A flexible hose high frequency micro-vibration durability test system

Abstract

The present application relates to the technical field of vibration test, in particular to a flexible hose high-frequency micro-vibration durability test system, the system comprises: obtaining service micro-vibration signal, processing the service micro-vibration signal by using fatigue damage statistical algorithm to obtain target fatigue damage spectrum; processing the service micro-vibration signal by using extreme value envelope scanning algorithm to obtain high-frequency wave peak segment; constructing a double-target constraint model containing damage equivalent boundary and limit impact boundary; combining the double-target constraint model to calculate the optimal scaling coefficient of the high-frequency wave peak segment, adjusting the high-frequency wave peak segment according to the optimal scaling coefficient to obtain a scaling segment. In the present application, the closed-loop correction mechanism ensures that the finally generated test signal is highly consistent with the actual service damage, and a target test driving signal with extremely compressed test time and real and reliable results is finally obtained, thereby improving the efficiency of flexible hose high-frequency micro-vibration durability test.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of vibration test, and in particular to a flexible hose high-frequency micro-vibration durability test system. BACKGROUND

[0002] Vibration test technology is to study and evaluate the mechanical response, fatigue life and structural integrity of various products, structural parts or materials under complex dynamic mechanical environment.

[0003] The existing vibration test technology has limitations in processing specific high-frequency micro-vibration durability evaluation. The traditional method often uses the way of lifting the overall vibration level or compressing the time history in proportion to accelerate the test, but this simple amplification processing mode lacks deep insight into the damage mechanism, and is easy to cause the test load to deviate from the real working condition. On the one hand, in order to pursue test efficiency, the load is excessively amplified, which may cause plastic deformation or instantaneous impact fracture of the test piece that will not occur in actual service, resulting in a conservative or even wrong failure conclusion, and it is difficult to truly reflect the durability of the product. On the other hand, if a lower acceleration coefficient is used to ensure authenticity, the test period is long and the cost is high, which is difficult to meet the rapid iteration of research and development requirements. Therefore, it needs to be improved. SUMMARY

[0004] The purpose of the present application is to solve the problems existing in the prior art and to provide a flexible hose high-frequency micro-vibration durability test system.

[0005] In order to achieve the above-mentioned purpose, the present application adopts the following technical scheme: a flexible hose high-frequency micro-vibration durability test system comprises:

[0006] Obtain the service micro-vibration signal, process the service micro-vibration signal by using the fatigue damage unified algorithm to obtain the target fatigue damage spectrum;

[0007] Process the service micro-vibration signal by using the extreme value envelope scanning algorithm to obtain the high-frequency peak segment;

[0008] A double-objective constraint model containing damage equivalent boundary and limit impact boundary is constructed;

[0009] The optimal scaling coefficient of the high-frequency peak segment is calculated in combination with the double-objective constraint model, and the high-frequency peak segment is adjusted according to the optimal scaling coefficient to obtain a scaled segment;

[0010] Splice each scaled segment to generate an accelerated driving signal, and calculate the accelerated damage spectrum of the accelerated driving signal;

[0011] Adjust the peak value of the accelerated driving signal according to the error between the accelerated damage spectrum and the target fatigue damage spectrum to obtain a target test driving signal;

[0012] The target test drive signal is used to control the test console to perform a durability test on the flexible hose.

[0013] Preferably, the steps of acquiring the service micro-vibration signal and processing the service micro-vibration signal using a fatigue damage statistical algorithm to obtain the target fatigue damage spectrum are as follows:

[0014] Long-term vibration data of the flexible hose is collected, and continuous time dimension features and continuous acceleration dimension features are extracted from the long-term vibration data. The continuous time dimension features and the continuous acceleration dimension features are combined to generate the service micro-vibration signal.

[0015] The high-frequency band range parameters for testing the flexible hose are set. Based on these parameters, the dynamic response of the flexible hose is equivalent to a set of single-degree-of-freedom spring systems. For each single-degree-of-freedom spring system in the set, the reference displacement data corresponding to the service micro-vibration signal is calculated. The reference displacement data is processed using a standard rainflow counting model to extract the cumulative number of stress cycles and the corresponding amplitude of stress cycles. The cumulative number of stress cycles and the corresponding amplitude of stress cycles are processed using a material fatigue damage model to calculate the basic frequency band fatigue damage value for each single-degree-of-freedom spring system. All the basic frequency band fatigue damage values ​​are collected to generate an initial damage data set. The initial damage data set is mapped to a target frequency domain coordinate system. The target fatigue damage spectrum is generated based on the mapped initial damage data set.

[0016] Preferably, the step of processing the service micro-vibration signal using the extreme value envelope scanning algorithm to obtain the high-frequency peak segment is as follows:

[0017] Obtain the fatigue extreme value data of the base material of the flexible hose, calculate the product of the fatigue extreme value data of the base material and the preset small proportional coefficient, and obtain the high-frequency stress damage judgment threshold parameter.

[0018] The service micro-vibration signal is extracted using a time-domain sliding window model to generate multiple candidate segments. For each of the multiple candidate segments, the extreme value envelope scanning algorithm is run to analyze the absolute envelope maxima of each candidate segment.

[0019] The absolute envelope maximum of each candidate segment is compared with the high-frequency stress damage determination threshold parameter. If the absolute envelope maximum of each candidate segment is lower than the high-frequency stress damage determination threshold parameter, each candidate segment is marked as a redundant segment. If the absolute envelope maximum of each candidate segment is not lower than the high-frequency stress damage determination threshold parameter, each candidate segment is marked as a high-frequency peak segment. All redundant segments are removed from the service micro-vibration signal, and each high-frequency peak segment in the service micro-vibration signal is retained. The initial occurrence timestamp data of each high-frequency peak segment in the service micro-vibration signal is extracted. A temporal correlation relationship for each high-frequency peak segment is constructed based on the initial occurrence timestamp data. All high-frequency peak segments are stored using the temporal correlation relationship.

[0020] Preferably, the steps for constructing a dual-objective constraint model that includes both damage equivalent boundaries and ultimate impact boundaries are as follows:

[0021] Obtain the material yield limit stress parameter of the flexible hose, evaluate the displacement critical state parameter of the flexible hose to undergo plastic deformation based on the material yield limit stress parameter, set the maximum critical acceleration spectrum limit parameter based on the displacement critical state parameter, generate the ultimate impact boundary using the maximum critical acceleration spectrum limit parameter, and obtain the equivalent fatigue damage mapping function.

