Dual Reliability Evaluation Method Based on Conditional Control

Abstract

This invention discloses a condition-controlled dual reliability evaluation method for structural health monitoring, belonging to the field of structural health monitoring technology. It includes first evaluating the reliability of the online monitoring sensors and the monitoring system itself under approximately repeatable conditions by controlling service condition influencing factors; then, by introducing key service influencing factors, evaluating the online monitoring reliability under intermediate evaluation conditions. The condition-controlled dual reliability evaluation method proposed in this invention can comprehensively evaluate the reliability of this novel structural health monitoring method.

Description

Technical Field

[0001] This invention relates to the field of engineering structural health monitoring technology, and in particular to a dual reliability evaluation method based on conditional control. Background Technology

[0002] Structural health monitoring (SHM) methods utilize sensors integrated into the structure's interior or surface to acquire signals related to the structure's health status in real time. Through specific signal processing and mechanical modeling analysis methods, damage-related signal features are extracted. Damage diagnosis methods are then used to assess the structural health status and predict damage propagation and remaining life, ultimately guiding structural design, mission decisions, and logistical maintenance. Therefore, structural health monitoring methods have received widespread attention and research in recent decades.

[0003] To promote the large-scale engineering application of structural health monitoring methods, reliability assessment of structural health monitoring is a critical issue that urgently needs to be addressed. Most existing studies on the reliability assessment of structural health monitoring directly reference conventional non-destructive testing reliability assessment methods, such as using hit / miss data and... â vs a "Data analysis techniques, when evaluating the reliability of damage detection, use metrics such as Probability of Detection (POD), Probability of False Alarm (PFA), and Receiver Operating Characteristic (ROC) curves. However, existing evaluation techniques lack attention to the evaluation conditions that are crucial for applying these metrics. In the reliability evaluation of structural health monitoring, since structural health monitoring methods are typically applied online during structural service, numerous uncertainties exist. Furthermore, different application scenarios may have different key uncertainties. How to evaluate the reliability of structural health monitoring and provide users with useful reliability information under these varying circumstances requires further in-depth research. The limited evaluation methods developed to date are only applicable to specific objects or application scenarios and attempt to cover all factors in structural health monitoring applications, conflating numerous factors in service and neglecting the unified control and standardization of evaluation conditions during reliability assessment." Summary of the Invention

[0004] This invention aims to provide a dual reliability evaluation method based on condition control, which includes first evaluating the reliability of the online structural health monitoring sensor and the monitoring system itself under approximately repeatable conditions by controlling the influencing factors of service conditions; and further evaluating the reliability of the online structural health monitoring under intermediate evaluation conditions by introducing key service influencing factors.

[0005] To achieve the above objectives, the technical solution of the present invention is as follows:

[0006] Dual reliability evaluation methods based on conditional control include,

[0007] Step S1: By controlling the influencing factors of service conditions, under approximately repeatable conditions, first evaluate the reliability of the online monitoring sensor and the monitoring system itself for structural health monitoring.

[0008] Step S2 introduces key service influencing factors and evaluates the reliability of online structural health monitoring under intermediate evaluation conditions.

[0009] Further, step S1 above includes,

[0010] Step S11: Design and manufacture identical structural components in the same batch;

[0011] Step S12: Control damage to specific morphology and size;

[0012] Step S13: Control the consistency of batch sensor performance and sensor / structure integration method;

[0013] Step S14: Control the room temperature and room pressure monitoring environment, use the same structural health monitoring system and its internal fixed diagnostic algorithm to monitor structural damage, obtain the structural health monitoring and diagnostic results corresponding to the actual damage size under approximately repeatable test conditions, and construct a reliability evaluation dataset.

[0014] Step S15: Under approximately repeatability conditions, evaluate the monitoring accuracy of structural health monitoring based on the reliability evaluation dataset obtained in step S14.

[0015] Step S16: Under approximately repeatability conditions, evaluate the monitoring uncertainty of structural health monitoring based on the reliability evaluation dataset obtained in step S14.

[0016] Further, step S2 includes,

[0017] Step S21: Based on the conditional control requirements for stand-alone applications, the same control conditions as in step S1 are used to perform consistent control on the specific monitored structure, damage, sensor performance, sensor / structure integration method, structural health monitoring system and its internal fixed diagnostic algorithm.

