A method and system for evaluating the interoperability of a UAV system based on Kalman filtering

Abstract

The application discloses a kind of based on Kalman filtering's unmanned plane system interoperability evaluation method, comprising: constructing including structural attribute, application attribute, facility attribute, data attribute and knowledge attribute hierarchical model;Based on the above attribute respectively constructs secondary index tree and is merged to obtain overall attribute index tree;Based on overall attribute index tree, score is obtained to each attribute corresponding secondary index score and weight operator;Based on corresponding secondary index score and weight operator, corresponding attribute state equation is obtained;Based on structural attribute state equation and application attribute state equation, first attribute value is obtained by fusion;Based on first attribute value and facility attribute state equation, second attribute value is obtained by fusion;The above fusion process is executed cyclically, and data attribute state equation and knowledge attribute state equation are sequentially fused to finally obtain fourth attribute value;Based on fourth attribute value and hierarchical model, interoperability grade is obtained.The accuracy of unmanned plane system interoperability evaluation is improved.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) system interoperability assessment technology, and more specifically to a UAV system interoperability assessment method and system based on Kalman filtering. Background Technology

[0002] Unmanned aerial vehicle (UAV) systems typically consist of unmanned aerial vehicles (UAVs) and ground control stations. Compared to manned aircraft, they offer two main advantages: first, they eliminate the risk to the pilot's life; and second, they possess extraordinary aviation capabilities unrestricted by human intervention, such as endurance. UAVs utilize stable design improvements to low-altitude detection technology, which were originally dangerous for humans, to perform "tedious, dirty, or dangerous" missions while protecting pilots' lives, and their procurement and operating costs are lower than those of manned aircraft. Based on this, the purchase of UAVs has shown a significant upward trend in recent years. During missions, UAV systems need to be integrated into their respective operations and interoperate with manned aircraft, other UAV systems, and communication data link operators. Therefore, developing UAV systems capable of synergistic interaction among these systems and evaluating their interoperability are particularly important.

[0003] A typical system interoperability measurement model is the LISI model proposed in 1998. The LISI model divides system interoperability into five levels from low to high: isolation level, connectivity level, functional level, domain level, and cross-domain level, and identifies the interoperability factors affecting each level and the four attributes that constitute it: Procedure (P), Application (A), Infrastructure (I), and Data (D). However, its operational assessment is conducted under the old constructivist framework. The main characteristics of current UAV systems have shifted from constructivism to service-oriented, networked, GIG (Global Internet of Things) frameworks, "network-cloud," and "military cloud brain" frameworks. This has changed the connotation of the original attributes when conducting interoperability assessments of UAV systems. Furthermore, given the impact of new technologies such as the Internet of Things and deep learning on UAV systems, the accuracy of UAV system interoperability assessments no longer meets the requirements of the new framework.

[0004] Therefore, how to improve the accuracy of unmanned aerial vehicle (UAV) system interoperability assessment results while meeting the requirements of the new system is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0005] In view of this, the present invention provides a method and system for evaluating the interoperability of unmanned aerial vehicle (UAV) systems based on Kalman filtering. The invention expands the connotation of UAV interoperability under a new framework, adds knowledge attributes, expands the key features of each attribute at different levels, and divides the five attributes into two-level index trees, thereby improving the accuracy of UAV system interoperability evaluation.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] A method for evaluating the interoperability of unmanned aerial vehicle (UAV) systems based on Kalman filtering includes:

[0008] Construct an interoperability level model for unmanned aerial vehicle (UAV) systems that includes structural attributes, application attributes, facility attributes, data attributes, and knowledge attributes;

[0009] Based on the above attributes, construct secondary indicator trees respectively and merge them to obtain the overall attribute indicator tree;

[0010] Scoring is performed based on the overall attribute index tree to obtain multiple secondary index scores and multiple weight operators for each attribute;

[0011] The corresponding attribute state equation is obtained based on the corresponding secondary indicator score and the weight operator;

[0012] The first attribute value is obtained by fusing the structural attribute state equation and the applied attribute state equation.

[0013] The second attribute value is obtained by fusing the first attribute value and the facility attribute state equation.

[0014] The above fusion process is repeated cyclically, successively fusing the data attribute state equation and the knowledge attribute state equation to finally obtain the fourth attribute value;

[0015] Based on the fourth attribute value and the UAV system interoperability level model, the interoperability level is obtained.

[0016] Preferably, the unmanned aerial vehicle (UAV) system interoperability level model includes:

[0017] The isolation level is 0, the connectivity level is 1, the functionality level is 2, the integration level is 3, the collaboration level is 4, and the general level is 5.

[0018] Preferably, based on the above attributes, two secondary indicator trees are constructed respectively, resulting in structural attribute indicator trees, application attribute indicator trees, facility attribute indicator trees, data attribute indicator trees, and knowledge attribute indicator trees;

[0019] The attribute indicator trees mentioned above all include: target attribute, multiple primary indicators, and multiple secondary indicators;

[0020] Each of the primary indicators corresponds to multiple secondary indicators.

[0021] Preferably, the first-level indicators of the structural attribute indicator tree include: data structure, system structure, and management structure;

[0022] The first-level indicators of the application attribute indicator tree include: interaction model, implementation operation, and application scale;

[0023] The primary indicators of the facility attribute indicator tree include: facility controllability, protocol usability, and facility robustness;

[0024] The primary indicators of the data attribute indicator tree include: simplex, half-duplex, and full-duplex.

