution. The beta wolf is a subordinate wolf that assists the alpha wolf in decision-making. The delta wolf is a scout wolf or a guard wolf that follows the decision-making orders of the alpha wolf and the beta wolf. The position information of the alpha wolf, the beta wolf and the delta wolf can guide other wolves to update their positions, so that the wolf pack iteratively optimizes towards the optimal solution.

[0050] The characteristic parameters of the characteristic modal decomposition algorithm include core characteristic parameters, auxiliary parameters and evaluation parameters.

[0051] The core characteristic parameters include but are not limited to the number of decomposition layers, the stopping criterion parameter, and the feature extraction parameter.

[0052] The auxiliary parameters include but are not limited to signal preprocessing parameters, optimization algorithm parameters, and boundary processing parameters.

[0053] The evaluation parameters include but are not limited to modal orthogonality, energy concentration, and correlation coefficient.

[0054] The preferred embodiment of the present application optimizes the filter length and the number of frequency bands of the characteristic modal decomposition algorithm. The filter length belongs to the above-mentioned feature extraction parameter. The number of frequency bands is related to the number of decomposition layers.

[0055] The time-frequency correlation analysis of the above-mentioned reconstructed components can be but is not limited to: calculating the time-frequency energy entropy + impact factor, the time-frequency cross-correlation coefficient + inherent frequency offset, the fractal dimension + Lyapunov exponent based on the above-mentioned reconstructed components. The preferred embodiment of the present application performs time-frequency correlation analysis on the above-mentioned reconstructed components by calculating the fluctuation energy ratio + double-end time delay distortion degree, which can quantify the energy transmission efficiency at the two ends of the flange bolt, has the advantages of strong anti-noise ability and low computational complexity, which will be described later.

[0056] In summary, the above method provided by the embodiment obtains the ultrasonic echo signals of the two ends of the GIS flange bolt, decomposes the ultrasonic echo signals of each end into intrinsic mode components based on a feature mode decomposition algorithm, optimizes the feature parameters of the feature mode decomposition algorithm based on the grey wolf optimization algorithm and the intrinsic mode components, obtains the optimized feature parameters, decomposes the ultrasonic echo signals of each end into reconstruction components based on the feature mode decomposition algorithm using the optimized feature parameters, and performs time-frequency correlation analysis on the reconstruction components corresponding to the ultrasonic echo signals of the two ends respectively to obtain the pretightening force detection result of the flange bolt. The nonlinear attenuation characteristics of the micron-level pretightening deformation and the interface contact state can be synchronously perceived, the dependence on a multi-sensor array is eliminated, strong self-adaptive analysis capability is achieved in complex working conditions, and the detection accuracy and stability are high. The problem of low accuracy and stability of the flange bolt pretightening force detection in the related art is solved.

[0057] Through the above scheme, the problem of low accuracy and stability of the flange bolt pretightening force detection in the related art is solved.

[0058] In addition, the following explanations are given to the English abbreviations in the embodiment of the present application.

[0059] The full name of the gas insulated switchgear in English is GAS insulated SWITCHGEAR, which is abbreviated as GIS.

[0060] The full name of the feature mode decomposition algorithm in English is Feature Mode Decomposition, which is abbreviated as FMD.

[0061] The full name of the intrinsic mode component in English is Intrinsic Mode Function, which is abbreviated as IMF.

[0062] The full name of the reconstruction signal in English is Feature Mode Function, which is abbreviated as FMF.

[0063] The full name of the grey wolf optimization algorithm in English is Grey Wolf Optimizer, which is abbreviated as GWO.

[0064] The full name of the support vector regression in English is Support Vector Regression, which is abbreviated as SVR.

[0065] Preferably, the above flange bolt is a flange bolt with a single screw rod structure without a screw head, please refer to Figure 2 .

[0066] Before the step S101 of obtaining the ultrasonic echo signals of the two ends of the flange bolt of the gas insulated switchgear GIS, the above method further includes the following excitation process.

[0067] The hollow cylindrical piezoelectric ceramic 26 is installed on the outer surface of any one of the nuts 24 of the flange bolt as an excitation source by passing the screw rod 23 of the flange bolt, and piezoelectric ultrasonic transducers 22 are installed at both ends of the screw rod 23 as receiving ends.

[0068] The control pulse driver 27 generates a high-frequency narrow pulse driving excitation source according to a preset center frequency to generate an excitation signal to obtain an ultrasonic echo signal at the receiving end, wherein the pulse driver 27 is connected with the hollow cylindrical piezoelectric ceramic 26.

