Aviation equipment ground safety accident prediction method and system based on space-time factors

Abstract

The invention provides an aviation equipment ground safety accident prediction method and system based on space-time factors, and relates to the technical field of resource allocation and risk assessment, and the method comprises the steps: S1, recognizing a test environment safety accident failure mode, and obtaining a failure mode of an aviation equipment ground test environment; s2, setting a safety standard and establishing a safety accident statistical prediction distribution mathematical model according to historical safety accident data; s3, solving the safety accident statistical prediction distribution mathematical model based on test data to obtain safety accident occurrence condition probability model parameters; and S4, predicting the overall safety accident probability of the test environment to determine the safety of the test environment. According to the method, a high-risk safety accident probability estimation problem is converted into a test problem of key parameters, time, distance and the like under controllable application constraints, and quantitative prediction based on small sample data is realized by establishing a mathematical model of a safety accident occurrence condition probability density function and a safety accident probability density function.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of resource allocation and risk assessment, and particularly relates to an aviation equipment ground safety accident prediction method and system based on space-time factors. BACKGROUND

[0002] Ground testing is the basis for ensuring the safety of the whole life cycle of aviation equipment, such as project demonstration, engineering development, installation and finalization, batch production and service support. Test environment safety analysis is an important means of safety risk prevention in ground testing. According to GJB 1405A, safety refers to the ability to not cause injury, harm to health and environment, and damage or loss to equipment or property. Based on this, test environment safety can be set as: in the test process, through systematic design and management, the test process will not cause harm or loss to equipment, materials and environment.

[0003] Currently, PFEMA based on machine, material, method, and environment can identify and extract process items, process steps, process work elements, failure influences, failure modes, and failure causes. With the help of fault tree analysis, the probability of events affecting safety can be predicted under the condition of known single risk event probability. However, due to the complexity of ground test process, it is costly to obtain accident data, and it is difficult to obtain safety accident probability through statistics, especially when the failure event probability is lower than According to the NASA 2023 test safety report, the traditional PFMEA has an 82% miss rate for specific equipment or environmental failure accidents with a probability lower than , and the sample size needs to meet , and 300 million tests are needed when the target probability is Therefore, a large number of tests are needed for low failure probability prediction. However, due to the safety and high value of aviation equipment in ground testing, it is difficult to obtain the safety accident probability by conducting a large number of high-risk destructive tests, making it difficult to evaluate the safety of the environment without quantitative data support.

[0004] Therefore, in view of the current situation of small sample size, high cost, and difficult evaluation of ground test environment safety analysis of aviation equipment, a method is needed to quantitatively predict the test environment safety accident probability caused by abnormal equipment operating state or environmental parameters based on small sample test process data. SUMMARY

[0005] In order to solve the above-mentioned deficiencies of the prior art, the purpose of the present application is to provide a kind of aviation equipment ground safety accident prediction method and system based on space-time factor, by the influence relationship of safety accident with time, distance parameter, based on historical data and constraint test data, the time, distance distribution probability density function of test process is deduced, and the conditional probability density function corresponding to accident occurrence is obtained, the total probability expectation of each accident occurrence is obtained, to evaluate the environmental safety index of multiple accidents, thus, using the conditional probability density function of safety accident occurrence, the space-time influence model in mathematics in environment is established, so as to evaluate and compare the safety of test environment and test process scheme.

[0006] Specifically, in one aspect, the present application provides a kind of aviation equipment ground safety accident prediction method based on space-time factor, which comprises the following steps: S1: identifying the test environment safety accident failure mode related to time, displacement, distance, obtaining the failure mode of aviation equipment ground test environment; S2: according to the failure mode in step S1, set safety standard and establish safety accident statistical prediction distribution mathematical model according to historical safety accident data;Analysis of the failure cause of failure mode, classification is carried out based on safety accident test participation form, no contact leads to failure or contact leads to failure, set safety time standard and safety distance displacement standard, according to historical safety accident data, get safety accident occurrence condition probability density function For: ; Wherein, safety accident occurrence condition probability density function; safety accident occurrence condition probability density function output; safety accident occurrence probability expectation; discrete fitting parameter; constraint level parameter; constraint level; test operation key parameter average; test operation key parameter; According to safety accident occurrence condition probability density function safety accident statistical prediction distribution mathematical model is established; S3: based on test data, the safety accident statistical prediction distribution mathematical model obtained in step S2 is solved, and the safety accident occurrence condition probability model parameter is obtained;Safety accident occurrence condition data is collected by simulation or test, including normal operation historical safety accident occurrence condition data and safety accident occurrence condition enhancement data generated by applying constraint, and safety accident probability is solved; S4: determining the safety of the test environment by the overall safety accident probability of the test environment determined in step S3; determining the overall safety accident probability according to the safety accident probability density function under the actual environment constraint