[0025] The primary indicators of the knowledge attribute indicator tree include: transmitted information, understood information, and predicted information.

[0026] Preferably, the secondary indicators of the data structure include: basic data standards, data service standards, and data management standards;

[0027] The secondary indicators of the system architecture include: physical layer, data layer, and semantic layer;

[0028] The secondary indicators of the management structure include: the standardization, universality, and integration of regulations and ordinances;

[0029] The secondary metrics of the interaction model include: interaction latency, interaction method, and interaction direction;

[0030] The secondary performance indicators for the operation include: endurance, flight altitude, and speed;

[0031] The secondary indicators of the application scale include: number of systems, number of software, and number of hardware.

[0032] The secondary indicators of facility controllability include: command controllability, management controllability, and collaboration controllability;

[0033] The secondary indicators of the protocol's usability include: protocol architecture, supported scale, and service types;

[0034] The secondary indicators of facility robustness include: link integrity, information redundancy, and resilience against destruction;

[0035] The secondary indicators of the simplex communication include: communication confidentiality, communication timeliness, and information capacity.

[0036] The secondary indicators of duplex and half-duplex include: instruction correctness, feedback timeliness, and data confidentiality;

[0037] The secondary indicators of duplex communication include: data confidentiality, communication fluency, and information accessibility.

[0038] The secondary indicators of the transmitted information include: images, video, and voice;

[0039] The secondary indicators for understanding information include: knowledge network, environment model, and target estimation;

[0040] The secondary indicators of the forecast information include: situation forecasting, model correction, and command decision-making.

[0041] Preferably, multiple secondary indicator scores and multiple weight operators are obtained for each attribute, specifically including:

[0042] Based on the expert experience of relevant professionals, scores are assigned to the secondary indicators of all attributes to obtain multiple secondary indicator scores for each attribute.

[0043] Based on the expert experience of the relevant professionals, multiple first weight operators are given between the primary indicators of each attribute and the corresponding attributes, and multiple second weight operators are given between the secondary indicators of each attribute and the corresponding primary indicators.

[0044] The weight operator is composed of the first weight operator and the second weight operator.

[0045] Preferably, the corresponding attribute state equation is obtained, specifically including:

[0046] The corresponding attribute state matrix is ​​obtained based on the first weight operator and the second weight operator corresponding to each attribute;

[0047] Based on the attribute state matrix and the corresponding secondary index scores, the corresponding attribute state quantity prediction equation X(k) is calculated:

[0048] X(k)=F X X(k-1)X∈{P,A,I,D,K}

[0049] Where k represents the state at the next moment, F X Let X represent the state matrix of attribute X, where X(k-1) represents the vector composed of the scores of all secondary indicators of attribute X, k-1 represents the current state, and P, A, I, D and K represent the structural attribute, application attribute, facility attribute, data attribute and knowledge attribute, respectively.

[0050] Based on the attribute state variable prediction equation X(k), the corresponding attribute covariance prediction equation is obtained:

[0051]

[0052] Where, ∑ X(k-1) Let X represent the covariance matrix of attribute X at the previous time step, and T denote the transpose sign;

[0053] The attribute state quantity prediction equation and the attribute covariance prediction equation together constitute the attribute state equation.

[0054] Preferably, the score of the second-level indicator corresponding to the largest second-level indicator of the first-level indicator of each attribute is used in the calculation of the attribute state equation.

[0055] Preferably, the fusion process is as follows:

[0056] The Kalman gain M is calculated based on the aforementioned attribute state equation. k :

[0057]

[0058] Among them, F Y Y represents the state matrix, and R represents the overall noise.

[0059] Calculate the nth attribute value P(k+n) based on the Kalman gain:

[0060] P(k+n)=P(k+M k (Y(1)-F Y P(k))

[0061] Based on the nth attribute value, the attribute value covariance matrix ∑ is obtained. P(k+n) :

[0062] ∑ P(k+n) =(1-M) k F Y )∑ P(k)

[0063] Where n represents the number of fusions.

[0064] An interoperability evaluation system for unmanned aerial vehicle (UAV) systems based on Kalman filtering includes: a ranking model construction module, an index tree construction module, a scoring module, a state equation construction module, a cyclic fusion module, and a result output module;

[0065] The hierarchical model construction module is used to construct an interoperability hierarchical model for unmanned aerial vehicle systems, including structural attributes, application attributes, facility attributes, data attributes, and knowledge attributes.

[0066] The indicator tree construction module is used to construct secondary indicator trees based on the above attributes and merge them to obtain the overall attribute indicator tree.

[0067] The scoring module is used to score based on the overall attribute index tree to obtain multiple secondary index scores and multiple weight operators corresponding to each attribute.

[0068] The state equation construction module is used to obtain the corresponding attribute state equation based on the corresponding secondary index score and the weight operator;

[0069] The cyclic fusion module is used to fuse the structural attribute state equation and the application attribute state equation to obtain a first attribute value; fuse the first attribute value and the facility attribute state equation to obtain a second attribute value; and cyclically execute the above fusion process to fuse the data attribute state equation and the knowledge attribute state equation in turn to obtain a fourth attribute value.

[0070] The result output module is used to obtain the interoperability level based on the fourth attribute value and the UAV system interoperability level model.