[0069] The dynamic signal acquisition system 28 can synchronously capture the time-domain echo signals received at both ends of the flange bolt and , that is, the above-mentioned ultrasonic echo signals.

[0070] The preset center frequency and pulse width of the above-mentioned high-frequency narrow pulse can be determined by those skilled in the art according to prior values and actual conditions.

[0071] For example, the preset center frequency is 1.5 MHz, and the pulse width is 10 μs.

[0072] It can be understood that by using a single-end excitation method to generate a double-end ultrasonic echo signal, a longitudinal wave coupling propagation path can be established in the axial direction of the screw rod 23, which helps to synchronously perceive the nonlinear attenuation characteristics of micron-level pre-tightening deformation and interface contact state.

[0073] In addition, those skilled in the art can also refer to the above-mentioned ultrasonic detection method based on single-end excitation and double-end echo to analyze the composite characteristics of the transverse elastic wave conduction path of the bolt.

[0074] Preferably, in step S101, the ultrasonic echo signals at both ends of the flange bolt of the gas insulated switchgear GIS are obtained, including: under different levels of measured pre-tightening force, a plurality of ultrasonic echo signals at both ends of the flange bolt are repeatedly obtained according to a preset number of times. A sample library is constructed based on the plurality of ultrasonic echo signals, which is used to obtain a plurality of pre-tightening force detection results corresponding to different levels of measured pre-tightening force, and a pre-tightening force estimation curve of the flange bolt.

[0075] Different levels of pre-tightening force can be applied to the flange bolt by a high-precision torque wrench.

[0076] For example, a torque wrench with an accuracy of ±1% is used to apply a stepwise pre-tightening force to the flange bolt.

[0077] The preset number of times can be determined by those skilled in the art according to prior values and actual conditions.

[0078] For example, the preset number of times is 10 or 15.

[0079] The way of obtaining the pre-tightening force estimation curve can refer to the description of the mapping model part provided subsequently.

[0080] It can be understood that detecting different levels of pre-tightening force can improve the accuracy of the detection result. Repeatedly obtaining the ultrasonic echo signal according to the preset number of times is to take the average to improve the accuracy of the detection data. The pre-tightening force estimation curve can represent the performance of the flange bolt corresponding to other levels of pre-tightening force. The other levels of pre-tightening force are pre-tightening forces that have not been actually measured.

[0081] Preferably, in step S102, the ultrasonic echo signal at each end is decomposed into an intrinsic modal component based on a characteristic modal decomposition algorithm, including: inputting the ultrasonic echo signal into a constrained variational model of the characteristic modal decomposition algorithm to obtain a signal intrinsic modal. Based on the constraint condition of the characteristic modal decomposition algorithm, the signal intrinsic modal is subjected to modal aliasing suppression processing to obtain the intrinsic modal component.

[0082] The mathematical model of the constrained variational model can be expressed as:

[0083] ;

[0084] The objective function and constraint condition of the characteristic modal decomposition algorithm can be expressed as:

[0085] , ;

[0086] In the formula, represents the ultrasonic echo signal input into the constrained variational model; represents the signal intrinsic modal of the i-th order, which needs to satisfy the narrowband characteristic; represents the maximum decomposition layer number; represents the signal residual component; represents the center frequency of the signal intrinsic modal of the i-th order; represents the time derivative operation, which is used to measure the instantaneous frequency fluctuation of the mode through the gradient; is a Dirac delta function, which is used to represent the unit impulse response; represents the imaginary unit; represents the analytic signal generated by the Hilbert transform of represents the natural constant; represents the L2 norm, which is used to quantify the frequency domain concentration degree of the signal energy; s.t. represents limited by or satisfied by. It can be understood that through the above constraint condition, the signal residual component can be removed, and the signal intrinsic modal of the i-th order can be obtained.

[0087] It can be understood that through the above constraint condition, the signal residual component ​​In the presence of interference, the sum of the decomposed modal components is strictly equal to the input signal.

[0088] By minimizing the energy of the derivative of each modal analytic signal in the frequency domain, each intrinsic modal component is forced to form a narrow-band distribution around its center frequency, thereby suppressing modal aliasing and improving the subsequent optimization effect and detection accuracy.

[0089] Further, based on the constraint condition of the characteristic modal decomposition algorithm, the signal intrinsic modal is subjected to modal aliasing suppression processing to obtain the intrinsic modal component, including: based on the constraint condition of the characteristic modal decomposition algorithm, the signal intrinsic modal is subjected to modal aliasing suppression processing to obtain a constraint modal component. Based on the filter length and the number of frequency bands, the signal spectrum of the constraint modal component is uniformly segmented into multiple frequency bands. Each frequency band is occupied by a constraint modal component through frequency domain truncation optimization to obtain the intrinsic modal component.