[0007] Preferably, step S2 is specifically: S21: setting the safety standard of the key test operation parameter; S22: determining the key test operation parameter according to the historical safety accident data, to obtain the safety accident probability density function ; S23: setting the safety accident occurrence condition probability density function , to obtain the safety accident occurrence probability expectation ; S24: establishing a fitting model of the safety accident occurrence condition probability density function ; according to the actual production test safety accident data, inferring the conditional probability of the occurrence of the safety accident; S25: determining the safety accident statistical prediction distribution mathematical model.

[0008] Preferably, the method for obtaining the safety accident probability density function in step S22 is: when the key test operation parameter , the safety accident probability density function is established by using ; wherein, is the average value of the key test operation parameter; is the standard deviation of the key test operation parameter; when the key test operation parameter satisfies the distribution, the safety accident probability density function is established by using , wherein, is the shape parameter of the test operation; is the scale parameter of the test operation.

[0009] Preferably, the safety accident occurrence probability expectation in step S23 is specifically:​​​​​​​​​ ; wherein, is the expected probability of safety accident occurrence; is the conditional probability density function of safety accident occurrence; is the probability density function of safety accident.

[0010] Preferably, the safety accident statistical prediction distribution mathematical model in step S25 is: ; According to the output of the conditional probability density function of safety accident occurrence and the probability density function of safety accident , the parameter value of the conditional probability density function of safety accident occurrence is fitted, so as to calculate the expected probability of safety accident occurrence .

[0011] Preferably, step S3 is specifically: S31: increase the sample size of historical safety accident data by filling more data through experiments, verify the probability density function of safety accident and the average value of the key parameters of the experimental operation and the standard deviation of the key parameters of the experimental operation ; S32: actively induce failure to form safety accident occurrence condition data, and count the safety accident frequency S33: use maximum likelihood estimation to solve the expected probability of safety accident occurrence and the conditional probability density function of safety accident occurrence ; S34: repeat steps S31 to S33 until the conditional probability density function of safety accident occurrence and the probability density function of safety accident of all failure modes are obtained.

[0012] Preferably, step S4 is specifically: S41: calculate the probability of accidents caused by the device or environmental failure mode of the th safety accident type; S42: calculate the overall safety accident probability of at least one device or environmental failure accident in a single round of experiment ; S43: by adjusting the experimental environment arrangement, optimizing the component layout to increase the safety distance, installing protective devices to limit displacement, and optimizing the experimental process scheme, reduce the overall safety accident probability of accidents in a single round of experiment , compare the total probability of accidents in a single round of experiment Safety of the test environment arrangement and the test process scheme is evaluated.

[0013] Preferably, the probability of the first kind of safety accident caused by the equipment or environment failure mode of the first kind of safety accident is equal to the expected probability of the first kind of safety accident, and specifically is: wherein, is the expected probability of the first kind of safety accident.

[0014] In another aspect, the present application provides an aviation equipment ground safety accident prediction system based on space-time factors, which comprises a safety accident identification module, a safety accident statistical prediction distribution module, a safety accident probability analysis module and a safety accident probability prediction module. The safety accident identification module can identify the test environment safety accident failure mode related to time, displacement and distance, and obtain the failure mode of the aviation equipment ground test environment. The safety accident statistical prediction distribution module can analyze the failure cause of the failure mode, classify based on the safety accident test participation form, set safety standards according to the failure mode, and establish a safety accident statistical prediction distribution mathematical model according to historical safety accident data. The safety accident probability analysis module collects safety accident occurrence condition data through simulation or test, solves the safety accident statistical prediction distribution mathematical model based on the test data, obtains the safety accident occurrence condition probability model parameters, and solves the safety accident probability. The safety accident probability prediction module determines the safety of the test environment through the overall safety accident probability of the test environment, and predicts the overall safety accident probability according to the safety accident probability density function under the actual environmental constraints.

[0015] Compared with the prior art, the present application has the following beneficial effects: (1) The present application converts the high-risk safety accident probability estimation problem into a controllable test problem of key parameters such as time and distance under constraints, and realizes quantitative prediction based on small sample data by establishing a model of safety accident occurrence condition probability density function and safety accident probability density function.