[0071] As can be seen from the above technical solutions, compared with the prior art, the present invention discloses a method and system for evaluating the interoperability of UAV systems based on Kalman filtering. It expands the connotation of UAV interoperability under a new framework, adds knowledge attributes, expands the key features of each attribute at different levels, divides the five attributes into two-level index trees, and combines qualitative and quantitative evaluation to improve the accuracy of UAV system interoperability evaluation results.

[0072] By constructing a secondary evaluation index under the interoperability evaluation attribute of unmanned aerial vehicle (UAV) systems, the interoperability attribute score was specifically quantified. An interoperability evaluation method based on Kalman filtering was proposed, and the interoperability level was derived from the quantified value of the secondary index, thus establishing an interoperability evaluation system. The developed interoperability evaluation system has been applied to the interoperability evaluation of actual UAV systems, achieving ideal results and proving the feasibility and practicality of the system. Attached Figure Description

[0073] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0074] Figure 1 The present invention provides a flowchart of an interoperability evaluation method for unmanned aerial vehicle systems based on Kalman filtering.

[0075] Figure 2 This is a schematic diagram of the structural attribute index tree structure provided by the present invention.

[0076] Figure 3 This is a schematic diagram of the application attribute index tree structure provided by the present invention.

[0077] Figure 4 A schematic diagram of the facility attribute index tree structure provided by the present invention.

[0078] Figure 5This is a schematic diagram of the data attribute index tree structure provided by the present invention.

[0079] Figure 6 A schematic diagram of the knowledge attribute index tree structure provided for this invention.

[0080] Figure 7 A schematic diagram of the overall attribute index tree structure provided by the present invention.

[0081] Figure 8 A schematic diagram of the overall attribute index tree structure with weighted operators provided by the present invention.

[0082] Figure 9 This invention provides a schematic diagram of the interoperability evaluation system structure for unmanned aerial vehicle (UAV) systems based on Kalman filtering. Detailed Implementation

[0083] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0084] Example 1

[0085] like Figure 1 As shown in the figure, this invention discloses a method for evaluating the interoperability of unmanned aerial vehicle (UAV) systems based on Kalman filtering, including:

[0086] Construct an interoperability level model for unmanned aerial vehicle (UAV) systems that includes structural attributes, application attributes, facility attributes, data attributes, and knowledge attributes;

[0087] Based on the above attributes, construct secondary indicator trees respectively and merge them to obtain the overall attribute indicator tree;

[0088] Scoring is performed based on the overall attribute indicator tree to obtain multiple secondary indicator scores and multiple weight operators for each attribute;

[0089] The corresponding attribute state equation is obtained based on the corresponding secondary index scores and weight operators;

[0090] The first attribute value is obtained by fusing the structural attribute state equation and the applied attribute state equation.

[0091] The second attribute value is obtained by fusing the first attribute value and the facility attribute state equation;

[0092] The above fusion process is repeated cyclically, successively fusing the data attribute state equation and the knowledge attribute state equation to finally obtain the fourth attribute value;

[0093] The interoperability level is obtained based on the fourth attribute value and the UAV system interoperability level model.

[0094] Example 2

[0095] This invention discloses a method for evaluating the interoperability of unmanned aerial vehicle (UAV) systems based on Kalman filtering, comprising:

[0096] Construct an interoperability level model for unmanned aerial vehicle (UAV) systems that includes structural attributes, application attributes, facility attributes, data attributes, and knowledge attributes:

[0097] Preferably, the LISI model provides a reference framework for defining and evaluating system interoperability. However, with increasing practical application, the interoperability levels assessed by the LISI model have shown discrepancies with actual work. Although UAVs are termed "unmanned," significant human involvement is still required to achieve the desired interoperability efficiency. For example, human personnel implicitly present in ground control stations determine the characteristics of a UAV system and play a crucial role in system safety. Even when coordinating the entire UAV development process using systems engineering methods, its scope is limited by the technical capabilities of human personnel; the human element is not removed from the UAV system. Therefore, when evaluating the interoperability of UAV systems, the human factor must be considered. The extent of human capabilities is often determined by their knowledge reserves; therefore, directly considering human knowledge attributes is fundamental to a complete interoperability assessment, thus adding the knowledge attribute K.

[0098] Preferably, the unmanned aerial vehicle (UAV) system interoperability level model includes:

[0099] The isolation level is 0, the connectivity level is 1, the functionality level is 2, the integration level is 3, the collaboration level is 4, and the general level is 5.

[0100] Preferably, the interoperability level model of the unmanned system in this embodiment is shown in Table 1:

[0101] Table 1. Unmanned System Interoperability Level Model

[0102]

[0103] Based on the above attributes, construct secondary indicator trees respectively and merge them to obtain the overall attribute indicator tree:

[0104] Preferred, such as Figures 2-6 As shown, in order to quantitatively evaluate the interoperability of unmanned aerial vehicle systems, two-level indicator trees are constructed based on the above attributes, resulting in structural attribute indicator trees, application attribute indicator trees, facility attribute indicator trees, data attribute indicator trees, and knowledge attribute indicator trees.

[0105] The attribute indicator trees mentioned above all include: target attribute, multiple primary indicators, and multiple secondary indicators;

[0106] Each primary indicator corresponds to multiple secondary indicators.

[0107] Preferably, the target attributes include: structural attributes, application attributes, facility attributes, data attributes, and knowledge attributes.

