Quantitative comprehensive evaluation method for reliability of aviation flight control actuator based on data fusion

Abstract

The invention provides a data fusion aviation flight control actuator reliability quantitative comprehensive evaluation method, and relates to the field of mechanical and electrical product reliability, and the method comprises the steps: S1, collecting the design data of a flight control actuator; s2, developing a development test and an identification test; s3, converting the fault time of the reliability strengthening / HALT test into fault occurrence time under a conventional stress condition; s4, converting the fault time of the reliability acceleration test into fault occurrence time under a conventional stress condition; s5, performing MTBF evaluation in a development test stage based on results of the step S3 and the step S4; and S6, performing multi-stage MTBF comprehensive evaluation based on the Bayesian theory. According to the method, the advantages of low cost, high efficiency and repeatability of a simulation test and high confidence of a physical test are utilized, the one-sidedness and limitation of a single data source are made up, a potential fault mode is more comprehensively exposed, and the confidence level is improved; reliability verification of the flight control actuator which is long in service life and high in reliability can be achieved under the limited test time and sample size.

Description

Technical Field

[0001] This invention relates to the field of electromechanical product reliability, and specifically to a data fusion-based quantitative comprehensive evaluation method for the reliability of aviation flight control actuators. Background Technology

[0002] As the core actuator of the flight control system, the flight control actuator is extremely complex in structure. It integrates precision mechanical transmission, hydraulic circuits or electric drive units, forming a mechatronic system with deep coupling of multiple disciplines. This component directly drives the control surfaces to achieve real-time control of flight attitude and trajectory. It is a key airborne device to ensure aircraft safety and mission completion, and its reliability level is directly related to flight safety and mission success rate.

[0003] Reliability verification refers to a series of technical activities that use comprehensive techniques such as reliability analysis, calculation, simulation, and testing to verify the degree to which a product meets preset reliability requirements. However, as the service environment faced by aviation equipment becomes increasingly complex and harsh, and the reliability indicators allocated to electromechanical products continue to rise, traditional reliability verification technologies are no longer sufficient to meet the current dual demands of "rapid and agile verification" and "high-confidence verification." Specifically, these limitations are reflected in the following three aspects: First, flight control actuators, as typical electromechanical products, are developing towards a high degree of electromechanical coupling and intelligence, with increasingly complex internal structures and intricate interactive coupling effects between mechanical and electronic components. Traditional reliability verification methods (such as accelerated testing based on single stress conditions) have model assumptions that deviate significantly from the actual operating conditions of the product, making it difficult to comprehensively cover all potential failure mechanisms and only exposing some reliability vulnerabilities, resulting in significantly biased evaluation results.

[0004] Second, with the rapid iteration and upgrading of aviation equipment, the reliability and lifespan of aviation electromechanical products have been significantly improved, with some product indicators reaching thousands of hours. If the traditional reliability acceptance test process is strictly followed, the required test time is no longer feasible under the current development cycle and cost constraints. While reliability simulation tests can significantly shorten the test cycle, the reliability of the evaluation results cannot be effectively guaranteed due to the inherent differences between the model accuracy and real operating conditions, and cannot meet the requirements for high-confidence verification.

[0005] Third, the development cycle of current aviation electromechanical products has been significantly shortened, causing reliability verification to gradually shift towards the earlier stages of development. In the development of existing models, timed end-of-life tests can often only be conducted on small sample products, making it difficult to comprehensively and accurately verify and evaluate product reliability indicators. More importantly, there is currently a lack of a mature analytical method that can fully integrate various types of data, such as simulation tests and reliability tests from multiple stages of development, failing to provide rich and comprehensive data support for reliability verification, and thus making it difficult to form comprehensive and objective reliability evaluation conclusions. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention aims to provide a data-fusion-based quantitative comprehensive evaluation method for the reliability of aircraft flight control actuators. It establishes a technical process that integrates multiple types of data (reliability physical test data and reliability simulation test data) and multiple stages (development stage and qualification stage) to assess the quantitative requirements of reliability. The method proposed in this invention fully considers the increasing reliability trends in the development process of aircraft flight control actuators. It leverages the advantages of low cost, high efficiency, and repeatability of simulation tests, as well as the high confidence level of physical tests, to compensate for the limitations and biases of single data sources. This more comprehensively exposes potential failure modes and improves the confidence level of reliability evaluation. It can verify the reliability indicators of long-life, highly reliable flight control actuators within limited test time and sample size.

