Outlier recognition method based on multi-source data fusion and mutual consistency

By constructing the functional relationship between multi-source data in aerospace measurement and control and shooting range measurement, the consistency score and ratio of the measurement element are judged, and the problem of misjudgment of outliers identification in multi-source data fusion processing in the prior art is solved, and the problem of misjudgment of outliers is realized. Effective identification of continuous or discontinuous outliers and reduce the probability of misjudgment.

CN113743480BActive Publication Date: 2024-11-22NO 27 RES INST CHINA ELECTRONICS TECH GRP +1
View PDF 6 Cites 0 Cited by

Patent Information

Application Number
CN202110954195.3
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2021-08-19
Publication Date
2024-11-22
Estimated Expiration
2041-08-19

AI Technical Summary

Technical Problem

The existing outlier value recognition methods in the fields of aerospace measurement and control and shooting range measurement have a high probability of misjudgment during multi-source data fusion processing, and it is impossible to correctly identify continuous multi-point outliers.

Method used

A multi-source data fusion outlier recognition method based on mutual consistency is used to construct a functional relationship between multiple data sources, and the consistency score and ratio of the measurement data are identified that are difficult to identify in a single source data.

Benefits of technology

It effectively reduces the probability of misjudgment of outliers identification, can identify continuous or discontinuous outliers, improves the use of information between multi-source data, and reduces the situation of misjudgment and misjudgment.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN113743480B_ABST
    Figure CN113743480B_ABST
Patent Text Reader

Abstract

The present invention proposes a multi-source data fusion outlier identification method based on mutual consistency, which is used to solve the technical problems that the existing outlier identification method has a high probability of misjudgment and cannot correctly identify continuous multi-point anomalies. The steps are: first, divide into m different data sources according to different observation devices; second, construct a functional relationship between the m data sources; finally, judge whether the measurement element in the data source is abnormal according to the functional relationship. The present invention judges data consistency by determining whether the functional relationship between the measurement element and the measurement element of other data sources is within the normal range. It is applicable to three or more independent data sources, does not rely on the continuity of a single data, and can identify most of the outliers; it has a good supplementary function to the outlier judgment method of an isolated single data source, is effective for continuous or non-continuous outliers, and can reduce the probability of misjudgment of outlier identification.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the field of aerospace measurement and control and range measurement data processing, and in particular to a multi-source data fusion outlier recognition method based on mutual consistency. Background Art

[0002] Outlier identification is an important part of data processing in aerospace measurement and control and range measurement systems. Before further processing, the observation data generated by the measurement equipment at each station needs to be identified as outliers. Outliers refer to observations that seriously deviate from the actual motion trajectory of the macro target. Generally, outlier identification can be performed based on the law of continuous change of observation data over time.

[0003] At present, there are classic outlier identification and correction algorithms in the field of aerospace measurement and control and range measurement. Generally, they are judged based on the continuity of local data points, such as algorithms based on second-order polynomial fitting. However, there are situations where local continuity and overall abnormality occur, leading to misjudgment. Algorithms based on overall judgment have advantages in judging outliers in terms of trend. However, all of the above methods are based on the perspective of a single measurement element and the characteristics of the measurement element itself. When multi-source data is fused and processed, the information between multi-source data is not fully utilized. Summary of the invention

[0004] In view of the shortcomings existing in the above-mentioned background technology, the present invention proposes a multi-source data fusion outlier identification method based on mutual consistency. When performing measurement data preprocessing before real-time or post-trajectory solution, the method can identify measurement data anomalies that are difficult to identify with single-source data, thereby solving the technical problems that the existing outlier identification method has a high probability of misjudgment and cannot correctly identify continuous multi-point anomalies.

[0005] The technical solution of the present invention is achieved in this way:

[0006] A multi-source data fusion outlier identification method based on mutual consistency, the steps are as follows:

[0007] Step 1: Divide into m data sources according to different observation devices, and the m data sources are all different, m ≥ 3;

[0008] Step 2: Construct a functional relationship between any two of the m data sources;

[0009] Step 3: Determine whether the measurement element in the data source is abnormal based on the functional relationship.