[0022] The equivalent fatigue damage mapping function is used to define the damage equivalent boundary of the high-frequency peak segment during scaling. A first objective function to be optimized, including a nonlinear amplification coefficient dimension, is introduced, and a second objective function to be optimized, including a frequency domain time cyclic increase coefficient dimension, is introduced. The first and second objective functions to be optimized are combined to generate a basic bi-objective function. The maximum critical acceleration spectrum limit parameter is configured as a safety envelope inequality constraint condition of the basic bi-objective function, and the damage equivalent boundary is configured as a fatigue conservation equality constraint condition of the basic bi-objective function. A penalty function of the system Lagrange multiplier optimization algorithm is configured. If the basic bi-objective function violates the safety envelope inequality constraint condition, the cost weight of the first objective function to be optimized is modified using the penalty function. The safety envelope inequality constraint condition, the fatigue conservation equality constraint condition, the penalty function, and the basic bi-objective function are integrated to generate the bi-objective constraint model.

[0023] Preferably, the specific steps for calculating the optimal scaling factor of the high-frequency peak segment using the dual-objective constraint model, and adjusting the high-frequency peak segment according to the optimal scaling factor to obtain the scaled segment are as follows:

[0024] Extract the basic amplitude feature data and basic time cycle frequency data from the high-frequency peak segment, input the basic amplitude feature data and the basic time cycle frequency data into the dual-objective constraint model, and use the dual-objective constraint model to perform optimization calculation on the basic amplitude feature data and the basic time cycle frequency data to generate the optimal scaling coefficient;

[0025] The optimal amplitude amplification factor for the basic amplitude characteristic data is extracted from the optimal scaling factor, and the optimal time cycle increment factor for the basic time cycle frequency data is extracted from the optimal scaling factor. The basic amplitude characteristic data and the optimal amplitude amplification factor are then processed using a scaling calculation formula based on fatigue equivalence and dynamic stiffness compensation. The scaling calculation formula based on fatigue equivalence and dynamic stiffness compensation is as follows: ,in This represents the amplitude value after compensation. This represents the basic amplitude characteristic data. This represents the optimal amplitude amplification factor. This indicates the dominant frequency of the high-frequency peak segment. This represents the first-order natural frequency of the flexible hose;

[0026] The amplitude compensation adjustment segment is obtained by replacing the basic amplitude feature data in the high-frequency peak segment with the compensated amplitude value. The basic time cycle frequency data of the amplitude compensation adjustment segment on the time axis is amplified by the optimal time cycle increase coefficient to obtain the time-domain tolerance adjustment segment. The time-domain tolerance adjustment segment is output and configured as the scaling segment.

[0027] Preferably, the steps of splicing the various scaled segments to generate an acceleration driving signal and calculating the acceleration damage spectrum of the acceleration driving signal are as follows:

[0028] Read each of the scaling segments and the timing relationship stored in the system memory, sort all the scaling segments according to the initial occurrence timestamp data in the timing relationship, determine the global splicing order parameter of all the scaling segments, and connect two adjacent scaling segments using a cosine smooth transition window function according to the global splicing order parameter.

[0029] The cosine smooth transition window function is used to eliminate the endpoint phase abruptness data between two adjacent scaling segments to obtain the acceleration drive signal after continuous time series splicing. A standard single-degree-of-freedom spring system set is configured, and the acceleration drive signal is processed using an impact response calculation model based on the viscoelastic system difference equation. The test displacement data of the acceleration drive signal on each single-degree-of-freedom spring system in the standard single-degree-of-freedom spring system set is calculated. The test displacement data is processed using the standard rainflow counting model, and the cumulative number of acceleration stress cycles and the corresponding amplitude of acceleration stress cycles are extracted from the test displacement data. The cumulative number of acceleration stress cycles and the corresponding amplitude of acceleration stress cycles are processed using the material fatigue damage model, and the acceleration band fatigue damage value generated by the acceleration drive signal is calculated. Based on the acceleration band fatigue damage value, the acceleration damage spectrum of the acceleration drive signal is plotted in the target frequency domain coordinate system.

[0030] Preferably, the step of adjusting the peak value of the acceleration driving signal according to the error between the accelerated damage spectrum and the target fatigue damage spectrum to obtain the target test driving signal specifically includes:

[0031] The system inputs verification boundary parameters, extracts actual spectral line continuous trajectory data contained within the verification boundary parameters from the accelerated damage spectrum, extracts expected spectral line continuous trajectory data contained within the verification boundary parameters from the target fatigue damage spectrum, introduces a weighted least squares fitting model, and uses the weighted least squares fitting model to compare and process the actual spectral line continuous trajectory data and the expected spectral line continuous trajectory data to obtain the current squared continuous error integral value for the verification boundary parameters.

[0032] Calculate the peak deviation data between the actual continuous trajectory data of the spectral line and the expected continuous trajectory data of the spectral line, and use the peak deviation data to modify the global absolute peak distribution state parameter in the acceleration driving signal in reverse to generate the iteratively corrected acceleration driving signal.

[0033] The updated square continuous error integral value corresponding to the iteratively corrected acceleration driving signal is calculated using the impact response calculation model based on the viscoelastic system difference equation. The updated square continuous error integral value is compared with the global preset error threshold parameter. The global absolute peak distribution state parameter of the iteratively corrected acceleration driving signal is continuously iteratively modified according to the direction of the peak deviation data until the latest square continuous error integral value extracted by the system is lower than the global preset error threshold parameter. The iteratively corrected acceleration driving signal that meets the condition of the global preset error threshold parameter in the final convergence state is configured as the target test driving signal.

[0034] Preferably, the steps for using the target test drive signal to control the test console to perform durability testing on the flexible hose are as follows:

[0035] The target test drive signal is imported into the digital signal processing component, and the digital signal processing component performs digital-to-analog conversion on the target test drive signal to convert the target test drive signal into an analog drive signal, and establishes a communication link between the digital signal processing component and the test console.

[0036] The analog drive signal is input to the servo controller of the test console through the communication link, and the servo controller amplifies the analog drive signal to generate physical electromagnetic excitation force parameters.