[0090] wherein the formula of the frequency domain truncation optimization is:

[0091] ;

[0092] ;

[0093] ;

[0094] In the formula, represents the signal spectrum of the first intrinsic modal component, represents the spectral component of the first intrinsic modal component, represents the frequency of , and represents the selection weight of the first intrinsic modal component, , represents the filter length, represents the number of frequency bands before optimization based on GWO, represents in other cases.

[0095] It can be understood that uniformly segmenting the signal spectrum of the constraint modal component into multiple frequency bands and forcing each frequency band to be occupied by a constraint modal component can avoid high-frequency noise interference.

[0096] For the multi-frequency aliasing characteristics of the GIS bolt ultrasonic signal, adaptive spectral segmentation is performed through the above method, which can overcome the noise sensitivity problem of related technologies such as the empirical mode decomposition (EMD) method.

[0097] Preferably, in step S103, the feature parameters of the feature modal decomposition algorithm are optimized based on the grey wolf optimization algorithm and the intrinsic modal component to obtain optimized feature parameters, including steps S1031 to S1033.

[0098] Step S1031, the selection weight of the intrinsic modal component and the number of frequency bands (the number of decomposition layers) are encoded to obtain the corresponding grey wolf position vector , wherein represents the selection weight of the intrinsic modal component corresponding to the first position vector .

[0099] Step S1032, the grey wolf position vector is updated and iterated based on the fitness function to obtain a final position vector, wherein the optimization is stopped when the iteration rate of the fitness function is less than a preset threshold.

[0100] wherein the formula of the fitness function is determined according to the energy entropy minimization principle as follows:

[0101] ;

[0102] ;

[0103] In the formula, represents the energy entropy, represents the energy fluctuation ratio of the kth order signal intrinsic modal, and the feature parameters of the feature modal decomposition algorithm include the filter length and the number of frequency bands , represents the maximum number of frequency bands, represents the maximum number of sampling points, represents the current sampling point, and represent the intrinsic modal components corresponding to the ultrasonic echo signals received at the two ends of the flange bolt, respectively, and and come from the same excitation signal.

[0104] The energy entropy minimization principle is that the entropy value of the modal energy proportion is minimized.

[0105] wherein the update mechanism of the grey wolf position vector is as follows: the positions of a wolf, wolf, wolf are updated, and the positions of other wolves are updated. According to the above fitness function, the positions of a wolf, wolf, wolf can be determined.

[0106] Based on the position vector of a wolf Guide the first position vector , which can be expressed as:

[0107] ;

[0108] ;

[0109] ;

[0110] Where, express Wolf distance vector; Represents the position update coefficient, which decays linearly during iteration; and for Random numbers in .

[0111] Accordingly, based on The wolf's position vector leads the second position vector ,based on The wolf's position vector leads the third position vector , this embodiment will not be described in detail here.

[0112] The final position vector can be expressed as Of course, this is not the only option.

[0113] The optimization is stopped when the iterative change rate of the fitness function is less than the preset threshold. The preferred method for this iterative termination condition is: the iterative change rate of the function value of the fitness function in n consecutive iterations is less than the preset threshold. Stop optimization if .

[0114] The above number n and the preset threshold All of these can be determined by those skilled in the art based on prior values ​​and actual conditions.

[0115] For example, n=10, .

[0116] The above iteration termination condition is implemented by combining the dual constraints of energy convergence factor and residual monotonicity. The energy convergence factor is set to ensure that the energy difference between adjacent iterations is less than 1%.

[0117] Step S1033: Determine the optimized feature parameters based on the final position vector.

[0118] The optimized characteristic parameter in this embodiment is the filter length and the number of modes (number of frequency bands / number of decomposition levels) .

[0119] Filter length Affects the frequency band resolution, Too low will cause modal aliasing, Too high will result in excessive additional calculations. The main frequency components of the signal need to be covered to avoid under-decomposition or over-decomposition.

[0120] For example, the filter length is set to , the modal number is set to .

[0121] This embodiment adopts the filter length based on the above and modal number The dual-objective optimization strategy is helpful to obtain accurate reconstruction components, thereby improving the accuracy of preload test results.

