Deconstruction optimization method for design characteristics of aircraft cockpit

Abstract

The invention discloses an aircraft cockpit design feature deconstruction optimization method, which comprises the following steps of: dividing an aircraft cockpit man-machine interaction system into a plurality of feature domains and feature elements, collecting eye movement data of a pilot about the feature domains and subjective evaluation data about feedback information of the feature elements in a flight task, and calculating visual information amount of the feature elements; the method comprises the following steps: deconstructing design features of an aircraft cockpit, constructing clustering samples of a system layer, an attribute layer and a feature layer, calculating the average weight of each layer in the clustering samples, and calculating the perceptual quantity of each design feature of the clustering samples based on the average weight of each layer of the clustering samples; constructing a mapping model between the visual information quantity of the feature elements and the perceptual quantity of the design features; and based on a mapping result of the mapping model, establishing a design feature deconstruction optimization normal form. According to the method, the design characteristics of the human-computer interaction interface in the cockpit are analyzed and defined, and the dynamic complex flight environment is optimized, so that the adaptability and the intelligent level of the cockpit layout are improved, the perceptual requirements of pilots are met, the operation efficiency and the safety of the pilots in the complex environment are improved, and the method is suitable for popularization and application. And a foundation is laid for intelligent cockpit design and man-machine co-driving system development in the future.

Description

Technical Field

[0001] The present invention relates to the technical field of design optimization of cockpit systems, and in particular, to a method for deconstructing and optimizing design features of an aircraft cockpit. Background Art

[0002] Modern cockpit design is gradually getting rid of the passive adaptation mode of "human adapting to machine" and moving towards the intelligent and user-friendly design direction of "machine adapting to human". In this process, the airworthiness requirements in the perceptual dimension of the cockpit have been increasingly emphasized. The design of the modern cockpit human-machine interaction system not only needs to meet the basic functional requirements, but also needs to fully consider the habits and preferences of pilots to improve their comfort, perceptual feedback and emotional needs during the flight. In addition, the flight process involves a variety of dynamic and complex environments, including factors such as meteorological changes, airspace traffic, system failures and emergencies, which put higher requirements on the cognitive load and decision-making ability of pilots. The interaction between the pilot and the cockpit is a dynamically evolving process over time series in different flight phases (taxi - takeoff - climb - cruise - descent - approach - landing - taxi) and different scenarios (high workload, low workload, emergency). High workload scenarios are very important, however, low workload scenarios are not foolproof either. The cruise flight phase will increase effects such as boredom and fatigue of pilots, resulting in a decline in the alert level and situational awareness. Facing the dynamic and complex environment, the interaction touchpoints and focuses between the pilot and the cockpit are changing, affected by many factors comprehensively, and the interaction mechanism is complex. Therefore, how to consider the combined effects of various factors in cockpit design, face the complex dynamic environment, and explore the characterization mechanism of cockpit design attributes that integrates overall and local features based on the research of the pilot's interaction cognitive model is the key to optimizing the design of the aircraft cockpit.

[0003] Therefore, in order to better optimize the design of the aircraft cockpit, it is necessary to construct a method for deconstructing the design features of the aircraft cockpit human-machine interface. Deconstructing the design features of the cockpit is the basis for optimizing the design of the aircraft cockpit. Only by scientifically and systematically deconstructing the design features of the cockpit can effective design guidance be provided for cockpit design, thereby optimizing the overall interaction experience of the cockpit.

[0004] Currently, the deconstruction method of cockpit design features mainly relies on the subjective definition of experts and lacks a systematic parametric analysis method. This deficiency restricts the standardization and practicality of cockpit perceptual interaction design. At the same time, existing research mostly relies on the cockpit interior presented in the form of pictures, and it is difficult to restore the real flight process. This is the optimal method under the condition of non-comprehensive simulation considering experimental cost control, but there is still a gap from the perception evoked in the real flight state.

[0005] During flight, pilots need to process complex environmental variables within a short period of time, such as meteorological changes, airspace traffic, system failures, and emergencies. In such cases, simply relying on traditional static feature deconstruction methods is difficult to meet the actual operation requirements. Therefore, studying a parametric cockpit design feature deconstruction method that combines dynamic complex environments has important theoretical and practical value for optimizing the cockpit's perceptual interaction design. Summary of the Invention

[0006] Objective of the Invention: The objective of the present invention is to provide an optimization of the design feature deconstruction of an aircraft cockpit. By analyzing and defining the design features of the human-machine interaction interface in the cockpit, it is optimized for the dynamic complex flight environment, thereby enhancing the adaptability and intelligence level of the cockpit layout.

[0007] Technical Solution: To achieve the above objective, a method for optimizing the design feature deconstruction of an aircraft cockpit according to the present invention includes the following steps:

[0008] S1: Divide the human-machine interaction system of the aircraft cockpit into multiple feature domains and feature elements;

[0009] S2: Collect the eye movement data of the pilot regarding the feature domains and the subjective evaluation data of the feedback information regarding the feature elements during the flight mission;

[0010] S3: Based on the eye movement data and subjective evaluation data, calculate the visible information amount of the feature elements to generate a feature element visible information amount data set;

[0011] S4: Deconstruct the design features of the aircraft cockpit, including the system layer, attribute layer, and feature layer, and further construct a clustering sample of the system layer - attribute layer - feature layer;

[0012] S5: Calculate the average weight of each layer in the clustering sample;

[0013] S6: Based on the average weight of each layer of the clustering sample, calculate the perceived amount of each design feature of the clustering sample, thereby generating an aircraft cockpit feature perception data set;

[0014] S7: Based on the feature element visible information amount data set and the aircraft cockpit feature perception data set, construct a VH-DF mapping model between the visible information amount of the feature elements and the perceived amount of the design features;

[0015] S8: Based on the mapping result of the VH-DF mapping model, establish an optimization paradigm for the design feature deconstruction.

