Data-driven civil aircraft interior design expert self-confident modeling method and system
By constructing a two-dimensional information entropy calculation mechanism in the interior design of civil aircraft, the confidence level of experts is dynamically evaluated, which solves the problem that it is difficult to characterize the differences in expert judgment in traditional methods, and achieves high efficiency, stability and explanatory power of group consensus.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- NORTHWESTERN POLYTECHNICAL UNIV
- Filing Date
- 2026-04-03
- Publication Date
- 2026-05-01
AI Technical Summary
Existing technologies are insufficient to effectively characterize the differences in expert judgments across different scoring directions in civil aircraft interior design, resulting in insufficient convergence speed and stability of group consensus. Traditional methods that rely on external features or single-dimensional information entropy calculations lack objectivity and explanatory power.
By constructing a two-dimensional information entropy calculation mechanism based on expert rating data, information entropy is calculated on both the scheme dimension and the indicator dimension, and expert confidence is dynamically evaluated. The resulting two-dimensional information entropy is then input into the confidence assessment model to generate differentiated expert confidence parameters.
It enhances the explanatory power and discriminative power of expert confidence, dynamically adjusts expert opinions, improves the convergence efficiency and stability of group consensus, reduces data acquisition costs and avoids privacy risks, and is suitable for group decision-making scenarios with complex solutions and multiple indicators.
Smart Images
Figure CN121958746A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of group decision support and consensus modeling technology, and more specifically, to a data-driven method and system for modeling the confidence of civil aircraft interior design experts. Background Technology
[0002] In multi-expert group decision-making and consensus modeling problems, the confidence parameter is used to measure the degree to which experts retain their original views and is an important factor affecting the convergence speed and consistency of group consensus. Existing methods mainly fall into three categories: First, uniformly setting fixed parameters, which is simple to operate but lacks expression of individual behavioral differences. Parameter settings usually rely on experience or human adjustment and lack objective basis. Second, assigning values based on external characteristics such as job title and social influence, which introduces heterogeneous information but is highly dependent on external structural information, has high data acquisition costs, poses privacy risks, and has a weak correspondence with actual scoring behavior, making it difficult to adapt to scenarios lacking structured data. Third, building information entropy models based on scoring data, but these are mostly limited to single-dimensional calculations and cannot characterize the differences in experts' judgments in different scoring directions, resulting in insufficient parameter explanatory power and discrimination.
[0003] In complex engineering decision-making tasks such as civil aircraft interior design (civil aircraft cabin interior design), due to the complexity of the scheme structure, the large number of evaluation indicators, and the significant heterogeneity of expert scores, traditional methods are difficult to reveal the structural characteristics of scoring behavior, nor can they locate the source of disagreement and dimensional differences, thus limiting modeling capabilities. Therefore, there is an urgent need for an uncertainty measurement method based on scoring data and integrating scheme dimensional and indicator dimensional information to dynamically evaluate expert confidence parameters, realize behavior-driven individual difference modeling, and thus improve the convergence efficiency and stability of group consensus.
[0004] In view of this, the present invention proposes a data-driven method and system for modeling the confidence of civil aircraft interior design experts to solve the above problems. Summary of the Invention
[0005] To overcome the aforementioned shortcomings of the prior art and achieve the above objectives, the present invention provides the following technical solution: a data-driven method for modeling the confidence level of civil aircraft interior design experts, comprising: Step S1: Obtain Experts Candidate designs for civil aircraft interiors Initial scores under each evaluation metric, and constructing them respectively. The initial rating matrix corresponding to each expert; Step S2: On both the solution and indicator dimensions, respectively... The initial rating matrix of the experts was normalized to obtain... The solution dimension matrix and indicator dimension matrix corresponding to each expert; Step S3: Based on The solution dimension matrix corresponding to each expert is calculated separately for each solution dimension. The information entropy of the expert's solution dimension; Step S4: Based on The indicator dimension matrix corresponding to each expert is calculated separately for each indicator dimension. Information entropy of the indicator dimensions of each expert; Step S5: Combine the information entropy of each expert's solution dimension with the information entropy of the indicator dimension to form the two-dimensional information entropy of each expert, and then... The two-dimensional information entropy of each expert is input into a pre-built confidence assessment model to dynamically evaluate the expert confidence of each expert.
[0006] Furthermore, construct The methods for generating the initial rating matrix for each expert include: Let the expert be The candidate designs for civil aircraft interiors are The evaluation indicators are ;in, , , , , , All are integers greater than 1; organization The experts respectively commented on Candidate designs for civil aircraft interiors Each evaluation indicator is scored independently to obtain an initial score for each expert; based on the initial scores of each expert, a corresponding initial score matrix is constructed. .