[0018] Step S22: Based on the single-machine service scenario in the application, introduce intermediate evaluation conditions for the impact of key service factors.

[0019] Step S23: Control the influence of other service factors;

[0020] Step S24: Use the same structural health monitoring system and its internal fixed diagnostic algorithm to monitor structural damage, obtain the structural health monitoring and diagnostic results corresponding to the actual damage size under critical service conditions, and construct a reliability evaluation dataset.

[0021] Step S25: Under the intermediate evaluation conditions that introduce the influence of key service factors, evaluate the monitoring accuracy of structural health monitoring based on the reliability evaluation dataset obtained in step S24.

[0022] Step S26: Under the intermediate evaluation conditions that introduce the influence of key service factors, evaluate the monitoring uncertainty of structural health monitoring based on the reliability evaluation dataset obtained in step S24.

[0023] Furthermore, in steps S15 and S25, the POD curve is used to evaluate the monitoring accuracy of the structural health monitoring damage alarm capability.

[0024] Establish structural health monitoring and diagnostic results and actual damage size The relationship is as follows:

[0025]

[0026] in, f (.)and g (.) denotes the necessary transformation function to make... and The relationship is linear; β 1 represents the slope; β 0 represents the intercept; The noise term has a mean of 0 and a standard deviation of . The normal distribution;

[0027] Define monitoring thresholds POD indicates a diagnostic result greater than the monitoring threshold. The probability of is calculated as follows:

[0028] in, It follows a standard normal cumulative distribution, where the mean is: The standard deviation is , Indicates the standard deviation of the noise term; This indicates the diagnostic results of the structural health monitoring system.

[0029] Furthermore, in steps S16 and S26, a POD one-sided confidence interval is used. To evaluate the monitoring uncertainty of structural health monitoring damage alarm capability, among which The lower limit of the one-sided confidence interval for POD is calculated as follows:

[0030] ,

[0031] is the confidence probability.

[0032] Furthermore, in steps S16 and S26, the following steps are adopted: The index is used to characterize the minimum damage size that a structural health monitoring system can reliably detect.

[0033] Furthermore, in steps S15 and S25, the maximum absolute error is used. Root mean square error The indicators evaluate the accuracy of the ability to quantify damage size in structural health monitoring.

[0034] Among them, the maximum absolute error for:

[0035] ;

[0036] Among them, root mean square error for:

[0037]

[0038] In the formula: This represents the total number of quantitative diagnoses performed by the structural health monitoring system. Indicates the first i The actual size of structural damage at the time of diagnosis; Indicates the true damage size The corresponding structural health monitoring and diagnostic results are as follows.

[0039] Furthermore, in steps S16 and S26, the monitoring uncertainty of the ability to quantify damage size in structural health monitoring is evaluated using standard deviation, confidence probability, or confidence interval.

[0040] The formula for calculating the standard deviation is as follows:

[0041]

[0042] In the formula: This represents the total number of quantitative diagnoses performed by the structural health monitoring system. This represents the average of multiple diagnostic results from structural health monitoring.

[0043] Among them, confidence probability This indicates that the quantitative diagnostic results of structural health monitoring damage size appear in the mean value. A confidence interval centered The probability within, where, Indicates the width of the upper confidence interval. Indicates the width of the lower half confidence interval;

[0044] The calculation method for the upper and lower limits of the confidence interval is as follows:

[0045]

[0046] In the formula, Indicates the width of the upper confidence interval. Indicates the width of the lower half confidence interval. The probability density function represents the quantitative diagnostic results of damage size in structural health monitoring. This represents the confidence probability.

[0047] The present invention also provides an apparatus comprising: one or more processors; and a memory for storing one or more programs, which, when executed by the one or more processors, cause the one or more processors to perform the aforementioned condition-controlled dual reliability evaluation method.

[0048] The present invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-described condition-controlled dual reliability evaluation method.

[0049] Beneficial effects: This invention provides a standard framework for evaluating the reliability of structural health monitoring based on a condition-controlled dual reliability evaluation method. Under near-repeatability conditions, the reliability of the structural health monitoring sensor and system itself is evaluated by controlling monitoring influencing factors as much as possible. Under intermediate evaluation conditions, typical key service factors are introduced according to different application scenarios to evaluate the reliability of structural health monitoring under specific service conditions.