[0016] Among them, the method in S1 for dividing the aircraft cockpit human-machine interaction system into multiple feature domains and feature elements is as follows: Divide the display system, control system, and warning system in the aircraft cockpit human-machine interaction system into multiple feature domains respectively. Each feature domain contains several feature elements. A feature element is a specific component element in the corresponding feature domain and is the main source for pilots to obtain information. The feature domain matrix is expressed as:

[0017]

[0018] Each feature domain FAs(i) contains several feature elements FEs(i k ) = {i1, i2, …, i k ,...i K}, where K is the number of feature elements in the feature domain FAs(i).

[0019] Among them, the method in S2 for collecting the eye movement data of pilots regarding feature domains and the subjective evaluation data of feedback information regarding feature elements during flight missions is as follows:

[0020] (1) Collect the visual attention of pilots during flight missions through an eye tracker, including the number of fixation points, fixation duration, and fixation area; conduct data verification on the collected eye movement data, including data cleaning, time series alignment, and outlier removal;

[0021] (2) Collect the subjective evaluation data of the feedback information of pilots regarding feature elements through a subjective questionnaire method, including importance and salience. Importance refers to the role and contribution of the information feedback by a certain feature element during the pilot's mission execution. Salience is the prominence of the information feedback by a certain feature element in the pilot's perception, that is, whether a certain feature element has attracted the pilot's high attention or concern during the perception process.

[0022] Among them, the calculation method for the visible information quantity of feature elements in S3 is as follows:

[0023]

[0024] Among them, Nf(i) is the number of fixation points of the pilot on the feature domain FAs(i) within the time period [0, t B , VHW is the visual channel bandwidth, which is expressed as:

[0025]

[0026] In the formula, Afd(i) is the average fixation duration of the pilot on the feature domain FAs(i), and Ip(i k ) is the information perception degree of the pilot on the feature element FEs(i k ). The calculation method for the information perception degree is as follows:

[0027] Ip(i k ) = αIi(i k ) + βFs(i k ),

[0028] Wherein, Ii(i k ) represents the salience of the feature element, Fs(i k ) represents the importance of the feature element, and α and β represent weights;

[0029] H t is the fixation transition entropy of the feature domain FAs(i), which is expressed as:

[0030] H t = -∑ i∈I P i ∑ j∈J p ij log2(p ij ),

[0031] Wherein, P i is the fixation probability of the pilot on the feature domain FAs(i) during the entire flight mission, and p ij is the fixation transition probability between FAs(i) and FAs(j), which are respectively expressed as:

[0032]

[0033] Wherein, n i is the number of times the pilot fixates on the feature domain FAs(i), I is the total number of feature domains being fixated, is the total sum of the number of times all fixated feature domains are fixated; n ij is the number of fixation transitions between FAs(i) and FAs(j), is the total sum of the number of fixation transitions among all fixated feature domains.

[0034] Among them, the method for deconstructing the design features of the aircraft cockpit described in S4, including the system layer, attribute layer, and feature layer, and further constructing the clustering samples of the system layer - attribute layer - feature layer is:

[0035] (1) Take the display system, control system, and warning system in the aircraft cockpit human-machine interaction system as the system layer; the attribute layer further decomposes the macroscopic design elements of each feature element in the system layer into more specific design elements, including digital indication, text, symbols, graphics and images, swing controllers, slide controllers, press controllers, rotary controllers, main visual warning lights, process status, warning flashes, warning indications; the feature layer further refines the visual features of each design element on the basis of the attribute layer, including color, shape, size, proportion, dynamic characteristics, feature lines, brightness, contrast, transparency, texture, thickness, angle, height, shadow depth, motion trajectory, shape change, contour line, edge angle, brightness change, reflectivity, smoothness, distortion, spacing;

[0036] (2) Pair the vocabulary in the feature layer, the vocabulary in the attribute layer, and the pictures in the system layer to form multiple "system-attribute-feature" combinations, and invite multiple experts to evaluate the relevance of various combinations, evaluate the credibility of the evaluation results, and select the combinations with high credibility evaluation as clustering samples.

