Search

CN-121706265-B - Aircraft navigability monitoring method based on operation data

CN121706265BCN 121706265 BCN121706265 BCN 121706265BCN-121706265-B

Abstract

An aircraft navigability monitoring method based on operation data. The method comprises the steps of obtaining system safety evaluation data in an airworthiness evidence obtaining stage of an aircraft, determining importance of fault contribution, obtaining operation observation failure rate, obtaining a posterior failure rate estimation mean value of each airworthiness monitoring period, calculating real-time failure probability of a roof event and the like. The invention has the effect that the safety of equipment is monitored by reasonably utilizing the effective operation data collected by the civil aircraft. And quantitatively analyzing through the fault tree model, and calculating the importance degree of fault contribution of each bottom event to the top event. The multi-period Bayesian updating method is adopted, so that initial priori information which is deviated from conservation is determined, the influence caused by extreme values under the condition of small samples is prevented, the stability of initial results of the navigable monitoring based on operation data is ensured, and the accuracy of equipment reliability calculation results is improved.

Inventors

  • XIAO NVE
  • WANG PENG
  • LI BOYANG
  • ZHU LIN

Assignees

  • 中国民航大学

Dates

Publication Date
20260512
Application Date
20260210

Claims (5)

  1. 1. The aircraft navigability monitoring method based on the operation data is characterized by comprising the following steps in sequence: 1) Acquiring system safety evaluation data of an aircraft in a navigable evidence obtaining stage, wherein the system safety evaluation data comprises a fault tree model, reliability data of each device/component, information of a system composition logic relationship and basic description of the system, and initial design reliability data of each bottom event of the device in the fault tree model; 2) Determining the importance of fault contribution of each bottom event in the fault tree model based on the system safety evaluation data, and determining key equipment as a key navigable monitoring object according to the importance of fault contribution; 3) Setting a navigable monitoring period, collecting operation data of the key equipment in the navigable monitoring period, and preprocessing and converting the data to obtain operation observation failure rate in the period; The method comprises the following steps: 3.1 Setting a navigable monitoring period, then collecting operation data of the key equipment in the navigable monitoring period, mainly comprising the total flight time of a fleet, the installed number of equipment and detailed unplanned replacement records, sequencing the unplanned replacement interval time in the collected unplanned replacement records, and then constructing a sample quantile statistic by using a formula (2) and a formula (3) Setting a critical value, judging abnormal data generated by recording errors or extreme conditions and eliminating the abnormal data if the calculated sample quantile statistic exceeds the critical value, and obtaining processed operation data; (2); (3); In the formula, The upper quartile value is represented by the value, The lower quartile value is represented by, 、 Respectively representing right and left abnormal test value statistics; 3.2 Using the processed operation data to evaluate the frequent degree of the disassembly and the replacement of the key equipment on the operation site, and calculating an average unplanned disassembly and replacement interval MTBUR in the navigable monitoring period through a formula (4); (4); In the formula, Indicating the total flight time of the fleet during the navigability monitoring period, Indicating the installed number of each machine of the key equipment, Representing all unplanned times of replacement of the key device in the navigable monitoring period; 3.3 Using empirical data correction to find out no fault rate Setting, no fault finding rate Typically empirical data based on service history of the system and equipment, expressed as a percentage, and then using equation (5) in combination with the failure-free discovery rate Converting the average unplanned replacement interval MTBUR into an average failure interval time MTBF which can reflect the inherent reliability of key equipment, and then taking the reciprocal of the average failure interval time MTBUR to obtain the operation observation failure rate in the navigation monitoring period ; (5); 4) Carrying out multi-period Bayesian updating on the equipment failure rate by combining the system security evaluation data obtained in the step 1) to obtain a posterior failure rate estimation mean value of each navigable monitoring period; 5) Substituting the posterior failure rate estimation mean value of each key device obtained in the step 4) into the fault tree model obtained in the step 1) as a new basic event probability input parameter, calculating the real-time failure probability of the top event, comparing the real-time failure probability with a set system airworthiness security target value to judge the system airworthiness state, and combining the failure contribution importance degree obtained in the step 2) and the operation observation failure rate obtained in the step 3) if the comparison result shows that the system security no longer meets the airworthiness requirement And performing early warning display to dynamically monitor the top event and the bottom event.
  2. 2. The method for aircraft airworthiness monitoring based on operational data of claim 1, wherein in step 1), the method for acquiring system security assessment data of the aircraft airworthiness evidence stage includes a fault tree model, reliability data of each device/component, information including a logical relationship of system composition and a basic description of the system, and initial design reliability data representing each bottom event of the device in the fault tree model is: Firstly, acquiring system safety evaluation data of a certain model airplane as a target in a seaworthiness evidence obtaining stage from an airplane manufacturer or a design department as an initial reference of seaworthiness monitoring, wherein the system safety evaluation data comprises a fault tree model, reliability data of each equipment/component, a system composition logic relationship and information of a system basic description, and simultaneously, acquiring initial design reliability data of each bottom event of the representing equipment in the fault tree model, wherein the initial design reliability data mainly comprises design failure rate Design parameters, among others.