[0007] Furthermore, to obtain The methods for using the solution dimension matrix corresponding to each expert include: right Each row of the initial rating matrix corresponding to each expert is normalized to obtain... The standard scores for each expert's proposed solutions are then aggregated to obtain the solution dimension matrix for each expert. The expression for the standard score is as follows: ; In the formula, Experts Candidate designs for civil aircraft interiors In evaluation indicators The following are the standard scoring criteria for the proposed solutions. Experts Candidate designs for civil aircraft interiors In evaluation indicators The initial score is given below.
[0008] Furthermore, to obtain The methods for using an indicator dimension matrix corresponding to each expert include: right Each column of the initial rating matrix corresponding to each expert is normalized to obtain... The indicator standard scores for each expert are summarized to obtain the indicator dimension matrix for each expert; the expression for the indicator standard score is as follows: ; In the formula, Experts Candidate designs for civil aircraft interiors In evaluation indicators The following are the scoring criteria for indicators.
[0009] Furthermore, calculations are performed separately at the scheme dimension. The methods for calculating the dimensional information entropy of an expert's solution include: Based on the solution dimension matrix corresponding to each expert, the entropy of the scoring row vector for each expert is calculated; the mean of the scoring row vector entropy for the same expert is then calculated to obtain the solution dimension information entropy for each expert; the expression for the scoring row vector entropy is: ; In the formula, Represents the entropy of the rating row vector, when When, define .
[0010] Furthermore, calculations are performed separately along the indicator dimensions. The methods for using the information entropy of an expert's indicator dimensions include: Based on the indicator dimension matrix corresponding to each expert, the entropy of the rating column vector for each expert is calculated; the mean of the entropy of the rating column vectors for the same expert is calculated to obtain the indicator dimension information entropy for each expert; the expression for the rating column vector entropy is: ; In the formula, Represents the entropy of the rating column vector, when When, define .
[0011] Furthermore, methods for generating the two-dimensional information entropy for each expert include: The maximum theoretical entropy values are calculated separately for the scheme dimension and the indicator dimension. Based on the maximum theoretical entropy values, the information entropy of the scheme dimension and the information entropy of the indicator dimension for each expert are normalized to obtain the standard information entropy of the scheme and the standard information entropy of the indicator for each expert. The standard information entropy of the scheme and the standard information entropy of the indicator for the same expert are integrated to form the two-dimensional information entropy for each expert.
[0012] Furthermore, the expressions for the standard information entropy of the scheme and the standard information entropy of the indicator are as follows: , ; In the formula, Experts The standard information entropy of the scheme, Experts The indicator standard information entropy, Indicating individual civil aircraft interior design candidate schemes in The maximum theoretical entropy value of the score distribution on each evaluation index. Indicates a single evaluation indicator in The maximum theoretical entropy value of the score distribution on the candidate interior design schemes for civil aircraft.
[0013] Furthermore, methods for dynamically assessing each expert's expert confidence include: A confidence assessment model is constructed, in which the two-dimensional information entropy of each expert is input into the model to dynamically evaluate the expert confidence level of each expert; the expression of the confidence assessment model is as follows: In the formula, Experts The confidence of experts This is the weighting adjustment coefficient. .
[0014] A data-driven modeling system for expert confidence in civil aircraft interior design, implementing the aforementioned data-driven modeling method for expert confidence in civil aircraft interior design, includes: The rating collection module is used to obtain... Experts Candidate designs for civil aircraft interiors Initial scores under each evaluation metric, and constructing them respectively. The initial rating matrix corresponding to each expert; The dual-dimensional normalization module is used to perform analysis on both the solution dimension and the indicator dimension. The initial rating matrix of the experts was normalized to obtain... The solution dimension matrix and indicator dimension matrix corresponding to each expert; The scheme entropy module is used to base on... The solution dimension matrix corresponding to each expert is calculated separately for each solution dimension. The information entropy of the expert's solution dimension; The index entropy module is used to base on... The indicator dimension matrix corresponding to each expert is calculated separately for each indicator dimension. Information entropy of the indicator dimensions of each expert; The confidence assessment module combines the information entropy of each expert's plan dimension with the information entropy of the indicator dimension to form a two-dimensional information entropy for each expert, and then... The two-dimensional information entropy of each expert is input into a pre-built confidence assessment model to dynamically evaluate the expert confidence of each expert.