[0037] The high-frequency vibration table device, electrically connected to the test control console, is driven by the physical electromagnetic excitation force parameters. The high-frequency vibration table device is controlled to apply a dynamic high-frequency micro-vibration cyclic load to the flexible hose under test according to the acceleration time history corresponding to the target test drive signal. During the continuous operation of the high-frequency vibration table device, the physical failure deformation state parameters of the flexible hose are collected in real time. When the physical failure deformation state parameters reach the fracture critical value, the high-frequency micro-vibration durability test of the flexible hose is terminated.

[0038] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0039] In this invention, by utilizing a fatigue damage statistical algorithm to process micro-vibration signals under actual service conditions, a target fatigue damage spectrum can be generated as the sole benchmark for subsequent tests. Simultaneously, an extreme value envelope scanning algorithm is employed to accurately identify and extract key high-frequency peak segments in the signal that cause damage, eliminating redundant data that contributes little to fatigue accumulation, thus improving the targeting and efficiency of signal processing. Furthermore, a dual-objective constraint, including damage equivalence boundaries and ultimate impact boundaries, is constructed, providing a rigorous mathematical framework and physical constraints for accelerating the signal scaling process in testing. This ensures that the fatigue damage accumulated in the test is equivalent to that in the real service environment, while avoiding the introduction of unrealistic instantaneous impact damage due to excessive signal amplification, thereby achieving an optimal balance between testing speed and testing realism. The optimal scaling factor for each peak segment is calculated based on this constraint, and then the segments are spliced ​​together to form a highly condensed acceleration drive signal. The acceleration damage spectrum of this signal is then compared with the target fatigue damage spectrum for error and peak value is iteratively adjusted. This closed-loop correction mechanism ensures that the final generated test signal is highly consistent with the actual service damage, ultimately obtaining a target test drive signal with extremely compressed test time and reliable results, thus improving the efficiency of high-frequency micro-vibration durability testing of flexible hoses. Attached Figure Description

[0040] Figure 1 This is a schematic diagram of the principle of the present invention. Detailed Implementation

[0041] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0042] Please see Figure 1 This invention provides a technical solution: a high-frequency micro-vibration durability testing system for flexible hoses, comprising:

[0043] The service micro-vibration signal is acquired, and the service micro-vibration signal is processed using a fatigue damage statistical algorithm to obtain the target fatigue damage spectrum;

[0044] The extreme value envelope scanning algorithm was used to process the service micro-vibration signal to obtain high-frequency peak segments;

[0045] Construct a dual-objective constraint model that includes damage equivalent boundary and ultimate impact boundary;

[0046] The optimal scaling factor of the high-frequency peak segment is calculated by combining the dual-objective constraint model, and the high-frequency peak segment is adjusted according to the optimal scaling factor to obtain the scaled segment;

[0047] The various scaled segments are spliced ​​together to generate an acceleration driving signal, and the acceleration damage spectrum of the acceleration driving signal is calculated.

[0048] The peak value of the acceleration drive signal is adjusted according to the error between the accelerated damage spectrum and the target fatigue damage spectrum to obtain the target test drive signal;

[0049] The durability test of the flexible hose was carried out using a test console controlled by the target test drive signal.

[0050] In this embodiment, the steps for acquiring service micro-vibration signals and processing them using a fatigue damage statistical algorithm to obtain the target fatigue damage spectrum are as follows: Long-history vibration data of the flexible hose is collected; continuous time dimension features and continuous acceleration dimension features are extracted from the long-history vibration data; the continuous time dimension features and continuous acceleration dimension features are combined to generate service micro-vibration signals; a test high-frequency band range parameter for the flexible hose is set; the dynamic response of the flexible hose is equivalent to a set of single-degree-of-freedom spring systems based on the test high-frequency band range parameter; for each single-degree-of-freedom spring system in the set of single-degree-of-freedom spring systems, the reference displacement data corresponding to each single-degree-of-freedom spring system is calculated; the reference displacement data is processed using a standard rainflow counting model; the cumulative number of stress cycles and the corresponding amplitude of stress cycles are extracted from the reference displacement data; the cumulative number of stress cycles and the corresponding amplitude of stress cycles are processed using a material fatigue damage model; the basic frequency band fatigue damage value of each single-degree-of-freedom spring system is calculated; all basic frequency band fatigue damage values ​​are collected to generate an initial damage data set; the initial damage data set is mapped to the target frequency domain coordinate system; and the target fatigue damage spectrum is generated based on the mapped initial damage data set.

[0051] Specifically, the vibration acceleration time history of the flexible hose under actual working conditions for at least 1000 consecutive hours is collected as long-history vibration data. The continuous time dimension and continuous acceleration dimension features are extracted from the long-history vibration data. Acceleration data is paired with corresponding timestamps to form a two-dimensional array, generating a service micro-vibration signal. The high-frequency band range parameter for testing the flexible hose is set, for example, 50 Hz to 2000 Hz based on the vibration characteristics of the equipment connected to the hose. Based on the high-frequency band range parameter, the dynamic response of the flexible hose is equivalent to a set of 195 single-degree-of-freedom spring systems, where the natural frequency of each system corresponds to a logarithmically divided frequency point within the band range. For each single-degree-of-freedom spring system in the set, its second-order ordinary differential equation of motion under the service micro-vibration signal as the basic excitation is solved to calculate the relative reference displacement data time history corresponding to each single-degree-of-freedom spring system. Then, the standard rainflow counting model is used to analyze the relative reference displacement data time history of each single-degree-of-freedom spring system. The reference displacement data of the spring system is used for cyclic counting. The cumulative number of stress cycles and the corresponding amplitude of stress cycles under different amplitude-mean combinations are statistically extracted. Then, based on the SN curve of the flexible hose material, the Palmgren-Miner linear cumulative damage criterion is used as the material fatigue damage model to process the cumulative number of stress cycles and the corresponding amplitude of stress cycles extracted in the previous steps. By calculating the damage ratio under each stress level and summing them, the basic frequency band fatigue damage value of each single-degree-of-freedom spring system under the entire long-term vibration data is calculated. The basic frequency band fatigue damage values ​​of all 195 single-degree-of-freedom spring systems are collected to generate an initial damage data set containing 195 damage values. Each damage value in the initial damage data set is mapped to its corresponding natural frequency of the single-degree-of-freedom system. The points are plotted and connected in the target frequency domain coordinate system with frequency as the abscissa and fatigue damage value as the ordinate. The target fatigue damage spectrum is generated based on the mapped initial damage data set.

