Target identification basic probability assignment generation method based on evidence theory
By using triangular fuzzy numbers based on evidence theory and sensor detection reliability correction, the basic probabilistic assignment problem of target recognition detection values being interval values in information warfare environments is solved, thereby improving the accuracy and reliability of target recognition.
Patent Information
- Application Number
- CN202511546551.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-10-28
- Publication Date
- 2026-02-17
AI Technical Summary
In an information warfare environment, when the target recognition detection value is an interval value, existing technologies struggle to convert the detection results into basic probability assignments of evidence focal elements.
A method based on evidence theory is adopted, which uses triangular fuzzy numbers to calculate the membership degree of interval endpoint values, and corrects the basic probability assignment by the detection reliability of the sensor under interference conditions. This includes normalization and correction steps to generate the basic probability assignment of the target.
In an information warfare environment, the target recognition results with detection values of interval type are effectively transformed into basic probability assignments of evidence focal elements, which improves the accuracy and reliability of target recognition.
Smart Images
Figure CN121542884A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of basic probability assignment generation technology, and in particular to a method for generating basic probability assignments for target recognition based on evidence theory. Background Technology
[0002] Dempser-Shafer evidence theory, as a generalization of probability theory, assigns probability values to the power set of the sub-proposition set. When the evidence provided by one piece of information cannot distinguish between two propositions, the probability value is directly assigned to the composite proposition formed by these two propositions, expressing the uncertainty of the information in a more natural way. The synthesis of multiple pieces of evidence is achieved through Dempser's combination formula, and the final decision is made based on the calculated confidence and likelihood. Since confidence and likelihood are the lower and upper bounds of probability, respectively, Dempser-Shafer evidence theory generalizes the precise point probability in probability theory to an interval value to better express uncertainty. Practical application results have also demonstrated the superiority of this theoretical tool.
[0003] Furthermore, there are many types of fuzzy numbers, such as the common triangular, trapezoidal, and Gaussian types. The specific type of fuzzy number to be used needs to be determined in conjunction with the actual problem. The principle is that the selected fuzzy number should be reasonable and easy to process later.
[0004] In the generation of basic probability assignment for target recognition in an information warfare environment, the technical solution for converting the detection result into the basic probability assignment problem of evidence focal elements when the target recognition detection value is an interval value still needs further improvement.
[0005] Therefore, it is necessary to improve one or more of the problems existing in the above-mentioned related technical solutions.
[0006] It should be noted that this section is intended to provide background or context for the technical solutions of this disclosure as set forth in the claims. The description herein does not constitute an admission that it is prior art simply because it is included in this section. Summary of the Invention
[0007] The purpose of this invention is to provide a method for generating basic probability assignments for target recognition based on evidence theory, thereby overcoming, to at least to some extent, one or more problems caused by the limitations and defects of related technologies.
[0008] This invention provides a method for generating basic probability assignments for target recognition based on evidence theory, comprising: S1. Based on the target attribute interval values observed by the sensor, the membership degree corresponding to the interval endpoint values is calculated using triangular fuzzy numbers: the membership function of the target attribute is represented as a triple, and the initial interval membership degree of the target attribute interval value to each target is obtained according to the numerical relationship between the interval endpoint values and the triple and the membership degree corresponding to the interval endpoint values. S2, normalize the initial interval membership of various types of targets to obtain the basic probability assignment; S3. Based on the detection reliability of each sensor under interference conditions, the basic probability assignment is corrected to obtain the corrected basic probability assignment of each sensor for various types of targets.
[0009] In this invention, the membership function described in S1 It is expressed as follows:
[0010] Where a < b < c, and (a, b, c) is a triplet.
[0011] In this invention, the target attribute range value The membership values for the target attribute interval are as follows: (1) When or hour,
[0012] (2) When hour,
[0013] (3) When hour,
[0014] (4) When hour,
[0015] (5) When hour,
[0016] (6) When hour,
[0017] (7) When hour, .
[0018] In this invention, in S2, using Indicates the target iAttributes At Sensor observation values are In the case of the target i The membership functions are compared as follows: .