[0108] Preferably, the first-level indicators of the structural attribute indicator tree include: data structure P1(k), system structure P2(k), and management structure P3(k);

[0109] The first-level indicators of the application attribute indicator tree include: interaction model A1(k), implementation operation A2(k), and application scale A3(k);

[0110] The primary indicators of the facility attribute indicator tree include: facility controllability I1(k), protocol usability I2(k), and facility robustness I3(k);

[0111] The first-level indicators of the data attribute indicator tree include: simplex D1(k), half-duplex D2(k), and full-duplex D3(k);

[0112] The first-level indicators of the knowledge attribute indicator tree include: transmission information K1(k), understanding information K2(k), and prediction information K3(k).

[0113] Preferably, the secondary indicators of the data structure include: basic data standard P 11 (k-1), Data Service Standard P 12 (k-1) and data management standard P 13 (k-1);

[0114] The secondary indicators of system architecture include: physical layer P 21 (k-1), Data Layer P 22 (k-1) and semantic layer P 23 (k-1);

[0115] The secondary indicators of management structure include: Standardization of regulations and rules (P) 31 (k-1), Generality P 32 (k-1) and integration P 33 (k-1);

[0116] The secondary metrics of the interaction model include: interaction latency A 11 (k-1), Interaction Method A 12 (k-1) and interaction direction A 13 (k-1);

[0117] The secondary indicators for the implementation of the operation include: Durability A21 (k-1), Flight altitude A 22 (k-1) and velocity A 23 (k-1);

[0118] The secondary indicators of application scale include: Number of systems A 31 (k-1), Number of software A 32 (k-1) and the number of hardware A 33 (k-1);

[0119] The secondary indicators of facility controllability include: Command Controllability I 11 (k-1), Management Controllability I 12 (k-1) and cooperative controllability I 13 (k-1);

[0120] The secondary indicators of protocol usability include: Protocol Architecture I 21 (k-1), Support Size I 22 (k-1) and service type I 23 (k-1);

[0121] Secondary indicators of infrastructure robustness include: Link Integrity I 31 (k-1), Information Redundancy I 32 (k-1) and resistance to destruction I 33 (k-1);

[0122] The secondary indicators of simplex include: communication confidentiality (D) 11 (k-1), Communication Timeliness D 12 (k-1) and information capacity D 13 (k-1);

[0123] The secondary indicators of duplex computing include: instruction correctness (D) 21 (k-1), Feedback timeliness D 22 (k-1) and data confidentiality D 23 (k-1);

[0124] The secondary indicators of duplex include: Data confidentiality (D) 31 (k-1), Communication fluency D 32 (k-1) and information availability D 33 (k-1);

[0125] Secondary indicators of transmitted information include: image K 11 (k-1), Video K 12 (k-1) and the sound K 13 (k-1);

[0126] Secondary indicators for understanding information include: Knowledge Network K21 (k-1), Environmental Model K 22 (k-1) and target estimate K 23 (k-1);

[0127] The secondary indicators of forecast information include: Situation forecast K 31 (k-1), Model Correction K 32 (k-1) and command decision K 33 (k-1).

[0128] Preferably, for ease of calculation, the above five attribute index trees are merged and simplified to obtain the overall attribute index tree as follows: Figure 7 As shown, the overall attribute indicator tree is divided into 4 levels:

[0129] The first layer is the interoperability level of the unmanned aerial vehicle (UAV) system: after quantification, the average value of the isolation level is 0, the average value of the connectivity level is 1, the average value of the functional level is 2, the average value of the integration level is 3, the average value of the collaboration level is 4, and the average value of the general level is 5. In actual evaluation, the boundaries between the levels are blurred, and the evaluation results may fluctuate around the average value.

[0130] The second layer consists of target attributes: structural attribute P(k), application attribute A(k), facility attribute I(k), data attribute D(k), and knowledge attribute K(k);

[0131] The third layer consists of first-level indicators: corresponding to the data structure P1(k), system structure P2(k), management structure P3(k), interaction model A1(k), ... in the attribute indicator tree, where k represents the state at the next moment;

[0132] The fourth layer consists of secondary indicators: these are represented by vectors, showing the basic data standard P in the indicator tree. 11 (k-1), Data Service Standard P 12 (k-1), Data Management Standard P 13 (k-1), ..., where k-1 represents the current state.

[0133] Scoring is performed based on the overall attribute indicator tree to obtain multiple secondary indicator scores and multiple weight operators for each attribute;

[0134] Preferably, the process of evaluating the system interoperability level using the LISI model can be summarized as follows: Create a questionnaire based on the secondary interoperability indicators; interview technical personnel or domain experts to obtain data and complete the questionnaire; map the questionnaire onto the interoperability level model to form a system interoperability profile; perform attribute analysis on the profile to finally derive the system interoperability level.

[0135] Preferably, the questionnaire serves as the data source, directly impacting the evaluation results. Human intervention introduces new forms of uncertainty into the provided data. Firstly, there is a lack of sufficient knowledge about the implicit statistical phenomena in reality; secondly, there is inherent fuzziness in the process of constructing operable data. A generally accepted standard for modeling fuzzy information is formal Bayesian modeling. This uses a formal motion model (likelihood function) describing the transient behavior of the target to modify the realization-independent formal observation model (prior probability), resulting in a posterior probability with smaller variance, thereby reducing noise and improving uncertainty. However, the calculation of infinite integrals in formal Bayesian models typically lacks analytical solutions. This invention uses a method generalized from formal Bayesian models—Kalman filtering—to evaluate system interoperability, solving the problem of the lack of analytical solutions for infinite integrals in formal Bayesian models.