[0037] Among them, the method for calculating the average weight of each level in the clustering sample described in S5 is as follows:

[0038] (1) First, construct a relative importance scoring matrix, evaluate the relative importance between elements within each level in the clustering sample, and conduct a consistency test. The scoring matrix is expressed as:

[0039]

[0040] Among them, the factors in the matrix represent the relative importance between two elements, and z represents the number of elements within the level in the clustering sample;

[0041] (2) Normalize the evaluation matrix for the system layer, attribute layer, and feature layer with the best consistency test results. Divide each element in the matrix by the sum of the corresponding column, and then calculate the weight of each factor in the normalized scoring matrix. Add up the weights of each level and take the average to obtain the average weights of the system layer, attribute layer, and feature layer respectively

[0042] Among them, the method for calculating the perceived quantity of each design feature in the clustering sample described in S6 is as follows:

[0043]

[0044] In the formula, i', j', and k' are the indexes of the design features of the system level, attribute layer, and feature layer respectively, R xy is the correlation relationship between the design feature x and the design feature y in the attribute layer, R ygFor the association relationship between the design feature y of the attribute layer and the design feature g of the feature layer; when the design feature y of the attribute layer belongs to the design feature x of the system layer, R xy = 1, otherwise R xy = 0; when the design feature g of the feature layer belongs to the design feature y of the attribute layer, R yg = 1, otherwise R yg = 0.

[0045] Among them, the VH-DF mapping model described in S7 includes the data type t, data objects, and the mapping relationship m and mapping strength s between data objects. The construction method is as follows:

[0046] (1) First, construct the model views View-VH and View-DF, which contain the data objects as the visual information dataset of feature elements and the feature perception dataset of the aircraft cockpit respectively; let the set of data objects in the View-VH view be D = {d1, d2, d3…d n’}, where each data object represents a specific data type observable in the View-VH view; let the set of data objects in the View-DF view be F = {f1, f2, f3…f m'}, where each data object corresponds to another data representation form in the View-DF view; use t1d1 and t1d1 to represent the projection expressions of the data object d1 of the data type t1 and the data object d2 of the data type t1 in the views View-VH and View-DF respectively.

[0047] (2) The mapping relationship m judgment between t i d j and t i' d j' :

[0048] When i = i', m = 1, forming an explicit mapping, and the mapping relationship is m1,

[0049] When i ≠ i', m = 0, forming an implicit mapping, and the mapping relationship is m0;

[0050] (3) Use the cosine distance as a similarity measurement tool to calculate the mapping strength s between the data objects d i and f j , which is expressed as:

[0051]

[0052] Among them, respectively represent the feature vectors of the data objects d i and f i , They are the norms of vectors respectively. The similarity value ranges from [0, 1]. The closer it is to 1, the higher the similarity between the two indicates.

[0053] Based on the calculation of the mapping strength between all data objects, the VH-DF mapping matrix S is obtained, which is expressed as:

[0054]

[0055] Among them, s ij represents the mapping strength between the data object d i and f i Under the premise that the mapping relationship is m1, if s ij ≠0, it means that there is an effective mapping relationship between d i and f i When the mapping relationship is m0, All values are invalid mappings.

[0056] Among them, the method for establishing the design feature deconstruction and optimization paradigm based on the mapping result of the VH-DF mapping model is as follows: According to the effective mapping relationship and its mapping strength between the visual information amount of the feature element and the design feature perception amount revealed by the VH-DF mapping model, the feature elements in the aircraft cockpit are hierarchically divided, and optimized according to the mapping strength and the level where each feature element is located. Specifically:

[0057] For display type feature elements, adjust the color, transparency and appearance time according to the mapping strength, so as to achieve the distinction between primary and secondary. For feature elements with high mapping strength and high priority, maintain prominent visual features to attract the attention of pilots. For feature elements with lower mapping strength and secondary level, gradually weaken their visual features to avoid interference;

[0058] For rotating type and swinging type feature elements, adjust the color and contour change according to the mapping strength and the priority of the feature element to ensure that pilots can quickly identify and operate key controllers;

[0059] For warning type feature elements, adjust the color, dynamic appearance time and font ratio according to the mapping strength and the priority of the feature element to optimize the pilot's perception and response ability to abnormal information.

[0060] Beneficial effects: The present invention has the following advantages: By analyzing and defining the design features of the human-machine interaction interface in the cockpit, and based on the visual information volume of the feature elements and the perception volume of the design features, a VH-DF mapping model is constructed. Based on the mapping results of this model, the dynamic and complex flight environment is optimized. By reasonably allocating the visual features of the feature elements, interference from excessive redundant information to the pilot is avoided, the cognitive load is reduced, thereby improving the adaptability and intelligent level of the cockpit layout, helping to meet the emotional needs of the pilot, and improving the operation efficiency and safety in complex environments. In addition, the proposed method will provide new theoretical support and practical paths for the emotional interaction design of the cockpit, and lay a foundation for the design of future intelligent cockpits and the development of human-machine co-driving systems. Description of the Drawings

[0061] Figure 1 It is a schematic flow diagram of this method;

[0062] Figure 2 (a)(b) are schematic diagrams of the regional division of the human-machine interaction interface;

[0063] Figure 3 It is a schematic diagram of clustering samples;

[0064] Figure 4 It is a visualized VH-DF mapping matrix diagram. Specific Embodiments

[0065] The technical solutions of the present invention will be described in detail below in conjunction with the embodiments and the drawings.

[0066] As Figure 1 shown, the method for deconstructing the design features of the aircraft cockpit of the present invention includes the following:

[0067] I. Divide the human-machine interaction system of the aircraft cockpit into multiple feature domains and feature elements.