[0007] Specifically, the present invention provides a data fusion-based quantitative comprehensive evaluation method for the reliability of aircraft flight control actuators, which includes the following steps: S1. Collect design data for the flight control actuator; S2. Conduct research and development tests and evaluation tests; S3. Convert the failure time of the reliability enhancement / HALT test to the failure time under normal stress conditions. : ;in, For reliability enhancement / HALT testing, the failure time is... The acceleration factor between hardening / HALT stress and conventional stress; S4. Convert the failure time in accelerated reliability testing to the failure time under normal stress conditions. : ;in, To accelerate the failure time in reliability testing, This is the acceleration factor between accelerated stress and normal stress; S5. Based on the results of steps S3 and S4, conduct an MTBF evaluation during the research and development testing phase. S6. Multi-stage comprehensive evaluation of MTBF based on Bayesian theory, specifically: S61. Collect the total test time and number of failures of single-unit equipment reliability simulation data, subsystem reliability verification test data and field test data, as well as the evaluation results of MTBF in the development and testing phase obtained in step S5. S62. Quantify the uncertainty parameters of the data: S621. Calculate the uncertainty of reliability simulation data for single-machine equipment. : ; in, To determine the degree of conformity between the simulation model and the real product; To assess the reliability of the source of the simulation input parameters; and These are the weighting coefficients; S622, Uncertainty in reliability verification test data of the calculation subsystem And the uncertainty of the evaluation results of MTBF during the research and development testing phase. : ; in, This refers to the actual test time; For reference test time; The coefficient of variation between the test environment and the actual use environment; To assess the completeness of data collection; Subscript, representing or ; S623, Calculate the uncertainty of field test data : ; in, This represents the actual number of field failures. The minimum effective number of faults threshold; For field data collection coverage; To record accuracy scores; S63. Perform Bayesian fusion on the uncertainty data obtained in step S62 to obtain a comprehensive evaluation result of multi-type and multi-stage MTBF.

[0008] Preferably, step S5 specifically includes the following sub-steps: S51. Arrange the n reliability enhancement / HALT tests and m reliability acceleration tests in chronological order to obtain the test time and number of failures for each test converted to normal stress level. S52. Sum the results of each test to obtain the cumulative test time T at the normal stress level for each test. iand cumulative number of faults N i And obtain the cumulative MTBF i The calculation formula is: ; S53, with As the independent variable, The parameters of the Duane model were obtained by fitting the data as the dependent variable. and The value is calculated at the time of the last data point and used as the evaluation result of the MTBF of the research and development test.

[0009] Preferably, the development tests in step S2 include extreme condition reliability simulation verification, comprehensive stress condition reliability simulation verification, reliability enhancement / HALT test, and reliability acceleration test, and the qualification tests include single-unit equipment reliability simulation verification, subsystem reliability verification test, and field test.

[0010] Preferably, the design data in step S1 includes the environmental profile and mission profile of the flight control actuator, comprehensive stress conditions, limit condition range, product design parameters and interfaces, and material and component properties.

[0011] Preferably, step S2 specifically includes the following sub-steps: S21. Establish a flight control actuator performance simulation model, conduct limit condition reliability simulation with limit conditions as input, and obtain the limit condition simulation results. S22. Improve the design based on the simulation results of the limit conditions, and design a reliability enhancement / HALT test based on the boundary conditions obtained from the reliability simulation of the limit conditions. S23. Conduct reliability enhancement / HALT tests and revise the flight control actuator performance simulation model based on the reliability enhancement / HALT tests; S24. If the reliability enhancement / HALT test does not fail within the product design limits, proceed to step S25. If a failure occurs, conduct a design iteration and improvement, and then conduct the reliability enhancement / HALT test again. S25. Conduct reliability simulation verification of comprehensive stress conditions as input, and improve the design based on the results of the comprehensive stress condition reliability simulation verification. S26. Design an accelerated reliability test based on the reliability simulation verification results under comprehensive stress conditions. S27. Conduct accelerated reliability tests and revise the flight control actuator performance simulation model based on the results of the accelerated reliability tests; S28. If the reliability accelerated test meets the failure occurrence requirement within the specified time, proceed to step S29; otherwise, conduct the reliability accelerated test again after design iteration and improvement. S29. Conduct single-machine equipment reliability simulation tests, subsystem reliability verification tests, and field tests in sequence.

