Methods and systems for assessing the autonomous capabilities of ground-based unmanned equipment, and computing devices

Abstract

This invention discloses a method and system for assessing the autonomous capabilities of ground unmanned equipment, a computing device, and a storage medium. The method first constructs an assessment system for the autonomous capabilities of ground unmanned equipment, then evaluates the system using a fuzzy optimization model, a grey relational model, and fuzzy interaction entropy. The assessment results of the three models are then used as different information sources and fused using D-S evidence theory to obtain the final assessment result. The advantages of this invention are that it can accurately and quantitatively evaluate the grey, fuzzy, and difficult-to-quantify factors present in the assessment of autonomous capabilities, thereby providing a basis for objective analysis of the autonomous capability level of ground unmanned equipment.

Description

Technical Field

[0001] This invention relates to the field of ground unmanned equipment technology, specifically to a method and system for assessing the autonomous capabilities of ground unmanned equipment, and a computing device. Background Technology

[0002] Ground-based unmanned equipment needs to perform tasks in dangerous environments that are inaccessible, undesirable, or unthinkable to humans. Therefore, it must possess capabilities such as environmental perception, mission planning, and autonomous action—in other words, a high degree of autonomy. By comparing the autonomy of different unmanned equipment, we can not only clarify the autonomy requirements for specific missions and select suitable equipment for those missions, but also identify the current weaknesses and strengths of unmanned equipment, further determining the future development direction of its autonomy. This is of great significance to policymakers, researchers, designers, and users of unmanned equipment.

[0003] The general approach to classifying the autonomous capabilities of ground-based unmanned equipment (UAVs) is based on the degree of human involvement in the mission. For example, UAVs whose missions are entirely completed independently by humans have the lowest level of autonomy, while those whose missions are entirely completed autonomously have the highest level. The specific number of levels in the intermediate tiers needs to be determined based on the application scenario of the UAV. However, autonomous capability assessment involves a large number of subjective factors, most of which are ambiguous, vague, and difficult to quantify, leading to a lack of intuitiveness in the overall performance and increasing the complexity of the evaluation process. Therefore, there is an urgent need for a method that can accurately and reasonably evaluate the autonomous capabilities of ground-based UAVs. Summary of the Invention

[0004] This invention proposes a method and system for assessing the autonomous capabilities of ground-based unmanned equipment, as well as a computing device and storage medium, which solves the problem that existing technologies are unable to accurately assess gray, ambiguous, and difficult-to-quantify factors in autonomous capability assessment.

[0005] According to one aspect of the present invention, a method for evaluating the autonomous capability of ground unmanned equipment is provided, characterized by comprising: constructing an evaluation system for the autonomous capability of ground unmanned equipment, wherein the evaluation system is characterized by evaluation indicators of the environmental complexity of the ground unmanned equipment, the complexity of the ground unmanned equipment in completing the task, the degree of human intervention received by the ground unmanned equipment, and the task completion degree of the ground unmanned equipment, and obtaining the weight of each indicator.

[0006] The system to be evaluated is evaluated using the grey relational model, fuzzy optimization model, and symmetric interaction entropy model to obtain different evaluation results. These evaluation results are then used as different information sources and fused according to the merging rules of the basic probability allocation function using the Dempster evidence synthesis formula to obtain the final evaluation result.

[0007] Furthermore, the grey relational model is used to evaluate the system to be evaluated, including: calculating and comparing the grey relational degree between the system to be evaluated and the standard grade sample, and evaluating the sample to be evaluated as the grade corresponding to the standard grade sample.

[0008] Furthermore, the system to be evaluated is evaluated using a fuzzy optimization model, including: calculating the relative membership degree of the similarity between the system to be evaluated and the standard grade sample, and selecting the standard grade with the largest relative membership degree as the evaluation grade of the sample to be evaluated.

[0009] Furthermore, the system to be evaluated is evaluated using a symmetric interaction entropy model, including: calculating the symmetric interaction entropy between the system to be evaluated and the standard grade samples, and determining the similarity between the system to be evaluated and the standard grade samples based on the symmetric interaction entropy.

[0010] Furthermore, the evidence is fused according to the merging rules of the basic probability allocation function using the Dempster evidence synthesis formula, including using the Pignistic probability to make decisions on the evidence. The decision-making process consists of two steps: first, obtaining the basic probability allocation function of each piece of evidence; and second, converting the basic probability allocation function into a traditional probability for decision-making.

