FADS system data processing method, data processing system and FADS system

Abstract

The invention discloses an FADS system data processing method, a data processing system and an FADS system. The FADS system data processing method comprises the following steps: S1, acquiring atmospheric pressure data measured by each measurement module of the FADS system; s2, performing fault analysis on the acquired data by adopting a fault detection method based on a principal component analysis method, and removing fault point data in the data; and S3, evaluating the credibility of the measurement data of each measurement module by adopting a trust calculation method based on an evidence theory, selecting the test data of the measurement module with high membership degree and high mutual support degree after being processed in the step S2, and carrying out atmospheric calculation to obtain atmospheric parameter data. By processing the atmospheric pressure measurement data, automatically removing fault point data in the atmospheric pressure measurement data and screening out data with high credibility for atmospheric calculation, atmospheric parameter information data obtained by calculation is more accurate, and the problem of FADS system failure caused by faults of an embedded pressure sensor in an FADS system is well solved.

Description

Technical Field

[0001] The present invention belongs to the field of data processing technology, and specifically relates to a FADS system data processing method capable of realizing autonomous fault tolerance, a data processing system and a FADS system Background Art

[0002] The atmospheric data system of traditional aircraft usually uses pitot tubes to measure total pressure and static pressure, uses head-on and side-angle wind vanes to measure head-on and side-angle, and completes the solution and correction of various atmospheric parameters in the atmospheric data computer. In this traditional atmospheric data system, pitot tubes and head-on and side-angle wind vanes all protrude from the aircraft's skin surface. However, as modern aircraft have higher and higher requirements for stealth performance, sensors protruding from the aircraft's skin surface will destroy the aircraft's radar stealth performance. Therefore, embedded atmospheric data systems built into the skin have obvious advantages.

[0003] The FADS system (Flush Air Data Sensing, embedded atmospheric data sensing system) usually includes an air intake device, an air bleed pipe, a pressure measurement unit, a data preprocessing unit, an embedded atmospheric data system algorithm, and a data cable. It measures the pressure distribution on the aircraft surface and obtains the angle of attack, sideslip angle, static pressure, dynamic pressure, and Mach number through solution. Due to the need for measurement and the actual situation on the aircraft, multiple pressure measuring points are usually designed for pressure measurement. These pressure measuring points are arranged on the upper and lower surfaces of the fuselage and on the wings. These pressure measuring devices need to be connected to the atmospheric data solution computer by cables for data transmission. Due to the limitations of the on-board environment, especially the pressure measuring holes on the wing part, it is difficult to lay and connect the cables. In order to improve the accuracy of atmospheric parameter solution, it is necessary to select as many pressure measuring holes as possible. The more positions selected, the more dispersed the position distribution, and the total weight of the cable, the space occupied on the aircraft, and the difficulty of laying will also increase exponentially.

[0004] In the layout of the FADS system, the pressure measuring holes of the embedded pressure sensors arranged on the upper surface of the aircraft wing are set upward, which is easy to accumulate dust or be attacked by rain, which will cause the pneumatic pipeline of the pressure measuring holes to be blocked, resulting in a large error in the pressure measurement. Although the pressure measuring holes of the embedded pressure sensors arranged on the lower surface of the aircraft wing are less affected by external environmental factors, considering the aircraft radar scattering cross-sectional area, the design of the pressure measuring hole structure distributed on the lower side of the fuselage needs to consider the influence of stealth performance, but the complexity of the structure will lead to an increase in its failure rate. Therefore, in the FADS system, the failure of any embedded pressure sensor will cause the embedded atmospheric data system to be unavailable, which will cause the atmospheric data system to fail and directly affect the flight safety of the aircraft. Summary of the invention

[0005] The object of the present invention is to provide a FADS system data processing method, a data processing system and a FADS system with autonomous fault tolerance capability.