[0019] In this invention, the interval membership degrees of various types of targets are normalized to obtain basic probability assignments. As shown below: ( i =1,……,n) Where n is the number of target types.
[0020] In this invention, in step S3, the corrected basic probability assignments for the i-th type of target by each sensor are obtained. It is expressed as follows:
[0021] in, The detection reliability of each sensor under interference conditions.
[0022] In this invention, the basic probability assignment to the entire set is as follows: (j=1,……,k) Where k is the number of sensors.
[0023] In this invention, the reliability of detection The possible values are as follows: When the interference received by the sensor is weak, Values range from 0.9 to 1; When the sensor experiences weak to moderate interference, Values range from 0.8 to 0.9. When the sensor experiences strong interference, Values range from 0.7 to 0.8.
[0024] The technical solution provided by this invention may include the following beneficial effects: This invention presents a method for generating basic probability assignments for target recognition based on evidence theory. When detection equipment in an information warfare environment experiences complex interference, resulting in interval-type values for target recognition, the method calculates the interval membership degree of each target detection value relative to its attribute feature value. Then, based on the detection reliability of each sensor under interference conditions, the normalized interval membership degree is corrected to obtain the basic probability assignments of each sensor to various target types. This method solves the problem of how to convert detection results into basic probability assignments of evidence focal elements when target recognition detection values are interval-type values in an information warfare environment. Attached Figure Description
[0025] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this disclosure and, together with the description, serve to explain the principles of this disclosure. It is obvious that the drawings described below are merely some embodiments of this disclosure, and those skilled in the art can obtain other drawings based on these drawings without any inventive effort.
[0026] Figure 1 A flowchart illustrating the target identification basic probability assignment generation method based on evidence theory in an exemplary embodiment of this disclosure is shown. Figure 2 This illustrates the representation of the target membership function triplet of a feature attribute in an exemplary embodiment of this disclosure. Detailed Implementation
[0027] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided so that this disclosure will be more comprehensive and complete, and will fully convey the concept of the exemplary embodiments to those skilled in the art. The described features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.
[0028] Furthermore, the accompanying drawings are merely illustrative diagrams of embodiments of this disclosure and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and therefore repeated descriptions of them will be omitted. Some block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities.
[0029] First, the theoretical basis of this application will be explained.
[0030] (1) Recognition Framework Suppose we have a decision problem. For this problem, let U represent the set of all possible outcomes that can be recognized. Then, any proposition of interest corresponds to a subset of U. Associating propositions with subsets transforms a more abstract logical concept into a more intuitive set-theoretic one. The disjunction, composition, and implication of any two propositions correspond to the union, intersection, and inclusion of their respective sets, and the negation of any proposition corresponds to the complement of the set it corresponds to. Shafer points out that the choice of U depends on knowledge, on the level of understanding, and on what is known and what is desired. To emphasize this epistemological characteristic of the possibility set U, Shafer calls it a recognition frame. When a proposition corresponds to a subset of this frame, the frame is said to recognize the proposition. The choice of U should be sufficiently rich so that any particular set of propositions considered can correspond to a class of sets in the power set of U.
[0031] (2) Basic concepts of evidence theory Definition 1: Let U be a recognition frame, then the function m: 2 U →[0,1] satisfies the following conditions: (1) m( )=0;(2) When m(A) is called the basic probability assignment of A, m(A) is called the basic probability assignment of A.
[0032] Where m(A) represents the degree of exact confidence in proposition A, and represents direct support for A.
[0033] Definition 2: Let U be a recognition frame, m: 2 U →[0,1] is the basic probability assignment on U, defined function: BEL:2 U →[0,1];BEL(A)= (A The function is called a trust function on U.
[0034] BEL(A) = Let A be the sum of the probability measures of all subsets of A, which represents the total trust in A. Therefore, we can conclude that: BEL( )=0,BEL(U)=1 Definition 3: If a subset A of the recognition frame U has m(A)>0, then A is called the focal element of the trust function BEL, and the union of all focal elements is called the kernel.