[0010] The functional relationship between any two of the m data sources is:

[0011]

[0012] Among them, C ik is the i-th data source Di The kth measurement element in , is the jth data source, i,j=1,2,…,m,i≠j,k=1,2,…,n i , n i is the i-th data source D i The number of test elements, n j is the jth data source D j The number of measured elements.

[0013] The method for judging whether the measured element in the data source is abnormal according to the functional relationship is:

[0014] S3.1. Set measurement element C ik The consistency score S ik =0, total number of consistency judgments P ik =0;

[0015] S3.2. For data source D i , list the data source D i The functional relationship between all measurement elements and measurement elements from other data sources is as follows:

[0016] S3.3. Judgment Is it true, where δ ik is a given threshold;

[0017] S3.4, P ik =P ik +1; if the inequality holds, then measure element C ik and The consistency score of S ik =S ik +1, S j1 =S j1 +1,S j2 =S j2 +1,...,S jn =S jn +1; if the inequality does not hold, the consistency score remains unchanged;

[0018] S3.5, traverse m data sources, traverse different measurement elements in the data source, and judge whether it is established;

[0019] S3.6. The final consistency ratio S ik / P ik Make a judgment, if S ik / P ik If it is not less than 50%, the measurement is considered normal; otherwise, the measurement is considered abnormal.

[0020] When m=3, the observation devices are a single pulse device, a multi-speed measuring device and a GNSS device, and there is one single pulse device, three multi-speed measuring devices and one GNSS device.

[0021] At time t, the measurement element of the single pulse device is R t , A t 、E t , the measuring element of the multi-speed measuring device I is The measuring elements of the multi-speed measuring device II are The measuring elements of the multi-speed measuring device III are The measurement element of the GNSS device is x t ,y t 、z t ,

[0022] Establish the functional relationship between the measurement elements of the monopulse device and the measurement elements of the GNSS device:

[0023]

[0024] Among them, px, py, pz are the station addresses of the single pulse equipment, and m′ represents the time number;

[0025] Establish the functional relationship between the measurement elements of multiple speed measuring devices and the measurement elements of GNSS devices:

[0026]

[0027] in, represents the measurement element of the i′th multi-speed measuring device, px i′ ,py i′ ,pz i′ is the site address of the i′th multi-speed measuring device, px0 represents the x-direction coordinate of the transmitting station site, ys0 represents the y-direction coordinate of the transmitting station site, zs0 represents the z-direction coordinate of the transmitting station site, R 0t represents the distance between the target and the transmitting station at time t, R i′t represents the distance between the target and the i′th multi-speed measuring device at time t;

[0028] When the master station signal transmitting station and receiving station of multiple speed measuring devices are the same station, take xs0 = xs1, ys0 = ys1, zs0 = zs1, R 0t =R 1t ;

[0029] Establish the functional relationship between the measurement elements of the single pulse device and the measurement elements of the multi-speed measurement device:

[0030]

[0031] Among them, x ct ,yct ,z ct is the target position in the station coordinate system of multiple speed measuring devices, which can be converted to the x t ,y t ,z t , Same coordinate system, denoted as x dt ,y dt ,z dt , use the position difference to get the velocity:

[0032]

[0033] in, It represents the velocity in the x direction obtained by using the single pulse information difference at time t, It represents the velocity in the y direction obtained by using the single pulse information difference at time t, represents the velocity in the z direction obtained by using the single pulse information difference at time t, x dt+1 represents the x coordinate of the target at time t+1, x dt-1 represents the target y coordinate at time t-1, y dt+1 represents the target y coordinate at time t+1, y dt-1 represents the target y coordinate at time t-1, z dt+1 represents the z coordinate of the target at time t+1, z dt-1 represents the target z coordinate at time t-1;

[0034] Then we have:

[0035]

[0036] To sum up, the formula is as follows:

[0037]

[0038] There are 9 equations in the above formula. Considering the measurement error of the observation equipment, the following inequalities exist:

[0039]

[0040] For the measurement element R t , the number of inequalities involved is 4, so the measurement variable R t The total number of consistent judgments P it =4; if the number of inequalities that hold is less than 2, then the measurement element R t abnormal;