[0136] Preferably, multiple secondary indicator scores and multiple weight operators are obtained for each attribute, specifically including:

[0137] Based on the expert experience of relevant professionals, scores are assigned to the secondary indicators of all attributes to obtain multiple secondary indicator scores for each attribute.

[0138] Based on the expert experience of relevant professionals, multiple first weight operators are given between the primary indicators of each attribute and the corresponding attributes, as well as multiple second weight operators between the secondary indicators of each attribute and the corresponding primary indicators.

[0139] The weight operator is composed of the first weight operator and the second weight operator.

[0140] Preferably, each attribute's primary indicator corresponds to a different primary weight operator. X∈{P,A,I,D,K}, i∈{1,2,3}, i.e., the first weight operator The total weight of the attribute was allocated; each attribute's primary indicator has a corresponding secondary indicator with a different secondary weight operator. j∈{1,2,3}, i.e., the second weight operator The total weights of the corresponding primary indicators have been allocated; the overall attribute indicator tree with weighted operators is as follows: Figure 8 As shown, a larger weight operator indicates that the index has a greater component in the evaluation.

[0141] Preferably, the score of the second-level indicator corresponding to the largest second-level indicator in the first-level indicator of each attribute is used in the calculation of the attribute state equation.

[0142] Based on the corresponding secondary index scores and weight operators, the corresponding attribute state equation is obtained:

[0143] Preferably, the corresponding attribute state equation is obtained, specifically including:

[0144] The corresponding attribute state matrix is ​​obtained based on the first and second weight operators corresponding to each attribute;

[0145] The corresponding attribute state quantity prediction equation X(k) is calculated based on the attribute state matrix and the corresponding secondary index scores:

[0146]

[0147] Where k represents the state at the next moment, F X Represents the state matrix of attribute X. This represents a vector consisting of the scores of all secondary indicators of attribute X, where k-1 represents the current state, and P, A, I, D, and K represent structural attributes, application attributes, facility attributes, data attributes, and knowledge attributes, respectively.

[0148] Based on the attribute state variable prediction equation X(k), the corresponding attribute covariance prediction equation is obtained:

[0149]

[0150] Where, ∑ X(k-1) Let X represent the covariance matrix of attribute X at the previous time step, and T denote the transpose sign;

[0151] The attribute state prediction equation and the attribute covariance prediction equation together constitute the attribute state equation.

[0152] Preferably, taking structural properties as an example, the structural property state variable prediction equation P(k) is:

[0153]

[0154]

[0155] Among them, P 11 (k-1) represents the score of the baseline data standard, and similarly, other P values... ij The meaning of (k-1) is the same as the above correspondence. i∈{1,2,3} represents the different first-level weight operators corresponding to the first-level indicators of structural attributes. j∈{1,2,3} represents the different secondary weight operators corresponding to the secondary indicators of structural attributes.

[0156] Will The structural property state matrix F P The above formula can be simplified to:

[0157]

[0158] in, This represents a vector consisting of the scores of all secondary indicators of the structural attributes.

[0159] Similarly, the prediction equation for the applied attribute state variables can be obtained:

[0160]

[0161] Among them, F A Represents the application attribute state matrix. This represents a vector consisting of the scores of all secondary indicators of the application attributes;

[0162] Facility attribute state quantity prediction equation:

[0163]

[0164] Among them, F I Represents the facility attribute state matrix. This represents a vector consisting of the scores of all secondary indicators of the facility attributes.

[0165] Data attribute state quantity prediction equation:

[0166]

[0167] Among them, F D Represents the data attribute state matrix. This represents a vector consisting of the scores of all secondary indicators of the data attribute.

[0168] Knowledge attribute state quantity prediction equation:

[0169]

[0170] Among them, F K Represents the knowledge attribute state matrix. This represents a vector consisting of the scores of all secondary indicators of the knowledge attribute.

[0171] Preferably, taking structural attributes as an example, the structural attribute covariance prediction equation is:

[0172]

[0173]

[0174] Where, ∑ P(k-1) Let T represent the covariance matrix of the structural properties at the previous time step, and let T denote the transpose sign.

[0175] Preferably, the initial values ​​of any two different secondary indicators in the covariance matrix are given by the technicians. When the initial values ​​of any two different secondary indicators tend to be consistent, the covariance is greater than 0, and when they are independent, the covariance is equal to 0. Similarly, the covariance matrices of attributes A, I, D, and K can be obtained.

[0176] Similarly, the prediction equation for applied attribute covariance can be obtained:

[0177]

[0178] Where, ∑ A(k-1) This represents the covariance matrix of the applied attribute at the previous time step.

[0179] Facility attribute covariance prediction equation:

[0180]

[0181] Where, ∑ I(k-1) The covariance matrix representing the facility attributes at the previous moment;

[0182] Data attribute covariance prediction equation:

[0183]

[0184] Where, ∑ D(k-1) The covariance matrix representing the data attributes at the previous time step;

[0185] Knowledge attribute covariance prediction equation:

[0186]

[0187] Where, ∑ K(k-1) This represents the covariance matrix of the knowledge attribute at the previous time step.