[0068] The display system, control system, and warning system in the human-machine interaction system of the aircraft cockpit are respectively divided into multiple feature domains (such as the steering wheel area, instrument panel area). Each feature domain contains several feature elements. The feature element is the specific component element in the corresponding feature domain and is the main source for the pilot to obtain information. For example, the display system includes an instrument panel feature domain, and the instrument panel feature domain includes a primary flight display feature element, as Figure 2 shown.

[0069] Among them, the display system feature domain is represented by the set A, A = {a1, a2,..., a n}, with a total of n display feature domains; the control system feature domain is represented by the set B, B = {b1, b2,..., b m}, with a total of m control feature domains. The feature domains of the alarm system are represented by the set C, where C = {c1, c2, …, c p}}, with a total of p control feature domains (if there are other areas, they can be defined as D, etc.). The feature domain matrix FAs of the aircraft cockpit human-machine interaction system is expressed as follows.

[0070]

[0071] Each feature domain FAs(i) contains several feature elements FEs(i k ) = {i1, i2, …, i k ,...i K}}, where K represents the number of feature elements in the feature domain FAs(i).

[0072] II. Construct the aircraft cockpit feature information dataset, including eye movement data and subjective evaluation data.

[0073] (1) Based on the divided feature domains and feature elements, collect the eye movement data of the pilot's visual attention during the flight mission. In this embodiment, based on the Tobii Pro Glasses 2 eye tracker, data collection is carried out in an A320 full-size flight simulator that meets international standards, comprehensively recording the pilot's visual attention in different flight phases (such as takeoff, cruise, and landing). The main dimensions of data collection include the number of fixation points, fixation duration, and fixation area.

[0074] After data collection is completed, data verification is first performed, including data cleaning, time series alignment, and outlier removal, to ensure the accuracy and availability of the data. Data cleaning is mainly used to remove invalid data such as dropped frames and extreme saccades. Time series alignment ensures the synchronization of eye movement data with the flight mission phase, and outlier removal is used to exclude invalid records caused by equipment jitter or environmental interference.

[0075] In the data processing stage, through the analysis of regions of interest (AOI), the fixation duration and number of times of the pilot in different feature domains are statistically analyzed, and the number of jumps between each feature domain is recorded.

[0076] (2) Collect the subjective evaluation data of the information feedback by the pilot on the feature elements, including collecting the pilot's evaluation of the importance and salience of the information feedback on the feature elements;

[0077] After the flight, the information perception of the participants is evaluated through a subjective questionnaire survey, focusing on collecting their subjective evaluations of the importance and salience of the feature elements involved in the task, preparing for calculating the information perception in the following text.

[0078] Specifically, importance refers to the role and contribution of the information fed back by a certain feature element during the task execution, and is used to evaluate the influence of the feature element on the achievement of the overall goal. By investigating the feedback of the subjects in a specific task, it is possible to quantitatively judge the priority and necessity of certain feature elements in task completion, so as to provide a basis for subsequent data processing and decision-making.

[0079] Salience, on the other hand, is related to the prominence of the information fed back by a certain feature element in the perception of the subject. It measures whether the feature element has attracted the subject's high attention or focus during the perception process. Feature elements with higher salience often can trigger strong reactions from the subjects, or stand out among multiple pieces of information and become the core basis for decision-making or judgment. Therefore, during the evaluation process, salience, as another important dimension, acts together with importance on the result of task execution.

[0080] To ensure accurate collection of the subjects' feedback on feature elements, the present invention uses the Likert scale for data collection. By scoring each subject on the "importance" and "salience" of feature elements, the influence of different information can be quantified. Specifically, the subjects evaluate each feature element according to the following criteria:

[0081] Importance evaluation: Using a five-point Likert scale, the subjects rate the importance of each feature element, ranging from 1 (completely unimportant) to 5 (very important). This scale is designed to evaluate the relative importance of each feature element in task execution.

[0082] Salience evaluation: Also using a five-point Likert scale, the significance of each piece of information in the perception of the subject is evaluated, with scores ranging from 1 (completely insignificant) to 5 (very significant). This evaluation reflects the prominence of the feature element in individual perception and helps to determine which feature elements play a more important role in the decision-making process.

[0083] III. Based on eye movement data and subjective evaluation data, by constructing an information theory model of the visible features of the aircraft cockpit, calculate the visible information volume of the feature elements, and generate a dataset of the visible information volume of the feature elements;

[0084] The information theory model consists of a source, a channel, a receiver, and a noise source. The source generates an information sequence to be transmitted, which can be discrete or continuous and has a certain probability distribution to describe the randomness and uncertainty of information generation; the channel is the medium for information transmission and may be affected by various noises and interferences, resulting in distortion or error of information during transmission; the receiver receives and understands the information transmitted by the source through the channel to achieve the ultimate purpose of information transmission; the noise source is the random interference factor existing in the channel, which may come from the environment, equipment defects or other signals and have a negative impact on the accuracy of information transmission.

[0085] Applied to the cockpit environment, the aircraft cockpit itself can be regarded as a source of information, which generates a large amount of information, such as instrument readings, warning light status, etc. This information is transmitted to the pilot (the receiver) through various channels (such as vision, hearing, etc.). There are some interference factors during the transmission process (such as visual chaos, too much information display making it difficult for the pilot to quickly identify key information), which may affect the accuracy and efficiency of information transmission from the source (cockpit) through the channels (such as vision, hearing) to the receiver (pilot).