[0012] Preferably, in step S53, weighted least squares regression analysis is used to fit the parameters of the Duane model. and The specific process for calculating the MTBF at the time of the last data point is as follows: ; in, Let MTBF be the time of the last data point, T be the cumulative experimental time, a be the scale parameter, and m be the growth rate.

[0013] Preferably, the design improvement in step S22 involves adjusting the geometry, increasing structural strength, or changing the material.

[0014] Preferably, the modification of the flight control actuator performance simulation model in steps S23 and S27 specifically involves: inputting the same stress conditions into the flight control actuator performance simulation model and performing simulation; comparing and analyzing the actual fault data observed in the experiment with the simulated fault data; identifying the differences and deviations; and using the measured data as a benchmark to reverse-calibrate the key parameters in the flight control actuator performance simulation model.

[0015] Preferably, the acceleration factor between the strengthening / HALT stress and the conventional stress in step S3 is... The acceleration factor between the accelerated stress and the normal stress in step S4 The calculation method is as follows: Based on the modified flight control actuator performance simulation model, both conventional stress and enhanced / HALT stress were applied, and the simulation model was run until a failure occurred. The time of failure in the simulation model was recorded, and the ratio of the failure time under conventional stress to the failure time under enhanced / HALT stress was taken as the acceleration factor between enhanced / HALT stress and conventional stress. ; Apply both conventional and accelerated stresses, and run the flight control actuator performance simulation model until a failure occurs. Record the time of failure in the simulation model. The ratio of the failure time under conventional stress to the failure time under accelerated stress is taken as the acceleration factor between the accelerated and conventional stresses. .

[0016] Preferably, step S63 specifically includes the following steps: S631. Sort the uncertainty data obtained in step S62 from smallest to largest, as follows: ; S632, with As prior information, construct the prior distribution and The data obtained after the first fusion is used as the first fusion data; S633. Using the first fusion data as prior information, construct a prior distribution and... The data after the second fusion is used as the second fusion data; S634. Using the second fusion data as prior information, construct the prior distribution and... A third fusion is performed to obtain the posterior distribution of the multi-source data fusion, which serves as the comprehensive evaluation result of the multi-type, multi-stage MTBF.

[0017] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) This invention proposes a comprehensive evaluation method for quantitative verification of the reliability of aircraft flight control actuators by fusing multi-stage and multi-type data. It establishes a process for simulation test, reliability acceleration test and reliability enhancement / HALT test in the development stage of aircraft flight control actuators, fully considers the characteristics of different types of data in each stage, and studies a comprehensive evaluation method for the mean time between failures by fusing simulation test data and physical test data. It has very high feasibility and effectiveness in carrying out quantitative comprehensive evaluation of the reliability of aircraft flight control actuators.

[0018] (2) The method of the present invention can support the implementation process of product design simulation test and physical test, and give full play to the advantages of simulation test and physical test. In the early stage of development, when the physical prototype has not yet been manufactured, the high-fidelity simulation model can be fully utilized to identify potential failure modes and cover the boundary of reliability assessment under extreme conditions. In the later stage of development, a small number of highly realistic physical test data can be used as anchors to calibrate and verify the simulation model, thereby improving the credibility of the simulation model.

[0019] (3) The method of the present invention breaks through the limitations of previous assessments and realizes the full integration of data throughout the entire life cycle. The system integrates the test data of the development and evaluation process, takes into account the reliability growth trend of the development process, takes into account the uncertainty of various tests, takes the test data with the least uncertainty as the prior distribution, and integrates it with the data of other stages, fully explores the value of various types of data in each stage of development, and increases the confidence of the overall assessment.

[0020] (4) The method of the present invention can realize the "forward" and "parallel" verification work, thereby greatly improving the credibility of the evaluation results. In view of the high reliability requirements of aircraft flight control actuators, it is not necessary to carry out long-term physical tests to obtain a comprehensive evaluation result of the reliability quantitative requirements with a high level of credibility. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of the overall process of the present invention; Figure 2 This is a schematic block diagram of the test process of the present invention; Figure 3 This outlines the experimental procedures for simulation and physical tests prior to the comprehensive reliability evaluation of this invention. Figure 4 This invention provides an MTBF assessment for the identification stage based on Bayes theory. Detailed Implementation

[0022] Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings.