[0052] In this embodiment, the steps of processing the service micro-vibration signal using the extreme value envelope scanning algorithm to obtain high-frequency peak segments are as follows: First, acquire the fatigue extreme value data of the base material of the flexible hose; second, calculate the product of the fatigue extreme value data of the base material and a preset small proportional coefficient to obtain the high-frequency stress damage judgment threshold parameter; third, use a time-domain sliding window model to extract the service micro-vibration signal, generating multiple candidate segments; for each candidate segment, run the extreme value envelope scanning algorithm to analyze the absolute envelope maximum value of each candidate segment; fourth, compare the absolute envelope maximum value of each candidate segment with the high-frequency stress damage judgment threshold parameter, and perform high-frequency stress damage judgment on each candidate segment. If the absolute envelope maximum value is lower than the high-frequency stress damage judgment threshold parameter, each candidate segment is marked as a redundant segment. If the absolute envelope maximum value of each candidate segment is not lower than the high-frequency stress damage judgment threshold parameter, each candidate segment is marked as a high-frequency peak segment. All redundant segments are removed from the service micro-vibration signal, and each high-frequency peak segment in the service micro-vibration signal is retained. The initial occurrence timestamp data of each high-frequency peak segment in the service micro-vibration signal is extracted. The temporal correlation relationship for each high-frequency peak segment is constructed based on the initial occurrence timestamp data. All high-frequency peak segments are stored using the temporal correlation relationship.

[0053] Specifically, obtain the fatigue extreme value data of the basic material of the flexible hose from the material handbook or previous tests, such as its performance in... The fatigue limit stress under the second cycle is 25 MPa. The high-frequency stress damage threshold parameter is obtained by multiplying the fatigue limit data of the base material by a preset small proportional coefficient based on engineering experience. For example, by statistically analyzing historical failure data of similar hoses, it was found that stress cycles below 5% of the fatigue limit contribute less than 0.1% to the total damage; therefore, this coefficient is set to 0.05. Thus, the high-frequency stress damage threshold parameter is calculated as 25 MPa multiplied by 0.05, which is 1.25 MPa. Using a time-domain sliding window model with a window length of 2048 data points and a 50% overlap, the entire service micro-vibration signal is segmented to generate multiple candidate segments. For each candidate segment, a Hilbert transform is applied to obtain the analytical signal, and its magnitude is taken to obtain the instantaneous envelope of the signal. An extreme value envelope scanning algorithm is then run to analyze and locate the absolute envelope extreme value of each candidate segment. The maximum absolute envelope value of each candidate segment is compared with the previously calculated high-frequency stress damage judgment threshold parameter of 1.25 MPa. If the maximum absolute envelope value of each candidate segment is lower than the high-frequency stress damage judgment threshold parameter, each candidate segment is marked as a redundant segment. If the maximum absolute envelope value of each candidate segment is not lower than the high-frequency stress damage judgment threshold parameter, each candidate segment is marked as a high-frequency peak segment. All data segments marked as redundant segments are removed from the service micro-vibration signal, and each high-frequency peak segment in the service micro-vibration signal is retained. The timestamp corresponding to the starting point of each high-frequency peak segment in the service micro-vibration signal is extracted as the initial occurrence timestamp data. A data structure is constructed based on the initial occurrence timestamp data. This structure associates the data array of each high-frequency peak segment with its initial occurrence timestamp data to form a temporal association relationship for each high-frequency peak segment. All high-frequency peak segments are stored using the temporal association relationship.

[0054] In this embodiment, the steps for constructing a dual-objective constraint model including a damage equivalent boundary and an ultimate impact boundary are as follows: Obtain the material yield strength stress parameter of the flexible hose; based on the material yield strength stress parameter, evaluate the displacement critical state parameter of the flexible hose undergoing plastic deformation; set the maximum critical acceleration spectrum limit parameter based on the displacement critical state parameter; generate the ultimate impact boundary using the maximum critical acceleration spectrum limit parameter; obtain the equivalent fatigue damage mapping function; define the damage equivalent boundary of the high-frequency peak segment during scaling using the equivalent fatigue damage mapping function; introduce a first objective function to be optimized that includes a nonlinear amplification coefficient dimension; and introduce a dimension that includes a frequency domain time cycle increase coefficient dimension. The second objective function to be optimized is combined with the first and second objective functions to generate a basic bi-objective function. The maximum critical acceleration spectrum limit parameter is configured as a safety envelope inequality constraint condition of the basic bi-objective function, and the damage equivalent boundary is configured as a fatigue conservation equality constraint condition of the basic bi-objective function. The penalty function of the system Lagrange multiplier optimization algorithm is configured. When the basic bi-objective function violates the safety envelope inequality constraint condition, the cost weight of the first objective function to be optimized is modified using the penalty function. The safety envelope inequality constraint condition, the fatigue conservation equality constraint condition, the penalty function, and the basic bi-objective function are integrated to generate a bi-objective constraint model.

[0055] Specifically, the yield strength stress parameter of the flexible hose material is obtained, for example, 350 MPa. Based on the yield strength stress parameter, combined with the hose geometry and the stress-strain curve of the material, the critical displacement state parameter when the flexible hose undergoes irreversible plastic deformation is evaluated and calculated, for example, the maximum allowable relative displacement is 1.5 mm. Based on the critical displacement state parameter, the maximum acceleration value that can produce a 1.5 mm displacement response at different frequencies is calculated in reverse using impact response spectrum theory. Thus, a function curve that varies with frequency is set as the maximum critical acceleration spectrum limit parameter. Using the maximum critical acceleration spectrum limit parameter, a spectral envelope that must not be exceeded in accelerated testing is generated, i.e., the ultimate impact boundary. An equivalent fatigue damage mapping function based on the Miner criterion is obtained. This function describes the mathematical relationship between stress amplitude, number of cycles, and fatigue damage. Using the equivalent fatigue damage mapping function, a mathematical constraint is defined that the high-frequency peak segment must maintain its original fatigue damage value during scaling, i.e., the damage equivalent boundary. A first objective function to be optimized is introduced with a nonlinear amplification factor as the variable. Its goal is to minimize this factor to approximate the actual working condition. Then, further... A second objective function, with a frequency-domain time-cycle increment coefficient as the variable, aims to maximize this coefficient to shorten the testing time. Combining the first and second objective functions, a fundamental bi-objective function is generated through weighted summation, seeking the optimal balance between the amplification coefficient and the time coefficient. The maximum critical acceleration spectrum boundary parameter is configured as a safety envelope inequality constraint condition for the fundamental bi-objective function, while the damage equivalent boundary is configured as a fatigue conservation equality constraint condition for the fundamental bi-objective function. A penalty function for the system's Lagrange multiplier optimization algorithm is then configured. The penalty coefficient is set, for example, the initial penalty coefficient is set to 1000, and if the constraint is still violated in each iteration, the coefficient is multiplied by 1.5. If the calculated response spectrum violates the safety envelope inequality constraint during the optimization process of the basic bi-objective function, a penalty term proportional to the excess is added to the objective function using the penalty function, thereby modifying the cost weight of the first objective function to be optimized, forcing the optimization direction to reduce the amplification coefficient, and integrating the safety envelope inequality constraint, the fatigue conservation equality constraint, the penalty function and the basic bi-objective function to generate a bi-objective constraint model.