[0086] This method mainly focuses on the information exchange through the visual channel and calculates the visible information quantity of aircraft cockpit features. The visible information quantity of aircraft cockpit features refers to the effective information quantity transmitted through the visual channel and related to the cockpit state and flight mission, which quantifies the amount and complexity of information received by the pilot from the cockpit human-machine interaction system during the flight.

[0087] The calculation process of the visible information quantity of aircraft cockpit features is as follows:

[0088] Calculate the fixation transition entropy H of the feature domain FAs(i) t ;

[0089] According to the number of fixations of the pilot on the cockpit feature domain FAs(i), calculate the fixation probability P of the pilot on the feature domain FAs(i) during the entire flight mission i :

[0090]

[0091] In the formula: n i is the number of fixations of the pilot on the feature domain FAs(i), I is the total number of feature domains being fixated, is the total number of fixations on all the feature domains being fixated.

[0092] Secondly, calculate the fixation transition probability p between FAs(i) and FAs(j) ij , and the calculation method is

[0093]

[0094] In the formula: n ij is the number of fixation transitions between FAs(i) and FAs(j), i, j are different feature domain codes, is the total number of fixation transitions among all the feature domains being fixated.

[0095] Combining P i and p ij , calculate the fixation transition entropy H t :

[0096] Ht = -∑ i∈I P i ∑ j∈J p ij log2(p ij )。

[0097] The information sources in the aircraft cockpit often face multi-dimensional and multi-variable scenarios, and the visual feature output signals received by the pilot sink also show different states due to the differences in their experience and cognitive level. Such differences can be classified as the additive noise of the channel and need to be processed for noise reduction specifically.

[0098] Taking the ratio of the fixation transition entropy to time as the power index of the noise component, t B is the time for collecting eye movement data, that is, the time required for the pilot to complete a scenario. When t B tends to infinity, the quantization of the cockpit feature information can be independent of the time dimension, and the power index is:

[0099]

[0100] Define the calculation method of the visual channel bandwidth VHW as:

[0101]

[0102] In the formula, Afd(i) is the average fixation duration of FAst(i), and Ip(i k ) is the information perception degree of the pilot for FEs(i k ).

[0103] The calculation method of the information perception degree is:

[0104] Ip(i k ) = αIi(i k ) + βFs(i k )

[0105] Ii(i k ) represents the saliency of the feature element, Fs(i k ) represents the importance of the feature element, and α and β represent weights.

[0106] In the time period of [0, t B , P l is the average power of the input signal sample value, and in the visual channel, it can be defined as

[0107] P l = Nf(i) / t B

[0108] In the formula, Nf(i) is the value within [0, t BThe number of fixation points on the feature domain FAs(i) within a time period.

[0109] The calculation method of the channel feature information amount is as follows:

[0110]

[0111] Substituting the above indexes, we can get:

[0112]

[0113] Because, within the time period [0, t B , L = 2VHWt B , and further derivation gives:

[0114]

[0115] When t B →∞, the visible information amount VHC of the feature element can be calculated t :

[0116]

[0117] IV. Decompose the design features of the feature elements of the aircraft cockpit, construct the aircraft cockpit design feature perception hierarchy framework, including the system layer, the attribute layer, and the feature layer, and further construct the clustering samples of the system layer - attribute layer - feature layer;

[0118] (1) By decomposing the design features of the feature elements of the aircraft cockpit, each component of the cockpit design can be analyzed more meticulously, thereby deeply understanding the functions, roles of each design feature and its impact on the pilot. This in - depth understanding also provides a basis for optimizing the interaction mode between the pilot and the cockpit interface. Designers can optimize the design features according to the needs, habits and cognitive characteristics of different pilots, so as to design a more pilot - demand - compliant human - machine interaction interface.

[0119] System layer: Consider the display system, control system and warning system in the aircraft cockpit human - machine interaction system as the system layer. The design at this level mainly focuses on the coordination and functionality between systems, ensuring that each system can cooperate efficiently and seamlessly, providing an intuitive and easy - to - operate interface for the pilot. At this level, the design goal is to ensure that each system can accurately transmit key information and avoid function overlap or information redundancy, thus simplifying the pilot's decision - making process.

[0120] Attribute layer: Further break down the macroscopic design elements of each feature element in the system layer (such as display system, control system, warning system) into more specific design elements (such as digital indicators, text, symbols, graphics, etc.), ensuring that each design element can convey information in a clear and unambiguous manner, reducing redundant and interfering information, highlighting key information, avoiding information overload, so that the pilot can quickly identify key information in the flight environment, enhancing the overall experience of the pilot's interaction with the cockpit, and making operations more intuitive.

[0121] Feature layer: On the basis of the attribute layer, further refine the visual features of each design element, such as color, proportion, dynamic changes, and feature lines, etc., to enhance the expressiveness and intuitiveness of information, enabling the driver to understand and respond faster and more accurately. The feature layer uses means such as dynamic feedback and color changes to clarify the priority of the feedback information of each feature element, highlighting the visual guiding role of emergency information, enhancing the driver's perception ability in complex situations, and reducing the cognitive burden. The feature layer improves the interaction experience between the pilot and each feature element and flight safety by optimizing the effect of feature element information transmission.