[0015] The technical effects and advantages of the data-driven modeling method and system for expert confidence in civil aircraft interior design of this invention are as follows: This data-driven approach, based entirely on expert rating data, eliminates reliance on external characteristics such as expert job level, social influence, and social network structure. It avoids the subjectivity and arbitrariness of manually set parameters in traditional methods, thus reducing data acquisition costs and effectively mitigating privacy risks. Compared to existing methods that only employ a single dimension of information entropy, this approach, by simultaneously constructing a two-way information entropy calculation mechanism encompassing both the scheme and indicator dimensions, can distinguish the specific sources of expert confidence, identify structural biases in the rating distribution, and accurately characterize the differences in expert judgment across different rating directions, thereby enhancing the explanatory power and discriminative power of expert confidence. Furthermore, it can tailor the calculation to each expert's actual situation. The scoring behavior dynamically generates differentiated confidence parameters, assigning higher expert confidence to experts with concentrated score distributions to maintain the stability of their original opinions, and assigning lower expert confidence to experts with dispersed score distributions to accelerate their opinion adjustments. This effectively overcomes the limitation of traditional fixed-parameter methods that cannot reflect the differences in individual expert behavioral characteristics. In addition, it has good versatility and scalability, and can be embedded in DeGroot models, Friedkin-Johnsen models, or other group opinion dynamics models. It is suitable for various group decision-making scenarios with complex scheme structures, numerous evaluation indicators, and significant heterogeneity in expert score distribution, and has broad application value. Attached Figure Description
[0016] Figure 1 This is a schematic diagram of the data-driven civil aircraft interior design expert confidence modeling system according to Embodiment 1 of the present invention; Figure 2 This is a schematic diagram comparing the convergence speed of the models in Embodiment 1 of the present invention; Figure 3 This is a flowchart of the data-driven civil aircraft interior design expert confidence modeling method according to Embodiment 2 of the present invention. Detailed Implementation
[0017] 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. Example 1:
[0018] Please see Figure 1 As shown in this embodiment, the data-driven modeling method for expert confidence in civil aircraft interior design includes: Step S1: Obtain Experts Candidate designs for civil aircraft interiors Initial scores under each evaluation metric, and constructing them respectively. The initial rating matrix corresponding to each expert.
[0019] Build The methods for generating the initial rating matrix for each expert include: Let the expert be The candidate designs for civil aircraft interiors are The evaluation indicators are ;in, , , , , , All are integers greater than 1; organization The experts respectively commented on Candidate designs for civil aircraft interiors Each evaluation indicator is scored independently to obtain an initial score for each expert. Each evaluation indicator is pre-set by experts in accordance with the evaluation requirements of civil aircraft interior design, such as the rationality of attitude assumptions, the matching of line-of-sight information, the coherence of structural generation, and semantic recognition. The initial score is a quantitative evaluation score given by experts to each candidate civil aircraft interior design scheme under each evaluation indicator based on their professional knowledge and experience. An initial score matrix is constructed based on each expert's initial score. ; where the initial rating matrix The specific expression is: ; In the formula, the element Experts Candidate designs for civil aircraft interiors In evaluation indicators The initial score is given below.
[0020] Step S2: On both the solution and indicator dimensions, respectively... The initial rating matrix of the experts was normalized to obtain... The solution dimension matrix and indicator dimension matrix corresponding to each expert.
[0021] get The methods for using the solution dimension matrix corresponding to each expert include: right Each row of the initial rating matrix corresponding to each expert is normalized to obtain... The standard scores for each expert's proposed solution are used to ensure that the sum of the standard scores for each candidate civil aircraft interior design solution under different evaluation indicators is 1. The standard scores for solutions from the same expert are then aggregated to obtain a solution dimension matrix for each expert. The standard score represents... Experts Candidate designs for civil aircraft interiors The initial scores under each evaluation indicator are the scores obtained after normalization based on the scheme dimension; the specific expression for the standard scheme score is: ; In the formula, Experts Candidate designs for civil aircraft interiors In evaluation indicators The following are the standard scoring criteria for the proposed solutions. Experts Candidate designs for civil aircraft interiors In evaluation indicators The initial score is given below.
[0022] get The methods for using an indicator dimension matrix corresponding to each expert include: right Each column of the initial rating matrix corresponding to each expert is normalized to obtain... The indicator standard scores for each expert are determined to ensure that the sum of the indicator standard scores for each evaluation indicator under different civil aircraft interior design candidate schemes is 1; the indicator standard scores for the same expert are aggregated to obtain the indicator dimension matrix for each expert; where the indicator standard score represents Experts Candidate designs for civil aircraft interiors The initial scores under each evaluation indicator are the scores obtained after normalization of the indicator dimensions; the specific expressions for the standard scores of the indicators are: ; In the formula, Experts Candidate designs for civil aircraft interiors In evaluation indicators The following are the scoring criteria for indicators.