[0006] The present invention is achieved through the following technical solutions:

[0007] The FADS system data processing method comprises the following steps:

[0008] S1, obtaining atmospheric pressure data measured by each measurement module of the FADS system;

[0009] S2. Use a fault detection method based on principal component analysis to perform fault analysis on the acquired data and remove the fault point data;

[0010] S3. Use a trust calculation method based on evidence theory to evaluate the credibility of the measurement data of each measurement module, select the test data of the measurement module with high membership and mutual support after processing in step S2, and perform atmospheric solution to obtain atmospheric parameter data.

[0011] As a further improvement to the above technical solution, the FADS system includes a plurality of measurement modules and an atmospheric data computer distributed on the fuselage, and the measurement modules are wirelessly connected to the atmospheric data computer.

[0012] As a further improvement to the above technical solution, the fault detection method adopted in step S2 is:

[0013] Obtain the sequence values ​​of the atmospheric pressure data measured by each measurement module in any time period and construct them into an atmospheric pressure data matrix X;

[0014] Standardize the obtained matrix X to obtain a matrix Z* with a mean of 0 and a variance of 1;

[0015] Construct the Lagrangian function, obtain the right eigenvector of the maximum characteristic of each column in the matrix Z*, and simplify and reduce the dimension of the matrix Z*;

[0016] Obtain the fault detection index Q and the control limit of the confidence level under the set value. When the fault detection index Q is greater than the control limit, it means that the system fails.

[0017] As a further improvement to the above technical solution, the method for determining the fault point data in step S2 is: when the fault detection index Q is greater than the control limit, the variance contribution at each moment in the time period is obtained to obtain the moment when the fault occurs, and the measurement data corresponding to the moment is the fault point data.

[0018] As a further improvement to the above technical solution, the trust calculation method based on evidence theory in step S3 evaluates the credibility of the measurement data of each measurement module, including the following steps:

[0019] S31, obtaining the membership relationship and mutual support between the various test modules. If the membership relationship and mutual support of a test module are lower than a set threshold, the test data of the test module is isolated and converted into evidence;

[0020] S32. Use the DS evidence theory synthesis rule to combine evidence, obtain the weight distribution coefficient of each measurement data, fuse the analysis results, and evaluate the credibility of the measurement data of each measurement module.

[0021] As a further improvement to the above technical solution, the credibility of the test data of each measurement module is evaluated in step S32, including the following steps:

[0022] S311, taking all the acquired measurement data as the identification framework and converting them into evidence for synthesis;

[0023] S312, obtaining the measurement value Z of the jth measurement module j The membership degree η ij , where η ij The larger the value is, the closer the results of the j-th measurement module are to those of the i-th measurement module, the greater the possibility that it belongs to the recognition target, and the higher the support for the target; otherwise, the lower the support;

[0024] S313, constructing a membership matrix U corresponding to the identification targets of n test modules in the system;

[0025] Get the i-th row vector P of the membership matrix U i And its mean square Euclidean distance D to other measurement modules i , according to D i The mutual support of the measurement module is determined by the size of . When the mutual support is greater than the set value, it means that the feasibility of the measurement module is high. When the mutual support is less than the set value, it will be isolated.

[0026] As a further improvement to the above technical solution, a dropped half Gaussian distribution function is used as the membership function in step S311.

[0027] On the other hand, the present invention also provides a FADS system data processing system, comprising:

[0028] A fault detection unit, wherein the fault detection unit adopts a fault detection method based on principal component analysis to perform fault analysis on the acquired data and remove the fault point data therein;

[0029] A credibility evaluation unit, wherein the credibility evaluation unit adopts a trust calculation method based on evidence theory to obtain the membership and mutual support of the test data of the measurement module, judge the size of the membership and mutual support of the test data of the measurement module, and evaluate the credibility of the measurement data of each measurement module;

[0030] A solution unit is used to perform atmospheric solution on the acquired atmospheric pressure measurement data to obtain atmospheric parameter data.

[0031] On the other hand, the present invention also provides a FADS system, comprising:

[0032] A plurality of measuring modules, each comprising a pressure sensor array, a data acquisition and processing module, and a wireless transceiver module respectively arranged at each pressure measuring hole;

[0033] The atmospheric data computer comprises a system data processing module and a wireless receiving module. The system data processing module adopts the data processing method to obtain atmospheric parameter data.