[0056] In this embodiment, the steps of calculating the optimal scaling factor of the high-frequency peak segment using a dual-objective constraint model and adjusting the high-frequency peak segment according to the optimal scaling factor to obtain the scaled segment are as follows: Extracting the basic amplitude characteristic data and basic time cycle frequency data from the high-frequency peak segment; inputting the basic amplitude characteristic data and basic time cycle frequency data into the dual-objective constraint model; using the dual-objective constraint model to perform optimization calculations on the basic amplitude characteristic data and basic time cycle frequency data to generate the optimal scaling factor; extracting the optimal amplitude amplification factor for the basic amplitude characteristic data from the optimal scaling factor; extracting the optimal time cycle increase factor for the basic time cycle frequency data from the optimal scaling factor; and processing the basic amplitude characteristic data and optimal amplitude amplification factor using a scaling calculation formula based on fatigue equivalence and dynamic stiffness compensation. The scaling calculation formula based on fatigue equivalence and dynamic stiffness compensation is as follows: ,in This represents the amplitude value after compensation. This represents the basic amplitude characteristic data. This represents the optimal amplitude amplification factor. This indicates the dominant frequency of the high-frequency peak segment. The first natural frequency of the flexible hose is represented. The basic amplitude characteristic data in the high-frequency peak segment is replaced with the compensated amplitude value to obtain the amplitude compensation adjustment segment. The basic time cycle frequency data of the amplitude compensation adjustment segment on the time axis is amplified by the optimal time cycle increase coefficient to obtain the time-domain tolerance adjustment segment. The time-domain tolerance adjustment segment is output and configured as a scaling segment.

[0057] Specifically, for each high-frequency peak segment, the peak value or root mean square value in its time-domain waveform is extracted as the basic amplitude characteristic data. The dominant frequency and cycle number are obtained through Fast Fourier Transform analysis as the basic time cycle frequency data. The basic amplitude characteristic data and basic time cycle frequency data extracted from each high-frequency peak segment are used as initial variables and input into the dual-objective constraint model constructed in the previous steps. Using the dual-objective constraint model, the basic amplitude characteristic data and basic time cycle frequency data are iteratively optimized by calling the Sequence Quadratic Programming (SQP) solver until the objective function converges, thereby generating a set of optimal solutions that simultaneously satisfy damage equivalence and impact boundary constraints, i.e., the optimal scaling factor. The optimal amplitude amplification factor for the basic amplitude characteristic data is extracted from the optimal scaling factor, and the optimal time cycle increment factor for the basic time cycle frequency data is extracted from the optimal scaling factor. The scaling calculation formula based on fatigue equivalence and dynamic stiffness compensation is used to process the basic amplitude characteristic data and the optimal amplitude amplification factor. The scaling calculation formula based on fatigue equivalence and dynamic stiffness compensation is as follows: ,in This represents the amplitude value after compensation. This represents the fundamental amplitude characteristic data extracted from high-frequency peak segments. This represents the optimal amplitude amplification factor obtained through optimization calculation. This indicates the dominant frequency of the high-frequency peak segment. The first natural frequency of the flexible hose is represented by the formula. The compensated amplitude value is calculated according to this formula and used to replace the basic amplitude characteristic data in the original waveform of the high-frequency peak segment to complete the amplitude adjustment and obtain the amplitude compensation adjustment segment. The optimal time cycle increase coefficient obtained by optimization calculation is used to amplify the basic time cycle frequency data on the time axis by resampling or periodically repeating the amplitude compensation adjustment segment on the time axis, so as to obtain the time-domain tolerance adjustment segment. The time-domain tolerance adjustment segment is output and configured as a scaling segment.

[0058] In this embodiment, the steps of splicing various scaling segments to generate an acceleration drive signal and calculating the acceleration damage spectrum of the acceleration drive signal are as follows: Each scaling segment and its temporal correlation stored in the system memory is read; all scaling segments are sorted according to the initial occurrence timestamp data in the temporal correlation; the global splicing order parameters of all scaling segments are determined; adjacent scaling segments are connected using a cosine smooth transition window function based on the global splicing order parameters; the endpoint phase abruptness data between adjacent scaling segments is eliminated using the cosine smooth transition window function to obtain the acceleration drive signal after continuous time series splicing; and a standard single-degree-of-freedom spring system set is configured. The acceleration drive signal is processed using an impact response calculation model based on the difference equation of a viscoelastic system. The test displacement data of the acceleration drive signal on each single-degree-of-freedom spring system in the standard single-degree-of-freedom spring system set is calculated. The test displacement data is processed using a standard rainflow counting model. The cumulative number of acceleration stress cycles and the corresponding amplitude of acceleration stress cycles are extracted from the test displacement data. The cumulative number of acceleration stress cycles and the corresponding amplitude of acceleration stress cycles are processed using a material fatigue damage model. The acceleration frequency band fatigue damage value generated by the acceleration drive signal is calculated. Based on the acceleration frequency band fatigue damage value, the acceleration damage spectrum of the acceleration drive signal is plotted in the target frequency domain coordinate system.