[0023] Step S3: Based on The solution dimension matrix corresponding to each expert is calculated separately for each solution dimension. The information entropy of the expert's solution dimension.
[0024] Calculate separately at the scheme dimension The methods for calculating the dimensional information entropy of an expert's solution include: Based on the scheme dimension matrix corresponding to each expert, the scoring row vector entropy corresponding to each expert is calculated; the mean of the scoring row vector entropy corresponding to the same expert is calculated to obtain the scheme dimension information entropy corresponding to each expert; where the scoring row vector entropy represents the information entropy of each expert under different civil aircraft interior design candidate schemes, and is used to quantify the uncertainty of the scheme standard score in the scheme dimension; the scheme dimension information entropy is used to comprehensively characterize the overall uncertainty of the experts' opinions in the scheme dimension. Specifically, the expression for the entropy of the rating row vector is: ; In the formula, Represents the entropy of the rating row vector, when When, define To avoid numerical instability; the scoring row vector entropy is used to measure whether the standard scores of experts on a single civil aircraft interior design candidate scheme are concentrated. The larger the scoring row vector entropy, the more dispersed the standard scores of the corresponding civil aircraft interior design candidate scheme, and the higher the uncertainty of opinions.
[0025] The specific expression for the information entropy of the solution dimension is: ; In the formula, This represents the entropy of the scheme's dimensional information.
[0026] Step S4: Based on The indicator dimension matrix corresponding to each expert is calculated separately for each indicator dimension. Information entropy of the indicator dimensions of each expert.
[0027] Calculate separately along the indicator dimensions The methods for using the information entropy of an expert's indicator dimensions include: Based on the indicator dimension matrix corresponding to each expert, the entropy of the scoring column vector for each expert is calculated; the mean of the scoring column vector entropy for the same expert is calculated to obtain the indicator dimension information entropy for each expert; where the scoring column vector entropy represents the information entropy of each expert under different evaluation indicators, and is used to quantify the uncertainty of the indicator standard score in the indicator dimension; the indicator dimension information entropy is used to comprehensively characterize the overall uncertainty of the experts' opinions in the indicator dimension. Specifically, the expression for the entropy of the rating column vector is: ; In the formula, Represents the entropy of the rating column vector, when When, define To avoid numerical instability; the scoring column vector entropy is used to characterize the consistency of the indicator standard scores of experts for different civil aircraft interior design candidate schemes under the same evaluation indicator. The larger the scoring column vector entropy, the more dispersed the indicator standard scores under the corresponding evaluation indicator, and the higher the uncertainty of opinions.
[0028] The specific expression for the information entropy of the indicator dimension is: ; In the formula, This represents the entropy of the indicator dimension.
[0029] Step S5: Combine the information entropy of each expert's solution dimension with the information entropy of the indicator dimension to form the two-dimensional information entropy of each expert, and then... The two-dimensional information entropy of each expert is input into a pre-built confidence assessment model to dynamically evaluate the expert confidence of each expert.
[0030] Methods for generating the two-dimensional information entropy for each expert include: The maximum theoretical entropy values are calculated separately for the scheme dimension and the indicator dimension. Based on the maximum theoretical entropy values, the information entropy of the scheme dimension and the indicator dimension for each expert are normalized to obtain the standard information entropy of the scheme and the standard information entropy of the indicator for each expert. The standard information entropy of the scheme and the standard information entropy of the indicator for the same expert are integrated to form the two-dimensional information entropy for each expert. The specific expressions for the standard information entropy of the scheme and the standard information entropy of the indicator are as follows: , ; In the formula, Experts The standard information entropy of the scheme, Experts The indicator standard information entropy, Indicating individual civil aircraft interior design candidate schemes in The maximum theoretical entropy value of the score distribution on each evaluation index. Indicates a single evaluation indicator in The maximum theoretical entropy value of the score distribution on the candidate interior design schemes for civil aircraft.
[0031] Methods for dynamically assessing each expert's expert confidence include: A confidence assessment model is constructed, in which the two-dimensional information entropy of each expert is input into the model to dynamically evaluate the expert confidence level of each expert. The specific expression of the confidence assessment model is as follows: In the formula, Experts The confidence of experts This is a weighting adjustment coefficient used to balance the influence of the scheme's standard information entropy and the indicator's standard information entropy. The specific values can be preset based on factors such as task characteristics, indicator importance, and data distribution characteristics; expert confidence directly reflects the stability and uncertainty level of experts' scores in the scheme dimension and indicator dimension; the higher the expert confidence, the stronger the experts maintain their original opinions during the consensus evolution process; the lower the expert confidence, the more easily the experts are influenced by the opinions of the group.