[0034] Compared with the prior art, the present invention has the following beneficial effects:

[0035] The FADS system data processing method and system processes the wirelessly received atmospheric pressure measurement data, automatically removes the fault point data therein, and further screens the data on this basis to select the data with high credibility for atmospheric solution, so that the atmospheric parameter information data obtained by solution is more accurate, and well solves the problem of FADS system failure caused by the failure of the embedded pressure sensor in the FADS system.

[0036] This data processing method performs autonomous fault-tolerant processing on multiple measurement data information, analyzes each measurement module, selects measurement modules with a credibility higher than 80%, and automatically eliminates faulty pressure measuring holes or those with excessive deviations, thereby reducing data redundancy. By intelligently screening high-precision measurement module data, the calculation accuracy of atmospheric parameter data is improved, and the stability and reliability of the FADS system are improved.

[0037] Based on the above data processing method and system, the FADS system has the ability of active fault tolerance, thereby reducing the complexity of the FADS system. Wireless data transmission can be better applied between each measurement module and the atmospheric data computer, which facilitates the functional expansion of the FADS system, and can realize the fault location and replacement of the measurement module, thereby extending the service life and work efficiency of the FADS. It is of great significance to promote the sustainable development of aircraft in the direction of long life, high reliability, strong scalability, plug and play, etc. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0039] Figure 1 This is a structural block diagram of the FADS system of the present invention.

[0040] Figure 2 The figure is a flow chart of the FADS system data processing method of the present invention. DETAILED DESCRIPTION

[0041] In order to make the purpose, technical scheme and advantages of the present application clearer, the specific embodiments of the present application are further described in detail below in conjunction with the accompanying drawings. It is understood that the specific embodiments described herein are only used to explain the present application, rather than to limit the present application. It should also be noted that, for the convenience of description, only the part related to the present application but not all the contents are shown in the accompanying drawings. Before discussing the exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flow charts. Although the flow chart describes each operation (or step) as a sequential process, many of the operations therein can be implemented in parallel, concurrently or simultaneously. In addition, the order of each operation can be rearranged. The process can be terminated when its operation is completed, but it can also have additional steps not included in the accompanying drawings. The process can correspond to a method, a function, a procedure, a subroutine, a subprogram, etc.

[0042] The FADS system in this embodiment includes 5 to 9 measurement modules for atmospheric data measurement and an atmospheric data computer.

[0043] Each measurement module includes pressure measuring holes, pressure sensor arrays, data acquisition and processing modules, wireless transceiver modules, etc. distributed at various positions on the fuselage, which are used to realize multi-point acquisition of atmospheric pressure data and pre-process the acquired signals and data, and can wirelessly transmit the pre-processed data to the atmospheric data computer, solving the problems of complex laying lines, occupying space on the aircraft, and increasing the weight of the aircraft when laying cross-linked cables of embedded atmospheric data sensors on the aircraft. The location of the measurement module on the fuselage can be selected according to CFD simulation and wind tunnel tests.

[0044] The atmospheric data computer generally includes a system data processing module, a wireless receiving module, a power supply module and an external communication interface module, which are used to receive data from each measuring module and perform calculation processing on the data. The atmospheric data computer receives the atmospheric measurement data of each measuring module through the wireless receiving module, and the system data processing module uses a central processing module to achieve calculation redundancy, and processes the received measurement data based on the data processing method in this embodiment, performs fault detection, and screens pressure measuring holes (measurement modules), and finally calculates and obtains atmospheric parameter data.

[0045] The FADS system data processing method comprises the following steps:

[0046] S1, the atmospheric pressure data measured by each measuring module is pre-processed and sent to the atmospheric data computer through the wireless transceiver module;

[0047] In this step, when sending data to the atmospheric data computer, the atmospheric data computer determines whether all data information has been received. When it is determined that all data information has not been received, it sends query information to the corresponding measurement module and re-receives the corresponding measurement data. When it is determined that all data information has been received, time calibration is performed, and then the subsequent data processing operation is entered.