[0122] Preferably, step S105 performs a time-frequency correlation analysis on the reconstructed components corresponding to the ultrasonic echo signals at both ends to obtain the flange bolt preload test result. This includes calculating the corresponding fluctuation energy ratio and double-end delay distortion based on the reconstructed components. The fluctuation energy ratio and double-end delay distortion are input into a support vector regression mapping model to obtain a preload value and determine the preload test result. The kernel function of the mapping model is determined through cross-validation.

[0123] The physical meaning of the fluctuation energy ratio is that the increase in preload leads to an increase in contact stiffness and an improvement in the efficiency of high-frequency vibration energy transmission. The formula can be expressed as:

[0124] ;

[0125] Where, is the high-sensitivity modal index selected by the Grey Wolf Optimization Algorithm GWO, represents the number of modes after GWO optimization, Represents The corresponding reconstruction component, Represents The corresponding reconstruction component.

[0126] The double-ended delay distortion represents the cumulative effect of signal distortion caused by wave velocity variations and contact nonlinearity. Those skilled in the art can refer to relevant calculation methods for delay distortion to calculate the double-ended delay distortion. This embodiment also provides a preferred method for calculating the double-ended delay distortion, which will be described in detail later.

[0127] The mapping model based on support vector regression SVR can be expressed as:

[0128] ;

[0129] Where, Indicates the preload value output by the mapping model, represents the above-mentioned double-end delay distortion, Represents the iterative energy difference between adjacent components.

[0130] For example, the mapping model is used to apply preloads to flange bolts in 5kN intervals within a range of 0-60kN. 13 types of measured signal samples are obtained, which are then expanded to 120 sets of samples by injecting noise signals. These 120 sets of samples are divided into 70% training samples and 30% testing samples. Preload estimation is performed using the mapping model using SVR hyperparameter optimization. The noise signal can be, but is not limited to, white noise with an amplitude of ±5%.

[0131] Furthermore, the kernel function of the above mapping model is a radial basis function RBF determined by Bayesian optimization, and the optimal RBF kernel is determined by grid search.

[0132] Furthermore, 5-fold cross-validation was used to determine the kernel function, calculate the coefficient of determination (R²), and the root mean square error (RMSE) to verify the accuracy of the mapping model, making the overfitting risk of the mapping model controllable.

[0133] Preferably, calculating the corresponding two-end delay distortion based on the reconstructed components includes: performing zero-phase filtering on the reconstructed components to obtain a two-end received signal; calculating a normalized cross-correlation function based on the two-end received signals to determine a delay value corresponding to a point with a maximum correlation coefficient; and calculating the two-end delay distortion based on the delay value and a preset nonlinear term.

[0134] Perform zero-phase filtering on the reconstructed component to obtain a double-ended received signal, including: The corresponding reconstructed signal , and the reconstruction component The corresponding reconstructed signal Perform frequency domain filtering to obtain the dual-end receiving signal . It can retain sensitive frequency bands and eliminate low-frequency vibration noise.

[0135] The formula for zero-phase filtering can be expressed as:

[0136] ;

[0137] Where, represents the filtered double-ended received signal, b and represents the filter coefficient; express Function, used to achieve zero-phase filtering to avoid group delay distortion; Represents the reconstruction component.

[0138] Furthermore, based on the dual-end receiving signal A normalized cross-correlation function NCCF is calculated to determine a time delay value corresponding to a maximum correlation coefficient point For:

[0139] ;

[0140] ;

[0141] wherein, denotes a time difference between modal components, denotes a sampling rate of a double-end signal.

[0142] wherein, the time delay value is in units of .

[0143] Further, a cubic spline interpolation is performed on the cross-correlation peak value region, so that the time delay resolution can be improved to 0.1 μs.

[0144] Further, a preset nonlinear term is introduced to describe a physical dependence relationship between the time delay and the pretightening force, so that the calculation of the double-end time delay distortion degree is simplified as the following formula:

[0145] ;

[0146] wherein, denotes a reference pretightening force derived according to the material stiffness of the flange bolt.

[0147] For example, .

[0148] The embodiments of the present application also provide experimental data of the above-mentioned scheme to verify the technical effects, and the specific data are as follows:

[0149] The experimental platform can refer to Figure 2 . The sample bolt is a standard GIS flange bolt with a size of M16*200 mm, which is installed on a GIS flange surface and an axial pretightening force of 5 kN is applied. A hollow column PZT-5H piezoelectric ceramic with a size of Φ18*5*20 mm and a resonance frequency of 1.5 MHz is used as an excitation source, and the excitation source is fixed to the outer surface of the nut through a coupling agent. Ultrasonic transducers are installed at both ends of the screw rod, and the ultrasonic echo signals at both ends are synchronously captured through a sampling rate of 10 MSa / s and a 16-bit resolution acquisition card, and the corresponding double-channel voltage response waveforms are recorded to construct an ultrasonic signal feature analysis sample library.