[0011] Furthermore, the weights of each indicator are obtained, including: calculating the entropy weight of each indicator based on the degree of variation of each indicator value using entropy.

[0012] Furthermore, entropy is used to calculate the entropy weight of each indicator, including: weighting all indicators using the entropy weight of each indicator.

[0013] According to another aspect of the present invention, a ground-based unmanned equipment autonomous capability assessment system is provided, characterized in that it comprises:

[0014] The analysis unit constructs an evaluation system for the autonomous capabilities of ground unmanned equipment. The evaluation system is characterized by evaluation indicators such as the environmental complexity of the ground unmanned equipment, the complexity of the task completed by the ground unmanned equipment, the degree of human intervention received by the ground unmanned equipment, and the task completion rate of the ground unmanned equipment, and obtains the weight of each indicator.

[0015] The processing unit evaluates the system to be evaluated using a grey relational model, a fuzzy optimization model, and a symmetric interaction entropy model, obtaining different evaluation results. These evaluation results are then used as different information sources and fused according to the merging rules of the basic probability allocation function using the Dempster evidence synthesis formula to obtain the final evaluation result.

[0016] According to another aspect of the present invention, a computing device is provided, characterized in that the device includes: a processor and a memory storing computer program instructions; the processor reads and executes the computer program instructions to implement the aforementioned method for assessing the autonomous capabilities of unmanned ground equipment.

[0017] According to another aspect of the present invention, a computer storage medium is provided, wherein at least one executable instruction is stored therein, the executable instruction causing a processor to perform an operation corresponding to the ground unmanned equipment autonomous capability assessment method.

[0018] As can be seen from the technical inventions provided by the present invention, the method and system for evaluating the autonomous capabilities of ground unmanned equipment provided by the present invention first constructs an evaluation system for the autonomous capabilities of ground unmanned equipment, evaluates the system using a fuzzy optimization model, a grey relational model, and a fuzzy interaction entropy, and then uses the evaluation results of the three models as different information sources. After fusion using DS evidence theory, the final evaluation result is obtained. This method can accurately and quantitatively evaluate the grey, fuzzy, and difficult-to-quantify factors in the evaluation of the autonomous capabilities of ground unmanned equipment, thereby providing a basis for objective analysis of the autonomous capability level of ground unmanned equipment.

[0019] The above description is merely an overview of the technical invention of this invention. In order to better understand the technical means of this invention and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this invention more apparent and understandable, specific embodiments of this invention are described below. Attached Figure Description

[0020] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:

[0021] Figure 1 The flowchart is a method for evaluating the autonomous capabilities of unmanned ground equipment according to the present invention.

[0022] Figure 2 This is a schematic diagram of the ground unmanned equipment evaluation system of the present invention;

[0023] Figure 3 This is a schematic diagram of the autonomous capability assessment system for ground unmanned equipment of the present invention;

[0024] Figure 4 This is a structural diagram of an exemplary hardware architecture for a computing device used to assess the autonomous capabilities of ground-based unmanned equipment in an embodiment of the present invention. Detailed Implementation

[0025] Exemplary embodiments of the invention will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the invention are shown in the drawings, it should be understood that the invention may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this invention will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art.

[0026] The following is a detailed description of a method and system for assessing the autonomous capabilities of unmanned ground equipment provided by this invention. Contents not described in detail in the embodiments of this invention are prior art known to those skilled in the art. Where specific conditions are not specified in the embodiments of this invention, they shall be performed according to conventional conditions in the art or conditions recommended by the manufacturer.

[0027] See Figure 1 The method for assessing the autonomous capabilities of ground-based unmanned equipment in this invention includes:

[0028] An evaluation system for the autonomous capabilities of ground unmanned equipment is constructed. The evaluation system is characterized by evaluation indicators such as the complexity of the environment in which the ground unmanned equipment is located, the complexity of the tasks completed by the ground unmanned equipment, the degree of human intervention received by the ground unmanned equipment, and the task completion rate of the ground unmanned equipment, and the weight of each indicator is obtained.

[0029] The system to be evaluated is evaluated using the grey relational model, fuzzy optimization model, and symmetric interaction entropy model to obtain different evaluation results. These evaluation results are then used as different information sources and fused according to the merging rules of the basic probability allocation function using the Dempster evidence synthesis formula to obtain the final evaluation result.