[0050] To make the above-mentioned features and advantages of the invention more apparent and understandable, specific embodiments are described below, and detailed descriptions are provided in conjunction with the accompanying drawings. Attached Figure Description

[0051] Figure 1 This is a flowchart of the dual reliability evaluation method based on conditional control according to the present invention.

[0052] Figure 2 This is a schematic diagram of the ear piece structure in a specific embodiment.

[0053] Figure 3 This is a schematic diagram of the integrated piezoelectric sensor and ear structure.

[0054] Figure 4 This is a schematic diagram of the diagnostic results during the reliability evaluation process of structural health monitoring sensors and systems.

[0055] Figure 5 This diagram illustrates the relationship between the false alarm rate and the monitoring threshold during the reliability evaluation of structural health monitoring sensors and systems.

[0056] Figure 6 This is a schematic diagram of the POD curve and the lower bound of the 95% confidence interval during the reliability evaluation of structural health monitoring sensors and systems.

[0057] Figure 7 This is a schematic diagram illustrating the reliability evaluation results of the ability to quantify damage size during the reliability evaluation process of structural health monitoring sensors and systems.

[0058] Figure 8 This is a schematic diagram of the diagnostic results during the reliability evaluation of structural health monitoring under critical service conditions.

[0059] Figure 9 This diagram illustrates the relationship between the false alarm rate and the monitoring threshold during the reliability evaluation of structural health monitoring under critical service conditions.

[0060] Figure 10 This is a schematic diagram of the POD curve and the lower bound of the 95% confidence interval during the reliability evaluation of structural health monitoring under critical service conditions.

[0061] Figure 11 This is a schematic diagram of the reliability evaluation results for the ability to quantify damage size during the reliability evaluation process of structural health monitoring under critical service conditions. Detailed Implementation

[0062] To make the objectives and technical solutions of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the described embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0063] like Figure 1 As shown, the dual reliability evaluation method based on conditional control of the present invention includes the following steps.

[0064] Step S1: By controlling the influencing factors of service conditions, under approximately repeatable conditions, first evaluate the reliability of the online monitoring sensor and the monitoring system itself for structural health monitoring.

[0065] More specifically, under near-repeatable test conditions, evaluation indicators of the reliability of structural health monitoring are obtained, and the influencing factors of uncertainty are controlled as much as possible to examine the reliability of the integrated structural health monitoring sensor and the structural health monitoring system itself.

[0066] Furthermore, the approximate repeatability test conditions include: controlling the consistency of batch structures and damage as much as possible, the consistency of sensor performance, the consistency of sensor / structure integration methods, and the consistency of application environment, and using the same structural health monitoring system with a fixed internal damage diagnosis algorithm. At the same time, other uncertainties should be controlled as much as possible to minimize their impact.

[0067] Further, step S1 includes the following steps.

[0068] Step S11: Design and manufacture identical structural components in the same batch.

[0069] More specifically, based on parameters that are completely consistent with the form and size of the specific monitored structural object of a single machine, the same batch of identical structural components are designed and manufactured as test objects for reliability evaluation, thereby achieving consistency control of batch structural components.

[0070] Step S12: Control damage to specific morphology and size.

[0071] More specifically, in order to control the consistency of damage morphology and damage size, damage with fixed morphology and size is manufactured, and the actual damage size under approximately repeatable test conditions is denoted as... , representing the first under approximately repeatable test conditions i The actual size of the structural damage at the time of diagnosis.

[0072] Alternatively, the damage can be caused manually or by machine, and the present invention is not limited thereto.

[0073] Step S13: Control the consistency of batch sensor performance and sensor / structure integration method.

[0074] More specifically, the sensor / structure integration method refers to the process of solidifying and integrating the sensor network onto the surface of the structure using a vacuum-temperature-pressure co-curing process after the structural components are manufactured and before assembly, while simultaneously integrating a flexible protective layer on the outermost side for protection and encapsulation.