[0023] This invention provides a quantitative comprehensive evaluation method for the reliability of aerospace flight control actuators based on data fusion, such as... Figure 1 As shown, it includes the following steps: S1. Collect design data for the flight control actuator; the design data includes environmental and mission profiles of the flight control actuator, comprehensive stress conditions, limit condition ranges, product design parameters and interfaces, and material and component properties.

[0024] S2. Conduct development and qualification tests. Development tests include extreme condition reliability simulation verification, comprehensive stress condition reliability simulation verification, reliability enhancement / HALT testing, and accelerated reliability testing. Qualification tests include single-unit equipment reliability simulation verification, subsystem reliability verification testing, and field testing. The comprehensive reliability evaluation data obtained during the development process includes extreme condition simulation test data, reliability enhancement / HALT test data, multi-environment and multi-condition simulation data, accelerated reliability test data, single-unit equipment reliability simulation data, subsystem reliability test data, and field test data. The test procedure design for simulation tests and physical tests before the comprehensive reliability evaluation is attached. Figure 2 As shown.

[0025] In its specific implementation, step S2 includes the following sub-steps: S21. Establish a performance simulation model for the flight control actuator. Conduct extreme condition reliability simulations using extreme conditions as input to identify the product's weaknesses and failure modes under extreme conditions, and analyze the failure mechanisms of these weaknesses. Obtain the simulation results under extreme conditions. In practical applications, a mature simulation model can be selected for the flight control actuator performance simulation model.

[0026] S22. Improve the design based on the simulation results of the limit conditions, and conduct reliability enhancement / HALT test design based on the boundary conditions obtained from the reliability simulation of the limit conditions; the design improvement includes adjusting the geometry, increasing the structural strength, changing the material or other designs.

[0027] S23. Conduct reliability enhancement / HALT tests and revise the flight control actuator performance simulation model based on the reliability enhancement / HALT tests.

[0028] S24. If the reliability enhancement / HALT test does not fail within the product design limits, proceed to step S25. If a failure occurs, conduct the reliability enhancement / HALT test again after design iteration and improvement.

[0029] S25. Conduct reliability simulation verification of comprehensive stress conditions as input, and improve the design based on the results of the comprehensive stress condition reliability simulation verification.

[0030] S26. Design an accelerated reliability test based on the reliability simulation verification results under comprehensive stress conditions.

[0031] S27. Conduct accelerated reliability tests and revise the flight control actuator performance simulation model based on the results of the accelerated reliability tests.

[0032] S28. If the reliability acceleration test meets the failure occurrence requirements within the specified time, proceed to step S29; otherwise, conduct the reliability acceleration test again after design iteration and improvement.

[0033] S29. Conduct single-machine equipment reliability simulation tests, subsystem reliability verification tests, and field tests in sequence.

[0034] Specifically, the modification of the flight control actuator performance simulation model in steps S23 and S27 involves: measuring and calculating the product failure sequence, failure threshold, and performance parameter response data during reliability enhancement / HALT testing of the aircraft flight control actuator; inputting the same stress conditions into the simulation model and performing simulation; comparing and analyzing the actual failure data observed in the test with the simulated failure data to identify any differences and deviations; and then, using the measured data as a benchmark, reverse-calibrating the key parameters in the simulation model to ensure that the model's response is consistent with the actual test results.

[0035] S3, such as Figure 3 As shown, the failure time of the reliability enhancement / HALT test is converted to the failure time under normal stress conditions. : ;in, To convert the failure time of the reliability enhancement / HALT test to the failure time under normal stress conditions, For reliability enhancement / HALT testing, the failure time is... This represents the acceleration factor between hardening / HALT stress and conventional stress. Specifically, it refers to the acceleration factor between hardening / HALT stress and conventional stress. The calculation method is as follows: Based on the modified flight control actuator performance simulation model, apply both conventional stress and enhanced / HALT stress, run the simulation model until a failure occurs, record the time of failure in the simulation model, and use the ratio of the failure time under conventional stress to the failure time under enhanced / HALT stress as the acceleration factor between enhanced / HALT stress and conventional stress. .

[0036] S4. Convert the failure time in accelerated reliability testing to the failure time under normal stress conditions. : ;in, To convert the failure time in accelerated reliability testing to the failure time under normal stress conditions, To accelerate the failure time in reliability testing, This is the acceleration factor between accelerated stress and normal stress.