[0041] For the measurement element A t , the number of inequalities involved is 4, so the measurement element A t The total number of consistent judgments P it=4; if the number of inequalities that hold is less than 2, then the measurement element A is determined t abnormal;

[0042] For the measurement element E t , the number of inequalities involved is 4, so the measurement element E t The total number of consistent judgments P it =4; if the number of inequalities that hold is less than 2, then the measurement element E is determined t abnormal;

[0043] For the measurement element The number of inequalities involved is 2, so the measurement element The total number of consistent judgments P it =2; if the number of inequalities that hold is less than 1, then the measurement element is determined abnormal;

[0044] For the measurement element x t , the number of inequalities involved is 6, so the measured variable x t The total number of consistent judgments P it =6; if the number of established inequalities is less than 3, then determine the measured element x t abnormal;

[0045] For the measurement element z t , the number of inequalities involved is 6, so the measurement element z t The total number of consistent judgments P it =6; if the number of established inequalities is less than 3, then the measurement element z is determined t abnormal;

[0046] For the measurement element y t , the number of inequalities involved is 5, so the measured variable y t The total number of consistent judgments P it =5; if the number of established inequalities is less than 3, then the measured element y is determined t abnormal;

[0047] For the measurement element The number of inequalities involved is 3, so the measurement element The total number of consistent judgments P it =3; if the number of inequalities that hold is less than 2, then the measurement element is determined abnormal;

[0048] For the measurement element The number of inequalities involved is 3, so the measurement element The total number of consistent judgments P it =3; if the number of inequalities that hold is less than 2, then the measurement element is determined abnormal;

[0049] For the measurement element The number of inequalities involved is 3, so the measurement element The total number of consistent judgments P it =3; if the number of inequalities that hold is less than 2, then the measurement element is determined abnormal.

[0050] Compared with the prior art, the present invention has the following beneficial effects:

[0051] 1) Data consistency is determined by judging whether the functional relationship between the measured element and the measured elements of other data sources is within the normal range, and the consistency scores of different functional relationships are accumulated. According to the consistency ratio, whether the measured element is normal is determined; it is applicable to three or more independent data sources, does not rely on the continuity of a single data, and can identify most of the outliers; it has a good supplementary function to the outlier judgment method of an isolated single data source, is effective for continuous or discontinuous outliers, and can reduce the probability of misidentification of outliers.

[0052] 2) The multi-source data consistency judgment technology has been added to enable it to have the ability to judge multi-source data from each other. From the perspective of data interrelationships, it supplements the classic outlier recognition method. Even continuous multi-point anomalies or trend anomalies can be identified, and misjudgment and miscorrection are extremely rare. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0054] Figure 1 It is a flow chart of the present invention. DETAILED DESCRIPTION

[0055] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0056] like Figure 1 As shown, the embodiment of the present invention provides a multi-source data fusion outlier identification method based on mutual consistency, and the steps are as follows:

[0057] Step 1: Divide into m data sources according to different observation devices, and the m data sources are all different, m ≥ 3; multi-source data is different source observations of the same target. Divide data according to the independence of multi-source data, generally divided into different sources according to different observation devices. Assume that there are m different data sources, namely D1, ..., D m . D i The data source has n i There are three measurement elements (referred to as measurement elements), which are D j The data source has n j The measurement elements are

[0058] Step 2: Construct the functional relationship between the m data sources. The corresponding functional relationship between the m different data sources is as follows:

[0059]

[0060] because is the measured value, there is a certain error, so when When With C ik are consistent, that is, they can verify the correctness of the measured elements. ik For a given threshold, we can ik and The error characteristics and functional relationship are determined. i,j=1,2,…,m,i≠j,k=1,2,…,n i , n i is the i-th data source D i The number of test elements, n j is the jth data source D j The number of measured elements.