[0030] Specifically, the autonomous capability assessment system for ground-based unmanned equipment of this invention is as follows: Figure 2 As shown, the main indicators are the environmental complexity of the ground-based unmanned equipment, the complexity of completing the task, the degree of human intervention, and the task completion rate. Each coordinate axis represents one aspect, forming a four-axis evaluation model for unmanned equipment.

[0031] Specifically, based on the degree of variation of each indicator value, entropy can be used to calculate the entropy weight of each indicator, and the entropy weight of each indicator can be used to weight all indicators to obtain a more objective evaluation result.

[0032] Optionally, the entropy weight method can be used to calculate the weights, determining the final weight of each indicator based on both its importance and the amount of information it provides. Given m items to be evaluated and n evaluation indicators, the original data matrix R = (r ij ) m×n :

[0033]

[0034] Where r ij Let be the evaluation value of the i-th item under the j-th indicator.

[0035] The process of calculating the weights of each indicator value is as follows:

[0036] a. Calculate the weight p of the indicator value of the i-th item under the j-th indicator. ij :

[0037]

[0038] b. Calculate the entropy value e of the j-th index. j :

[0039] Where, k = 1 / lnm

[0040] c. Calculate the entropy weight w of the j-th index. j :

[0041]

[0042] d. Determine the comprehensive weight β of the indicators. j

[0043] Suppose the evaluator determines the weight of the indicator's importance as α based on their own objectives and requirements. j j = 1, 2, ..., n, combined with the entropy weight w of the index j This allows us to obtain the overall weight of index j:

[0044]

[0045] When all candidate items have the same value on index j, the entropy of the index reaches its maximum value of 1, and its entropy weight is zero.

[0046] like Figure 3 As shown, the present invention provides a ground-based unmanned equipment autonomous capability assessment system, comprising:

[0047] Analysis unit 101 constructs an evaluation system for the autonomous capabilities of ground unmanned equipment. The evaluation system is characterized by evaluation indicators such as the environmental complexity of the ground unmanned equipment, the complexity of the ground unmanned equipment in completing the task, the degree of human intervention received by the ground unmanned equipment, and the task completion rate of the ground unmanned equipment, and obtains the weight of each indicator.

[0048] Processing unit 102 uses grey relational model, fuzzy optimization model and symmetric interaction entropy model to evaluate the system to be evaluated, and obtains different evaluation results. Then, the evaluation results are used as different information sources and fused according to the merging rules of the basic probability allocation function and the Dempster evidence synthesis formula to obtain the final evaluation result.

[0049] This invention first uses fuzzy optimization model, grey relational model and symmetric interaction entropy to evaluate the system, and then uses the evaluation results of the three models as different information sources, and obtains the final evaluation result after fusion by DS evidence theory.

[0050] Specifically, grey relational analysis refers to the degree of correlation between the system to be evaluated and the standard system. Let the standard level sample be: X0 = {x 01 ,…,x 09 ,x 10}, where x 0k (k = 1, 2, ..., 10) represent the corresponding 10 standard levels; x 0k (j)(j=1,2,...,n) represent x respectively 0k The j-th indicator, and the individual indicators corresponding to each standard level are determined by experts.

[0051] Let the set of samples to be evaluated be X = {x1, x2, ..., xn}. n}, where x i For the i-th system to be evaluated, x i (j) represents the j-th feature value of the i-th sample to be evaluated. X0 and X can be expressed as:

[0052]

[0053] The task of assessing the autonomous capabilities of ground-based unmanned equipment is to calculate and compare the correlation between the system to be assessed and the standard-level samples, and to compare the samples to be assessed with the standard-level samples. i The sample was assessed as a standard grade. 0k The corresponding level.

[0054] Sample x to be evaluated i Compared with each standard level sample x 0k The correlation degree at the j-th feature parameter point is:

[0055]

[0056] Where, Δ ik (j)=|x ok (j)-x i (j)|,Δ|x 0k (j)-x i (j)| min , Δ|x 0k (j)-x i (j)| max i = 1, 2, ..., m; k = 1, 2, ..., 10; j = 1, 2, ..., n. ρ is the resolution coefficient, which is generally taken as ρ = 0.5.