[0037] Among them, the acceleration factor between accelerated stress and normal stress The calculation method is as follows: Based on the modified flight control actuator performance simulation model, both conventional and accelerated stresses were applied, and the simulation model was run until a fault occurred. The time of the fault in the simulation model was recorded, and the ratio of the fault time under conventional stress to the fault time under accelerated stress was taken as the acceleration factor between the accelerated stress and the conventional stress. .

[0038] S5. Based on the results of steps S3 and S4, conduct an MTBF evaluation during the research and development testing phase; step S5 specifically includes the following sub-steps: S51. Sort the n reliability enhancement / HALT tests and m reliability acceleration tests in chronological order to obtain the test time and number of failures converted to normal stress level for each test; the specific conversion process is carried out according to the methods in steps S3 and S4 respectively.

[0039] S52. Add up the converted test results to obtain the cumulative test time T at the normal stress level for each test. i and cumulative number of faults N i And obtain the cumulative MTBF i The calculation formula is: ; S53. The Duane model is selected to describe the growth trend of MTBF. As the independent variable, As the dependent variable, weighted least squares regression analysis was performed to fit the parameters of the Duane model. and The value is calculated at the time of the last data point in the fitted Duane model, and used as the evaluation result of the MTBF of the research and development experiment. The specific process is as follows: ; in, Let T be the cumulative MTBF after the i-th trial, and T be the cumulative trial time. The cumulative number of failures observed up to time T; a is the scale parameter; m is the growth rate; Take the logarithm of both sides of the above expression: ; The final state's MTBF value for: .

[0040] S6. A multi-stage comprehensive evaluation of MTBF based on Bayesian theory is conducted, with the following specific steps: S61. Collect the total test time and number of failures of single-unit equipment reliability simulation data, subsystem reliability verification test data and field test data, as well as the evaluation results of MTBF in the development and testing phase obtained in step S5.

[0041] S62. Quantify the uncertainty parameters of the data: S621. Calculate the uncertainty of reliability simulation data for single-machine equipment. : ; in, The value ranges from 0 to 1, representing the degree of conformity between the simulation model and the real product. The value ranges from 0 to 1, indicating the reliability of the simulation input parameter source. and These are weighting coefficients, and their sum is 1. Relevant data can be obtained from an expert database.

[0042] S622, Uncertainty in reliability verification test data of the calculation subsystem And the uncertainty of the evaluation results of MTBF during the research and development testing phase. : ; in, This refers to the actual test time; For reference test time; The coefficient of variation between the test environment and the actual use environment; The value ranges from 0 to 1, representing the completeness of the data collection. Subscript, representing or . , All data were obtained from an expert database.

[0043] S623, Calculate the uncertainty of field test data : ; in, This represents the actual number of field failures. The minimum effective number of faults threshold; For field data collection coverage; To record accuracy scores; and All data were obtained from an expert database.

[0044] S63. Perform Bayesian fusion on the uncertainty data obtained in step S62 to obtain a comprehensive evaluation result of multi-type and multi-stage MTBF.

[0045] Preferably, step S63 specifically includes the following steps: S631. Sort the uncertainty data obtained in step S62 from smallest to largest, as follows: ; S632, with As prior information, construct the prior distribution and The data obtained after the first fusion is used as the first fusion data. The specific fusion formula is as follows: Prior information: Total test time Number of faults Assume it follows a gamma distribution Then the prior parameters are: The prior density is: Data to be fused: Experiment time is Number of faults Then the likelihood function is: Assume the posterior distribution follows a gamma distribution. Then the posterior density function is: Then there is Therefore, the posterior distribution point estimate of MTBF obtained from the Bayesian distribution is: The confidence level of MTBF is The interval estimate is: in, , Gamma distribution and Quantiles. This value represents the data from the first fusion. Subsequent fusions follow the same method and will not be described further.

[0046] S633. Using the first fusion data as prior information, construct a prior distribution and... The data after the second fusion is used as the second fusion data.

[0047] S634. Using the second fusion data as prior information, construct the prior distribution and... A third fusion is performed to obtain the posterior distribution of the multi-source data fusion, which serves as the comprehensive evaluation result of the multi-type, multi-stage MTBF. Once the comprehensive evaluation result of the multi-type, multi-stage MTBF is obtained, the MTBF evaluation can be carried out.