[0061] Step 3: Determine whether the measured element in the data source is abnormal based on the functional relationship; the specific steps for outlier identification are:

[0062] S3.1. Set measurement element C ik The consistency score S ik =0, total number of consistency judgments P ik =0;

[0063] S3.2. For data source D i , list the data source D i The functional relationship between all measurement elements and measurement elements from other data sources is as follows:

[0064] S3.3. Judgment Is it true, where δ ik is a given threshold;

[0065] S3.4, P ik =P ik +1; if the inequality holds, then measure element C ik and The consistency score of S ik =S ik +1, S j1 =S j1 +1,S j2 =S j2 +1,...,S jn =S jn +1; if the inequality does not hold, the consistency score remains unchanged;

[0066] S3.5, traverse m data sources, traverse different measurement elements in the data source, and judge whether it is established;

[0067] S3.6. The final consistency ratio S ik / P ik Make a judgment, if S ik / P ik If it is not less than 50%, the measurement is considered normal; otherwise, the measurement is considered abnormal.

[0068] Specific Examples

[0069] Assuming that there are three types of observation equipment, m=3, and the observation equipment are a single pulse equipment, a multi-speed measuring equipment and a GNSS equipment, respectively. There is one single pulse equipment, three multi-speed measuring equipment and one GNSS equipment.

[0070] At time t, the measurement element of the single pulse device is R t , A t 、E t , the measuring element of the multi-speed measuring device I is The measuring elements of the multi-speed measuring device II are The measuring elements of the multi-speed measuring device III are The measurement element of the GNSS device is x t ,y t 、z t , Three relationships can be established based on the above measurements.

[0071] Establish the functional relationship between the measurement elements of the monopulse device and the measurement elements of the GNSS device:

[0072]

[0073] Among them, px,py,pz are the site addresses of single pulse equipment, and m′ represents the time number.

[0074] Establish the functional relationship between the measurement elements of multiple speed measuring devices and the measurement elements of GNSS devices:

[0075]

[0076] in, represents the measurement element of the i′th multi-speed measuring device, px i′ ,py i′ ,pz i′ is the site address of the i′th multi-speed measuring device, px0 represents the x-direction coordinate of the transmitting station site, ys0 represents the y-direction coordinate of the transmitting station site, zs0 represents the z-direction coordinate of the transmitting station site, R 0t represents the distance between the target and the transmitting station at time t, R i′t Indicates the distance between the target and the i′th multi-speed measuring device at time t. When the main station signal transmitting station and the receiving station of the multi-speed measuring device are the same station, take xs0 = xs1, ys0 = ys1, zs0 = zs1, R 0t =R 1t .

[0077] Establish the functional relationship between the measurement elements of the single pulse device and the measurement elements of the multi-speed measurement device:

[0078]

[0079] Among them, x ct ,y ct ,z ct is the target position in the station coordinate system of multiple speed measuring devices, which can be converted to the x t ,y t ,z t , Same coordinate system, denoted as x dt ,y dt ,z dt , use the position difference to get the velocity:

[0080]

[0081] in, It represents the velocity in the x direction obtained by using the single pulse information difference at time t, It represents the velocity in the y direction obtained by using the single pulse information difference at time t, represents the velocity in the z direction obtained by using the single pulse information difference at time t, x dt+1 represents the x coordinate of the target at time t+1, x dt-1 represents the target y coordinate at time t-1, y dt+1 represents the target y coordinate at time t+1, y dt-1 represents the target y coordinate at time t-1, z dt+1represents the z coordinate of the target at time t+1, z dt-1 Indicates the target z coordinate at time t-1.

[0082] Then we have:

[0083]

[0084] To sum up, the formula is as follows:

[0085]

[0086] There are 9 equations in the above formula. Considering the measurement error of the observation equipment, the following inequalities exist:

[0087]

[0088] For the measurement element R t , the number of inequalities involved is 4, so the measurement variable R t The total number of consistent judgments P it =4; if the number of inequalities that hold is less than 2, then the measurement element R t Abnormal; for measurement element A t , the number of inequalities involved is 4, so the measurement element A t The total number of consistent judgments P it =4; if the number of inequalities that hold is less than 2, then the measurement element A is determined t Abnormal; for the measurement element E t , the number of inequalities involved is 4, so the measurement element E t The total number of consistent judgments P it =4; if the number of inequalities that hold is less than 2, then the measurement element E is determined t abnormal.