[0057] Assuming that the weights of different feature points are different, the weight vector can be expressed as: w = (w1, w2, ..., w j ), where j = 1, 2, ..., n; the weights used in this invention are the weights calculated using the aforementioned combined weighting method. Therefore, x can be calculated. i With x 0k The degree of difference is:

[0058]

[0059] This formula is called the Minkowski distance, where p is the distance parameter, typically taken as 2. Then, the sample x to be evaluated... i Compared with standard grade sample x 0k The grey relational degree (GRD) is:

[0060]

[0061] This formula represents the sample x to be evaluated. i Compared with standard grade sample x 0k Grey relational analysis model for similarity; sample x to be evaluated i The grey relational vector ri = {r1, r2, ..., r1} with 10 standard grade samples 10}, r ik The larger the value, the more likely the system being evaluated belongs to the standard level sample x. 0k The higher the probability of the corresponding level.

[0062] As mentioned earlier, the assessment levels can be divided into 10 levels; let the standard level sample set be: X0 = {x 01 x 02 x 03 x 04 , ..., x 10}, where x 0k k = 1, 2, ..., 10 represent the corresponding 10 levels, x Ok(j), j = 1, 2, ..., n represent x respectively. 0k The j-th characteristic value is the various evaluation indicators established above.

[0063] Let the sample to be evaluated be X = {x1, x2, ..., x}. n}, where x i For the i-th sample to be evaluated, x i (j) is x i The purpose of fuzzy optimization is to evaluate the level of the sample to be evaluated, given the j-th feature value.

[0064] Standard grade sample x 0k Compared with the sample to be evaluated x i Regarding the similarity of a single index j, if x i (j)=x 0k (j), then x i with x 0k The similarity of a single indicator j is 1. Using r... ik ={r ik (1), r ik (2), ..., r ik (j)} represents x i With x 0k Similarity among j individual indicators.

[0065]

[0066] Let the system to be evaluated be x. i Compared with standard grade sample x 0k The relative membership degree of similarity is u ik ,but

[0067]

[0068] In the formula, i = 1, 2, ..., m; k = 1, 2, ..., 10; the sample to be evaluated is x. i The relative membership vector u with respect to k standard-level samples i ={u i1 u i2 ,…,u ik}, u ik The larger the value, the more likely the sample to be evaluated belongs to the standard grade sample x. 0k The greater the probability, the higher the standard grade with the highest membership degree will be the evaluation grade of the sample to be evaluated.

[0069] The aforementioned standard fuzzy optimization theory does not consider the weights of each indicator. However, in practice, the importance of each indicator varies. Therefore, this invention proposes a weighted fuzzy optimization theory. The main idea is to combine fuzzy optimization theory with the previously obtained combined weights to obtain a relative membership calculation model for the weighted fuzzy optimization theory.

[0070]

[0071] In the formula, i = 1, 2, ..., m; k = 1, 2, ..., 10; m is the number of samples to be evaluated; x is the number of samples to be evaluated. i The relative membership vector u with respect to k standard-level samples i ={u i1 u i2 ,…,u ik}, u ik The larger the value, the more likely the sample to be evaluated belongs to the standard grade sample x. 0k The greater the probability, the higher the standard grade with the highest membership degree will be the evaluation grade of the sample to be evaluated.

[0072] This invention utilizes symmetric interaction entropy for system evaluation, where the eigenvector of the system to be evaluated is x. i The standard system indicator sample set is x 0k First, for x i and x 0k Processing is performed to eliminate the difference in magnitude between the feature parameters:

[0073]

[0074]

[0075] Then for x i and x 0k Normalization is performed to meet the probability requirements for calculating symmetric interaction entropy:

[0076]

[0077]

[0078] Then x can be calculated. i and x 0k Symmetric cross-entropy D(x) between i x 0k Since D(x) i x 0k ) reflects x i and x 0k To address the differences between x, the inventors have improved the calculation formula by incorporating the aforementioned combined weights, enabling it to reflect the differences between x. i and x0k The similarity between them is calculated using the following steps:

[0079] Let x i and x 0k The symmetric interaction entropy at the j-th feature point is:

[0080] d(j)=x i (j)log x i (j)+x 0k (j)log x 0k (j)-x i (j)log x 0k (j)-x 0k (j)log x i (j)

[0081] Then x i and x 0k The similarity is:

[0082]

[0083] In the formula, w = (w1, w2, ..., w n D(x) represents the combined weights of n indicators; i x 0k ) reflects x i and x 0k The similarity can therefore be used as a criterion for judging the system x to be evaluated. i Belonging to standard grade sample x 0k The corresponding level.