[0075] Step S14: Control the room temperature and room pressure monitoring environment, and use the same structural health monitoring system and its internal fixed diagnostic algorithm to monitor structural damage and obtain the actual damage size under approximately repeatable test conditions. The corresponding structural health monitoring and diagnostic results Construct a reliability evaluation dataset .

[0076] Among them, superscript This represents the total number of diagnoses performed by the structural health monitoring system under approximately repeatable testing conditions.

[0077] Optionally, under standard environments such as normal temperature and pressure, the same structural health monitoring system and its internally fixed diagnostic algorithm can be used to monitor structural damage.

[0078] Step S15: Under approximately repeatability conditions, evaluate the monitoring accuracy of structural health monitoring based on the reliability evaluation dataset obtained in step S14.

[0079] Step S16: Under approximately repeatability conditions, evaluate the monitoring uncertainty of structural health monitoring based on the reliability evaluation dataset obtained in step S14.

[0080] Step S2 introduces key service influencing factors and evaluates the reliability of online structural health monitoring under intermediate evaluation conditions.

[0081] More specifically, under intermediate test conditions, considering the special online application mode of aircraft structural health monitoring, the influence of key service factors is introduced according to specific application scenarios and user needs to evaluate the reliability of structural health monitoring under key service conditions.

[0082] Furthermore, the intermediate test conditions include: the structure and damage type, sensor performance, and the sensor / structure integration method, structural health monitoring system, and diagnostic algorithm, all using the same control conditions as in step S1. Further, based on the control requirements of the intermediate test conditions, critical service conditions are not controlled, and the reliability of structural health monitoring is examined under the influence of these service conditions. Other service conditions not examined need to be controlled and kept as constant as possible.

[0083] Furthermore, step S2 includes the following steps.

[0084] Step S21: Based on the conditional control requirements for stand-alone applications, the same control conditions as in step S1 are used to perform consistent control on the specific monitored structure, damage, sensor performance, sensor / structure integration method, structural health monitoring system and its internal fixed diagnostic algorithm.

[0085] Step S22: Based on the single-machine service scenario in the application, introduce intermediate evaluation conditions for the impact of key service factors.

[0086] More specifically, the reliability of structural health monitoring is examined under the influence of this service factor.

[0087] Step S23: Control the influence of other service factors.

[0088] More specifically, other service factors that are not being considered need to be controlled and kept as constant as possible.

[0089] Step S24: Use the same structural health monitoring system and its internally fixed diagnostic algorithm to monitor structural damage and obtain the actual damage size under critical service conditions. The corresponding structural health monitoring and diagnostic results Construct a reliability evaluation dataset .

[0090] in, Represented as the first under critical service conditions i At the time of diagnosis, the actual size of the structural damage, superscript This indicates the total number of diagnostics performed by the structural health monitoring system under critical service conditions.

[0091] Step S25: Under the intermediate evaluation conditions that introduce the influence of key service factors, evaluate the monitoring accuracy of structural health monitoring based on the reliability evaluation dataset obtained in step S24.

[0092] Step S26: Under the intermediate evaluation conditions that introduce the influence of key service factors, evaluate the monitoring uncertainty of structural health monitoring based on the reliability evaluation dataset obtained in step S24.

[0093] Furthermore, considering the different monitoring objects of structural health monitoring, such as damage alarm, dimensional quantification, location, and prediction functions, specific evaluation indicators are used.

[0094] Optionally, the ability to alarm damage and quantify damage size in structural health monitoring can be used as monitoring objects, and evaluation indicators for monitoring accuracy and uncertainty can be introduced.

[0095] Furthermore, in steps S15 and S25, the POD curve is used to evaluate the monitoring accuracy of the structural health monitoring damage alarm capability.

[0096] First, establish the structural health monitoring and diagnostic results. and actual damage size The relationship is as follows:

[0097]

[0098] in, f (.)and g (.) denotes the necessary transformation function to make... and The relationship is linear; β 1 represents the slope; β 0 represents the intercept; The noise term has a mean of 0 and a standard deviation of . It follows a normal distribution.

[0099] More specifically, regarding the reliability evaluation dataset obtained in step S1, for For the reliability evaluation dataset obtained in step S2, for .