[0089] For the measurement element The number of inequalities involved is 2, so the measurement element The total number of consistent judgments P it =2; if the number of inequalities that hold is less than 1, then the measurement element is determined abnormal.

[0090] For the measurement element x t , the number of inequalities involved is 6, so the measured variable x t The total number of consistent judgments P it =6; if the number of established inequalities is less than 3, then determine the measured element x t Abnormal; for the measurement element z t , the number of inequalities involved is 6, so the measurement element z t The total number of consistent judgments P it =6; if the number of established inequalities is less than 3, then the measurement element z is determined t Abnormal; for the measurement element yt , the number of inequalities involved is 5, so the measured variable y t The total number of consistent judgments P it =5; if the number of established inequalities is less than 3, then the measured element y is determined t Abnormal; for the measurement element The number of inequalities involved is 3, so the measurement element The total number of consistent judgments P it =3; if the number of inequalities that hold is less than 2, then the measurement element is determined Abnormal; for the measurement element The number of inequalities involved is 3, so the measurement element The total number of consistent judgments P it =3; if the number of inequalities that hold is less than 2, then the measurement element is determined Abnormal; for the measurement element The number of inequalities involved is 3, so the measurement element The total number of consistent judgments P it =3; if the number of inequalities that hold is less than 2, then the measurement element is determined abnormal.

[0091] Outlier identification and correction is an indispensable step in the process of ballistic measurement data solution and processing. When processing data, the outliers of the observed data must first be identified and processed to ensure the quality of the external measurement data processing results. The present invention is an identification method that can effectively identify continuous outlier points and reduce the probability of misjudgment and miscorrection. The best way to implement this technology is to generalize and modularize the multi-source data fusion outlier identification algorithm based on mutual consistency, unify the input parameters and output parameters of the algorithm, and modularly package the algorithm to make the algorithm easy to understand and use, providing effective preprocessing for ballistic measurement data solution and processing.

[0092] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A multi-source data fusion outlier identification method based on mutual consistency, characterized in that: The steps are as follows: Step 1: Divide into m data sources according to different observation devices, and the m data sources are all different, m ≥ 3; Step 2: Construct a functional relationship between any two of the m data sources; The functional relationship between any two of the m data sources is: Among them, C ik is the i-th data source D i The kth measurement element in , is the jth data source, i,j=1,2,…,m,i≠j,k=1,2,…,n i , n i is the i-th data source D i The number of test elements, n j is the jth data source D j The number of measured elements; Step 3: Determine whether the measured element in the data source is abnormal based on the functional relationship; The method for judging whether the measured element in the data source is abnormal according to the functional relationship is: S3.

1. Set measurement element C ik The consistency score S ik =0, total number of consistency judgments P ik =0; S3.

2. For data source D i , list the data source D i The functional relationship between all measurement elements and measurement elements from other data sources is as follows: S3.

3. Judgment Is it true, where δ ik is a given threshold; S3.4, P ik =P ik +1; if the inequality holds, then measure element C ik and The consistency score of S ik =S ik +1, S j1 =S j1 +1, If the inequality does not hold, the consistency score remains unchanged; S3.5, traverse m data sources, traverse different measurement elements in the data source, and judge whether it is established; S3.

6. The final consistency ratio S ik / P ik Make a judgment, if S ik / P ik If it is not less than 50%, the measurement is considered normal; otherwise, the measurement is considered abnormal.