[0032] It should be understood that expert confidence is calculated entirely based on expert scoring behavior, without relying on social network structure, identity characteristics, or external auxiliary data. It is characterized by data-driven approach, strong universality, and high computational efficiency. Furthermore, expert confidence can be embedded in Friedkin-Johnsen models, DeGroot models, or other consensus evolution methods to regulate the updating and convergence process of group opinions. Compared to existing technologies that uniformly set fixed parameters, assign values based on external features, or calculate based on a single-dimensional information entropy, this embodiment, by constructing a two-way information entropy mechanism for both the scheme dimension and the indicator dimension, can dynamically generate differentiated expert confidence based on scoring data. This effectively characterizes the differences in expert judgments across different scoring directions, improves the objectivity, explanatory power, and discriminative power of expert confidence, and simultaneously enhances the convergence efficiency and stability of group consensus. For example, taking the evaluation of civil aircraft interior design schemes as an application scenario, the application scenario specifically includes 3 candidate civil aircraft interior design schemes, 9 evaluation indicators, and 5 experts; among them, the 9 evaluation indicators are visual functional zoning. Interactive layout coordination Reasonableness of attitude assumptions Eye contact information matching Structural generation coherence Semantic recognizability Clear spatial hierarchy The screen is simple and free of redundancy. , closedness of the solution expression The initial scoring matrix of the five experts is shown in Table 1-1. Table 1-1 Initial rating matrix of 5 experts
[0033] In terms of both the scheme dimension and the indicator dimension, the initial scoring matrices of the 5 experts were normalized to obtain the scheme dimension matrix and indicator dimension matrix corresponding to the 5 experts; the scheme dimension matrix of the 5 experts is shown in Table 1-2, and the indicator dimension matrix of the 5 experts is shown in Table 1-3. Table 1-2 Dimensional Matrix of Solutions from 5 Experts
[0034] Table 1-3 Index Dimension Matrix of 5 Experts
[0035] Based on the solution dimension matrix corresponding to the 5 experts, the entropy of the scoring row vector corresponding to the 5 experts is calculated; the entropy of the scoring row vector of the 5 experts is shown in Table 1-4. Table 1-4 Row vector entropy of scores from 5 experts
[0036] For the five experts, the mean of the row vector entropy of the scoring for each expert was calculated to obtain the information entropy of the scheme dimension for each expert. The information entropies of the scheme dimension for the five experts were 3.1360, 3.1519, 3.1493, 3.0839, and 3.1684, respectively. Among them, the three initial scores of expert 5 under the scheme dimension showed a highly consistent distribution characteristic, which became an equal probability distribution after normalization, resulting in the same row vector entropy of their scores on the three candidate schemes of civil aircraft interior design. This indicates that expert 5 lacked obvious preference expression in the scheme evaluation process, their initial score distribution tended to be balanced, their opinion uncertainty was high, and their expert confidence was correspondingly low. In addition, expert 4 had the lowest information entropy of the scheme dimension, indicating that their initial scores for the same candidate scheme of civil aircraft interior design were more concentrated under each evaluation index, and their opinion expression was relatively stable.
[0037] Based on the indicator dimension matrix corresponding to the 5 experts, the entropy of the rating column vector corresponding to the 5 experts is calculated; the entropy of the rating column vector of the 5 experts is shown in Table 1-5.
[0038] Table 1-5 Entropy of the column vector of ratings from 5 experts For the five experts, the mean of the column vector entropy of the scores for each expert was calculated to obtain the information entropy of the indicator dimension for each expert. The information entropies of the indicator dimensions for the five experts were 1.5273, 1.5660, 1.5630, 1.5112, and 1.5841, respectively. Among them, expert 4 had the lowest information entropy of the indicator dimension, indicating that his initial scores for the indicator dimension were relatively consistent. On the other hand, expert 5 had the highest information entropy of the indicator dimension, indicating that his initial scores for the same evaluation indicator varied greatly under different candidate schemes for civil aircraft interior design, and his opinions were not stable enough.
[0039] The maximum theoretical entropy values were calculated for both the scheme dimension and the indicator dimension. Based on the maximum theoretical entropy values, the information entropy values for the scheme dimension and the indicator dimension of the five experts were normalized to obtain the standard information entropy values for the scheme and the indicator for the five experts. The standard information entropy values for the scheme of the five experts are shown in Table 1-6, and the standard information entropy values for the indicator of the five experts are shown in Table 1-7. Table 1-6 Information Entropy of the Scheme Standards of 5 Experts
[0040] Table 1-7 Information Entropy of Indicator Standards for 5 Experts
[0041] The standard information entropy of the proposed solutions and the standard information entropy of the indicators from the five experts are respectively input into a pre-constructed confidence assessment model to dynamically evaluate the expert confidence level of each expert; among them, The value was set to 0.5; the final expert confidence scores for the five experts were 0.02353, 0.00882, 0.01017, 0.03683, and 0.00051, respectively.