[0048] S2. Establish a principal component model based on the received atmospheric pressure data, perform fault analysis on the received data, and eliminate the fault points according to the analysis results;

[0049] In step S2, a fault detection method based on principal component analysis is adopted, and the specific steps are as follows:

[0050] Statistical analysis is performed on the atmospheric pressure data obtained within a certain period of time, and the data samples are decomposed and transformed into two variable subspaces, one of which is the principal element subspace where the principal element is located, and the other is the corresponding residual subspace. These two subspaces respectively characterize the characteristics of the data. By processing the observed data and calculating the indicators of the offset model, the state of the fault can be identified.

[0051] The commonly used detection statistical methods include establishing the Hotelling T2 statistic in the principal component subspace for statistical testing and establishing the Q statistic in the residual subspace for statistical testing.

[0052] The data processing method in this embodiment is to first establish the principal component model of the observation equation, and then use the statistics at each moment in the calculation process to realize the detection of the fault; if it is found that the detection value exceeds the normal range, each variable in the process is statistically analyzed to determine the time and degree of the fault, thereby realizing the function of fault detection and diagnosis.

[0053] Specifically, assuming that within a period of time m, the sequence values ​​of the atmospheric pressure data measured by the measurement module within the period of time are obtained, and the sequence values ​​are used to form a matrix X of atmospheric pressure data. The matrix X consists of n state values, which is expressed as:

[0054] Z * =(z ij * )=[z * (1),z * (2),…z * (m)];

[0055] In the matrix X, each row represents a sample and each column represents a measured variable.

[0056] Normalize the matrix X:

[0057]

[0058] In the formula, E(z * i ), var(z * i ) represent the sample mean and variance of the measured value of the i-th measurement module respectively. Through standardization, a matrix with a mean of 0 and a variance of 1 is obtained. The matrix Z after standardization is:

[0059] Z=(z ij )=[z(1),z(2),…z(m)].

[0060] Then, by constructing the Lagrangian function, we can find the right eigenvector of the maximum feature of each column in the matrix Z. According to the degree of data change, we can get that the first principal component describes the largest amount of data information, and the importance of the data information of other principal components decreases in turn.

[0061] Through the above method, the standardized matrix Z is simplified and dimensionally reduced, and its mathematical model can be expressed as:

[0062]

[0063] Among them, t i , i = 1, 2, ..., d is the principal element vector; p i , i = 1, 2, ..., d is the projection vector of the principal element; d is the number of principal elements, T = [t1, t2, ..., t d ] is the score matrix; P = [p1, p2, ..., p d ] is the load matrix;

[0064] is the main component; are other error components, and their statistical mean is 0.

[0065] The fault detection index Q is defined to measure the projection change of the sample vector in the residual space, which is expressed as:

[0066]

[0067] in, is the control limit with confidence level α;

[0068] P=[p1,p2,…,p d ] is the load matrix obtained after principal component analysis, z(k) is the column vector of the standardized matrix Z, and I is the unit matrix.

[0069] The control limits are calculated as follows:

[0070]

[0071] in, c α is the threshold of the standard normal distribution at the confidence level α.

[0072] if This means that the FADS system is operating normally at this moment; if This indicates that the FADS system may have failed at that moment.

[0073] Since the Q statistic is the result of the combined effect of multiple variables in the system, the statistical graph generated by calculating the statistic at each moment can be used to identify and analyze the fault conditions of the entire system.

[0074] If the system does not fail, the measured distribution will be consistent with the principal component model; otherwise, if the system fails, the true model will deviate from the principal component model; by obtaining the variance contribution at each moment, the moment when the failure occurs can be obtained.

[0075] The measured data at this moment is removed from the calculated data as the fault point.

[0076] S3. Recalculate the data after removing the fault point, use the trust calculation method based on evidence theory to calculate the membership relationship and mutual support between the measurement data of each measurement module, evaluate the credibility of the measurement data of each measurement module, select the test data obtained by the measurement modules with high membership and mutual support for atmospheric solution, and obtain atmospheric parameter data.