[0122] The perception hierarchy framework of aircraft cockpit design features is shown in Table 1:

[0123] Table 1 Perception hierarchy framework of aircraft cockpit design features

[0124]

[0125] (2) Based on the perception hierarchy framework of aircraft cockpit design features, cluster the design features of the system layer, attribute layer, and feature layer to construct a clustering sample of system layer - attribute layer - feature layer.

[0126] Through clustering, the step-by-step refinement process from the system layer to the attribute layer and then to the feature layer can be clearly shown, and specific design elements that need to be modified or optimized can be quickly located according to the clustering relationship. The mapping method is as follows:

[0127] (1) Construct a clustering sample: Pair 24 feature layer words, 12 attribute layer words with 3 system layer pictures to form multiple "system - attribute - feature" combinations;

[0128] (2) Construct a questionnaire: Invite multiple senior cockpit industrial design engineering experts to evaluate the relevance of the clustering sample. The scoring standard uses the LIKERT five-point scale, and the scoring options are: not in line (1 point), less in line (2 points), average value (3 points), more in line (4 points), very in line (5 points).

[0129] (3) Evaluate the credibility of the questionnaire: Evaluate the internal consistency of the questionnaire through the Cronbach's Alpha coefficient. First, determine the number of items N in the questionnaire, then calculate the variance of the scores of each item and sum them to obtain the total variance; further calculate the Alpha coefficient:

[0130]

[0131] The Alpha coefficient ranges from 0 to 1. An Alpha coefficient greater than 0.7 indicates acceptable internal consistency, greater than 0.8 indicates good internal consistency, and greater than 0.9 indicates excellent internal consistency.

[0132] Based on the credibility evaluation results, the clustering samples at each level are finally as Figure 3 shown.

[0133] V. Calculate the average weight at each level in the clustering samples.

[0134] (1) First, construct a relative importance scoring matrix to evaluate the relative importance among elements within each level of the clustering samples. Invite multiple pilots with many years of flight experience to score the importance of the elements at three levels in the clustering samples. Based on the scores of each pilot, construct a scoring matrix of the relative importance between pairwise elements within each level of the clustering samples respectively. The factors in the matrix represent the relative importance between two elements, and the scoring matrix is expressed as:

[0135]

[0136] z represents the number of elements within each level of the clustering samples.

[0137] (2) Consistency test to ensure the consistent judgment logic in the scoring matrix and improve the reliability of the decision-making results. Calculate the consistency index CI of the scoring matrix:

[0138] CI = (λ max - z) / (z - 1)

[0139] where λ max is the maximum eigenvalue of the scoring matrix and the consistency index CI: The calculation formula is:

[0140] Calculate the consistency ratio CR of the normalized scoring matrix:

[0141] CR = CI / RI

[0142] where RI is the random consistency index, which is a preset value corresponding to the matrix order.

[0143] If the consistency ratio CR of the matrix is less than 0.1, it indicates that the judgment matrix has sufficient consistency and the decision result is reliable; otherwise, if CR is greater than 0.1, it means there are inconsistencies in the judgment matrix, which may affect the accuracy of the final decision and further adjustment is required. Therefore, only when the consistency test passes, the result of the judgment matrix is considered valid and can be used for subsequent weight calculation.

[0144] (3) Calculate the weights of each level in the clustering samples.

[0145] Normalize the scoring matrix by dividing each element in the matrix by the sum of the corresponding column, and then calculate the weight of each factor in the normalized scoring matrix;

[0146] Sum up the weights of each level and take the average to obtain the average weights of the system layer, attribute layer, and feature layer respectively

[0147] VI. Based on the average weights of each level in the clustering samples, calculate the perception amount of each design feature in the clustering samples, so as to generate the aircraft cockpit feature perception data set.

[0148] (1) The perception amount of a design feature refers to the ability or degree of a pilot to perceive, recognize, and interpret the design features of the clustering samples (such as color, proportion, dynamic characteristics, symbols, text, etc.) through the visual channel in the flight environment. It measures how effectively the design features are transmitted and understood, and thus affects the pilot's cognitive efficiency, operation decision-making, and reaction speed. Specifically, the perception amount of a design feature not only includes the accuracy and clarity of the perceived information, but also includes the guiding effect of the design feature on the pilot's attention in a specific context, as well as the timeliness of information transmission and the recognition of priorities. A high perception amount of a design feature means that the design feature can be quickly and accurately recognized by the pilot and helps the pilot make effective judgments and operations in a complex environment.

[0149] Define the association relationship between the design feature x of the system layer and the design feature y of the attribute layer as R xy , and define the association relationship between the design feature y of the attribute layer and the design feature g of the feature layer as R yg . When the design feature y of the attribute layer belongs to the design feature x of the system layer, R xy = 1, otherwise R xy = 0. Similarly, when the design feature g of the feature layer belongs to the design feature y of the attribute layer, R yg = 1, otherwise R yg = 0.

[0150] Calculate the design feature perception amount DFA of each design feature, so as to construct the aircraft cockpit feature perception data set:

[0151]

[0152] In the formula, i', j', and k' are the indices of the design features of the system level, attribute level, and feature level, respectively.

[0153] (2) Perform a consistency check on the aircraft cockpit feature perception data set. First, convert the data set into matrix form, calculate the maximum eigenvalue λ and consistency index CI of the matrix, and then calculate the consistency ratio CR according to the random consistency index RI. When CR < 0.1, it indicates that the matrix construction is reasonable and can better reflect the hierarchical structure and weight relationship of the decision-making target.