[0016] (2) The segmented discretization condition probability model and the enhanced data generation method proposed by the present application significantly reduce the difficulty of data acquisition; when appropriate discrete fitting parameters are selected, the present application only needs dozens of constraint tests to obtain the probability estimation result.

[0017] (3) The overall safety accident probability prediction method provided by the present invention can quantitatively evaluate the test environment layout and provide a quantifiable decision basis for optimizing test safety design based on the differences in safety impact caused by equipment operation or environmental anomalies. Attached Figure Description

[0018] Figure 1 This is a control block diagram of the aviation equipment ground safety accident prediction method based on spatiotemporal factors according to the present invention. Figure 2 This is an overall flowchart of the method for predicting ground safety accidents of aviation equipment according to the present invention; Figure 3 The time distribution probability density function obtained from the propeller wind tunnel test in this invention A schematic diagram; Figure 4 The probability density function of the propeller wind tunnel test obtained in this invention is the conditional probability density function for the occurrence of an accident. A schematic diagram; Figure 5 This is a schematic diagram of the original layout used in the wind tunnel test of the propeller in this invention. Figure 6 This is a schematic diagram illustrating the improved layout applied in a propeller wind tunnel test according to the present invention; Figure 7 The time distribution probability density function is obtained for the landing gear drop test in this invention. A schematic diagram; Figure 8 The probability density function for the accident condition obtained from the landing gear drop test in this invention. A schematic diagram. Detailed Implementation

[0019] Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings.

[0020] This invention proposes a method for predicting ground safety accidents of aviation equipment based on spatiotemporal factors, such as... Figure 1 As shown, the failure modes of safety accidents in the test environment are identified to obtain the failure modes of the ground test environment for aviation equipment; safety standards are set and a mathematical model for the statistical prediction distribution of safety accidents is established based on historical safety accident data; the mathematical model for the statistical prediction distribution of safety accidents is solved based on the test data to obtain the parameters of the conditional probability model for the occurrence of safety accidents; the overall safety probability of the test environment is predicted through step S3 to determine the safety of the test environment; as shown... Figure 2 The diagram shown is an overall flowchart of the method of the present invention, and the specific steps are as follows: Step S1: identify the test environment safety accident failure mode related to time, displacement, distance, and obtain the failure mode of the ground test environment of the aviation equipment. According to the ground test process of the aviation equipment, based on the process failure mode and effect analysis PFMEA, the process items of the test process are refined and decomposed to process steps, and the safety accident failure mode related to time, distance and displacement is extracted from the process steps. Key equipment, material, method, and environment. Specifically includes: Step S11: based on the process failure mode and effect analysis PFMEA, determine the safety accident type with potential risk.

[0021] Step S12: extract the test environment safety accident failure mode related to time, displacement, distance under the equipment operating state or environment parameter from the ground test environment safety accident of the aviation equipment. The selection basis is that the failure mode is caused by the fact that the key equipment action time exceeds the safety window, is too long or too short, the key component displacement exceeds the safety threshold, the actual distance between key components is less than the minimum safety distance, and the key environment parameter exceeds the safety range. It also includes failure modes related to operation distance and time that affect test safety, such as test near a certain distance for too long and near a certain distance for too short.

[0022] The embodiment of the application identifies four typical failure modes in the propeller wind tunnel test process, which are: disassembly failure, which is manifested as disassembly failure and injury. Adjustment failure, which is manifested as adjustment error leading to error. Installation failure, which is manifested as poor installation leading to flying propeller. Test and evaluation failure, which is manifested as injury due to too close distance.

[0023] Step S2: set safety standards according to the failure modes in step S1 and establish a safety accident statistical prediction distribution mathematical model according to historical safety accident data; analyze the failure cause of the failure mode, classify based on the safety accident test participation form, no contact leading to failure or contact leading to failure, set safety time standards and safety distance displacement standards, according to historical safety accident data, establish a safety accident statistical prediction distribution mathematical model of safety accidents, and determine the mathematical relationship that meets the safety standards.

[0024] Step S21: set the safety standards of the test operation key parameters, including the safety time standards, such as test operation time, equipment action time window greater than or equal to the threshold And the safety distance displacement standard, the test distance away from the dangerous source, the displacement of the key component, and the safety range distance of the key environment parameter should not be less than the safety threshold . Determine the minimum threshold corresponding to 3 safety time standards and 1 safety distance displacement standard, the first safety time standard is disassembly: . The second safety time standard is adjustment: The third safety time criterion is installation: Test safety distance displacement criterion: .