[0100] Then the monitoring threshold was defined. POD indicates a diagnostic result greater than the monitoring threshold. The probability of is calculated as follows:

[0101] in, It follows a standard normal cumulative distribution, where the mean is: The standard deviation is , where parameters β 0、 β 1. All are estimated using the maximum likelihood method; Indicates the standard deviation of the noise term; This represents the diagnostic results of the structural health monitoring system. More specifically, it refers to the reliability evaluation dataset obtained in step S1. for For the reliability evaluation dataset obtained in step S2, for .

[0102] Furthermore, in steps S16 and S26, the POD one-sided confidence interval is used to evaluate the monitoring uncertainty of the structural health monitoring damage alarm capability. This requires adhering to the specified confidence probability. Find the corresponding one-sided confidence interval of POD. Confidence probability This characterizes the probability that the actual POD of the structural health monitoring system falls within the confidence interval. The lower limit of the one-sided confidence interval for POD is calculated as follows:

[0103] .

[0104] Furthermore, in steps S16 and S26, the following methods may also be employed: The index is used to characterize the minimum damage size that a structural health monitoring system can reliably detect. The indicator indicates that, with a 95% confidence level, the structural health monitoring system can detect the presence of damage of this size with at least a 90% probability, i.e., the confidence probability. Lower limit of POD confidence interval .in, The metric can be obtained from the intersection of the 90% POD and the lower bound of the 95% confidence interval on the POD curve.

[0105] More specifically, regarding the reliability evaluation dataset obtained in step S1, for For the reliability evaluation dataset obtained in step S2, for .

[0106] Furthermore, in steps S15 and S25, the maximum absolute error is used. Root mean square error The indicators evaluate the accuracy of the ability to quantify damage size in structural health monitoring.

[0107] Among them, the maximum absolute error for:

[0108] .

[0109] Among them, root mean square error for:

[0110]

[0111] In the formula: This represents the total number of quantitative diagnoses performed by the structural health monitoring system. Indicates the first i The actual size of structural damage at the time of diagnosis; Indicates the true damage size The corresponding structural health monitoring and diagnostic results are then obtained. More specifically, for the reliability evaluation dataset obtained in step S1, for For the reliability evaluation dataset obtained in step S2, for For the reliability evaluation dataset obtained in step S1, for For the reliability evaluation dataset obtained in step S2, for .

[0112] Furthermore, in steps S16 and S26, the monitoring uncertainty of the ability to quantify damage size in structural health monitoring is evaluated using standard deviation, confidence probability, or confidence interval.

[0113] The formula for calculating the standard deviation is as follows:

[0114]

[0115] In the formula: This represents the total number of quantitative diagnoses performed by the structural health monitoring system; the dataset for reliability evaluation obtained in step S1, for For the reliability evaluation dataset obtained in step S2, for ; This represents the average of multiple diagnostic results from structural health monitoring, based on the reliability evaluation dataset obtained in step S1. for The mean of the reliability evaluation dataset obtained in step S2; for The mean.

[0116] Among them, confidence probability This indicates that the quantitative diagnostic results of structural health monitoring damage size appear in the mean value. A confidence interval centered The probability within, where, Indicates the width of the upper confidence interval. This indicates the width of the lower half confidence interval.

[0117] The calculation method for the upper and lower limits of the confidence interval is as follows:

[0118]

[0119] In the formula, Indicates the width of the upper confidence interval. Indicates the width of the lower half confidence interval. The probability density function represents the quantitative diagnostic results of damage size in structural health monitoring. This represents the confidence probability.

[0120] The present invention also provides an apparatus comprising: one or more processors; and a memory for storing one or more programs, which, when executed by the one or more processors, cause the one or more processors to execute the condition-controlled dual reliability evaluation method.

[0121] The present invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the condition-controlled dual reliability evaluation method.

[0122] The following specific embodiment illustrates the dual reliability evaluation method based on conditional control according to the present invention. In this specific embodiment, reliability evaluation is performed on the monitoring of cracks in a typical aerospace lug structure.

[0123] First, control the influencing factors when applying the system and calibrate the fixed waveguide structure health monitoring and diagnostic algorithm inside the structural health monitoring system.