[0048] In practical applications, assuming that single-unit equipment reliability simulation data is used as prior information, a priori analysis is constructed. The evaluation results of development test data, subsystem reliability verification test data, and field flight test data are used as sample information. The overall workflow diagram is as follows: Figure 4 As shown. Specific Implementation This embodiment takes a certain aircraft flight control actuator as the object to carry out a comprehensive assessment of the reliability quantitative verification requirements, and evaluates the mean time between failures and the probability of functional failure of the actuator jamming failure mode.

[0050] S1. Collect design data for the flight control actuator. Obtain basic evaluation data: During the development and evaluation phase of the actuator, the following multi-source data were obtained through simulation and physical tests: Extreme condition reliability simulation test: simulate extreme working conditions such as +85℃ high temperature, -55℃ low temperature, peak voltage, and maximum hydraulic shock.

[0051] Reliability Enhancement / HALT Test Data: Two aircraft flight control actuators were tested, with a cumulative test time of 100 hours, resulting in two failures (one overstress failure was eliminated).

[0052] Accelerated reliability testing: Three aircraft flight control actuators were tested, and two valid associated failures occurred during a cumulative 8,000 hours of testing.

[0053] Single-unit equipment reliability simulation test: On the modified simulation model, a 50,000-hour simulation test was conducted to simulate normal stress conditions, and one failure occurred.

[0054] Subsystem reliability verification test data: The subsystems underwent joint testing and accumulated 5,000 hours of operation without any failures.

[0055] Field test data: The test aircraft was used for test flights, accumulating 2000 hours of flight time without any malfunctions.

[0056] S2. Conduct research and development tests and evaluation tests, and improve the design and optimize the performance model based on the tests.

[0057] Design improvements based on ultimate stress simulation: Through extreme stress simulation, the stress concentration factor of a bearing housing inside the actuator exceeded the standard under the conditions of -55℃ and maximum load. After FMEA analysis, it was decided to change the material from ordinary alloy steel to high-toughness low-temperature alloy steel.

[0058] Performance simulation model correction based on reliability enhancement / HALT testing: The upgraded product underwent a reliability enhancement / HALT test, during which two faults were induced: one was a microcrack in the aforementioned bearing housing (a valid fault), and the other was a loose sensor wiring caused by resonance of the test fixture (an overstress fault, which was discarded). The fault data and stress conditions of the "microcrack in the bearing housing" were input into the simulation model, and the material fatigue life parameters in the simulation model were calibrated.

[0059] S3. Convert the failure time of the reliability enhancement / HALT test to the failure occurrence time under normal stress conditions, based on the step stress condition conversion of the simulation model: Using the modified performance simulation model, a failure time of 18,000 hours was obtained under normal stress. Inputting the same stress conditions as the reliability enhancement / HALT test, a failure time of 8,000 hours was obtained. The calculated acceleration factor was 18,000 / 80,000 = 22.5. Since the failure time in the reliability enhancement / HALT test was 100 hours, the equivalent normal stress cumulative test time was calculated to be 2,250 hours, with one effective failure. The calculated MTBF observation value was 2,250 hours.

[0060] S4. Convert the failure time of the accelerated reliability test to the failure time under normal stress conditions. Based on comprehensive stress simulation, the motor driver exhibits performance degradation under high temperature and high vibration conditions. The driver is improved, and accelerated reliability testing is conducted. The first accelerated test accumulated 1000 hours, resulting in two valid failures. Using the failure data from this accelerated test, the parameters in the fatigue life model of the solder joints and the servo valve model in the simulation model are corrected. Using the corrected model, the acceleration factor for this accelerated condition is calculated to be 18. Therefore, the MTBF (Mean Time Between Failures) of the above test data converted to normal stress conditions is 9000 hours.

[0061] A second accelerated test was conducted on the improved product, accumulating 1500 hours. One valid failure occurred. The acceleration factor under the second accelerated test conditions was calculated to be 20, and the MTBF under the equivalent conventional conditions was 30000 hours.

[0062] S5. Based on the results of steps S3 and S4, conduct an MTBF evaluation during the development and testing phase. Arrange the test data during the development phase in chronological order: ① Reliability Enhancement / HALT Test: Cumulative equivalent test time is 2250h, cumulative failure count is 1, so the cumulative MTBF is 2250h; ② First Accelerated Reliability Test: Cumulative equivalent test time is 2250+18000=20250h, cumulative failure count is 3, so the cumulative MTBF is 6750h; ③ Second Accelerated Reliability Test: Cumulative equivalent test time is 20250+30000=50250h, cumulative failure count is 4, so the cumulative MTBF is 12562.5h.