[0025] Step S22: According to the historical safety accident data, determine the test operation time And test operation distance , unified by variable As the test operation key parameter; the acquisition method of safety accident probability density function : When the test operation key parameter The acquisition method of safety accident probability density function is: ; Where, Safety accident probability density function when ; Test operation key parameter Test operation key parameter average Test operation key parameter standard deviation Pi parameter Natural logarithm

[0026] When the test operation key parameter Satisfy Distribution, the acquisition method of safety accident probability density function is: ; Where, Test operation shape parameter Test operation scale parameter Gamma function of test operation shape parameter .

[0027] In the embodiment, according to Kolmogorov-Smirnov test, the optimal safety accident distribution probability density function is selected based on the statistical results of historical safety accident data: the skewness and kurtosis are obtained by calculating the expected value of the cube and the fourth power of the difference between the historical safety accident data and the average value, if the skewness And kurtosis , use normal distribution, otherwise use Gamma distribution.

[0028] The historical operation data is collected in the embodiment to obtain the statistical parameters of the failure modes of the propeller test as shown in Table 1. The threshold of the safety standard in Table 1 is obtained by experience or by using the average value minus a certain variance, and it is considered that the occurrence probability of the safety accident above the threshold of the safety standard tends to the first occurrence probability of each failure mode The first occurrence probability of each failure mode is obtained by fitting the distribution parameters based on Table 1.

[0029] Table 1: Statistical results of the historical safety accident data of the failure modes of the propeller test Therefore, Figure 3 The schematic diagram of the first safety accident probability density function of the dismounting time failure mode of the propeller wind tunnel test in the embodiment is shown in FIG. 1. The probability density function of the time distribution of the dismounting time failure mode is shown in FIG. 2. Figure 3 The discrete fitting parameters are obtained by corresponding to the safety accident probability density function The safety accident occurrence condition probability density function The safety accident occurrence probability expectation The first safety accident occurrence probability expectation

[0030] Step S23: Set the safety accident occurrence condition probability density function The safety accident occurrence probability expectation is: ; Wherein, is the safety accident occurrence probability expectation; is the safety accident occurrence condition probability density function; is the safety accident probability density function.

[0031] Step S24: Establish the fitting model of the safety accident occurrence condition probability density function According to the actual production test safety accident data, the condition probability of the safety accident under the time and distance condition is inferred. This point is determined according to the actual situation, for example, when the test operation key parameter is large enough, for example, ,​​​​​​ Approaching the expected overall failure probability When the key parameters of the test operation Enough hours Approaching 1. Similarly, there are also key parameters in experimental operations. Sufficient hours, for example hour, Approaching the expected overall failure probability When the key parameters of the test operation When large enough, Approaching 1. The examples only use the key parameters of the first experimental operation. When large enough Approaching the expected probability of a safety accident Let's analyze this with examples.

[0032] A piecewise discretized conditional probability density function is used to fit the conditional probability density function of the occurrence of safety accidents. Let the discrete fitting parameters be... and constraint level parameters ,when hour, ;when At that time, set Thus, the conditional probability density function for the occurrence of a safety accident is obtained. for: ; in, This is the output of the conditional probability density function for the occurrence of a safety accident. These are discrete fitting parameters; For constraint level parameters; For constraint levels.

[0033] In specific embodiments of the present invention, based on the statistical results of the enhanced dataset constrained by experimental operation time and distance, an appropriate probability density function for the occurrence of safety accidents is selected. .

[0034] For example, using the conditional probability density function of safety accident occurrence. Defit The results are as follows: ; in, The parameter representing the probability of an accident occurring is used to characterize the expected probability of a safety accident occurring. ; The shape parameter is used to characterize the rate of decrease.

[0035] At this point, the conditional probability density function for the occurrence of a safety accident Expressed as: ; Step S25: Determine the mathematical model for the statistical prediction distribution of safety accidents as follows: ; At this point, based on the output of the conditional probability density function for the occurrence of a safety accident... and the probability density function of safety accidents Fit the conditional probability density function of the occurrence of a safety accident The parameter values ​​are used to calculate the expected probability of a safety accident occurring. .

[0036] Step S3: Solve the mathematical model of the statistical prediction distribution of safety accidents obtained in step S2 based on the experimental data to obtain the parameters of the probability model of safety accident occurrence conditions; collect safety accident occurrence condition data through simulation or experiment, including historical safety accident occurrence condition data under normal operation and enhanced safety accident occurrence condition data generated by applying constraints, and solve for the probability of safety accidents.

[0037] Step S31: Increase the sample size of historical safety accident data through experiments, enhance the data according to the conditions under which the safety accidents occurred, and verify the results. The probability density function of a safety accident Passing the exam Average value of key parameters for each test operation and the standard deviation of key parameters of the test operation .