[0059] Specifically, the system reads all scaled segments stored in system memory and records their original time sequence and temporal relationships. Based on the initial occurrence timestamp data in the temporal relationships, all scaled segments are sorted in ascending order to determine the global splicing order parameters. According to these parameters, a cosine smoothing transition window function with a window length of 64 data points is used to window and superimpose the tails and heads of two adjacent scaled segments after sorting. The cosine smoothing transition window function eliminates the step changes caused by direct splicing between adjacent scaled segments, smoothing the phase abrupt changes at the endpoints, resulting in a continuous and seamless signal waveform in the time domain—the acceleration drive signal after continuous time series splicing. A standard single-degree-of-freedom spring system set, identical to that used when generating the target fatigue damage spectrum, is configured, and a difference equation based on a viscoelastic system is used. The impact response calculation model uses the acceleration drive signal as input excitation and calculates the relative displacement response caused by it on each single-degree-of-freedom spring system in the standard single-degree-of-freedom spring system set. This yields test displacement data at each frequency point. The standard rainflow counting model is then used to process the test displacement data at each frequency point, extracting the cumulative number of acceleration stress cycles and the corresponding amplitude of the acceleration stress cycles under accelerated test conditions. Furthermore, the material fatigue damage model, combined with the material's SN curve, is used to process the cumulative number of acceleration stress cycles and the corresponding amplitude of the acceleration stress cycles, calculating the acceleration band fatigue damage value generated by the acceleration drive signal at each frequency point. Based on the acceleration band fatigue damage value, in the target frequency domain coordinate system identical to the target fatigue damage spectrum, with frequency as the abscissa and the calculated damage value as the ordinate, points are plotted and connected to obtain the acceleration damage spectrum of the acceleration drive signal.

[0060] In this embodiment, the step of adjusting the peak value of the acceleration driving signal based on the error between the accelerated damage spectrum and the target fatigue damage spectrum to obtain the target test driving signal is as follows: Obtain the system input verification boundary parameters; extract the actual spectral line continuous trajectory data contained within the verification boundary parameters from the accelerated damage spectrum; extract the expected spectral line continuous trajectory data contained within the verification boundary parameters from the target fatigue damage spectrum; introduce a weighted least squares fitting model; use the weighted least squares fitting model to compare and process the actual spectral line continuous trajectory data and the expected spectral line continuous trajectory data to obtain the current squared continuous error integral value for the verification boundary parameters; calculate the peak deviation data between the actual spectral line continuous trajectory data and the expected spectral line continuous trajectory data; and use... The global absolute peak distribution state parameters in the acceleration drive signal are modified in reverse using peak deviation data to generate an iteratively corrected acceleration drive signal. The updated square continuous error integral value corresponding to the iteratively corrected acceleration drive signal is calculated using an impact response calculation model based on the difference equation of a viscoelastic system. The updated square continuous error integral value is compared with the global preset error threshold parameter. The global absolute peak distribution state parameters of the iteratively corrected acceleration drive signal are continuously iteratively modified in the direction of peak deviation data until the latest square continuous error integral value extracted by the system is lower than the global preset error threshold parameter. The iteratively corrected acceleration drive signal that meets the final convergence state under the condition of the global preset error threshold parameter is configured as the target test drive signal.

[0061] Specifically, the system input, such as a frequency range defined by the test standard or user, is used as the verification boundary parameter. Actual spectral line continuous trajectory data within the verification boundary parameter is extracted from the accelerated damage spectrum generated in the preceding steps. Desired spectral line continuous trajectory data within the same frequency range is extracted from the original target fatigue damage spectrum. A weighted least squares fitting model is introduced, where the weighting function is set, for example, assigning a weight of 1.5 to the peak region of the damage spectrum and a weight of 0.8 to the valley region, to prioritize the consistency of key damage frequency bands. The weighted least squares fitting model is used to compare the actual and desired spectral line continuous trajectory data point by point. By calculating the sum of squares of the weighted errors, the current squared continuous error integral value for the verification boundary parameter is obtained. Simultaneously, the difference between the actual and desired spectral line continuous trajectory data at each frequency point is calculated as peak deviation data. Using the peak deviation data, a proportional feedback control logic is used to reverse-modify the contribution in the acceleration drive signal. The amplitude of the peak segment corresponding to the frequency error is adjusted, i.e., its global absolute peak distribution state parameter is adjusted to generate an iteratively corrected accelerated driving signal. The accelerated damage spectrum corresponding to the iteratively corrected accelerated driving signal is calculated again using the impact response calculation model based on the difference equation of the viscoelastic system, and the updated square continuous error integral value between it and the target spectrum is calculated. A global preset error threshold parameter is set, for example, requiring the total integral error to be less than 2% of the total damage energy of the target spectrum, i.e., set to 0.02. The updated square continuous error integral value is compared with the global preset error threshold parameter. The amplitude is increased according to the direction of the peak deviation data, i.e., the actual damage is lower than expected, and the amplitude is decreased otherwise. The global absolute peak distribution state parameter of the iteratively corrected accelerated driving signal is continuously modified iteratively until the latest square continuous error integral value extracted by the system is lower than the global preset error threshold parameter. The iteratively corrected accelerated driving signal that meets the final convergence state under the condition of the global preset error threshold parameter is configured as the target test driving signal.

[0062] In this embodiment, the steps for using the target test drive signal to control the test console to perform durability testing on the flexible hose are as follows: The target test drive signal is imported into a digital signal processing component; the digital signal processing component performs digital-to-analog conversion on the target test drive signal, converting it into an analog drive signal; a communication link is established between the digital signal processing component and the test console; the analog drive signal is input to the servo controller of the test console through the communication link; the servo controller amplifies the analog drive signal to generate physical electromagnetic excitation force parameters; the physical electromagnetic excitation force parameters are used to drive a high-frequency vibration table device electrically connected to the test console; the high-frequency vibration table device is controlled to apply a dynamic high-frequency micro-vibration cyclic load to the flexible hose under test according to the acceleration time history corresponding to the target test drive signal; the physical failure deformation state parameters of the flexible hose are collected in real time during the continuous operation of the high-frequency vibration table device; the high-frequency micro-vibration durability test on the flexible hose ends when the physical failure deformation state parameters reach the fracture critical value.