[0150] The step-by-step pre-tightening force is applied by a torque wrench with an accuracy of ±1%, and each 5kN in the range of 0-60kN is a test point. After 30 seconds of stabilization, the pulse excitation is triggered. The excitation signal is set to a 5-cycle Hanning window modulated sine wave with an amplitude of 5V, which is converted to 100V by a high-voltage pulse excitation circuit, a center frequency of 1.5 MHz, and a high-frequency narrow pulse with a pulse width of 10μs. Each working condition is repeated 10 times for averaging.

[0151] The frequency band segmentation method is used to segment the signal 10 times to obtain 10 frequency bands. The number of segments is set to 9, and 10 frequency bands are obtained.

[0152] The parameters of the gray wolf optimization algorithm GWO are set as follows: the population size sizepop=30, which can balance global search and computational efficiency; the maximum iteration Max_iter=30, which can balance convergence speed and parameter stability; the position update coefficient a is linearly decreased from 2 to 0, which can ensure smooth transition of the search process. When the fitness function value changes by less than a specified threshold for 10 consecutive iterations , the optimization is stopped.

[0153] In combination with the above step S103, the optimal parameter combination is determined as =15, =3.

[0154] Figure 3 is a schematic diagram of a double-end ultrasonic echo signal, and the orange signal is the near-end original ultrasonic echo signal, and the blue signal is the far-end original ultrasonic echo signal. As shown in Figure 2 , the piezoelectric ultrasonic transducer 22 closer to the hollow column piezoelectric ceramic 26 receives the above-mentioned orange signal, and the piezoelectric ultrasonic transducer 22 farther from the hollow column piezoelectric ceramic 26 receives the above-mentioned blue signal.

[0155] Taking the above-mentioned blue signal as an example, Figure 4 is a high-frequency intrinsic modal component obtained by decomposing the above-mentioned blue signal based on step S102, Figure 5 is a high-frequency reconstructed component obtained by decomposing the above-mentioned blue signal based on step S104, Figure 6 is a medium-frequency intrinsic modal component obtained by decomposing the above-mentioned blue signal based on step S102, Figure 7 is a medium-frequency reconstructed component obtained by decomposing the above-mentioned blue signal based on step S104.

[0156] Comparing Figure 4 and Figure 5 , it can be seen that the signal characteristics of the high-frequency reconstructed component are more obvious than those of the high-frequency intrinsic modal component. Comparing Figure 6 and Figure 7 , it can be seen that the signal characteristics of the medium-frequency reconstructed component are more obvious than those of the medium-frequency intrinsic modal component. In combination with Figure 4 to Figure 7It can be seen that, based on the grey wolf optimization algorithm and the characteristic parameter optimization of the intrinsic modal component decomposition algorithm, and then based on the optimized characteristic modal decomposition algorithm, the characteristic of the modal component obtained by decomposing the ultrasonic echo signal can be easily distinguished, which helps to improve the detection accuracy and stability.

[0157] Figure 8 The convergence curve of the grey wolf optimization algorithm GWO includes the energy entropy convergence curve of the near-end intrinsic modal component corresponding to the near-end original ultrasonic echo signal, and the energy entropy convergence curve of the far-end intrinsic modal component corresponding to the far-end original ultrasonic echo signal.

[0158] Taking the energy entropy convergence curve of the far-end intrinsic modal component as an example, the energy entropy decreases from the initial 095 to 0.44 and remains stable, and the convergence is completed within 20 iterations, indicating that the convergence speed is relatively fast.

[0159] Figure 9 The optimization schematic diagram of the pre-tightening force estimation by the mapping model using SVR hyperparameter optimization, the minimum target value of the estimation is highly consistent with the observed minimum target value, that is, the estimated value is basically consistent with the actual value, indicating that the estimation accuracy is relatively high.

[0160] In the range of 0~60kN, the corresponding pre-tightening force is applied to the flange bolt at an interval of 5kN, and a total of 13 types of measured signal samples are obtained, which are expanded to 120 groups of samples by injecting ±5% amplitude white noise. The above 120 groups of samples are divided into 70% training samples and 30% test samples, and the pre-tightening force is estimated by the mapping model using SVR hyperparameter optimization.

[0161] For example, under the condition of 5kN pre-tightening force, the corresponding time delay value calculated based on the above method is =0.233 The corresponding double-end time delay distortion degree is =0.0225.