  3. 3. The method for monitoring the airworthiness of an aircraft based on operation data according to claim 1, wherein in the step 2), the method for determining the importance of the fault contribution of each bottom event in the fault tree model based on the system security evaluation data and then determining the key equipment as the key airworthiness monitoring object according to the importance of the fault contribution is as follows: 2.1 Based on the fault tree model, taking the occurrence of a top event in the system as a precondition; 2.2 Quantitative analysis is carried out on the fault tree model, and the importance degree of fault contribution of each bottom event is calculated by using the formula (1) Fault contribution importance refers to the fact that a top event in the system has occurred Under the condition of (1) The individual bottom events are also in a disabled state Conditional probability of (2) Is marked as To quantify the number of times when the system fails, by Probability duty cycle of failure caused by the bottom events; (1); In the formula, Representing a top event in the system, Indicating that a top event occurred; Represent the first A single event of the bottom of the container, Represent the first Occurrence of a bottom event; 2.3 All bottom events are attributed to importance according to fault Sequencing from big to small in size; 2.4 Selecting a plurality of bottom events which are ranked forward, or selecting a bottom event set with accumulated fault contribution importance reaching a set threshold, and determining the bottom events as key equipment serving as key navigable monitoring objects so as to preferentially allocate monitoring resources.
  4. 4. The method for aircraft airworthiness monitoring based on operation data according to claim 1, wherein in the step 4), the method for performing multi-period Bayesian update on the equipment failure rate by combining the system security evaluation data obtained in the step 1) to obtain the posterior failure rate estimation mean value of each airworthiness monitoring period is as follows: 4.1 Assuming that the life of the determined key device obeys an exponential distribution, the failure rate of the device of the exponential distribution is based on the Bayesian conjugate theory Is gamma distribution Setting initial shape parameters of gamma distribution in operation initial stage And scale parameter ; Obtaining the design failure rate of the airworthiness evidence obtaining stage obtained in the step 1) ; 4.2 For the first Collecting failure flight time data of key equipment according to the airworthiness monitoring period, and setting a group of collected failure flight time data as Wherein , Is the total number of failures observed by the critical device, Is the failure flight time data at the end of the airworthiness monitoring period, assuming the equipment failure interval time The equipment failure interval time sample set in the navigable monitoring period is Calculating the equipment failure interval time by using a formula (6) Probability density functions of (2); (6); In the formula, Time between failure of a device ; Finally, collecting the obtained equipment failure interval time sample set Carrying out the method into a formula (6) to construct a likelihood function of the navigable monitoring period data; 4.3 According to Bayesian formula, multiplying the prior distribution parameter with the likelihood function and normalizing, and calculating the failure rate of the equipment by using formula (7) According to the property of gamma distribution, calculating the failure rate of the equipment by using a formula (8) Posterior failure rate estimation mean value updated by Bayes ; Failure rate of equipment The posterior distribution density function of (c) is calculated as follows: (7); Failure rate for equipment It can be seen that the posterior distribution density function is still in the form of a gamma distribution, i.e. the failure rate of the device Posterior distribution parameters of (2) ; Failure rate of equipment The calculation formula of the posterior failure rate estimation mean value updated by Bayes is as follows: (8); In the formula, , Respectively representing the shape parameter and the scale parameter of gamma distribution, wherein the first navigable monitoring period is an initial value, and the subsequent navigable monitoring period is an updated value of the previous navigable monitoring period; representing the total number of equipment failures observed during the navigable monitoring period; Representing the sum of all equipment failure interval times in the navigable monitoring period; 4.4 Using a rolling update strategy to accommodate continuous airworthiness monitoring, gamma distribution parameters determined by design parameters For the prior distribution parameter, the operation data of the 1 st navigable monitoring period is combined to output the shape parameter of the 1 st navigable monitoring period And scale parameter Average value of posterior failure rate estimation Then, collecting the operation data of the second airworthiness monitoring period to obtain the posterior distribution parameters of the first airworthiness monitoring period As a priori distribution parameter of the second navigability monitoring period, calculating a shape parameter of the second navigability monitoring period according to formulas (7), (8) And scale parameter Average value of posterior failure rate estimation Repeating the above operation using the first Posterior distribution parameters of individual navigability monitoring period as the first The prior distribution parameters of the individual airworthiness monitoring period are used for realizing the alternate transmission and knowledge accumulation of the data, and finally obtaining the first Posterior failure rate estimation mean value of individual navigable monitoring period 。
  5. 5. The aircraft airworthiness monitoring method based on operation data according to claim 1, wherein in step 5), the average value of the posterior failure rate estimates of each key device obtained in step 4) is used as a new basic event probability input parameter to replace the failure tree model obtained in step 1), the real-time failure probability of the top event is calculated and compared with a set system airworthiness safety target value to judge the system airworthiness state, and if the comparison result shows that the system safety no longer meets the airworthiness requirement, the failure contribution importance obtained in step 2) and the operation observation failure rate obtained in step 3) are combined The method for carrying out early warning display to dynamically monitor the top event and the bottom event comprises the following steps: 5.1 The posterior failure rate estimation mean value of each key device obtained in the step 4) in the current navigability monitoring period As new basic event probability input parameters, replacing the new basic event probability input parameters into the fault tree model obtained in the step 1), and recalculating the real-time failure probability of the top event in the current operation state from bottom to top by using Boolean logic operation of the fault tree model ; 5.2 Setting a target value of the system aviation security Then the real-time failure probability is calculated Comparing the target value with the target value of the system navigability safety if Judging that the current system has good navigability, maintaining the existing monitoring strategy, and automatically entering an operation data collection and monitoring cycle of the next navigability monitoring period; 5.3 If any) Determining that the system safety no longer meets the airworthiness requirement, immediately triggering a system-level early warning mechanism, carrying out attribution analysis by combining the importance of fault contribution obtained in the step 2), highlighting the top event state of a fault tree model by a system interface, marking a main cause source list causing risk rise, screening and displaying actual operation observation failure rates based on the key airworthiness monitoring object determined in the step 2) and the latest real-time failure probability calculated in combination with the 5.1) Significantly higher than the design expectations and contributing the highest to the probability of failure of the current system, resulting in a reduced level of overall system safety.