[0084] First, three different evaluation methods are used for evaluation, and then the DS evidence theory is used to integrate them: Let Ω be the set of all possible values ​​of variable X, the set is finite, Ω = {A1, A2, ..., A...} n}, and element A in Ω j (1≤j≤n) are mutually exclusive and independent. Ω is called an identification frame for the variable X. The set of all subsets of Ω is called the power set, denoted as 2. Ω That is, when the element in Ω is n, n∈N + When the power set has 2 elements, its power set has 2 elements. n Each element corresponds to a proposition about X.

[0085] Let Ω be a recognition frame, if the function m: 2 Ω →[0,1] satisfies the following condition:

[0086]

[0087]

[0088] Then m is called the basic probability assignment (BPA) on the recognition frame Ω, also known as the mass function. The BPA reflects the degree of support for propositions on the recognition frame. For a subset A of the recognition frame Ω, i.e. If m(A) > 0, then A is called the focal set, and m(A) is the basic probability number.

[0089] Several points to note about BPA: The function of BPA is to map any subset of the recognition frame Ω to the set [0, 1], where m(A) is the basic probability. And it consists of a single element, where m(A) represents the degree of confidence in the exactness of A; if And if A ≠ Ω, and A is not composed of a single element, then m(A) also represents the degree of trust in A, but it is unknown which elements in A should be assigned this degree of trust; if A = Ω, it means that it is unknown how to assign it.

[0090] This leads to important concepts such as trust function, likelihood function, and mode function, as well as their meanings and interrelationships.

[0091] Let m be a basic probability assignment, and let the function Bel be 2. Ω →[0,1] satisfies the following conditions:

[0092]

[0093] The function Bel is called the trust function induced by m on Ω. Of course, the trust function Bel may not be induced by BPA and may have its own definition.

[0094] If Ω is a recognition frame, then the function Bel: 2 Ω → [0, 1] is a trust function if and only if the following conditions are satisfied:

[0095] (a)

[0096] (b) Bel(Ω) = 1;

[0097] (c) For any positive integer n and any finite subset A1, ..., A n ∈2 Ω ,satisfy

[0098]

[0099] like Then A, B∈2 Ω We have Bel(A∪B)=Bel(A)+Bel(B), which means that the trust function and the classical probability are different, and it is semi-additive.

[0100] If the function Bel: 2 Ω →[0,1] is induced by the basic probability assignment m, then have

[0101] Let m be the basic probability assignment on the recognition frame Ω, and let the function P1: 2 Ω →[0, 1], satisfy:

[0102]

[0103] Pl is called the likelihood function.

[0104] The functions Pl and Bel have the following relationship: In fact, each of them defines the following:

[0105]

[0106] From the above relationship, it can be seen that functions Bel and Pl are conjugated.

[0107] The confidence function Bel represents the degree of confidence in proposition A under the current circumstances. Its value is the sum of the basic probabilities of all subsets of A, representing a "pessimistic estimate" and a lower bound function. The corresponding function Pl is an "optimistic estimate" and an upper bound function, representing the degree of confidence that A is not false. [Bel(A), Pl(A)] constitutes the uncertainty interval, representing a measure of the uncertainty of A. Furthermore, the uncertainty measure of A can be expressed as follows:

[0108]

[0109] f(A) has the following properties:

[0110]

[0111] Let m be the basic probability assignment on the recognition frame Ω, and let the function Q be: 2 Ω →[0, 1], satisfy:

[0112]

[0113] The function Q is called the modality function induced by m.

[0114] The relationship between the popularity function Q and the trust function Bel is as follows:

[0115]

[0116] The above formula can be derived from the definitions of the mode function and the trust function, but the detailed derivation will not be provided here.

[0117] set up have to

[0118]

[0119] If we know that Q(A) = Cq(A), function Given a known function, but C is an unknown positive constant, we can calculate C.

[0120]

[0121] This leads to the important property that the fusion of multiple pieces of evidence is independent of their order of arrangement.

[0122] If Ω is a recognition frame, then the function Bel: 2 Ω → [0, 1] is called the Bayesian trust function, which satisfies:

[0123] (a)

[0124] (b) Bel(Ω) = 1;

[0125] (c) Bel(A∪B)=Bel(A)+Bel(B). When and hour.

[0126] The trust function of the focal element set is a Bayesian trust function (with only one element) for all single subsets.

[0127] Bayesian trust functions are a special type of trust function. In addition to the upper and lower probability interpretations, Dempster's evidence theory can also be explained by generalized Bayesian theory. In the Dempster combination rule mentioned later, when the focal set of the mass function is a simple subset and the focal elements are independent, the Bayesian formula is a special case of Dempster's combination rule.