[0154] VII. Based on the feature element visual information amount data set and the aircraft cockpit feature perception data set, construct a VH-DF mapping model between the feature element visual information amount VHC and the perception amount DFA of the design features, providing quantitative data for the subsequent construction of the design feature paradigm, including determining the mapping explicit and implicit relationship, calculating the mapping strength, and generating the mapping matrix.

[0155] The VH-DF mapping model includes data type t, data objects, and the mapping relationship m and mapping strength s between data objects. First, construct model views View-VH and View-DF, which contain data objects as the feature element visual information amount data set and the aircraft cockpit feature perception data set, respectively.

[0156] Let the set of data objects in the View-VH view be D = {d1, d2, d3... d n’} where each data object represents a specific data type observable in the View-VH view. Let the set of data objects in the View-DF view be F = {f1, f2, f3... f m'}, where each data object corresponds to another data representation form in the View-DF view. Use t1d1 and t1d1 to represent the projection expressions of the data object d1 of data type t1 and the data object d2 of data type t1 in the View-VH view and the View-DF view, respectively.

[0157] (1) Determine the mapping explicit and implicit relationship

[0158] t i d j and t i' d j' The mapping relationship m has two cases:

[0159] When i = i', m = 1, forming an explicit mapping, and the mapping relationship is m1;

[0160] When i ≠ i', m = 0, forming an implicit mapping, and the mapping relationship is m0;

[0161] Among them, the source of the i index is the feature element or the system to which the design feature belongs. For example, if the data is the index of the display system, then i = 1. If the data is the index of the control system, then i = 2. The explicit mapping indicates that the relationship between data objects is valid and the process ends; the implicit mapping indicates that the relationship between data objects is invalid and it is necessary to continue to find the valid data type until a valid data object is matched.

[0162] (2) Mapping intensity calculation

[0163] In order to describe the correlation between data objects d i and f j , the mapping intensity s is introduced. This method uses the cosine distance as a similarity measurement tool to evaluate the correlation between data objects. For each pair of objects d i and f j screened by type matching, the similarity measurement formula is:

[0164]

[0165] Among them, respectively represent the feature vectors of objects d i and f i , are the norms of the vectors respectively. The value range of the similarity value is [0, 1]. The closer its value is to 1, the higher the similarity between the two.

[0166] During the mapping process, the cosine similarity is not only used to measure the matching degree between two objects, but also provides a quantitative basis for establishing the next mapping relationship.

[0167] (3) Generate the VH-DF mapping matrix based on the mapping intensity calculation result

[0168] The size of the mapping matrix S is n'×m', where n' is the number of data objects in the View-VH view and m' is the number of data objects in the View-DF view. Each element S[i][j] of the matrix is defined as follows:

[0169]

[0170] Among them, t(d i ) and t(f j ) respectively represent the data types of data objects d i and f j . Only when the data types of the two are the same, the cosine similarity is calculated and its value is assigned to the matrix element, otherwise the matrix element is 0.

[0171] ① Expression of similarity weight: Each element S[i][j] in the matrix represents the data object di and f j The similarity weight between them. The larger the value, the stronger the correlation between the two. The value range is [0, 1], where 1 indicates complete similarity and 0 indicates complete irrelevance or type mismatch.

[0172] ② The meaning of 0 value: When a certain element S[i][j] in the matrix = 0, there may be two cases: one is that the similarity between d i and f j is extremely low (i.e., the cosine similarity is close to 0), and the other is that their data types are different, so effective mapping cannot be performed.

[0173] Each data object d in set D i can be mapped to one or more data objects f in F i , and vice versa, that is, each f i can also be associated with one or more d i . Specifically, the value of s represents the mapping strength or degree of association between d i and f i . This many-to-many mapping method not only covers the relationship between data objects in the View-VH view and the View-DF view, but also reflects the interdependence and diversity of the two views when representing different data. To better illustrate the nature of this mapping, the mapping matrix S can be further expressed as:

[0174]

[0175] where s ij represents the mapping strength or correlation between data object d i and f i . On the premise that the mapping relationship is m1, if s ij ≠0, it means that there is an effective mapping relationship between d i and f i ; when the mapping relationship is m0, the values are all invalid mappings.

[0176] (4) Construct a visual VH-DF mapping matrix

[0177] Based on the mapping matrix S, the k-Nearest Neighbors (k-NN) method is used to select the k most similar relationship objects f i for each data object d j to perform mapping. The k-NN method is a common method for dealing with many-to-many mappings. Its core idea is to select the objects closest to the target object from multiple objects according to the similarity metric and regard these objects as mapping pairs. The specific process is as follows:

[0178] ① Type matching: First, only consider the object pairs (d i , f i ) screened out by data type matching. That is, only when t(d i ) = t(f j ), are they allowed to participate in the similarity ranking of the nearest neighbor method.

[0179] ② Similarity ranking: Sort all data object pairs that meet the type matching according to the similarity measure CosSim(d i , f j ) to determine the most similar object pairs.