[0188] Preferably, the structural property state equation is:

[0189]

[0190]

[0191] The applied attribute state equation is:

[0192]

[0193] The facility attribute state equation is:

[0194]

[0195]

[0196] The state equation for data attributes is:

[0197]

[0198] The state equation for knowledge attributes is:

[0199]

[0200] The first attribute value is obtained by fusing the structural attribute state equation and the applied attribute state equation:

[0201] Preferably, in the prediction equation for the state variables, the state at the current time (k-1) is used to predict the state variables at the next time k. In the prediction equation for the covariance, the uncertainty at time k is predicted. Besides data errors caused by expert evaluation, model uncertainty and parameter uncertainty can all cause deviations in the results, which are all handled by Kalman filtering. Furthermore, the evaluation system is a static system without external forces, so no control variables are set here. Since the prediction equation does not include true values, even with continuous prediction, the results may not be accurate. Updating the prediction equation by using the next attribute as the true observation value yields a more accurate estimate.

[0202] Preferably, the fusion process is as follows:

[0203] Calculating the Kalman gain M based on the attribute state equation k :

[0204]

[0205] Among them, F Y Y represents the state matrix, and R represents the overall noise.

[0206] Calculate the value of the nth attribute P(k+n) based on Kalman gain:

[0207] P(k+n)=P(k+M k (Y(1)-F Y P(k))

[0208] The attribute value covariance matrix ∑ is obtained based on the nth attribute value. P(k+n) :

[0209] ∑ P(k+n) =(1-M) k F Y )∑ P(k)

[0210] Where n represents the number of fusions.

[0211] Preferably, the structural property state equation is: As the prediction equation, the applied attribute state equation will be: As the observation equation; based on the first fusion of the structural property state equation and the applied property state equation, i.e. when k=1, the first attribute value P(2) is obtained:

[0212] When k=1, by giving ∑ P(0)P(1) is obtained by solving the secondary index scores of structural attributes and the first and second weight operators of structural attributes.

[0213] When k=1, A(1) and F are obtained by solving the secondary index scores of application attributes and the first and second weight operators of application attributes. A ;

[0214] Calculation of Kalman gain M1 based on attribute state equations:

[0215]

[0216] At this point, the overall noise R is 0;

[0217] Calculate the first attribute value P(2) based on Kalman gain M1 and P(1):

[0218] P(2)=P(1)+M1(A(1)-F A P(1))

[0219] The attribute value covariance matrix ∑ is obtained based on the first attribute value. P(2) :

[0220] ∑ P(2) =(1-M1F) A )∑ P(1) .

[0221] The second attribute value is obtained by fusing the first attribute value and the facility attribute state equation:

[0222] Preferably, the second attribute value P(3) is obtained by performing a second fusion based on the structural attribute state equation and the facility attribute state equation, i.e., when k=2:

[0223] When k=1, I(1) and F are obtained by solving the secondary index scores of facility attributes and the first and second weight operators of facility attributes. I ;

[0224] Calculation of Kalman gain M2 based on attribute state equations:

[0225]

[0226] The second attribute value P(3) is calculated based on the Kalman gain M2 and P(2):

[0227] P(3)=P(2)+M2(I(1)-F I P(2))

[0228] The attribute value covariance matrix ∑ is obtained based on the second attribute value. P(3) :

[0229] ∑ P(3)=(1-M2F) I )∑ P(2) .

[0230] The above fusion process is repeated cyclically, successively fusing the data attribute state equation and the knowledge attribute state equation, ultimately yielding the fourth attribute value:

[0231] Preferably, the third attribute value P(4) is obtained by performing a third fusion based on the structural attribute state equation and the data attribute state equation, i.e., when k=3:

[0232] When k=1, D(1) and F are obtained by solving the secondary index scores of data attributes and the first and second weight operators of data attributes. D ;

[0233] Calculation of Kalman gain M3 based on attribute state equations:

[0234]

[0235] The third attribute value P(4) is calculated based on the Kalman gain M2 and P(3):

[0236] P(4)=P(3)+M3(D(1)-F D P(3))

[0237] The attribute value covariance matrix ∑ is obtained based on the third attribute value. P(4) :

[0238] ∑ P(4) =(1-M3F D )∑ P(3) ;

[0239] The fourth fusion is performed based on the structural attribute state equation and the knowledge attribute state equation, i.e., when k=4, to obtain the fourth attribute value P(5):

[0240] When k=1, K(1) and F are obtained by solving the secondary index scores of knowledge attributes and the first and second weight operators of knowledge attributes. K

[0241] Calculation of Kalman gain M4 based on attribute state equations:

[0242]

[0243] The fourth attribute value P(5) is calculated based on the Kalman gain M4 and P(4):

[0244] P(5)=P(4)+M4(K(1)-F K P(4))

[0245] The attribute value covariance matrix Σ is obtained based on the second attribute value. P(5):

[0246] ∑ P(5) =(1-M4F K )Σ P(4) .

[0247] The interoperability level is obtained based on the fourth attribute value and the UAV system interoperability level model.

[0248] Preferably, the final interoperability level is obtained by searching the fourth attribute value P(5) in the UAV system interoperability level model.

[0249] Example 3

[0250] Based on expert experience, the secondary indicators of the attribute indicator tree for the evaluated UAV system are scored, and the corresponding primary and secondary weight operators are provided, as shown in Table 2:

[0251] Table 2 Scoring Table for Secondary Indicators

[0252]

[0253]

[0254] As shown in Table 1, the interoperability levels are quantified from 0 to 5, with integers representing the mean of the level. There are no clear boundaries between levels, so decimals in the scoring results represent uncertainty.