[0124] Based on the typical aerospace lug connection structure in engineering applications, this embodiment is designed as follows: Figure 2 The ear-shaped structural component shown is made of LY12 aluminum alloy with a thickness of 5mm. The typical aerospace ear-shaped connecting components used in this specific embodiment all employ the same materials, design, dimensions, and manufacturing process. This batch of test pieces was divided into three groups, named Group A, Group B, and Group C, for different applications, with five test pieces prepared in each group.

[0125] Among them, the specimens in Group A are numbered C1, C2, C3, C4, and C5, and are used to calibrate the fixed diagnostic algorithm inside the structural health monitoring system. Two types of crack damage were prepared for this purpose: three specimens with artificially cut crack damage and two specimens with naturally propagating fatigue cracks under fatigue load.

[0126] In addition, the test specimens in Group B are numbered IS1, IS2, IS3, IS4, and IS5, which are used to conduct the reliability evaluation of the integrated sensor structure health monitoring in step S1, in order to examine the reliability of the integrated structure health monitoring sensor and the structure health monitoring system itself.

[0127] In addition, the specimens in group C are numbered CSC1, CSC2, CSC3, CSC4, and CSC5, and are used to conduct structural health monitoring reliability evaluation under key service conditions in step S2.

[0128] Based on the characteristics of the structure and the initiation and propagation of damage, a sensor network consisting of five piezoelectric sensors is arranged on the structure for health monitoring and diagnosis of the waveguide structure. A consistency control method based on multidimensional feature-similarity distance is used to select piezoelectric sensors in batches for performance consistency, and a surface-mount co-curing integration method is used to achieve integrated integration of the piezoelectric sensors and the structure. Figure 3 This is a schematic diagram of the integrated piezoelectric sensor and ear structure.

[0129] A structural health monitoring system with an embedded guided wave-Gaussian process diagnostic algorithm was adopted, and the algorithm was calibrated using Group A specimens.

[0130] Secondly, under near repeatability conditions, the reliability of the online monitoring sensors and the monitoring system itself for structural health monitoring is evaluated.

[0131] In the reliability evaluation of the integrated sensor structure health monitoring, the waveguide structure health monitoring environment was controlled to be a laboratory environment with normal temperature and pressure. Cracks of fixed morphology and specific dimensions were created on one side of the through-hole of the ear-shaped specimen using wire electrical discharge machining. The same waveguide structure health monitoring system and a calibrated waveguide-Gaussian process diagnostic algorithm were used to monitor the crack lengths of specimens IS1 to IS5 in group B. The diagnostic results are as follows: Figure 4 As shown.

[0132] To evaluate the reliability of structural health monitoring damage alarm capabilities, POD curves, POD confidence intervals, and other parameters were used. Indicators. To determine the monitoring thresholds for POD evaluation. First, obtain the structural health monitoring and diagnostic results under healthy conditions. Second, calculate the relationship between the false alarm rate of the structural health monitoring method and the monitoring threshold, such as... Figure 5 As shown, the monitoring threshold was finally determined to be 0.3 mm by balancing the false alarm rate and POD. In this case, the false alarm rate is 0, while also ensuring a high POD. Based on the set monitoring threshold, the cumulative probability of structural health monitoring diagnostic results exceeding the threshold was calculated, and then the POD curve and the lower bound of the 95% confidence interval were calculated, as shown below. Figure 6 As shown. The minimum crack length that can be reliably identified by the integrated structural health monitoring sensor and structural health monitoring system. Indicators can be obtained through Figure 6 The intersection of the 90% POD and the lower bound of the 95% confidence interval was obtained, and the results indicate that... The value of 0.8 mm indicates that when the crack length is greater than or equal to 0.8 mm, the waveguide structure health monitoring method can detect the presence of the crack with at least 90% probability and 95% confidence.

[0133] To evaluate the reliability of the structural health monitoring's ability to quantify damage size, the maximum absolute error for evaluating monitoring accuracy was calculated based on the acquired structural health monitoring results. The root mean square error is 0.8 mm. The value is 0.3 mm; the calculated standard deviation of the evaluation monitoring uncertainty is 0.3 mm; the 95% confidence interval is as follows: Figure 7 As shown by the dotted line, this interval indicates a 95% probability that the structural health monitoring and diagnostic results will fall within this interval.