[0162] Further, the kernel function of the above mapping model is the radial basis function RBF determined by Bayesian optimization, and the optimal RBF kernel is determined by grid search, and the corresponding parameters are penalty factor C=12.5 and kernel width γ=0.8.

[0163] Further, the above kernel function is determined by 5-fold cross-validation, and the determination coefficient R²=0.934 and the root mean square error RMSE=0.82kN are calculated, indicating that the overfitting risk of the mapping model is controllable.

[0164] Figure 10A pre-tightening force estimation curve is shown in the figure, wherein the standard pre-tightening force is a measured pre-tightening force applied to the flange bolt at an interval of 5 kN in the range of 0-60 kN, and 13 pre-tightening force estimation values corresponding to the 13 types of measured signal samples are obtained based on the above method, that is, 13 numerical points in the pre-tightening force estimation curve. Figure 10 According to the numerical points, the pre-tightening force estimation curve of the flange bolt is fitted. It is shown that the above method provided by the embodiment can accurately and stably detect the pre-tightening force.

[0165] Further, through the above method provided by the embodiment, a pre-tightening force estimation model of bolts of different specifications can be constructed. In actual application, the measured ultrasonic signal is substituted into the corresponding model to obtain the corresponding pre-tightening force value, and then the loosening state of the to-be-measured bolt is judged.

[0166] In summary, the above method provided by the embodiment of the application breaks through the technical bottleneck of related bolt pre-tightening force detection, and realizes accurate quantization of the bolt connection state by constructing a self-adaptive estimation framework of feature modal decomposition parameter optimization, multi-dimensional feature correlation screening and nonlinear prediction model cooperation. Based on the internal physical characteristics of the vibration acoustic signal, a parameter-adaptive feature modal decomposition algorithm is used to extract a dynamic intrinsic mode strongly associated with the pre-tightening force; a hierarchical sensitivity screening is performed on the feature cluster by combining a bionic optimization algorithm to eliminate redundant interference; and finally, a general regression model of mechanical state and physical response is constructed by mapping modeling in a high-dimensional feature space enhanced by a kernel function. The method can effectively solve the industry pain points of strong subjectivity of traditional torque wrench manual calibration and insufficient generalization ability of ultrasonic guided wave mode recognition.

[0167] Compared with related detection methods relying on discrete artificial detection or single physical quantity discrimination, the accuracy and working condition adaptability of bolt loosening state recognition are significantly improved through hierarchical feature optimization and integrated learning mechanism, and the detection process is upgraded from discrete experience judgment to continuous data intelligent analysis, which overcomes the technical barriers of strong artificial dependence and poor recognition stability under complex load in the online detection of bolt connection state of power equipment, and provides reliable technical support for safe operation of GIS equipment.

[0168] The embodiment of the application also provides an electronic device, including: at least one processor; and a memory connected in communication with the at least one processor. The above-mentioned memory stores a computer program capable of being executed by the above-mentioned at least one processor, and the above-mentioned computer program is used to make the electronic device execute the method of the embodiment of the application when being executed by the above-mentioned at least one processor.

[0169] Reference Figure 11, a block diagram of an electronic device that can be a server or a client of an embodiment of the present invention will now be described, which is an example of a hardware device that can be applied to aspects of the present invention. The electronic device is intended to represent various forms of digital electronic computing devices such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computing devices. The electronic device can also represent various forms of mobile devices such as personal digital processing, cellular telephones, smart phones, wearable devices, and other similar computing devices. The components, their connections, and their functions, as shown in the figures, and their functions, are merely examples and are not intended to limit implementations of the present invention described and / or claimed herein.

[0170] As shown in FIG. 11, Figure 11 The electronic device includes a computing unit 1101 that can perform various appropriate actions and processes in accordance with a computer program stored in a read only memory (ROM) 1102 or a computer program loaded from a storage unit 1108 into a random access memory (RAM) 1103. Various programs and data required for operation of the electronic device can also be stored in the RAM 1103. The computing unit 1101, the ROM 1102, and the RAM 1103 are connected to each other through a bus 1104. An input / output (I / O) interface 1105 is also connected to the bus 1104.

[0171] Various components in the electronic device are connected to the I / O interface 1105, including an input unit 1106, an output unit 1107, the storage unit 1108, and a communication unit 1109. The input unit 1106 can be any type of device that can input information to the electronic device, and can receive inputted digital or character information, and generate key signal inputs related to user settings and / or function controls of the electronic device. The output unit 1107 can be any type of device that can present information, and can include, but is not limited to, a display, a speaker, a video / audio output terminal, a vibrator, and / or a printer. The storage unit 1108 can include, but is not limited to, a magnetic disk, an optical disk. The communication unit 1109 allows the electronic device to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks, and can include, but is not limited to, a modem, a network card, an infrared communication device, and / or a wireless communication transceiver, such as a Bluetooth device, a WiFi device, a WiMax device, a cellular communication device, and / or the like.