[0042] Based on the standard information entropy of the proposed solutions and the standard information entropy of the indicators from the five experts, it can be concluded that: Expert 4 has the lowest information entropy values in both dimensions, indicating that the initial score distribution at both the proposed solution and indicator levels is relatively concentrated, and the opinions expressed are highly consistent, reflecting strong opinion stability and confidence. In contrast, Expert 5's information entropy values in both dimensions are close to the theoretical maximum, indicating that the initial score distribution in both dimensions shows high uncertainty, the opinions expressed are relatively scattered, and the level of confidence is significantly low. The information entropy values of Experts 1 to 3 in both dimensions are in the middle range, showing a certain degree of difference in score distribution, and their expert confidence is between that of Experts 4 and 5.
[0043] Based on the expert confidence levels of the five experts, a diagonal matrix of expert confidence is constructed. A trust weight matrix among experts is also constructed, and both the expert confidence diagonal matrix and the trust weight matrix are introduced into the Friedkin-Johnsen model for consensus iteration and updating. The expert confidence diagonal matrix... The trust weight matrix among experts is shown in Table 1-8; the model converged after 21 iterations, and the consensus degree of each iteration is shown in Table 1-9; the final ranking of the three civil aircraft interior design candidate schemes is as follows: ; Table 1-8 Expert Trust Weight Matrix
[0044] Table 1-9 Consensus Level in the 21-Round Iteration Process
[0045] At the same time, the traditional Friedkin-Johnsen model (fixed) is adopted. A comparative experiment was conducted under the same initial conditions; at this point, the model converged in the 14th round, and the final ranking of the three candidate interior design schemes for civil aircraft remained as follows. However, the consensus level has decreased significantly. The consensus level of each iteration process is shown in Table 1-10. Table 1-10 Consensus Degree in 14 Iteration Rounds
[0046] It should be understood that the convergence speed of the models in the two comparative experiments is, for example, different. Figure 2 As shown, by dynamically evaluating the expert confidence of each expert, the consensus level increases rapidly in the early stage of model iteration, reaching a high consensus level (above 0.95) in a short period of time, demonstrating a strong ability to integrate opinions. Although the number of rounds required for the model to reach the strict convergence threshold is slightly higher than that of the traditional Friedkin-Johnsen model due to the differentiated setting of expert confidence, its final consensus level is significantly improved (about 0.97), which is significantly better than the traditional Friedkin-Johnsen model (about 0.81).
[0047] This embodiment relies on data-driven calculations, based entirely on expert rating data. It eliminates the need for external characteristics such as expert job level, social influence, or social network structure, avoiding the subjectivity and arbitrariness of manually set parameters in traditional methods. This reduces data acquisition costs and effectively mitigates privacy risks. Compared to existing methods that only use a single dimension of information entropy, this embodiment, by simultaneously constructing a two-way information entropy calculation mechanism for both the scheme and indicator dimensions, can distinguish the specific sources of expert confidence, identify structural biases in the rating distribution, and accurately characterize the differences in expert judgments across different rating directions, thereby enhancing the explanatory power and discriminative power of expert confidence. Furthermore, it can analyze each expert's... The system dynamically generates differentiated confidence parameters based on actual scoring behavior. Experts with concentrated scoring distributions are assigned higher confidence levels to maintain the stability of their original opinions, while those with dispersed scoring distributions are assigned lower confidence levels to accelerate their opinion adjustments. This effectively overcomes the limitation of traditional fixed-parameter methods that cannot reflect the differences in individual expert behavior characteristics. Furthermore, it has good versatility and scalability, and can be embedded in DeGroot models, Friedkin-Johnsen models, or other group opinion dynamics models. It is suitable for various group decision-making scenarios with complex scheme structures, numerous evaluation indicators, and significant heterogeneity in expert scoring distributions, and has broad application value. Example 2:
[0048] Please see Figure 3 As shown in the figure, the parts not described in detail in this embodiment are described in Embodiment 1. A data-driven civil aircraft interior design expert confidence modeling system is provided, including a scoring acquisition module, a two-dimensional normalization module, a scheme entropy value module, an indicator entropy value module, and a confidence assessment module. Each module is connected through wired and / or wireless means to realize data transmission between modules.