[0077] In step S3, the trust calculation method based on evidence theory is specifically as follows:

[0078] The idea of ​​basic trust distribution in evidence theory is adopted to calculate the membership relationship and mutual support between each pressure measuring hole, and evaluate the credibility of the measurement data of each measurement module. If the membership relationship and mutual support of a pressure measuring hole are lower than the set threshold, it means that the measurement value of the pressure measuring hole is too different from that of other pressure measuring holes, indicating that the system may have failed. The measurement data of the pressure measuring hole is isolated and converted into evidence.

[0079] Then, the DS evidence theory synthesis rule is used to combine the evidence to obtain the weight distribution coefficient β of each measurement data. i , and integrate the analysis results; the specific process is as follows:

[0080] All atmospheric pressure measurement data (Z0, Z1, ..., Z n-1 ) is used as the identification framework and converted into evidence for synthesis. Taking into account that all measurement values ​​include measurement errors that obey Gaussian distribution, the membership function is selected as the half-Gaussian distribution function. The selected membership function is a trust-assigned value, which is used to reflect the degree of trust in the measurement module.

[0081] Define the jth measurement module for the measurement value Z j The membership degree is η ij ,in:

[0082]

[0083] η ij The larger the value is, the closer the results of the jth measurement module are to those of the ith measurement module, the greater the possibility that it belongs to the recognition target, and the higher the support for the target; otherwise, the lower the support. The recognition target here refers to the measurement data output by the corresponding measurement module. By defining the membership degree of each measurement module, the correctness of the measurement data is evaluated and judged from the perspective of evidence theory.

[0084] Therefore, the membership matrix U of the n measurement modules in the system corresponding to the identification target is expressed as:

[0085]

[0086] The i-th row vector of the membership matrix U is P i , represents the membership vector of the i-th measurement module to the recognition target, and its mean squared Euclidean distance to all measurement modules is D i , its size reflects the difference between the i-th measurement module and other measurement modules. The smaller the value, the smaller the difference in the measurement data of the measurement module, the higher the mutual support, and the higher the credibility of the measurement module; otherwise, it means that the measurement value of the measurement module has a large deviation and can be isolated.

[0087] That is, the mutual support can be expressed as:

[0088]

[0089] In the formula, Represents the confidence distance between the i-th and j-th measurement modules.

[0090] At this point, the credibility of the measurement data of each measurement module can be evaluated.

[0091] On the other hand, some embodiments of the present invention also relate to a FADS system data processing system for implementing the above data processing method. Accordingly, the data processing system includes:

[0092] A fault detection unit, wherein the fault detection unit adopts a fault detection method based on principal component analysis to perform fault analysis on the acquired data and remove the fault point data therein;

[0093] A credibility evaluation unit, wherein the credibility evaluation unit adopts a trust calculation method based on evidence theory to obtain the membership and mutual support of the test data of the measurement module, judge the size of the membership and mutual support of the test data of the measurement module, and evaluate the credibility of the measurement data of each measurement module;

[0094] A solution unit is used to perform atmospheric solution on the acquired atmospheric pressure measurement data to obtain atmospheric parameter data.

[0095] The specific implementation methods used by the above-mentioned units to process data have been described in detail and will not be repeated here.

[0096] On the other hand, some embodiments of the present invention also relate to a FADS system, including:

[0097] Multiple measurement modules, each of which includes a pressure sensor array, a data acquisition and processing module, and a wireless transceiver module respectively arranged at each pressure measuring hole;

[0098] The atmospheric data computer comprises a system data processing module and a wireless receiving module. The system data processing module adopts the data processing method in the above embodiment to obtain atmospheric parameter data.

[0099] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any form. Any simple modification or equivalent change made to the above embodiment based on the technical essence of the present invention shall fall within the protection scope of the present invention.