[0038] Step S32: Generate an augmented dataset by applying constraints and setting the constraint levels. From 1 to Compress time or distance to Simulate constraint levels to actively induce failures, forming a range from large to small. Data on the conditions under which safety accidents occur, and statistics. accident frequency percentage The number of trials is greater than or equal to If the condition is not met, the subsequent fitted accident occurrence conditional probability density function is not considered. Constraints are applied in ways such as time parameters and action time; action time is changed by increasing load, decreasing hydraulic pressure, etc. For distance parameters, displacement and spacing are changed by applying additional force or setting physical limits.

[0039] Step S33: Solve the first step using maximum likelihood estimation. Expected probability of a safety incident occurring and the The conditional probability density function of a safety accident For example, based on the least square method, the statistical square difference between the fitting curve and the actual value is solved to be minimum.

[0040] Step S34: repeating steps S31 to S33 until obtaining the safety accident occurrence condition probability density function of all failure modes and the safety accident probability density function .

[0041] The embodiment of the present application takes the disassembly operation of the disassembly time failure mode as an example, generates an enhanced data set by artificially restricting the operation time, recording the proportion of the number of near-failure events under the test number in the operation time region, as shown in Table 2 of REF _Ref200809909 \ h, wherein corresponding to the test operation time range . It should be noted that here is not counted in statistics because the test number is not significantly greater than .

[0042] Table 2: Proportion of frequency of constrained test of propeller disassembly time failure mode Statistical table According to step S2, the safety accident occurrence condition probability density function is obtained, and the first safety accident occurrence probability expectation is solved by maximum likelihood estimation, so that the distribution of the accident occurrence condition probability density function is obtained as shown in Figure 4 The schematic diagram of the accident occurrence condition probability density function fitted according to the enhanced data statistical results in the propeller wind tunnel test disassembly blade time failure mode in the present application is shown. Similarly, the second safety accident occurrence probability expectation , the third safety accident occurrence probability expectation and the fourth safety accident occurrence probability expectation are obtained, and based on the relationship between the first accident occurrence condition probability density function and the first safety accident occurrence probability expectation , the safety accident occurrence condition probability density function of each failure mode in the test process is obtained.

[0043] Step S4: determining the safety of the test environment by the overall safety accident probability of the pre-test environment in step S3; and predicting the overall safety accident probability according to the safety accident probability density function under the actual environment constraint. ​​

[0044] Step S41: Calculate the safety accident probability of various failure modes in a single round of testing, such as the first... The probability that a certain type of equipment or environmental failure mode will lead to an accident is equal to the expected probability of the safety accident occurring, specifically: ; Step S42: Calculate the overall safety accident probability of at least one equipment or environmental failure occurring in a single test: ; in, This represents the total probability of an accident occurring in a single round of testing. This represents the total number of equipment or environmental failure incidents.

[0045] Step S43: Adjust the test environment layout, such as optimizing component layout to increase safety distances, adding protective devices to limit displacement, and the test process plan, such as optimizing the control logic action time window, setting stricter environmental parameter monitoring thresholds, and changing key test operation parameters. Safety accident probability density function For example, to average the key parameters of the experimental operation. Further away from the failure boundary and reduce the standard deviation of key test parameters This reduces the overall safety accident probability in a single test. Compare the total probability of an accident occurring in a single round of testing. Assess the safety of the test environment setup and test procedure.

[0046] This invention calculates the overall accident probability of a single test run under the current environment and testing process. Let the first... The conditional probability density function of an accident occurrence Unchanged, changed The probability density function of a safety accident , obtained the Expected probability of a safety incident occurring Therefore, in this propeller test, safety was improved by adjusting the workflow, reasonably increasing the test operation time and the distance from the hazard source. For example, […]. The translational safety accident probability density function increased from 132.4s to 142s A new probability density function for safety accidents is obtained. Assuming If the probability remains unchanged, then the expected probability of the first safety accident is recalculated. , Safety is improved by 41%. Correspondingly, for the two test setups, such as... Figure 5The figure shows the schematic diagram of the original layout in the propeller wind tunnel test of the application, and the operation table is arranged in the test area, which is the initial original design test environment arrangement. Figure 6 The figure shows the schematic diagram of the improved layout in the propeller wind tunnel test of the application, and the operation table is arranged outside the test table far away from the propeller blade, which is the new design test environment arrangement, used to prove the safety level under the test arrangement and process safety standard.