[0255] Determine the state equation and observation equation

[0256] Substituting the secondary indicator with the highest weight from the primary structural attribute indicator into the above formula, we get:

[0257]

[0258] Calculate the Kalman gain:

[0259]

[0260] Merge Update:

[0261]

[0262] Similarly, we can conclude that:

[0263]

[0264] Substituting these values ​​into the calculation of the Kalman gain, we obtain the following:

[0265]

[0266] and

[0267] The second attribute value P(3) is calculated based on the Kalman gain M2 and P(2):

[0268] The third attribute value P(4) is calculated based on the Kalman gain M3 and P(3):

[0269] The fourth attribute value P(5) is calculated based on the Kalman gain M4 and P(4):

[0270] The final P(5) vector consists of three elements, which have been weighted. The final evaluation result is the average: (4.35+3.12+3.22) / 3=3.5633, that is, the fourth attribute value P(5)=3.5633. Based on the unmanned system interoperability level model in Table 1, it can be seen that 3.5633 is between 3 (integration level) and 4 (cooperative level), and is close to 4. Therefore, it is judged to be cooperative level.

[0271] This invention first extends the LISI model by adding knowledge attributes, improving the accuracy of UAV interoperability assessment. Secondly, it introduces the classic Kalman filter algorithm for interoperability assessment of five attributes, proposing an attribute index tree. Finally, an example is provided, demonstrating how fusing the secondary index scores of the five attributes (P, A, I, D, K) of a UAV system to achieve more accurate estimates. Currently, the UAV system interoperability assessment device developed using this technology has been practically promoted and applied.

[0272] Example 4

[0273] like Figure 9 As shown, an interoperability evaluation system for unmanned aerial vehicle (UAV) systems based on Kalman filtering includes: a ranking model construction module, an index tree construction module, a scoring module, a state equation construction module, a cyclic fusion module, and a result output module.

[0274] The hierarchical model building module is used to construct an interoperability hierarchical model for unmanned aerial vehicle (UAV) systems, including structural attributes, application attributes, facility attributes, data attributes, and knowledge attributes.

[0275] The indicator tree construction module is used to construct secondary indicator trees based on the above attributes and merge them to obtain the overall attribute indicator tree.

[0276] The scoring module is used to score based on the overall attribute indicator tree, and obtain multiple secondary indicator scores and multiple weight operators for each attribute.

[0277] The state equation construction module is used to obtain the corresponding attribute state equation based on the corresponding secondary index scores and weight operators;

[0278] The iterative fusion module is used to fuse the structural attribute state equation and the application attribute state equation to obtain the first attribute value; to fuse the first attribute value and the facility attribute state equation to obtain the second attribute value; and to iteratively execute the above fusion process, successively fusing the data attribute state equation and the knowledge attribute state equation to finally obtain the fourth attribute value.

[0279] The results output module is used to obtain the interoperability level based on the fourth attribute value and the UAV system interoperability level model.

[0280] Example 5

[0281] Based on the same inventive concept, the present invention also provides a computer device, including a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus;

[0282] Memory, used to store computer programs;

[0283] When the processor executes a program stored in memory, it is able to implement a Kalman filter-based unmanned aerial vehicle system interoperability evaluation method as shown in Embodiment 1 or 2.

[0284] The electronic device may include a processor, a communications interface, a memory, and a communication bus, wherein the processor, communications interface, and memory communicate with each other via the communication bus. The processor can call logical instructions in the memory to execute a Kalman filter-based unmanned aerial vehicle system interoperability evaluation method as described in Embodiment 1 or 2.

[0285] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to the method section.

[0286] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for evaluating the interoperability of unmanned aerial vehicle (UAV) systems based on Kalman filtering, characterized in that, include: Construct an interoperability level model for unmanned aerial vehicle (UAV) systems that includes structural attributes, application attributes, facility attributes, data attributes, and knowledge attributes; Based on the above attributes, construct secondary indicator trees respectively and merge them to obtain the overall attribute indicator tree; Based on the above attributes, two secondary indicator trees are constructed respectively, resulting in structural attribute indicator trees, application attribute indicator trees, facility attribute indicator trees, data attribute indicator trees, and knowledge attribute indicator trees. The attribute indicator trees mentioned above all include: target attribute, multiple primary indicators, and multiple secondary indicators; Scoring is performed based on the overall attribute index tree to obtain multiple secondary index scores and multiple weight operators for each attribute; This yields multiple secondary indicator scores and multiple weight operators corresponding to each attribute, specifically including: Based on the expert experience of relevant professionals, scores are assigned to the secondary indicators of all attributes to obtain multiple secondary indicator scores for each attribute. Based on the expert experience of the relevant professionals, multiple first weight operators are given between the primary indicators of each attribute and the corresponding attributes, and multiple second weight operators are given between the secondary indicators of each attribute and the corresponding primary indicators. The weight operator is composed of the first weight operator and the second weight operator; The corresponding attribute state equation is obtained based on the corresponding secondary indicator score and the weight operator; The corresponding attribute state equations are obtained, specifically including: The corresponding attribute state matrix is ​​obtained based on the first weight operator and the second weight operator corresponding to each attribute; Based on the attribute state matrix and the corresponding secondary index scores, the corresponding attribute state quantity prediction equation is calculated. : ； Where k represents the state at the next moment, F X Represents the state matrix of attribute X. This represents a vector consisting of the scores of all secondary indicators of attribute X, where k-1 represents the current state, and P, A, I, D, and K represent structural attributes, application attributes, facility attributes, data attributes, and knowledge attributes, respectively. Based on the attribute state variable prediction equation X(k), the corresponding attribute covariance prediction equation is obtained: ； in, Let X represent the covariance matrix of attribute X at the previous time step, and T denote the transpose sign. The attribute state quantity prediction equation and the attribute covariance prediction equation together constitute the attribute state equation; The first attribute value is obtained by fusing the structural attribute state equation and the applied attribute state equation. The second attribute value is obtained by fusing the first attribute value and the facility attribute state equation. The above fusion process is repeated cyclically, successively fusing the data attribute state equation and the knowledge attribute state equation to finally obtain the fourth attribute value; The fusion process is as follows: Calculate the Kalman gain based on the attribute state equation. : ； Among them, F Y Y represents the state matrix, and R represents the overall noise. Calculate the nth attribute value based on the Kalman gain. : ； The attribute value covariance matrix is ​​obtained based on the nth attribute value. : ； Where n represents the number of fusions; Based on the fourth attribute value and the UAV system interoperability level model, the interoperability level is obtained.