2. The multi-source data fusion outlier identification method based on mutual consistency according to claim 1 is characterized in that: When m=3, the observation devices are a single pulse device, a multi-speed measuring device and a GNSS device, and there is one single pulse device, three multi-speed measuring devices and one GNSS device. At time t, the measurement element of the single pulse device is R t , A t 、E t , the first multi-speed measuring device's measuring element is The measuring element of the second multi-speed measuring device is The third multi-speed measuring device has the following measuring elements: The measurement element of the GNSS device is x t ,y t 、z t , Establish the functional relationship between the measurement elements of the monopulse device and the measurement elements of the GNSS device: Among them, px, py, pz are the station addresses of the single pulse equipment, and m′ represents the time number; Establish the functional relationship between the measurement elements of multiple speed measuring devices and the measurement elements of GNSS devices: in, represents the measurement element of the i′th multi-speed measuring device, px i , Py i′ , pz i′ is the site address of the i′th multi-speed measuring device, px0 represents the x-direction coordinate of the transmitting station site, py0 represents the y-direction coordinate of the transmitting station site, pz0 represents the z-direction coordinate of the transmitting station site, R 0t represents the distance between the target and the transmitting station at time t, R i′t Indicates the distance between the target and the i′th multi-speed measuring device at time t; the master station signal transmitting station and receiving station of the multi-speed measuring device are the same station; Establish the functional relationship between the measurement elements of the single pulse device and the measurement elements of the multi-speed measurement device: Among them, x ct ,y ct , z ct Convert the target position in the station coordinate system of multiple speed measuring devices to the GNSS device In the same coordinate system, denoted as x dt ,y dt , z dt , use the position difference to get the velocity: in, It represents the velocity in the x direction obtained by using the single pulse information difference at time t, It represents the velocity in the y direction obtained by using the single pulse information difference at time t, represents the velocity in the z direction obtained by using the single pulse information difference at time t, x dt+1 represents the x coordinate of the target at time t+1, x dt-1 represents the target y coordinate at time t-1, y dt+1 represents the target y coordinate at time t+1, y dt-1 represents the target y coordinate at time t-1, z dt+1 represents the z coordinate of the target at time t+1, z dt-1 represents the z coordinate of the target at time t-1; Then we have: To sum up, the formula is as follows: Considering the measurement error of the observation equipment, the following inequality exists: For the measurement element R t , the number of inequalities involved is 4, so the measurement variable R t The total number of consistent judgments P it =4; if the number of inequalities that hold is less than 2, then the measurement element R t abnormal; For the measurement element A t , the number of inequalities involved is 4, so the measurement element A t The total number of consistent judgments P it =4; if the number of inequalities that hold is less than 2, then the measurement element A is determined t abnormal; For the measurement element E t , the number of inequalities involved is 4, so the measurement element E t The total number of consistent judgments P it =4; if the number of inequalities that hold is less than 2, then the measurement element E is determined t abnormal; For the measurement element The number of inequalities involved is 2, so the measurement element The total number of consistent judgments P it =2; if the number of inequalities that hold is less than 1, then the measurement element is determined abnormal; For the measurement element x t , the number of inequalities involved is 6, so the measured variable x t The total number of consistent judgments P it =6; if the number of established inequalities is less than 3, then determine the measured element x t abnormal; For the measurement element z t , the number of inequalities involved is 6, so the measurement element z t The total number of consistent judgments P it =6; if the number of established inequalities is less than 3, then the measurement element z is determined t abnormal; For the measurement element y t , the number of inequalities involved is 5, so the measured variable y t The total number of consistent judgments P it =5; if the number of established inequalities is less than 3, then the measured element y is determined t abnormal; For the measurement element The number of inequalities involved is 3, so the measurement element The total number of consistent judgments P it =3; if the number of inequalities that hold is less than 2, then the measurement element is determined abnormal; For the measurement element The number of inequalities involved is 3, so the measurement element The total number of consistent judgments P it =3; if the number of inequalities that hold is less than 2, then the measurement element is determined abnormal; For the measurement element The number of inequalities involved is 3, so the measurement element The total number of consistent judgments P it =3; if the number of inequalities that hold is less than 2, then the measurement element is determined abnormal.

Citation Information

Patent Citations

  • Method for fault diagnosis of rotating mechanical equipment based on interval-type evidence fusion

    CN102033984A

  • Fault diagnosis method of random fuzzy fault characteristic fusion rotating mechanical device

    CN102662390A

  • Time-space relationship analysis system based on multi-source Internet-of-things location sensing

    CN106844585A

  • Scoring matrix-oriented multi-data-source fusion method

    CN111523590A

  • Power terminal error data identification method based on evidence theory

    CN112924812A