[0134] Finally, a reliability assessment of structural health monitoring under critical service conditions is conducted.

[0135] In the reliability evaluation of structural health monitoring under critical service conditions, the same guided-wave-Gaussian process structural health monitoring diagnostic algorithm was used to monitor the fatigue crack propagation of specimens CSC1 to CSC5 in group C under flight load spectrum conditions. Two important service uncertainties were introduced in this evaluation: the influence of service load and damage. Structural health monitoring was conducted under critical service load spectrum conditions, and the damage consisted of real fatigue cracks with uncertainties in angle, depth, width, and morphology. The crack propagation length of specimens CSC1 to CSC5 in group C was monitored, and the diagnostic results are as follows: Figure 8 As shown.

[0136] In evaluating the reliability of structural health monitoring damage alarm capabilities, in order to determine the monitoring threshold of POD... First, the structural health monitoring method was used to obtain multiple diagnostic results of the structure in the unpropagated crack state under the influence of key service load spectra. Second, the relationship between the false alarm rate of the structural health monitoring method and the monitoring threshold was calculated. Figure 9 As shown. The monitoring threshold was determined to be 0.9 mm by balancing the false alarm rate and POD. Based on the set monitoring threshold, the cumulative probability of structural health monitoring diagnostic results exceeding the threshold was calculated, resulting in the POD curve and the lower bound of the 95% confidence interval under critical service conditions, as shown. Figure 10 As shown. The minimum extended crack size reliably detected by structural health monitoring is obtained from the intersection of the 90% POD and the lower bound of the 95% confidence interval. The value of 1.2 mm indicates that under critical service loads and damage effects, when the crack propagation length at the hole edge of the lug structure is greater than or equal to 1.2 mm, the waveguide structure health monitoring method can detect the existence of this damage with at least a 90% probability and a 95% confidence probability.

[0137] Regarding the reliability evaluation of the ability to quantify damage size in structural health monitoring, the maximum absolute error for evaluating monitoring accuracy is first calculated based on the obtained structural health monitoring results. The root mean square error is 1.9 mm. The value is 0.6 mm. Furthermore, the standard deviation of the evaluation monitoring uncertainty is 0.4 mm, and the 95% confidence interval is as follows: Figure 11 As shown.

[0138] Although the present invention has been disclosed above by way of embodiments, it is not intended to limit the present invention. Anyone skilled in the art can make some modifications and refinements without departing from the spirit and scope of the present invention. Therefore, the scope of protection of the present invention shall be determined by the appended claims.

Claims

1. A dual reliability evaluation method based on conditional control, characterized in that, include, Step S1: By controlling the influencing factors of service conditions, under approximately repeatable conditions, first evaluate the reliability of the online monitoring sensor and the monitoring system itself for structural health monitoring. Step S2: Introduce key service influencing factors and evaluate the reliability of online structural health monitoring under intermediate evaluation conditions; Step S1 includes, Step S11: Design and manufacture identical structural components in the same batch; Step S12, control damage with specific morphology and size; wherein, damage with specific morphology and size is damage with fixed morphology and size; Step S13: Control the consistency of the performance of batch sensors and the sensor / structure integration method; wherein, the sensor / structure integration method refers to the process of using vacuum-temperature-pressure surface-mount co-curing integration process to solidify and integrate the sensor network on the surface of the structure after the structural parts are manufactured and before assembly, and at the same time, a flexible protective layer is integrated on the outermost side for protection and encapsulation. Step S14: Control the room temperature and room pressure monitoring environment, use the same structural health monitoring system and its internal fixed diagnostic algorithm to monitor structural damage, obtain the structural health monitoring and diagnostic results corresponding to the actual damage size under approximately repeatable test conditions, and construct a reliability evaluation dataset. Step S15: Under approximately repeatability conditions, evaluate the monitoring accuracy of structural health monitoring based on the reliability evaluation dataset obtained in step S14. Step S16: Under approximately repeatability conditions, evaluate the monitoring uncertainty of structural health monitoring based on the reliability evaluation dataset obtained in step S14. Step S2 includes, Step S21: Based on the conditional control requirements for stand-alone applications, the same control conditions as in step S1 are used to perform consistent control on the specific monitored structure, damage, sensor performance, sensor / structure integration method, structural health monitoring system and its internal fixed diagnostic algorithm. Step S22: Based on the single-machine service scenario in the application, intermediate evaluation conditions for the influence of key service factors are introduced; the intermediate evaluation conditions include: the structure and damage type, sensor performance, and the sensor / structure integrated method, structural health monitoring system and diagnostic algorithm adopt the same control conditions as in step S1, and the key service conditions are not controlled; Step S23: Control the influence of other service factors; Step S24: Use the same structural health monitoring system and its internal fixed diagnostic algorithm to monitor structural damage, obtain the structural health monitoring and diagnostic results corresponding to the actual damage size under critical service conditions, and construct a reliability evaluation dataset. Step S25: Under the intermediate evaluation conditions that introduce the influence of key service factors, evaluate the monitoring accuracy of structural health monitoring based on the reliability evaluation dataset obtained in step S24. Step S26: Under the intermediate evaluation conditions that introduce the influence of key service factors, evaluate the monitoring uncertainty of structural health monitoring based on the reliability evaluation dataset obtained in step S24.