2. The interoperability evaluation method for unmanned aerial vehicle systems based on Kalman filtering according to claim 1, characterized in that, The unmanned aerial vehicle system interoperability level model includes: The isolation level is 0, the connectivity level is 1, the functionality level is 2, the integration level is 3, the collaboration level is 4, and the general level is 5.

3. The interoperability evaluation method for unmanned aerial vehicle systems based on Kalman filtering according to claim 1, characterized in that, Each of the primary indicators corresponds to multiple secondary indicators.

4. The interoperability evaluation method for unmanned aerial vehicle systems based on Kalman filtering according to claim 3, characterized in that, The first-level indicators of the structural attribute indicator tree include: data structure, system structure, and management structure; The first-level indicators of the application attribute indicator tree include: interaction model, implementation operation, and application scale; The primary indicators of the facility attribute indicator tree include: facility controllability, protocol usability, and facility robustness; The primary indicators of the data attribute indicator tree include: simplex, half-duplex, and full-duplex. The primary indicators of the knowledge attribute indicator tree include: transmitted information, understood information, and predicted information.

5. The method for evaluating the interoperability of unmanned aerial vehicle (UAV) systems based on Kalman filtering according to claim 4, characterized in that, The secondary indicators of the data structure include: basic data standards, data service standards, and data management standards; The secondary indicators of the system architecture include: physical layer, data layer, and semantic layer; The secondary indicators of the management structure include: the standardization, universality, and integration of regulations and ordinances; The secondary metrics of the interaction model include: interaction latency, interaction method, and interaction direction; The secondary performance indicators for the operation include: endurance, flight altitude, and speed; The secondary indicators of the application scale include: number of systems, number of software, and number of hardware. The secondary indicators of facility controllability include: command controllability, management controllability, and collaboration controllability; The secondary indicators of the protocol's usability include: protocol architecture, supported scale, and service types; The secondary indicators of facility robustness include: link integrity, information redundancy, and resilience against destruction; The secondary indicators of the simplex communication include: communication confidentiality, communication timeliness, and information capacity. The secondary indicators of duplex and half-duplex include: instruction correctness, feedback timeliness, and data confidentiality; The secondary indicators of duplex communication include: data confidentiality, communication fluency, and information accessibility. The secondary indicators of the transmitted information include: images, video, and voice; The secondary indicators for understanding information include: knowledge network, environment model, and target estimation; The secondary indicators of the forecast information include: situation forecasting, model correction, and command decision-making.

6. The interoperability evaluation method for unmanned aerial vehicle systems based on Kalman filtering according to claim 1, characterized in that, The score of the second-level indicator corresponding to the largest second-level indicator of the first-level indicator of each attribute is used in the calculation of the attribute state equation.

7. A Kalman filter-based unmanned aerial vehicle (UAV) system interoperability evaluation system, applied to any one of the evaluation methods described in claims 1-6, characterized in that, include: The system includes a rating model construction module, an indicator tree construction module, a scoring module, a state equation construction module, a cyclic fusion module, and a result output module. The hierarchical model construction module is used to construct an interoperability hierarchical model for unmanned aerial vehicle systems, including structural attributes, application attributes, facility attributes, data attributes, and knowledge attributes. The indicator tree construction module is used to construct secondary indicator trees based on the above attributes and merge them to obtain the overall attribute indicator tree. The scoring module is used to score based on the overall attribute index tree to obtain multiple secondary index scores and multiple weight operators corresponding to each attribute. The state equation construction module is used to obtain the corresponding attribute state equation based on the corresponding secondary index score and the weight operator; The cyclic fusion module is used to fuse the structural attribute state equation and the application attribute state equation to obtain a first attribute value; fuse the first attribute value and the facility attribute state equation to obtain a second attribute value; and cyclically execute the above fusion process to fuse the data attribute state equation and the knowledge attribute state equation in turn to obtain a fourth attribute value. The result output module is used to obtain the interoperability level based on the fourth attribute value and the UAV system interoperability level model.

Images