[0047] The second embodiment of the application is a landing gear drop test, focusing on the failure mode of the safety of aviation equipment instruments. This method can handle failures caused by too small time, distance parameters, and can also handle failures caused by too large time, distance displacement parameters, only need to change the accident condition probability density function model.

[0048] Step S51: Identify the failure mode of the safety accident of the test environment related to time and distance. Through PFMEA analysis of the landing gear drop test process of a certain type of aircraft, three typical failure modes are identified: The main landing gear deviates from the central zero position, the displacement parameter fails, and the performance is that the lateral force of the main landing gear landing is too large to damage the equipment.

[0049] The lift of the lift simulation cylinder is insufficient, the time parameter fails, and the performance is that the lift of the lift simulation cylinder is insufficient due to insufficient inflation time.

[0050] The debris near the test table flies out of the protective net, the distance parameter fails, and the performance is that the protective net cannot block the debris at the set distance.

[0051] Step S52: Set safety standards according to the failure modes in step S1 and establish a safety accident statistical prediction distribution mathematical model according to historical safety accident data.

[0052] Based on historical test data, 1 safety time standard and 2 safety distance displacement standards are determined, and the minimum threshold corresponding to failure modes 1 and 3 is as follows: 1. Safety distance displacement standard, maximum displacement of landing deviation from zero point: .

[0053] 2. Safety time standard, minimum inflation time: .

[0054] 3. Safety distance displacement standard, minimum distance between protective net and test table: .

[0055] The 50 historical test data are collected, the parameter distribution is fitted, and the historical safety accident data statistics results are shown in Table 3. The safety threshold in Table 3 is obtained by historical test data or by using average value plus or minus certain variance, and the "maximum" describes the test operation key parameter The safety accident occurrence condition probability is close to the first The safety accident occurrence probability expectation The "minimum" describes the test operation key parameter The safety accident occurrence condition probability is close to the first The safety accident occurrence probability expectation Based on Table 3, REF _Ref200719561 \h and the measured data parameter condition, the safety accident probability density function of the failure mode is fitted .

[0056] Table 3 Historical safety accident data statistics table of landing gear drop test failure mode Taking the ground contact deviation distance failure mode as an example, the safety accident probability density function of the distance distribution is shown in Table 3 As Figure 7 shown, the safety accident probability density function of the time distribution is fitted according to the normal operation historical safety accident data statistics results in the landing gear drop test buffer overrun distance failure mode in the application The shape parameter , the scale parameter , , the maximum probability of near-loss event is determined when the buffer overrun distance is 0.01 m. In order to prevent the test from causing serious damage to the equipment, the test is selected to be carried out in the range of 0.002 m to 0.01 m, and certain additional buffer measures are added to prevent unacceptable test accidents.

[0057] Corresponding to , the model is established, corresponding to , and the specific is: ; When the stroke distance , corresponding to , the maximum selected in the test is 8.

[0058] Step S53: Solve the expected probability of safety accident occurrence by constraint test and the first safety accident occurrence condition probability density function .

[0059] The test scheme with constraints is designed to simulate severe working conditions, such as ground contact deviation. By adjusting the deviation position, the stress hazard event caused by deviation is measured by adding a strain gauge. The lifting cylinder is simulated by adjusting the limit pressure time, and the near miss event is determined by confirming whether the lifting force is reached. The protective net is based on the symmetry assumption to establish multiple layers at intervals, and the number of debris reaching each protective net in each test is counted to determine whether a near miss event occurs at a certain distance.

[0060] Taking the ground contact deviation failure mode as an example, the initial value of the displacement distance is changed by applying constraints, and the proportion of near miss events in the test number within the interval distance size range is recorded to generate an enhanced data set, as shown in Table 4 of REF _Ref202972217\h, wherein corresponds to the test distance range . It should be noted that here to are not counted in the statistics, as the test number is not significantly greater than .

[0061] Table 4 Enhanced data set for ground contact case collection According to the safety accident occurrence condition probability density function of step S52, the first expected probability of safety accident occurrence is solved by maximum likelihood estimation, and the distribution of the accident occurrence condition probability density function is obtained as shown in Figure 8 . The schematic diagram of the accident occurrence condition probability density function obtained by fitting the enhanced data statistics results in the landing gear drop test buffer overrun distance failure mode in the present application is shown. Similarly, the second expected probability of safety accident occurrence and the third expected probability of safety accident occurrence are obtained, and based on the relationship between the accident occurrence condition probability density function and the first expected probability of safety accident occurrence , the safety accident occurrence condition probability density function of each failure mode in the test process is obtained.