2. The dual reliability evaluation method based on conditional control as described in claim 1, characterized in that, In steps S15 and S25, the POD curve is used to evaluate the monitoring accuracy of the structural health monitoring damage alarm capability. Establish structural health monitoring and diagnostic results and actual damage size The relationship is as follows: in, f (.)and g (.) denotes the necessary transformation function to make... and The relationship is linear; β 1 represents the slope; β 0 represents the intercept; The noise term has a mean of 0 and a standard deviation of . The normal distribution; Define monitoring thresholds POD indicates a diagnostic result greater than the monitoring threshold. The probability of is calculated as follows: in, It follows a standard normal cumulative distribution, where the mean is: The standard deviation is , Indicates the standard deviation of the noise term; This indicates the diagnostic results of the structural health monitoring system.

3. The dual reliability evaluation method based on conditional control as described in claim 2, characterized in that, In steps S16 and S26, a POD one-sided confidence interval is used. To evaluate the monitoring uncertainty of structural health monitoring damage alarm capability, among which The lower limit of the one-sided confidence interval for POD is calculated as follows: ， is the confidence probability.

4. The dual reliability evaluation method based on conditional control as described in claim 3, characterized in that, In steps S16 and S26, the following methods are adopted: The index is used to characterize the minimum damage size that a structural health monitoring system can reliably detect.

5. The dual reliability evaluation method based on conditional control as described in claim 1, characterized in that, In steps S15 and S25, the maximum absolute error is used. Root mean square error The indicators evaluate the accuracy of the ability to quantify damage size in structural health monitoring. Among them, the maximum absolute error for: ； Among them, root mean square error for: In the formula: This represents the total number of quantitative diagnoses performed by the structural health monitoring system. Indicates the first i The actual size of structural damage at the time of diagnosis; Indicates the true damage size The corresponding structural health monitoring and diagnostic results are as follows.

6. The dual reliability evaluation method based on conditional control as described in claim 5, characterized in that, In steps S16 and S26, the monitoring uncertainty of the ability to quantify damage size in structural health monitoring is evaluated using standard deviation, confidence probability, or confidence interval. The formula for calculating the standard deviation is as follows: In the formula: This represents the total number of quantitative diagnoses performed by the structural health monitoring system. This represents the average of multiple diagnostic results from structural health monitoring. Among them, confidence probability This indicates that the quantitative diagnostic results of structural health monitoring damage size appear in the mean value. A confidence interval centered The probability within, where, Indicates the width of the upper confidence interval. Indicates the width of the lower half confidence interval; The calculation method for the upper and lower limits of the confidence interval is as follows: In the formula, Indicates the width of the upper confidence interval. Indicates the width of the lower half confidence interval. The probability density function represents the quantitative diagnostic results of damage size in structural health monitoring. This represents the confidence probability.

7. A device, characterized in that, The device includes: one or more processors; and a memory for storing one or more programs that, when executed by the one or more processors, cause the one or more processors to perform the condition-controlled dual reliability evaluation method as described in any one of claims 1-6.

8. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the dual reliability evaluation method based on conditional control as described in any one of claims 1-6.

Images