[0128] In practical problems, for the same piece of evidence, different fundamental probability assignment functions (BPAs) can be obtained due to different evaluation sources. Let m1, m2, ... m n They are all basic probability allocation functions on the same identification framework Ω. Since there are multiple and different ones, it is necessary to fuse the information evidence.

[0129] The following describes the merging rule of the trust function BPA: Let m1 and m2 be two basic probability assignment functions on the same recognition frame Ω, with focal sets A1, A2, ..., A... p and B1, B2, ..., B l p, l∈N + Basic probability allocation values:

[0130] m1={m1(A1), m1(A2),..., m1(A p )}, m2={m2(B1), m2(B2),..., m2(B l )},A i ∩B i The trust value is m1(A) i )m2(B j Let m be the BPA resulting from the fusion of m1 and m2. Then, for the proposition set... The confidence level after fusion can be expressed as However when Sometimes, That will appear The reason for this phenomenon is that the information provided by m1 and m2 conflicts with each other. In this case, the above synthesis result needs to be corrected, that is, when discarding... This part assigns the confidence scores of the empty set to the non-empty set, which is a normalization process.

[0131] Let Bel1 and Bel2 be trust functions on the same frame Ω, m1 and m2 be their basic probability assignment functions, and let the focal sets be A1, A2, ..., A p and B1, B2, ..., B1, p, l∈N + Assuming

[0132]

[0133] Then the function m: 2 Ω →[0, 1], and

[0134]

[0135] in That is, the normalization factor. It is a basic probability distribution.

[0136] m is the result of the fusion of m1 and m2, denoted as This is called an orthogonal sum.

[0137] Let m1 and m2 be two fundamental probability assignment functions on the recognition frame Ω. There exist functions of mode, where Q1, Q2, and Q represent the modes of m1, m2, and m, respectively.

[0138] Q(A) = KQ1(A)Q2(A)

[0139] Where K is the normalization factor.

[0140] And it can be deduced from the previous equation that

[0141]

[0142] Let m1, m2, ... m n There are n basic probability assignment functions on the same recognition framework Ω, which are pairwise orthogonal and exist. The result of their data fusion is m, denoted as m. The process can be represented as follows:

[0143]

[0144] That is, after merging in pairs, they are then merged with new evidence, satisfying the commutative and associative laws, regardless of their order of arrangement.

[0145] Dempster's rule of evidence combination can be viewed as a function operation, which maps n BPAs to one BPA.

[0146] While the Dempster evidence synthesis formula has some desirable properties and is not particularly complex, in practical applications, if the identification frame Ω dimension is large and the number of focal elements in the basic probability assignment is too large, applying the Dempster rule can lead to combinatorial explosion, resulting in a large computational load and reduced practicality.

[0147] This invention simplifies the synthesis by reducing the number of focal elements in the mass function, and provides a Bayesian approximation calculation formula:

[0148]

[0149] Where A is a single subset. m(A) The Bayesian approximation of m(A) for non-single subsets m(A) =0, These are called Bayesian coefficients. The Bayesian approximation formula above reduces the number of focal elements to at most |Ω|, which simplifies the calculation.

[0150] Bayesian approximation methods have the following properties: (a) m The number of focal elements is at most |Ω|; (b) The Bayesian approximation is equal to (c) If the desired synthesis is a Bayesian trust function, then the Bayesian approximation is accurate.

[0151] Using Bayesian approximation, property (a) reduces the number of focal elements. Property (b) shows that the Bayesian approximation of the Dempster synthesis result can be achieved by separately approximating each piece of evidence using Bayesian methods before synthesis, thus preventing combinatorial explosion and improving computational efficiency. Property (c) establishes a connection between the Dempster synthesis formula and the Bayesian formula, demonstrating that the Bayesian formula can be considered a special case of the Dempster synthesis formula. This invention uses pignistic probability to make decisions based on evidence. The decision-making process consists of two steps: first, obtaining the BPA of each piece of evidence; and second, converting the BPA into traditional probabilities for decision-making. The purpose of pignistic probability transformation is to redistribute the BPA values ​​of the evidence already obtained by the system for more efficient decision-making. The most common allocation method is the average allocation method, which distributes the BPA values ​​of the non-single subset focal element set equally among the individual elements it contains.