Claims

1. FADS system data processing method, characterized in that: The following steps are involved: S1, obtaining atmospheric pressure data measured by each measurement module of the FADS system; S2. Use a fault detection method based on principal component analysis to perform fault analysis on the acquired data and remove the fault point data; S3. Use a trust calculation method based on evidence theory to evaluate the credibility of the measurement data of each measurement module, select the test data of the measurement module with high membership and mutual support after processing in step S2, and perform atmospheric solution to obtain atmospheric parameter data.

2. The FADS system data processing method according to claim 1, characterized in that: The FADS system includes a plurality of measurement modules and an air data computer distributed on the fuselage, and the measurement modules are wirelessly connected to the air data computer.

3. The FADS system data processing method according to claim 1, characterized in that: The fault detection method adopted in step S2 is: Obtain the sequence values ​​of the atmospheric pressure data measured by each measurement module in any time period and construct them into an atmospheric pressure data matrix X; Standardize the obtained matrix X to obtain a matrix Z* with a mean of 0 and a variance of 1; Construct the Lagrangian function, obtain the right eigenvector of the maximum characteristic of each column in the matrix Z*, and simplify and reduce the dimension of the matrix Z*; Obtain the fault detection index Q and the control limit of the confidence level under the set value. When the fault detection index Q is greater than the control limit, it means that the system fails.

4. The FADS system data processing method according to claim 3, characterized in that: The method for determining the fault point data in step S2 is: when the fault detection index Q is greater than the control limit, the variance contribution at each moment in the time period is obtained to obtain the moment when the fault occurs, and the measurement data corresponding to the moment is the fault point data.

5. The FADS system data processing method according to claim 1, characterized in that: In step S3, the credibility of the measurement data of each measurement module is evaluated by a trust calculation method based on evidence theory, including the following steps: S31, obtaining the membership relationship and mutual support between the various test modules. If the membership relationship and mutual support of a test module are lower than a set threshold, the test data of the test module is isolated and converted into evidence; S32. Use the DS evidence theory synthesis rule to combine evidence, obtain the weight distribution coefficient of each measurement data, fuse the analysis results, and evaluate the credibility of the measurement data of each measurement module.

6. The FADS system data processing method according to claim 5, characterized in that: Step S32 evaluates the credibility of the test data of each measurement module, including the following steps: S311, taking all the acquired measurement data as the identification framework and converting them into evidence for synthesis; S312, obtaining the measurement value Z of the jth measurement module j The membership degree η ij , where the larger the value is, the closer the results of the j-th measurement module are to those of the i-th measurement module, the greater the possibility that it belongs to the recognition target, and the higher the support for the target; otherwise, the lower the support; S313, constructing a membership matrix U corresponding to the identification targets of n test modules in the system; Get the i-th row vector P of the membership matrix U i And its mean square Euclidean distance D to other measurement modules i , according to D i The mutual support of the measurement module is determined by the size of . When the mutual support is greater than the set value, it means that the feasibility of the measurement module is high. When the mutual support is less than the set value, it will be isolated.

7. The FADS system data processing method according to claim 6, characterized in that: In step S311, a half-Gaussian distribution function is used as the membership function. 8.FADS system data processing system, characterized in that: include: A fault detection unit, wherein the fault detection unit adopts a fault detection method based on principal component analysis to perform fault analysis on the acquired data and remove the fault point data therein; A credibility evaluation unit, wherein the credibility evaluation unit adopts a trust calculation method based on evidence theory to obtain the membership and mutual support of the test data of the measurement module, judge the size of the membership and mutual support of the test data of the measurement module, and evaluate the credibility of the measurement data of each measurement module; A solution unit is used to perform atmospheric solution on the acquired atmospheric pressure measurement data to obtain atmospheric parameter data.

9. FADS system, characterized in that, include: A plurality of measuring modules, each comprising a pressure sensor array, a data acquisition and processing module, and a wireless transceiver module respectively arranged at each pressure measuring hole; An atmospheric data computer comprises a system data processing module and a wireless receiving module. The system data processing module obtains atmospheric parameter data by using the data processing method described in any one of claims 1 to 7.

Images