[0063] Specifically, the final generated target test drive signal is imported into the digital signal processing component as a digital waveform array. A 16-bit digital-to-analog converter (DAC) within the DAC performs a digital-to-analog conversion operation on the target test drive signal at a sampling rate of 25.6 kHz, converting it into a continuously varying analog voltage signal, i.e., an analog drive signal. A communication link is established between the output port of the DAC and the input port of the servo controller on the test console via a physical cable. Through this communication link, the analog drive signal, with a voltage range of -10V to +10V, is input to the servo controller on the test console in real time. The servo controller's internal power amplifier circuit linearly amplifies the weak input analog drive signal, generating a physical electromagnetic excitation force parameter sufficient to drive the vibration table, i.e., a high-current drive signal. This physical electromagnetic excitation force parameter... The high-frequency vibration table, electrically connected to the test control console, controls the table surface to perfectly replicate the acceleration-time relationship defined by the target test drive signal. This means applying a dynamic high-frequency micro-vibration cyclic load to the flexible hose under test according to the acceleration time history corresponding to the target test drive signal. During continuous operation of the high-frequency vibration table, physical failure deformation parameters of the flexible hose, such as strain values ​​or surface crack lengths, are collected in real time via strain gauges mounted on the hose surface or through a high-frequency camera system. A fracture threshold is set; for example, according to relevant test standards, failure is determined when the surface crack length reaches 5 mm or the internal pressure of the hose drops by more than 30%. When the real-time collected physical failure deformation parameters reach the preset fracture threshold, the high-frequency vibration table is immediately stopped, and the high-frequency micro-vibration durability test of the flexible hose ends.

Claims

1. A high-frequency micro-vibration durability testing system for flexible hoses, characterized in that, The system includes: The service micro-vibration signal is acquired, and the service micro-vibration signal is processed using a fatigue damage statistical algorithm to obtain the target fatigue damage spectrum; The service micro-vibration signal was processed using an extreme value envelope scanning algorithm to obtain high-frequency peak segments; Construct a dual-objective constraint model that includes damage equivalent boundary and ultimate impact boundary; The optimal scaling factor of the high-frequency peak segment is calculated using the dual-objective constraint model, and the high-frequency peak segment is adjusted according to the optimal scaling factor to obtain the scaled segment; The various scaled segments are spliced ​​together to generate an acceleration driving signal, and the acceleration damage spectrum of the acceleration driving signal is calculated. The peak value of the acceleration driving signal is adjusted according to the error between the accelerated damage spectrum and the target fatigue damage spectrum to obtain the target test driving signal; The target test drive signal is used to control the test console to perform a durability test on the flexible hose.

2. The high-frequency micro-vibration durability testing system for flexible hoses according to claim 1, characterized in that, The specific steps for obtaining the service micro-vibration signal and processing it using a fatigue damage statistical algorithm to obtain the target fatigue damage spectrum are as follows: Long-term vibration data of the flexible hose is collected, and continuous time dimension features and continuous acceleration dimension features are extracted from the long-term vibration data. The continuous time dimension features and the continuous acceleration dimension features are combined to generate the service micro-vibration signal. The high-frequency band range parameters for testing the flexible hose are set. Based on these parameters, the dynamic response of the flexible hose is equivalent to a set of single-degree-of-freedom spring systems. For each single-degree-of-freedom spring system in the set, the reference displacement data corresponding to the service micro-vibration signal is calculated. The reference displacement data is processed using a standard rainflow counting model to extract the cumulative number of stress cycles and the corresponding amplitude of stress cycles. The cumulative number of stress cycles and the corresponding amplitude of stress cycles are processed using a material fatigue damage model to calculate the basic frequency band fatigue damage value for each single-degree-of-freedom spring system. All the basic frequency band fatigue damage values ​​are collected to generate an initial damage data set. The initial damage data set is mapped to a target frequency domain coordinate system. The target fatigue damage spectrum is generated based on the mapped initial damage data set.

3. The high-frequency micro-vibration durability testing system for flexible hoses according to claim 2, characterized in that, The specific steps for processing the service micro-vibration signal using the extreme value envelope scanning algorithm to obtain high-frequency peak segments are as follows: Obtain the fatigue extreme value data of the base material of the flexible hose, calculate the product of the fatigue extreme value data of the base material and the preset small proportional coefficient, and obtain the high-frequency stress damage judgment threshold parameter. The service micro-vibration signal is extracted using a time-domain sliding window model to generate multiple candidate segments. For each of the multiple candidate segments, the extreme value envelope scanning algorithm is run to analyze the absolute envelope maxima of each candidate segment. The absolute envelope maximum of each candidate segment is compared with the high-frequency stress damage determination threshold parameter. If the absolute envelope maximum of each candidate segment is lower than the high-frequency stress damage determination threshold parameter, each candidate segment is marked as a redundant segment. If the absolute envelope maximum of each candidate segment is not lower than the high-frequency stress damage determination threshold parameter, each candidate segment is marked as a high-frequency peak segment. All redundant segments are removed from the service micro-vibration signal, and each high-frequency peak segment in the service micro-vibration signal is retained. The initial occurrence timestamp data of each high-frequency peak segment in the service micro-vibration signal is extracted. A temporal correlation relationship for each high-frequency peak segment is constructed based on the initial occurrence timestamp data. All high-frequency peak segments are stored using the temporal correlation relationship.

4. The high-frequency micro-vibration durability testing system for flexible hoses according to claim 1, characterized in that, The specific steps for constructing a dual-objective constraint model that includes both damage equivalent boundaries and ultimate impact boundaries are as follows: Obtain the material yield limit stress parameter of the flexible hose, evaluate the displacement critical state parameter of the flexible hose to undergo plastic deformation based on the material yield limit stress parameter, set the maximum critical acceleration spectrum limit parameter based on the displacement critical state parameter, generate the ultimate impact boundary using the maximum critical acceleration spectrum limit parameter, and obtain the equivalent fatigue damage mapping function. The equivalent fatigue damage mapping function is used to define the damage equivalent boundary of the high-frequency peak segment during scaling. A first objective function to be optimized, including a nonlinear amplification coefficient dimension, is introduced, and a second objective function to be optimized, including a frequency domain time cyclic increase coefficient dimension, is introduced. The first and second objective functions to be optimized are combined to generate a basic bi-objective function. The maximum critical acceleration spectrum limit parameter is configured as a safety envelope inequality constraint condition of the basic bi-objective function, and the damage equivalent boundary is configured as a fatigue conservation equality constraint condition of the basic bi-objective function. A penalty function of the system Lagrange multiplier optimization algorithm is configured. If the basic bi-objective function violates the safety envelope inequality constraint condition, the cost weight of the first objective function to be optimized is modified using the penalty function. The safety envelope inequality constraint condition, the fatigue conservation equality constraint condition, the penalty function, and the basic bi-objective function are integrated to generate the bi-objective constraint model.