[0035] The corresponding level of uncertainty for A is available. It is measured by the degree of trust.
[0036] Definition 4: Let U be a recognition frame, and define PL: 2 U →[0,1]
[0037] PL is called the plausibility function.
[0038] PL(A) represents the degree of confidence in not rejecting A, and is the sum of the basic probability assignments of all sets intersecting with A, with BEL(A) ≤ PL(A). PL(A) - BEL(A) represents the degree of ignorance regarding A. The defined confidence interval (BEL(A), PL(A)) describes the uncertainty of A.
[0039] Definition 5: [BEL(A),PL(A)] is called the confidence interval of focal element A.
[0040] PL(A)-BEL(A) describes the uncertainty of A, and is called the uncertainty of the focal element A. PL(A) corresponds to the upper probability measure defined by Dempster, and BEL(A) corresponds to the lower probability measure.
[0041] (3) The combination rules of evidence theory Dempster's rule of composition is a rule that reflects the combined effect of evidence. Given several confidence functions based on different pieces of evidence within the same identification framework, if these pieces of evidence are not completely conflicting, the rule of composition can be used to calculate a single confidence function. This single confidence function can then be used as the confidence function under the combined effect of those pieces of evidence. This confidence function is called the direct sum of the original confidence functions.
[0042] Let BEL1 and BEL2 be two trust functions on the same recognition frame U, and let m1 and m2 be their corresponding basic probability assignments, with focal elements A1, ..., A k And B1,……,B r And let it be:
[0043] but:
[0044] In the above formula, if K1 If K1=1, then m1 and m2 are considered contradictory, and the basic probability assignments cannot be combined. For combinations of multiple pieces of evidence, the above rules can be used for pairwise synthesis.
[0045] (4) Decision-making based on evidence theory How to make decisions after combining evidence using the DS rules is a problem closely related to the application. Let U be the recognition frame, and m be the basic probability assignment obtained based on the combination of DS rules.
[0046] 1) Decision-making based on trust function Based on the m obtained after combination, the trust function BEL is calculated. This trust function is our decision result, which is actually a soft decision.
[0047] 2) Decision-making based on basic probability assignment Let A1, A2 U, satisfy
[0048]
[0049] Rule 1: The decision category should have the highest basic probability assignment m(A1); Rule 2: The difference between the basic probability assignments of the decision category and other categories must be greater than a certain threshold;
[0050] Rule 3: The uncertainty probability assignment m(X) must be less than a certain threshold;
[0051] Rule 4: The basic probability assignment for the decision category must be greater than the uncertain probability assignment;
[0052] (5) Operations on interval numbers remember ,say Let be a positive closed interval number, or simply an interval number. When , a degenerates into a positive real number.
[0053] set up and Let be two positive closed intervals, then ① ; ② In particular, ; ③ In particular, ; ④ If and only if , .
[0054] Definition 6: Let and Given two interval numbers A and B, find the square distance between them. for:
[0055]
[0056] (6) Basic concepts of fuzzy numbers Any mapping from the domain X to the closed interval [0, 1]:
[0057] Determine a fuzzy set on X , yes membership function, yes right The membership degree. A fuzzy number. It is a normal convex fuzzy set of membership functions defined on the real number field R, and satisfies the following conditions: 1) There exists a unique point It has a membership function ; 2) Membership function It is continuous from left to right.
[0058] Fuzzy Number The general form can be written as:
[0059] In the formula, It is an increasing function, right continuous, and ; It is a decreasing function, left continuous, and .
[0060] This application uses interval membership based on fuzzy functions to assign values to basic probability assignments.