[0049] The rating collection module is used to obtain... Experts Candidate designs for civil aircraft interiors Initial scores under each evaluation metric, and constructing them respectively. The initial rating matrix corresponding to each expert; The dual-dimensional normalization module is used to perform analysis on both the solution dimension and the indicator dimension. The initial rating matrix of the experts was normalized to obtain... The solution dimension matrix and indicator dimension matrix corresponding to each expert; The scheme entropy module is used to base on... The solution dimension matrix corresponding to each expert is calculated separately for each solution dimension. The information entropy of the expert's solution dimension; The index entropy module is used to base on... The indicator dimension matrix corresponding to each expert is calculated separately for each indicator dimension. Information entropy of the indicator dimensions of each expert; The confidence assessment module combines the information entropy of each expert's plan dimension with the information entropy of the indicator dimension to form a two-dimensional information entropy for each expert, and then... The two-dimensional information entropy of each expert is input into a pre-built confidence assessment model to dynamically evaluate the expert confidence of each expert. Example 3:
[0050] This application also provides an electronic device. The electronic device may include one or more processors and one or more memories. The memories store computer-readable code that, when executed by the one or more processors, can perform the data-driven confidence modeling method for civil aircraft interior design experts as described above.
[0051] The method or system according to the embodiments of this application can also be implemented using the architecture of the electronic device shown in this application. The electronic device may include a bus, one or more CPUs, ROM, RAM, a communication port connected to a network, input / output, a hard disk, etc. The storage device in the electronic device, such as a ROM or hard disk, may store the data-driven confidence modeling method for civil aircraft interior design experts provided in this application. Furthermore, the electronic device may also include a user interface. Of course, the architecture shown in this application is merely exemplary; when implementing different devices, one or more components in the electronic device shown in this application may be omitted according to actual needs. Example 4:
[0052] One embodiment of this application discloses a computer-readable storage medium. The computer-readable storage medium stores computer-readable instructions. When the computer-readable instructions are executed by a processor, a data-driven civil aircraft interior design expert confidence modeling method according to an embodiment of this application, as described with reference to the above figures, can be executed. The storage medium includes, but is not limited to, volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and cache memory. Non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc.
[0053] Furthermore, according to embodiments of this application, the processes described in the above-referenced flowcharts can be implemented as computer software programs. For example, this application provides a non-transitory machine-readable storage medium storing machine-readable instructions that can be executed by a processor to perform instructions corresponding to the method steps provided in this application, such as a data-driven method for modeling the confidence level of civil aircraft interior design experts. When this computer program is executed by a central processing unit (CPU), it performs the functions defined in the method of this application.
[0054] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
[0055] All formulas in this manual are dimensionless and calculated numerically. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters and thresholds in the formulas are set by those skilled in the art according to the actual situation.
[0056] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.
Claims
1. A data-driven method for modeling the confidence level of civil aircraft interior design experts, characterized in that, include: Step S1: Obtain Experts Candidate designs for civil aircraft interiors Initial scores under each evaluation metric, and constructing them respectively. The initial rating matrix corresponding to each expert; Step S2: On both the solution and indicator dimensions, respectively... The initial rating matrix of the experts was normalized to obtain... The solution dimension matrix and indicator dimension matrix corresponding to each expert; Step S3: Based on The solution dimension matrix corresponding to each expert is calculated separately for each solution dimension. The information entropy of the expert's solution dimension; Step S4: Based on The indicator dimension matrix corresponding to each expert is calculated separately for each indicator dimension. Information entropy of the indicator dimensions of each expert; Step S5: Combine the information entropy of each expert's solution dimension with the information entropy of the indicator dimension to form the two-dimensional information entropy of each expert, and then... The two-dimensional information entropy of each expert is input into a pre-built confidence assessment model to dynamically evaluate the expert confidence of each expert.
2. The data-driven method for modeling the confidence level of civil aircraft interior design experts according to claim 1, characterized in that, Build The methods for generating the initial rating matrix for each expert include: Let the expert be The candidate designs for civil aircraft interiors are The evaluation indicators are ;in, , , , , , All are integers greater than 1; organization The experts respectively commented on Candidate designs for civil aircraft interiors Each evaluation indicator is scored independently to obtain an initial score for each expert; based on the initial scores of each expert, a corresponding initial score matrix is constructed. .