[0062] Step S54: Determine the safety of the test environment by predicting the overall safety accident probability of the test environment in step S3. Calculate the overall accident probability of a single test run under the current environment and test process. Let the conditional probability density function for the occurrence of a safety accident be... Unchanged, Changed , to obtain a new first Expected probability of a safety incident occurring Therefore, in this landing gear test, by further improving the accuracy of the landing contact points, Reducing the value from 0.002m to 0.0015m lowers the expected probability of the first safety accident. For the inflation process of the simulated lifting actuator, improving or monitoring the inflation time can reduce the probability of a second safety accident. For the protective net, based on the expected probability of the third safety incident. Size and appropriate distance settings are selected to ensure the safety of the test environment.

[0063] The second aspect of this invention provides a ground safety accident prediction system for aviation equipment based on spatiotemporal factors, which includes: a safety accident identification module, a safety accident statistical prediction distribution module, a safety accident probability analysis module, and a safety accident probability prediction module.

[0064] The safety incident identification module can identify the failure modes of test environment safety incidents related to time, displacement, and distance, and obtain the failure modes of the ground test environment of aviation equipment.

[0065] The safety accident statistics prediction distribution module can analyze the failure causes of failure modes, classify them based on the participation form of safety accident tests, set safety standards according to failure modes, and establish a mathematical model for safety accident statistics prediction distribution based on historical safety accident data.

[0066] The safety accident probability analysis module collects data on the conditions for the occurrence of safety accidents through simulation or experimentation, solves the mathematical model of the statistical prediction distribution of safety accidents based on the experimental data, obtains the parameters of the probability model of the conditions for the occurrence of safety accidents, and solves the probability of safety accidents.

[0067] The safety accident probability prediction module determines the safety of the test environment by the overall safety accident probability of the test environment, and predicts the overall safety accident probability based on the safety accident probability density function under the actual environmental constraints.

[0068] The present application has the following advantages: the present application converts the high-risk safety accident probability estimation problem into a controllable test problem of key parameters, time, distance and the like under constraints, realizes quantitative prediction based on small sample data by establishing a mathematical model of a safety accident occurrence condition probability density function and a safety accident probability density function, and realizes the quantitative prediction based on small sample data by using an accelerated life test method; the data acquisition difficulty is significantly reduced by using a segmented discrete condition probability model and an enhanced data generation method; it can be proved by the embodiments that, when appropriate discrete fitting parameters are selected, the present application can obtain a probability estimation of the order of magnitude of tens of groups of constraint tests, compared with the order of magnitude of billions of sample quantities required by a traditional method The overall safety accident probability prediction method of the present application can quantitatively evaluate the influence of test environment arrangement, such as equipment layout, protection setting and process design, such as control logic and parameter threshold, on the safety caused by equipment operation or environmental abnormality, and provides a quantifiable decision basis for optimizing the safety design of the test.

[0069] The above-described embodiments are only used to describe the preferred embodiments of the present application, and do not limit the scope of the present application. Various modifications and improvements to the technical solutions of the present application made by those skilled in the art without departing from the design spirit of the present application shall fall within the protection scope of the present application defined by the claims.

Claims

1. A method for predicting ground safety accidents of aviation equipment based on space-time factors, characterized in that: It comprises: S1: identifying the test environment safety accident failure mode related to time, displacement and distance, obtaining the failure mode of the ground test environment of the aviation equipment; S2: Set safety standards according to the failure mode in step S1 and establish a safety accident statistical prediction distribution mathematical model according to historical safety accident data; analyze the failure causes of the failure mode, classify based on the safety accident test participation form, set safety time standards and safety distance displacement standards, and obtain the safety accident occurrence condition probability density function according to historical safety accident data For: ； wherein, is the safety incident occurrence conditional probability density function; is the output of the safety incident occurrence conditional probability density function; is the safety incident occurrence probability expectation; is the discrete fitting parameter; is the constraint level parameter; is the constraint level; is the test operating key parameter average value; is the test operating key parameter; According to the safety accident occurrence condition probability density function A safety accident statistical prediction distribution mathematical model is established; S3: solving the safety accident statistical prediction distribution mathematical model obtained in step S2 based on the test data, obtaining the safety accident occurrence condition probability model parameters; collecting safety accident occurrence condition data through simulation or test, including normal operation historical safety accident occurrence condition data and safety accident occurrence condition enhancement data generated by applying constraints, and solving the safety accident probability; S4: determining the safety of the test environment according to the overall safety accident probability of the test environment determined by step S3 and the safety accident probability density function under the actual environment constraint , the overall safety accident probability is predicted as : ； wherein, is the total probability of accidents in a single trial; is the total number of equipment or environmental failure accidents; is the number of the th safety accident probability expectation; is the safety accident type number.