5. The high-frequency micro-vibration durability testing system for flexible hoses according to claim 4, characterized in that, The specific steps for calculating the optimal scaling factor of the high-frequency peak segment using the dual-objective constraint model, and adjusting the high-frequency peak segment according to the optimal scaling factor to obtain the scaled segment are as follows: Extract the basic amplitude feature data and basic time cycle frequency data from the high-frequency peak segment, input the basic amplitude feature data and the basic time cycle frequency data into the dual-objective constraint model, and use the dual-objective constraint model to perform optimization calculation on the basic amplitude feature data and the basic time cycle frequency data to generate the optimal scaling coefficient; The optimal amplitude amplification factor for the basic amplitude characteristic data is extracted from the optimal scaling factor, and the optimal time cycle increment factor for the basic time cycle frequency data is extracted from the optimal scaling factor. The basic amplitude characteristic data and the optimal amplitude amplification factor are then processed using a scaling calculation formula based on fatigue equivalence and dynamic stiffness compensation. The scaling calculation formula based on fatigue equivalence and dynamic stiffness compensation is as follows: ,in This represents the amplitude value after compensation. This represents the basic amplitude characteristic data. This represents the optimal amplitude amplification factor. This indicates the dominant frequency of the high-frequency peak segment. This represents the first-order natural frequency of the flexible hose; The amplitude compensation adjustment segment is obtained by replacing the basic amplitude feature data in the high-frequency peak segment with the compensated amplitude value. The basic time cycle frequency data of the amplitude compensation adjustment segment on the time axis is amplified by the optimal time cycle increase coefficient to obtain the time-domain tolerance adjustment segment. The time-domain tolerance adjustment segment is output and configured as the scaling segment.

6. The high-frequency micro-vibration durability testing system for flexible hoses according to claim 3, characterized in that, The steps of splicing the various scaled segments to generate an acceleration driving signal and calculating the acceleration damage spectrum of the acceleration driving signal are as follows: Read each of the scaling segments and the timing relationship stored in the system memory, sort all the scaling segments according to the initial occurrence timestamp data in the timing relationship, determine the global splicing order parameter of all the scaling segments, and connect two adjacent scaling segments using a cosine smooth transition window function according to the global splicing order parameter. The cosine smooth transition window function is used to eliminate the endpoint phase abruptness data between two adjacent scaling segments to obtain the acceleration drive signal after continuous time series splicing. A standard single-degree-of-freedom spring system set is configured, and the acceleration drive signal is processed using an impact response calculation model based on the viscoelastic system difference equation. The test displacement data of the acceleration drive signal on each single-degree-of-freedom spring system in the standard single-degree-of-freedom spring system set is calculated. The test displacement data is processed using the standard rainflow counting model, and the cumulative number of acceleration stress cycles and the corresponding amplitude of acceleration stress cycles are extracted from the test displacement data. The cumulative number of acceleration stress cycles and the corresponding amplitude of acceleration stress cycles are processed using the material fatigue damage model, and the acceleration band fatigue damage value generated by the acceleration drive signal is calculated. Based on the acceleration band fatigue damage value, the acceleration damage spectrum of the acceleration drive signal is plotted in the target frequency domain coordinate system.

7. The high-frequency micro-vibration durability testing system for flexible hoses according to claim 6, characterized in that, The specific steps for adjusting the peak value of the acceleration driving signal based on the error between the accelerated damage spectrum and the target fatigue damage spectrum to obtain the target test driving signal are as follows: The system inputs verification boundary parameters, extracts actual spectral line continuous trajectory data contained within the verification boundary parameters from the accelerated damage spectrum, extracts expected spectral line continuous trajectory data contained within the verification boundary parameters from the target fatigue damage spectrum, introduces a weighted least squares fitting model, and uses the weighted least squares fitting model to compare and process the actual spectral line continuous trajectory data and the expected spectral line continuous trajectory data to obtain the current squared continuous error integral value for the verification boundary parameters. Calculate the peak deviation data between the actual continuous trajectory data of the spectral line and the expected continuous trajectory data of the spectral line, and use the peak deviation data to modify the global absolute peak distribution state parameter in the acceleration driving signal in reverse to generate the iteratively corrected acceleration driving signal. The updated square continuous error integral value corresponding to the iteratively corrected acceleration driving signal is calculated using the impact response calculation model based on the viscoelastic system difference equation. The updated square continuous error integral value is compared with the global preset error threshold parameter. The global absolute peak distribution state parameter of the iteratively corrected acceleration driving signal is continuously iteratively modified according to the direction of the peak deviation data until the latest square continuous error integral value extracted by the system is lower than the global preset error threshold parameter. The iteratively corrected acceleration driving signal that meets the condition of the global preset error threshold parameter in the final convergence state is configured as the target test driving signal.

8. The high-frequency micro-vibration durability testing system for flexible hoses according to claim 1, characterized in that, The specific steps for performing durability testing on flexible hoses using the target test drive signal to control the test console are as follows: The target test drive signal is imported into the digital signal processing component, and the digital signal processing component performs digital-to-analog conversion on the target test drive signal to convert the target test drive signal into an analog drive signal, and establishes a communication link between the digital signal processing component and the test console. The analog drive signal is input to the servo controller of the test console through the communication link, and the servo controller amplifies the analog drive signal to generate physical electromagnetic excitation force parameters. The high-frequency vibration table device, electrically connected to the test control console, is driven by the physical electromagnetic excitation force parameters. The high-frequency vibration table device is controlled to apply a dynamic high-frequency micro-vibration cyclic load to the flexible hose under test according to the acceleration time history corresponding to the target test drive signal. During the continuous operation of the high-frequency vibration table device, the physical failure deformation state parameters of the flexible hose are collected in real time. When the physical failure deformation state parameters reach the fracture critical value, the high-frequency micro-vibration durability test of the flexible hose is terminated.

Images