3. The data-driven method for modeling the confidence level of civil aircraft interior design experts according to claim 2, characterized in that, get The methods for using the solution dimension matrix corresponding to each expert include: right Each row of the initial rating matrix corresponding to each expert is normalized to obtain... The standard scores for each expert's proposed solutions are then aggregated to obtain the solution dimension matrix for each expert. The expression for the standard score is as follows: ； In the formula, Experts Candidate designs for civil aircraft interiors In evaluation indicators The following are the standard scoring criteria for the proposed solutions. Experts Candidate designs for civil aircraft interiors In evaluation indicators The initial score is given below.
4. The data-driven method for modeling the confidence level of civil aircraft interior design experts according to claim 3, characterized in that, get The methods for using an indicator dimension matrix corresponding to each expert include: right Each column of the initial rating matrix corresponding to each expert is normalized to obtain... The indicator standard scores for each expert are summarized to obtain the indicator dimension matrix for each expert; the expression for the indicator standard score is as follows: ； In the formula, Experts Candidate designs for civil aircraft interiors In evaluation indicators The following are the scoring criteria for indicators.
5. The data-driven modeling method for expert confidence in civil aircraft interior design according to claim 4, characterized in that, Calculate separately at the scheme dimension The methods for calculating the dimensional information entropy of an expert's solution include: Based on the solution dimension matrix corresponding to each expert, the entropy of the scoring row vector for each expert is calculated; the mean of the scoring row vector entropy for the same expert is then calculated to obtain the solution dimension information entropy for each expert; the expression for the scoring row vector entropy is: ； In the formula, Represents the entropy of the rating row vector, when When, define .
6. The data-driven modeling method for expert confidence in civil aircraft interior design according to claim 5, characterized in that, Calculate separately along the indicator dimensions The methods for using the information entropy of an expert's indicator dimensions include: Based on the indicator dimension matrix corresponding to each expert, the entropy of the rating column vector for each expert is calculated; the mean of the entropy of the rating column vectors for the same expert is calculated to obtain the indicator dimension information entropy for each expert; the expression for the rating column vector entropy is: ； In the formula, Represents the entropy of the rating column vector, when When, define .
7. The data-driven method for modeling the confidence level of civil aircraft interior design experts according to claim 6, characterized in that, Methods for generating the two-dimensional information entropy for each expert include: The maximum theoretical entropy values are calculated separately for the scheme dimension and the indicator dimension. Based on the maximum theoretical entropy values, the information entropy of the scheme dimension and the information entropy of the indicator dimension for each expert are normalized to obtain the standard information entropy of the scheme and the standard information entropy of the indicator for each expert. The standard information entropy of the scheme and the standard information entropy of the indicator for the same expert are integrated to form the two-dimensional information entropy for each expert.
8. The data-driven method for modeling the confidence level of civil aircraft interior design experts according to claim 7, characterized in that, The expressions for the standard information entropy of the scheme and the standard information entropy of the indicator are as follows: ， ； In the formula, Experts The standard information entropy of the scheme, Experts The indicator standard information entropy, Indicating individual civil aircraft interior design candidate schemes in The maximum theoretical entropy value of the score distribution on each evaluation index. Indicates a single evaluation indicator in The maximum theoretical entropy value of the score distribution on the candidate interior design schemes for civil aircraft.
9. The data-driven method for modeling expert confidence in civil aircraft interior design according to claim 8, characterized in that, Methods for dynamically assessing each expert's expert confidence include: A confidence assessment model is constructed, in which the two-dimensional information entropy of each expert is input into the model to dynamically evaluate the expert confidence level of each expert; the expression of the confidence assessment model is as follows: In the formula, Experts The confidence of experts This is the weighting adjustment coefficient. .
10. A data-driven expert confidence modeling system for civil aircraft interior design, implementing the data-driven expert confidence modeling method for civil aircraft interior design as described in any one of claims 1-9, characterized in that, include: The rating collection module is used to obtain... Experts Candidate designs for civil aircraft interiors Initial scores under each evaluation metric, and constructing them respectively. The initial rating matrix corresponding to each expert; The dual-dimensional normalization module is used to perform analysis on both the solution dimension and the indicator dimension. The initial rating matrix of the experts was normalized to obtain... The solution dimension matrix and indicator dimension matrix corresponding to each expert; The scheme entropy module is used to base on... The solution dimension matrix corresponding to each expert is calculated separately for each solution dimension. The information entropy of the expert's solution dimension; The index entropy module is used to base on... The indicator dimension matrix corresponding to each expert is calculated separately for each indicator dimension. Information entropy of the indicator dimensions of each expert; The confidence assessment module combines the information entropy of each expert's plan dimension with the information entropy of the indicator dimension to form a two-dimensional information entropy for each expert, and then... The two-dimensional information entropy of each expert is input into a pre-built confidence assessment model to dynamically evaluate the expert confidence of each expert.