[0172] The computing unit 1101 can be various general and / or special purpose processing components with processing and computing capabilities. Some examples of the computing unit 1101 include, but are not limited to, CPUs, graphics processing units (GPUs), various special-purpose artificial intelligence (AI) computing units, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 1101 performs various methods and processes described above. For example, in some embodiments, the method embodiments of the present application can be implemented as a computer program tangibly embodied in a machine-readable medium, such as the storage unit 1108. In some embodiments, parts or all of the computer program can be loaded and / or installed onto the electronic device via the ROM 1102 and / or the communication unit 1109. In some embodiments, the computing unit 1101 can be configured to perform the methods described above by other appropriate means, such as with the aid of firmware.

[0173] The embodiments of the present application also provide a non-transitory machine-readable medium storing a computer program, wherein the computer program, when executed by a processor of a computer, causes the computer to perform the method of the embodiments of the present application.

[0174] The embodiments of the present application also provide a computer program product comprising a computer program, wherein the computer program, when executed by a processor of a computer, causes the computer to perform the method of the embodiments of the present application.

[0175] The computer program for implementing the method of the embodiments of the present application can be written in any combination of one or more programming languages. The computer program can be provided to a processor or controller of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the computer program, when executed by the processor or controller, enables the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The computer program can be executed entirely on a machine, partially on a machine, partially on a machine as part of a standalone software package, and partially on a remote machine or server.

[0176] In the context of embodiments of the present invention, a machine-readable medium can be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable signal medium can include, but is not limited to, electronic, magnetic, optical, electromagnetic, or infrared signals, for example. A machine-readable storage medium can include, but is not limited to, volatile memory, non-volatile memory, or any suitable combination thereof. More specific examples of machine-readable storage media will now be given with reference to the following description.

[0177] It is noted that the terms "comprises" and variations thereof do not have a limiting meaning where "comprises" is used in the context of what a feature comprises. Also, the terms "based on" and "one embodiment" are not intended to mean "based only on" or "one and only embodiment." Rather, the term "one embodiment" means "at least one embodiment." The terms "another embodiment" and "some embodiments" mean "at least one additional embodiment." The terms "one embodiment" and "some embodiments" do not mean "one and only embodiment" or "some embodiments only." The terms "a or an" are used herein to mean "one or more" unless specified otherwise.

[0178] The various steps of the method embodiments provided by embodiments of the present invention can be performed in different orders and / or in parallel. Furthermore, the method embodiments can include additional steps and / or omit performing the steps shown. The scope of protection of the present invention is not limited in this respect.

[0179] The word "comprise" or variations such as "comprises" or "comprising" within this description is used throughout to mean the term "include" or "including" but not limited to. The word "example" is used exclusively herein to mean "an example of. As used herein, the term "exemplary" or "for example" means "an example of" or "an example," and is merely intended for illustration. The terms "a cause of" and "a cause for" are used interchangeably and mean "at least part of the cause of" or "at least part of the cause for." The terms "another" and "an additional" are defined as "at least one more." The terms "first" and "second" are used to describe a difference and are not used to designate an order of importance or to imply a quantity of one or more of the technical features indicated.

[0180] The above embodiments only express several implementation manners of the present application, the description is more specific and detailed, but it cannot be understood as the limitation of the protection scope. It should be pointed out that for ordinary skilled in the art, without departing from the concept of the present application, several modifications and improvements can be made, which belong to the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the appended claims.

Claims

1. A method for detecting the preload force of GIS flange bolts, characterized in that: include: Obtain ultrasonic echo signals from both ends of the flange bolts of the gas-insulated switchgear GIS; Decomposing the ultrasonic echo signal at each end into eigenmode components based on an eigenmode decomposition algorithm; Optimizing characteristic parameters of the characteristic mode decomposition algorithm based on the Grey Wolf optimization algorithm and the eigenmode components to obtain optimized characteristic parameters; Decomposing the ultrasonic echo signal at each end into reconstructed components based on an eigenmode decomposition algorithm using the optimized eigenvalues; Performing time-frequency correlation analysis on the reconstructed components corresponding to the ultrasonic echo signals at both ends to obtain the preload force detection result of the flange bolt, including: Calculating the corresponding fluctuation energy ratio and double-end delay distortion based on the reconstructed components; Inputting the fluctuation energy ratio and the double-end delay distortion into a support vector regression mapping model to obtain a preload value and determine the preload detection result, wherein the kernel function of the mapping model is determined by cross-validation; Calculating the corresponding two-end delay distortion based on the reconstructed component includes: performing zero-phase filtering on the reconstructed component to obtain a two-end received signal; Calculating a normalized cross-correlation function based on the dual-end received signals, and determining a time delay value corresponding to a point with a maximum correlation coefficient; The double-end delay distortion is calculated based on the delay value and a preset nonlinear term.