Description

Aircraft navigability monitoring method based on operation data Technical Field The invention belongs to the technical field of civil aviation, and particularly relates to an aircraft navigability monitoring method based on operation data. Background During service of a particular model of civil aircraft, the actual airworthiness risk level may be higher than the safety standards set at the beginning of the design due to the complexity of the operating environment, the limitations of the standard formulation, and the potentially unprovisioned changes in compliance methods. For this reason, commercial aircraft put into use should be subjected to continuous safety evaluations to maintain the aeronautical performance of the aircraft, thereby improving the overall safety level of the aircraft. In terms of operation data acquisition and management, the international aviation world has formed a relatively mature specification. Operators typically collect operation data according to the reliability data collection and exchange of ATA Spec 2000 chapter 11 or the S5000F standard "International specification for in-SERVICE DATA feedback" issued by the european aviation defense industry association, and feed back the original data to a host manufacturer such as the AnalytX platform of boeing or the Skywise platform of air, etc. digital big data systems. The whole life cycle safety data collection, integration and analysis system of the matched aircraft is gradually perfected in the field of China. Although the data acquisition standards are increasingly perfect, in the early stage of aircraft in use, actual navigable monitoring still faces core technical difficulties: 1. The operational data reflects the true inherent reliability of the device, which is far higher than the design data. However, the cumulative number of operating hours and failure samples are generally very limited at the beginning of the aircraft's use, especially for high reliability on-board systems. 2. Classical statistical methods based on large samples are no longer applicable when the data volume is insufficient, and it is therefore difficult to assess the current real risk. 3. The existing management system focuses on data storage and report forms, lacks a dynamic evaluation method capable of realizing safe real-time navigability monitoring of short-term operation and timely finding potential safety hazards based on system safety data in a evidence obtaining stage and combining limited operation data. In summary, developing a navigable monitoring method capable of integrating design priori knowledge and operation posterior data has important engineering application value for improving flight safety, meeting reliability requirements of aviation operation and perfecting a national civil aircraft navigable management system. In addition, for the security data acquisition process, the operational data is used as a reflection device to reflect the true inherent reliability of the device, and the data result is more reliable than the security data acquired in the design and verification stage. Therefore, it is necessary to acquire and monitor the safety data of domestic airborne equipment in the operation stage, so that the reliability condition of the operation fleet equipment can be known, and the reliability condition can be fed back to a provider for analysis, so that potential problems are found, help is provided for improvement of equipment design, and the overall aviation industry safety level of China is further improved. Yanyun discloses an air operation data acquisition system based on big data technology, which can acquire original data reflecting the reliability condition of an operation fleet, such as ACARS data, QAR data, text data and the like, upload the big data analysis platform for analysis every month, promote automation of operation data acquisition, solve the problem of data island after traditional data acquisition (an air operation data acquisition system [ P ] based on big data technology, shanghai city: CN201810489479.8,2019-11-29.), provide a comprehensive safety risk early warning model based on deep learning by Guo et al, describe the risk development condition caused by system faults by adopting a failure mode, influence analysis, cause chain analysis, long-term memory and other methods in combination with the characteristics of QAR data and aircraft system faults, predict the trend of flight parameters of each system, thereby improving the prediction capability （A Data-Driven integrated safety risk warning model based on deep learning for Civil Aircraft[J].IEEE TRANSACTIONS ON AEROSPACE AND ELECTRONIC SYSTEMS,2023,59(2):1707-1719）;Sun of aviation system risk and risk severity and the like based on a large number of specific aircraft state monitoring system (ACMS) reports, monitor the health of an Air Conditioning System (ACS) by adopting a non-parameter modeling technology, and effectively monitor the state degradation