[0061] Secondly, this example implementation provides a method for generating basic probability assignments for target recognition based on evidence theory. Please refer to [link / reference]. Figure 1 This method may include: S1-S3, as follows: S1. Based on the target attribute interval values observed by the sensor, the membership degree corresponding to the interval endpoint values is calculated using triangular fuzzy numbers: the membership function of the target attribute is represented as a triple, and the initial interval membership degree of the target attribute interval value to each target is obtained according to the numerical relationship between the interval endpoint values and the triple and the membership degree corresponding to the interval endpoint values. S2, normalize the initial interval membership of various types of targets to obtain the basic probability assignment; S3. Based on the detection reliability of each sensor under interference conditions, the basic probability assignment is corrected to obtain the corrected basic probability assignment of each sensor for various types of targets.
[0062] This embodiment provides a method for generating basic probability assignments for target recognition based on evidence theory in an information warfare environment. When the detection value of target recognition is an interval value due to complex interference with the detection equipment in an information warfare environment, the method calculates the interval membership degree of the attribute interval value relative to the attribute feature value for each target detection value. Then, based on the detection reliability of each sensor under interference conditions, the normalized interval membership degree is corrected to obtain the basic probability assignment of each sensor to various types of targets. This method solves the problem of how to convert the detection result into a basic probability assignment of evidence focal elements when the target recognition detection value is an interval value in an information warfare environment.
[0063] The specific process of each step in the above embodiments will be described below.
[0064] First, basic probabilistic assignment based on evidence theory is generally built upon a relatively complete information foundation. However, in airborne target detection and identification systems, due to limited detection methods or secrecy requirements, observations of enemy targets are limited and uncertain, resulting in a small number of available samples for establishing descriptive models of target attributes. Furthermore, advancements in radar frequency agility and frequency hopping technologies, along with equipment measurement errors and various jamming techniques, mean that the radiation source parameters acquired by electronic reconnaissance equipment are no longer limited to scalar forms but are often interval values. Interval values require both a lower and upper limit, making them suitable for applications with limited descriptive information and high uncertainty.
[0065] Let Θ = {T1, T2, ..., T} n} represents the recognition framework. A sensor observes a target's attribute At and generates a measurement value D, where D is an interval value. The degree to which this measurement value belongs to each target attribute template At is _____. , , ..., .in This represents the degree to which an attribute At of target i belongs to that target when the sensor observation value is D.
[0066] S1, based on the target attribute interval values observed by the sensor. Calculate using the following triangular fuzzy number. and Corresponding membership degree and .
[0067] The target membership function of a feature attribute can be expressed as follows: Figure 2 The membership function of the triple (a, b, c) shown is:
[0068] 1) When or hour,
[0069] 2) When hour,
[0070] 3) When hour,
[0071] 4) When hour,
[0072] 5) When hour,
[0073] 6) When hour,
[0074] 7) When hour,
[0075] S2, using Let At represent the attribute of target i, and the sensor observation value be... In the case of comparison with the membership function of target i, the degree to which it belongs to the target is determined.
[0076]
[0077] The interval membership degrees of various target types are normalized to obtain the basic probability assignment. : (i=1,……,n) Where n is the number of target types.
[0078] S3, based on the detection reliability of each sensor under interference conditions The normalized interval membership (basic probability assignment) is corrected to obtain the basic probability assignment of each sensor to the i-th type of target. :
[0079] The basic probability assignment to the entire set is: (j=1,……,k) Where k is the number of sensors.
[0080] Detection reliability The range of values for is shown in Table 1: Table 1. Range of detection reliability values
[0081] In electronic warfare environments, the detection accuracy and range of various detection facilities are generally significantly affected. Considering the uncertainty and ambiguity of target information, the following example illustrates the operation process and results of the basic probability assignment method for target recognition based on evidence theory in complex interference environments.
[0082] In this basic probability assignment model, due to complex interference, the output detection values of each detection facility are not ordinary real values, but interval values.
[0083] To verify the effectiveness of this basic probability assignment method, the following example is provided.
[0084] Using low frequency S LF , intermediate frequency S MF and high frequency S HF Three types of sensors identify targets. Let the three types of targets be A1, A2, and A3, and their characteristic attribute parameters are shown in Table 2. The first item in the table is the characteristic value, and the second item is the characteristic value deviation under the information warfare environment.