[0063] The Duane model was fitted using three data points (ln2250, ln2250), (ln20250, ln6750), and (ln50250, ln12562.5) to obtain the parameters. , The cumulative equivalent test time of 50250h at the end of the development phase was evaluated, and the evaluation value of MTBF for the development phase was 11693h.

[0064] S6. Multi-stage MTBF comprehensive evaluation based on Bayesian theory. The gamma distribution is used as the prior distribution of failure rate. The parameters in Gamma(a, b) are calculated using single-machine equipment reliability simulation test data. Then, three fusions are performed to obtain the estimated value of MTBF in the qualification stage. First fusion: The MTBF evaluation value obtained during the development phase is equivalent to the equivalent test time corresponding to a failure. This value is used as the likelihood function and fused with the prior distribution to update the parameters in Gamma(a, b). Second fusion: Subsystem reliability verification. Since there is no fault data, the shape parameter remains unchanged, the scale parameter time is increased, and the parameters in the updated Gamma(a, b) are obtained. The third fusion: Since there is no fault data in the field test, the shape parameters are kept unchanged, the scale parameter time is increased, and the parameters in the updated Gamma(a, b) are obtained.

[0065] The final posterior distribution is Gamma (a=2, b=68693). At a 90% confidence level, the lower confidence limit of MTBF is 11000h, which is the final comprehensive evaluation result.

[0066] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made by those skilled in the art to the technical solutions of the present invention without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A data fusion method for quantitatively synthesizing reliability of an aircraft flight control actuator, characterized in that: It comprises the following steps: S1, collecting the design data of the flight control actuator; S2, carrying out development test and identification test; S3, convert the failure time of the reliability enhancement / HALT test to a failure occurrence time under normal stress conditions : ; wherein, is the failure time of the reliability enhancement / HALT test, is an acceleration factor between the enhancement / HALT stress and normal stress; S4, converting the failure time of the reliability accelerated test into a failure occurrence time under regular stress conditions : ; wherein is the failure time of the reliability acceleration test, is the acceleration factor between the acceleration stress and the regular stress; S5, based on the results of step S3 and step S4, carrying out the development test phase MTBF evaluation; S6, based on the Bayesian theory, carrying out the multi-phase MTBF comprehensive evaluation, specifically: S61, collecting the test total time and the number of failures of the single machine device reliability simulation data, the subsystem reliability verification test data and the field test data, and the evaluation results of the development test phase MTBF obtained in step S5; S62, quantifying the uncertainty parameters of the data: S621, calculating the single-computer device reliability simulation data uncertainty : ； wherein, a degree of conformity of the simulation model to the real product; a degree of trustworthiness of the simulation input parameter source; and a weight coefficient; S622, calculating the data uncertainty of the system reliability verification test and the data uncertainty of the evaluation result of the development test phase MTBF : ； wherein, is the actual test time; is the reference test time; is the test environment and actual use environment variation coefficient; is the data collection completeness; is the index, representing or ; S623, calculate the uncertainty of the field test data : ； wherein, is the actual number of field failures; is the minimum number of valid failures threshold; is the field data collection coverage; is the record accuracy score; S63, carrying out Bayesian fusion on the uncertainty data obtained in step S62 to obtain the comprehensive evaluation results of the multi-type multi-phase MTBF.

2. The data fusion based quantitative comprehensive reliability evaluation method for aviation flight control actuator according to claim 1, characterized in that: Step S5 specifically comprises the following sub-steps: S51, sorting the n times of reliability enhancement / HALT test and the m times of reliability acceleration test according to the order of the tests to obtain the test time and the number of failures of each test converted to the normal stress level; S52, the cumulative test time T of each test normal stress level is obtained by accumulating and adding the results of each test i and the cumulative failure number N i , and the cumulative MTBF is obtained i , and the calculation formula is: ； S53、with As independent variable, the number of cycles to failure As dependent variable, the number of cycles to failure and and evaluated at the time of the last data point as an estimate of the development test MTBF.