2. The method according to claim 1, characterized in that Decomposing the ultrasonic echo signal at each end into eigenmode components based on an eigenmode decomposition algorithm includes: Inputting the ultrasonic echo signal into the constrained variational model of the eigenmode decomposition algorithm to obtain the signal eigenmode; Based on the constraints of the eigenmode decomposition algorithm, modal aliasing suppression processing is performed on the eigenmode of the signal to obtain the eigenmode component.

3. The method according to claim 2, characterized in that Based on the constraints of the eigenmode decomposition algorithm, modal aliasing suppression processing is performed on the eigenmode of the signal to obtain the eigenmode component, including: Based on the constraint conditions of the eigenmode decomposition algorithm, modal aliasing suppression processing is performed on the eigenmode of the signal to obtain constrained modal components; Based on the filter length and the number of frequency bands, the signal spectrum of the constrained modal component is evenly divided into a plurality of frequency bands; Constraining each of the frequency bands to be occupied by one constrained modal component by frequency domain truncation optimization to obtain the eigenmode component; The formula for frequency domain truncation optimization is expressed as: ； ； ； Where, Indicates the The signal spectrum of the eigenmode components, ; Indicates the The spectral components of the eigenmode components, for frequency, Indicates the The selection weights of the eigenmode components, , represents the filter length, represents the number of frequency bands, In other cases.

4. The method according to claim 1, wherein Optimizing the characteristic parameters of the characteristic mode decomposition algorithm based on the Grey Wolf optimization algorithm and the eigenmode components to obtain optimized characteristic parameters includes: Encoding the selection weights and the number of frequency bands of the intrinsic mode components to obtain a corresponding gray wolf position vector; Iteratively updating the gray wolf position vector based on a fitness function to obtain a final position vector, wherein the optimization is stopped when an iterative change rate of the fitness function is less than a preset threshold; determining the optimized feature parameters based on the final position vector; The formula for determining the fitness function according to the principle of minimizing energy entropy is: ； ； Where, Represents energy entropy. The characteristic parameters of the eigenmode decomposition algorithm include filter length Sum frequency band number , Indicates the maximum number of frequency bands, Indicates the maximum number of sampling points, Indicates the current sampling point, and represents the eigenmode components corresponding to the ultrasonic echo signals received at both ends of the flange bolt, and and from the same stimulus signal.

5. The method according to claim 1, wherein The flange bolt is a flange bolt with a single screw structure without a screw head; Before obtaining ultrasonic echo signals from both ends of the flange bolts of the gas-insulated switchgear GIS, the method further includes: A hollow cylindrical piezoelectric ceramic is passed through the screw of the flange bolt and installed on the outer surface of any nut of the flange bolt as an excitation source, and piezoelectric ultrasonic transducers are respectively installed at both ends of the screw as receiving ends; The pulse driver is controlled to generate a high-frequency narrow pulse according to a preset center frequency to drive the excitation source to generate an excitation signal, so as to obtain the ultrasonic echo signal at the receiving end, wherein the pulse driver is connected to the hollow cylindrical piezoelectric ceramic.

6. The method according to claim 1, characterized in that Acquire ultrasonic echo signals from both ends of the flange bolts of the gas-insulated switchgear (GIS), including: Under different levels of measured preload, repeatedly obtaining multiple ultrasonic echo signals from both ends of the flange bolt according to a preset number of times; A sample library is constructed based on the multiple ultrasonic echo signals, for obtaining multiple preload force detection results corresponding to the different levels of measured preload forces, and a preload force estimation curve of the flange bolt.

7. An electronic device comprising: A processor and a memory storing a program, wherein the program comprises instructions which, when executed by the processor, cause the processor to perform the method according to any one of claims 1 to 6.

8. A non-transitory machine-readable medium storing computer instructions, characterized in that: The computer instructions are used to cause the computer to execute the method according to any one of claims 1 to 6.

Images