2. The spatio-temporal factor based method for predicting ground safety incidents of aerial equipment according to claim 1, characterized in that: Step S2 is specifically: S21: setting the safety standard of the test operation key parameter; S22: determining the key parameters of the test operation according to the historical safety accident data , obtaining a safety accident probability density function ; S23: set the safety accident occurrence condition probability density function , get the safety accident occurrence probability expectation ; S24: Establishing the safety accident occurrence condition probability density function of the fitting model; according to the actual production test safety accident data, inferring the safety accident occurrence condition probability; S25: determining the safety accident statistical prediction distribution mathematical model.

3. The spatio-temporal factor based method for predicting ground safety incidents of aerial equipment according to claim 2, wherein: The safety accident probability density function in step S22 is obtained by the following method: The obtaining method is as follows: When the test operating key parameter is used to establish a safety accident probability density function ; wherein, is the test operating key parameter average value; is the test operating key parameter standard deviation; When the test operating key parameters satisfy distribution, use to establish the safety accident probability density function , wherein is the test operating shape parameter; is the test operating scale parameter.

4. The spatio-temporal factor based method for predicting ground safety incidents of aerial equipment according to claim 2, wherein: The expected probability of a safety accident in step S23 , specifically: ； wherein, is the expected probability of a safety incident; is the conditional probability density function of a safety incident; is the probability density function of a safety incident.

5. The spatio-temporal factor based method for predicting ground safety incidents of aerial equipment according to claim 2, wherein: The safety accident statistical prediction distribution mathematical model in step S25 is: ； According to the output of the safety accident occurrence condition probability density function and the safety accident probability density function , the parameter value of the safety accident occurrence condition probability density function is fitted, thereby calculating the safety accident occurrence probability expectation .

6. The spatio-temporal factor based method for predicting ground safety incidents of aerial equipment according to claim 1, wherein: Step S3 is specifically: S31: increase the sample size of historical safety accident data by filling more data through experiments, and verify the safety accident probability density function and the average value of the key parameters of the test operation and the standard deviation of the key parameters of the test operation ; S32: actively inducing failure to form safety accident occurrence condition data and count the safety accident frequency; S33: Solving the expected probability of safety accident occurrence by using maximum likelihood estimation and the conditional probability density function of safety accident occurrence ; S34: repeating steps S31 to S33 until the safety accident occurrence condition probability density function of all failure modes is obtained and the safety accident probability density function .

7. The spatio-temporal factor based method for predicting ground safety incidents of aerial equipment according to claim 1, wherein: Step S4 is specifically: S41: calculate the probability of the type of safety accident of the device or environment failure mode leading to an accident; S42: Calculate the overall safety accident probability of at least one device or environmental failure accident occurring in a single round of testing ; S43: By adjusting the test environment arrangement, optimizing the component layout, increasing the safety distance, adding protective devices to limit displacement, and optimizing the test process scheme, the overall safety accident probability of accidents occurring in single-wheel tests is reduced , the total probability of accidents occurring in single-wheel tests is compared The safety of the test environment arrangement and the test process scheme is evaluated.

8. The spatio-temporal factor based method for predicting ground safety incidents of aerial equipment according to claim 1, wherein: The probability of the accident caused by the device or environment failure mode of the first kind of safety accident type is equal to the probability expectation of the first kind of safety accident, specifically: ； wherein, is the probability of a safety incident.

9. A space-time factor based ground safety incident prediction system for aerial equipment, as claimed in claim 1, wherein, It comprises: a safety accident identification module, a safety accident statistical prediction distribution module, a safety accident probability analysis module and a safety accident probability prediction module; The safety accident identification module can identify the test environment safety accident failure mode related to time, displacement and distance, and obtain the failure mode of the ground test environment of the aviation equipment; The safety accident statistical prediction distribution module can analyze the failure cause of the failure mode, classify based on the safety accident test participation form, set the safety standard according to the failure mode, and establish a safety accident statistical prediction distribution mathematical model according to the historical safety accident data; The safety accident probability analysis module collects safety accident occurrence condition data through simulation or test, solves the safety accident statistical prediction distribution mathematical model based on the test data, obtains the safety accident occurrence condition probability model parameters, and solves the safety accident probability; The safety accident probability prediction module determines the safety of the test environment through the overall safety accident probability of the test environment, and predicts the overall safety accident probability according to the safety accident probability density function under the actual environmental constraints.

Images