3. The data fusion based quantitative comprehensive reliability evaluation method for aviation flight control actuator according to claim 1, characterized in that: The development test in step S2 includes limit condition reliability simulation verification, comprehensive stress condition reliability simulation verification, reliability enhancement / HALT test and reliability acceleration test, and the identification test includes single machine device reliability simulation verification, subsystem reliability verification test and field test.

4. The data fusion based quantitative comprehensive reliability evaluation method for aviation flight control actuator according to claim 1, characterized in that: The design data in step S1 includes the environment profile and task profile of the flight control actuator, the comprehensive stress condition, the limit condition range, the product design parameters and interface, and the material and component attributes.

5. The data fusion based quantitative comprehensive reliability evaluation method for aviation flight control actuator according to claim 3, characterized in that: Step S2 specifically comprises the following sub-steps: S21, establishing a flight control actuator performance simulation model, carrying out limit condition reliability simulation with limit condition as input to obtain limit condition simulation results; S22, based on the limit condition simulation results, improving the design, and designing the reliability enhancement / HALT test according to the boundary conditions obtained from the limit condition reliability simulation; S23, carrying out the reliability enhancement / HALT test, and correcting the flight control actuator performance simulation model according to the reliability enhancement / HALT test; S24, if no failure occurs in the reliability enhancement / HALT test within the product design limit, then step S25 is performed, if a failure occurs, then the design is iteratively improved and the reliability enhancement / HALT test is carried out again; S25, carrying out comprehensive stress condition reliability simulation verification with comprehensive stress condition as input, and improving the design according to the comprehensive stress condition reliability simulation verification results; S26, designing the reliability acceleration test according to the comprehensive stress condition reliability simulation verification results; S27, carrying out the reliability acceleration test, and correcting the flight control actuator performance simulation model according to the reliability acceleration test results; S28, if the reliability acceleration test meets the failure occurrence requirement within the specified time, then step S29 is entered, otherwise the design is iteratively improved and the reliability acceleration test is carried out again; S29, sequentially carrying out single machine device reliability simulation test, subsystem reliability verification test and field test.

6. The data fusion based quantitative comprehensive reliability evaluation method for aviation flight control actuator according to claim 2, characterized in that: In step S53, regression analysis is performed using the weighted least squares method, and parameters of the Duane model are fitted and The evaluation process of the MTBF at the time of the last data point is specifically as follows: ； wherein, MTBF at the time of the last data point, T is the cumulative test time, a is the scale parameter; m is the growth rate.

7. The data fusion based quantitative comprehensive reliability evaluation method for aviation flight control actuator according to claim 5, characterized in that: The design improvement in step S22 is to adjust the geometric shape, increase the structural strength or replace the material.

8. The data fusion based quantitative comprehensive reliability evaluation method for aviation flight control actuator according to claim 5, characterized in that: The correction of the flight control actuator performance simulation model in steps S23 and S27 is specifically as follows: the same stress conditions are input into the flight control actuator performance simulation model and simulation is performed, the real fault data observed in the test are compared with the simulation fault data, differences and deviations existing are identified, and the key parameters in the flight control actuator performance simulation model are inversely calibrated based on the measured data.

9. The method of claim 1, wherein: Acceleration factor between the accelerated stress in step S3 and the regular stress Acceleration factor between the accelerated stress in step S4 and the regular stress The calculation method is as follows: Based on the modified flight control actuator performance simulation model, apply the conventional stress and the same stress as the strengthening / HALT, run the simulation model until the failure occurs, record the time of failure in the simulation model, and divide the failure time under the conventional stress by the failure time under the same stress as the strengthening / HALT to obtain the acceleration factor between the strengthening / HALT stress and the conventional stress ; The ratio of the failure time under the normal stress to the failure time under the accelerated equivalent stress is taken as the acceleration factor between the accelerated stress and the normal stress .

10. The method of claim 1, wherein: The specific steps of step S63 are specifically as follows: S631, the uncertainty data obtained in step S62 is sorted from small to large, respectively ; S632、with As prior information, the prior distribution is constructed with After the first fusion, it is taken as the first fusion data; S633, taking the first fused data as prior information, constructing a prior distribution and the second fused data after the second fusion is performed; S634, taking the second fused data as prior information, constructing a prior distribution and A third fusion is performed to obtain a posterior distribution of the multi-source data fusion, as a comprehensive evaluation result of the multi-type and multi-stage MTBF.

Images