[0152] The definition of the pignistic probability function is given below.

[0153] Let m be the basic probability assignment function on the recognition frame Ω, and define the Pignistic probability function BetP. m Ω→[0,1]: satisfies:

[0154]

[0155] BetP m It can be extended to 2 Ω Above, there is

[0156]

[0157] BetP m (A) represents the total mass value of set A.

[0158] BetP m It is based on the mass function, BetP m The transformation aligns better with human intuition and effectively amplifies the BPA difference between propositions, eliminating some candidate targets below the decision threshold. Compared to probability-based methods such as subjective Bayesian methods, it does not require prior probabilities, can combine evidence at different levels, and can handle uncertainty caused by ignorance or inaccurate knowledge.

[0159] Figure 4This is a structural diagram of an exemplary hardware architecture for a computing device for assessing the autonomous capabilities of ground unmanned equipment, as described in an embodiment of the present invention. The device 900 for assessing the autonomous capabilities of ground unmanned equipment includes an input device 901, an input interface 902, a central processing unit 903, a memory 904, an output interface 905, and an output device 906. The input interface 902, the central processing unit 903, the memory 904, and the output interface 905 are interconnected via a bus 910. The input device 901 and the output device 906 are connected to the bus 910 via the input interface 902 and the output interface 905, respectively, and are thus connected to other components of the device 900 for assessing the autonomous capabilities of ground unmanned equipment.

[0160] Specifically, input device 901 receives input information from the outside and transmits the input information to central processing unit 903 through input interface 902; central processing unit 903 processes the input information based on computer-executable instructions stored in memory 904 to generate output information, temporarily or permanently stores the output information in memory 904, and then transmits the output information to output device 906 through output interface 905; output device 906 outputs the output information to the outside of ground unmanned equipment autonomous capability assessment device 900 for user use.

[0161] This invention also provides a computer storage medium storing computer program instructions, which, when executed by a processor, implement the user behavior recognition method of this invention.

[0162] In an exemplary embodiment, the computing device 900 may be implemented by one or more application-specific integrated circuits (ASICs), DSPs, programmable logic devices (PLDs), complex programmable logic devices (CPLDs), field-programmable gate arrays (FPGAs), general-purpose processors, controllers, microcontrollers (MCUs), microprocessors, or other electronic components to perform the aforementioned method.

[0163] It is understood that the memory 904 in this embodiment can be volatile memory or non-volatile memory, or both. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), ferromagnetic random access memory (FRAM), flash memory, magnetic surface memory, optical disc, or compact disc read-only memory (CD-ROM); the magnetic surface memory can be disk storage or magnetic tape storage. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of RAM are available, such as Static Random Access Memory (SRAM), Synchronous Static Random Access Memory (SSRAM), Dynamic Random Access Memory (DRAM), Synchronous Dynamic Random Access Memory (SDRAM), Double Data Rate Synchronous Dynamic Random Access Memory (DDRSDRAM), Enhanced Synchronous Dynamic Random Access Memory (ESDRAM), SyncLink Dynamic Random Access Memory (SLDRAM), and Direct Rambus Random Access Memory (DRRAM).The memories described in the embodiments of this application are intended to include, but are not limited to, these and any other suitable types of memories.

[0164] In an exemplary embodiment, this application also provides a storage medium, namely a computer storage medium, specifically a computer-readable storage medium, such as a memory 904 storing a computer program, which can be executed by the central processing unit 903 of the computing device 900 to complete the steps described in the aforementioned method. The computer-readable storage medium may be a memory such as FRAM, ROM, PROM, EPROM, EEPROM, Flash Memory, magnetic surface memory, optical disc, or CD-ROM.

[0165] It should be clarified that the present invention is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present invention is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of the present invention.

[0166] The functional blocks shown in the above structural diagram can be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, they can be, for example, electronic circuits, application-specific integrated circuits (ASICs), appropriate firmware, plug-ins, function cards, etc. When implemented in software, the elements of this invention are programs or code segments used to perform the required tasks. The programs or code segments can be stored on a machine-readable medium or transmitted over a transmission medium or communication link via data signals carried on a carrier wave. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, fiber optic media, radio frequency (RF) links, etc. Code segments can be downloaded via computer networks such as the Internet, intranets, etc.

[0167] It should also be noted that the exemplary embodiments mentioned in this invention describe methods or systems based on a series of steps or apparatus. However, this invention is not limited to the order of the steps described above; that is, the steps can be performed in the order mentioned in the embodiments, or in a different order, or several steps can be performed simultaneously.