[0085] Table 2 Target Feature Parameter Table
[0086] The target membership function of a feature attribute can be represented as a triple (a, b, c) as shown in Table 3.
[0087] Table 3. Triples of Target Membership Functions for Feature Attributes
[0088] Assuming an electronic warfare environment, all three types of sensors are subject to varying degrees of interference, and the output detection values are interval values: S LF =[12,23]、S MF =[98,110]、S HF =[312,343].
[0089] Normalization of interval membership for various types of targets: Table 4 Normalized values of membership degree for target intervals
[0090] Based on the different levels of interference experienced by various sensors, a low-frequency S is set. LF, intermediate frequency S MF and high frequency S HF The detection reliability of the three types of sensors is 0.9, 0.85, and 0.75, respectively.
[0091] Table 5 Basic Probability Assignments for Various Targets
[0092] For low frequency S LF , intermediate frequency S MF and high frequency S HF The three types of sensors combine the basic probability values assigned to the target as evidence to obtain the final fusion result.
[0093] Table 6 Results after fusion
[0094] Based on the results in Table 5, a decision can be made by selecting... , Clearly, the decision result indicates that the target is type A2.
[0095] It should be noted that although several modules of the system for executing actions are mentioned in the detailed description above, this division is not mandatory. In fact, according to embodiments of the present invention, the features and functions of two or more modules described above can be embodied in one module. Conversely, the features and functions of one module described above can be further divided into multiple modules for embodiment. Components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of the present invention according to actual needs. Those skilled in the art can understand and implement this without any inventive effort.
[0096] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of the present invention without departing from the spirit and scope of the claims. All of these forms are within the protection scope of the present invention.
Claims
1. A method for generating basic probability assignments for target recognition based on evidence theory, characterized in that, include: S1. Based on the target attribute interval values observed by the sensor, the membership degree corresponding to the interval endpoint values is calculated using triangular fuzzy numbers: the membership function of the target attribute is represented as a triple, and the initial interval membership degree of the target attribute interval value to each target is obtained according to the numerical relationship between the interval endpoint values and the triple and the membership degree corresponding to the interval endpoint values. S2, normalize the initial interval membership of various types of targets to obtain the basic probability assignment; S3. Based on the detection reliability of each sensor under interference conditions, the basic probability assignment is corrected to obtain the corrected basic probability assignment of each sensor for various types of targets.
2. The target recognition basic probability assignment generation method based on evidence theory according to claim 1, characterized in that, Membership function described in S1 It is expressed as follows: Where a < b < c, and (a, b, c) is a triplet.
3. The target recognition basic probability assignment generation method based on evidence theory according to claim 2, characterized in that, Target attribute range value The membership values for the target attribute interval are as follows: (1) When or hour, (2) When hour, (3) When hour, (4) When hour, (5) When hour, (6) When hour, (7) When hour, 。 4. The target recognition basic probability assignment generation method based on evidence theory according to claim 3, characterized in that, In S2, using Indicates the target i Attributes At Sensor observation values are In the case of the target i The membership functions are compared as follows: 。 5. The target recognition basic probability assignment generation method based on evidence theory according to claim 4, characterized in that, The interval membership degrees of various target types are normalized to obtain the basic probability assignment. As shown below: ( i =1,……，n) Where n is the number of target types.
6. The target recognition basic probability assignment generation method based on evidence theory according to claim 5, characterized in that, In S3, the corrected basic probability assignments for the i-th type of target are obtained for each sensor. It is expressed as follows: in, The detection reliability of each sensor under interference conditions.
7. The target recognition basic probability assignment generation method based on evidence theory according to claim 6, characterized in that, The basic probability assignment to the entire set is: (j=1,……,k) Where k is the number of sensors.
8. The target recognition basic probability assignment generation method based on evidence theory according to claim 7, characterized in that, Detection reliability The possible values are as follows: When the interference received by the sensor is weak, Values range from 0.9 to 1; When the sensor experiences weak to moderate interference, Values range from 0.8 to 0.
9. When the sensor experiences strong interference, Values range from 0.7 to 0.8.