Claims

1. A method for evaluating the autonomous capabilities of ground-based unmanned equipment, characterized in that, include: An evaluation system for the autonomous capabilities of ground unmanned equipment is constructed. The evaluation system is characterized by evaluation indicators such as the complexity of the environment in which the ground unmanned equipment is located, the complexity of the tasks completed by the ground unmanned equipment, the degree of human intervention received by the ground unmanned equipment, and the task completion rate of the ground unmanned equipment, and the weight of each indicator is obtained. The system to be evaluated is evaluated using the grey relational model, the fuzzy optimization model, and the symmetric interaction entropy model to obtain different evaluation results. These evaluation results are then used as different information sources and fused according to the merging rules of the basic probability allocation function and the Dempster evidence synthesis formula to obtain the final evaluation result. The grey relational model is used to evaluate the system to be evaluated, including: calculating and comparing the grey relational degree between the system to be evaluated and the standard grade sample, and evaluating the sample to be evaluated as the grade corresponding to the standard grade sample; The evaluation of the system to be evaluated is carried out using a fuzzy optimization model, including: calculating the relative membership degree of the similarity between the system to be evaluated and the standard level sample, and selecting the standard level with the largest relative membership degree as the evaluation level of the sample to be evaluated; The evaluation of the system to be evaluated is carried out using a symmetric interaction entropy model, including: calculating the symmetric interaction entropy between the system to be evaluated and the standard grade samples, and determining the similarity between the system to be evaluated and the standard grade samples based on the symmetric interaction entropy.

2. The method for evaluating the autonomous capability of ground-based unmanned equipment according to claim 1, characterized in that, The evidence is fused according to the merging rules of the basic probability assignment function using the Dempster evidence synthesis formula. This includes using Pignistic probability to make decisions about the evidence. The decision-making process consists of two steps: first, obtaining the basic probability assignment function for each piece of evidence; and second, converting the basic probability assignment function into a traditional probability for decision-making.

3. The method for evaluating the autonomous capabilities of ground-based unmanned equipment according to claim 1, characterized in that, Obtain the weights of each indicator, including: calculating the entropy weight of each indicator based on the degree of variation of each indicator value using entropy.

4. The method for evaluating the autonomous capability of ground-based unmanned equipment according to claim 3, characterized in that, Calculating the entropy weight of each indicator using entropy includes: weighting all indicators using the entropy weight of each indicator.

5. A ground-based unmanned equipment autonomous capability assessment system, characterized in that, include: The analysis unit is used to construct an evaluation system for the autonomous capabilities of ground unmanned equipment. The evaluation system is characterized by evaluation indicators such as the environmental complexity of the ground unmanned equipment, the complexity of the ground unmanned equipment in completing the task, the degree of human intervention received by the ground unmanned equipment, and the task completion rate of the ground unmanned equipment, and the weight of each indicator is obtained. The processing unit uses the grey relational model, fuzzy optimization model and symmetric interaction entropy model to evaluate the system to be evaluated, obtain different evaluation results, and then uses the evaluation results as different information sources. After merging them according to the merging rules of the basic probability allocation function and the Dempster evidence synthesis formula, the final evaluation result is obtained. The grey relational model is used to evaluate the system to be evaluated, including: calculating and comparing the grey relational degree between the system to be evaluated and the standard grade sample, and evaluating the sample to be evaluated as the grade corresponding to the standard grade sample; The evaluation of the system to be evaluated is carried out using a fuzzy optimization model, including: calculating the relative membership degree of the similarity between the system to be evaluated and the standard level sample, and selecting the standard level with the largest relative membership degree as the evaluation level of the sample to be evaluated; The evaluation of the system to be evaluated is carried out using a symmetric interaction entropy model, including: calculating the symmetric interaction entropy between the system to be evaluated and the standard grade samples, and determining the similarity between the system to be evaluated and the standard grade samples based on the symmetric interaction entropy.

6. A computing device, characterized in that, The device includes: a processor and a memory storing computer program instructions; the processor reads and executes the computer program instructions to implement the autonomous capability assessment method for ground unmanned equipment as described in any one of claims 1-4.

7. A computer storage medium, characterized in that, The storage medium stores at least one executable instruction, which causes the processor to perform the operation corresponding to the autonomous capability assessment method for ground unmanned equipment as described in any one of claims 1-4.

Images