[0180] ③ Select the nearest neighbors: For each data object d i select k relational objects f j with the highest cosine similarity, that is, select the k nearest neighbor objects closest to it to construct a many-to-many mapping relationship. If k = 1, select one most similar object for one-to-one mapping; if k > 1, allow a mapping to be established between an object and multiple relational objects to form a many-to-many mapping relationship.

[0181] Through the use of the nearest neighbor method, this process can effectively screen out the objects with the highest similarity, ensuring the accuracy and rationality of the mapping.

[0182] To visually display the mapping relationship between data objects in the View-VH view and the View-DF view, this method uses matrix visualization technology. Similar to the symbol matrix method, the mapping relationships of different degrees are presented in a symbolic form in the visualized two-dimensional matrix, as Figure 4 shown. Figure 4 The horizontal axis of [] represents the set of data objects in the View-VH view, and the vertical axis represents the work items or feature areas in the View-DF view. The position of each element is represented by different symbols to indicate the similarity and relevance between objects.

[0183] In the visualized two-dimensional matrix, three symbols are introduced to represent different similarity intensities:

[0184] Black dot (●): Indicates a strong association between the data object and the view object. Specifically, when the cosine similarity CosSim(d i , f j ) ≥ 0.8, it is considered that there is a strong mapping relationship between them.

[0185] Grey circle (◎): Indicates a weak association between the data object and the view object. When the similarity value is in the range of 0.5 ≤ CosSim(d i , f j) When it is between <0.8, the symbol is a gray circle, indicating a weak association.

[0186] Hollow circle (○): indicates that there is no association between the data object and the view object. At this time, the cosine similarity value CosSim(d i , f j ) <0.5, it is considered that there is no significant association between these two objects.

[0187] VIII. Based on the mapping structure of the VH-DF mapping model, establish an optimization paradigm for the deconstruction of design features.

[0188] The optimization paradigm for the deconstruction of design features is a systematic design method aimed at achieving the distinction between primary and secondary design information, reducing visual interference, and improving user operation efficiency and experience by carefully deconstructing and analyzing design features (such as color, contrast, proportion, transparency, etc.) and hierarchically processing and optimizing them according to their importance.

[0189] In specific applications, this paradigm adopts a strategy of decreasing information hierarchy, maintaining prominent visual features (such as high contrast, original proportion) for high-priority information, while differentiating secondary information by gradually weakening its visual features (such as reducing transparency, shrinking proportion, adjusting dynamic characteristics). This hierarchical processing method has been systematically applied in the design of display, control, and alarm systems, effectively optimizing the user's ability to perceive, recognize, and respond to information.

[0190] For example, in the design of the cockpit, display elements such as digital indicators, text, and symbol markings achieve the distinction between primary and secondary information by adjusting color, transparency, and appearance time; control features such as rotation and swing types ensure that users can quickly identify and operate key controllers through hierarchical color and contour changes; the alarm system adjusts color, dynamic appearance time, and font proportion according to priority to optimize the user's ability to perceive and respond to abnormal information.

[0191] The following provides some examples of the optimization of the deconstruction of design features based on the cockpit display system

[0192] (1) Digital indicator optimization: As the main way to transmit key flight parameters (such as speed, altitude, etc.), the design of digital indicators strictly follows the similarity principle and the priority stratification strategy. High-priority digital indicators use high-contrast colors and a stroke width-font proportion of 1:8 to ensure the highest visual priority. Secondary information is weakened by gradually reducing transparency (to 50%, 20%) and shrinking the proportion (to 1:10, 1:12), guiding the pilot's attention to key data and avoiding visual interference. This design meets the requirements in the airworthiness certification guidelines regarding "highlighting key information and avoiding interference from secondary information".

[0193] (2) Optimization of text information presentation: The presentation of text information adopts a decreasing strategy, involving color transparency, proportion, and font size. High-priority text maintains high contrast and a larger font size, while secondary text gradually reduces transparency (to 50%, 20%) and shrinks the font proportion (to 1:10, 1:12). Based on the closure principle and proximity principle of Gestalt psychology, this design makes high-priority text easy to perceive, while low-priority text naturally blends into the background, reducing visual interference and improving the information acquisition efficiency of pilots in complex environments.

[0194] (3) Optimization of symbol identification: Symbol identification undertakes the functions of status indication and warning in the display system, and its design deconstruction is based on hierarchical adjustment of form, dynamic characteristics, and color. The first-level symbol identification retains the original form, sets a 3-second display time, and maintains a high-contrast color to ensure the highest priority. As the priority decreases, the display time of the second-level symbol is shortened to 1.5 seconds, and the transparency is reduced to 50%; the display time of the third-level symbol is further shortened to 0.75 seconds, and the transparency is increased to 20% to reduce interference to the pilot. This design follows the closure principle of Gestalt psychology to ensure that high-priority symbols are prominent, while low-priority symbols gradually blend into the background.

[0195] (4) Optimization of the deconstruction of graphics and images: The deconstruction design of graphics and images follows the principle of gradual decrease in size, thickness of feature lines, and transparency to ensure the layering and readability of information. High-priority images use a 1.5-point solid line, maintain the original size and high-saturation primary colors to ensure clear visibility. Secondary information is weakened by making the feature lines thinner (to 0.75 points), reducing transparency (to 50%), and shrinking the size (to 75%); the lowest-priority images are further reduced to 50% of the size, the feature lines are adjusted to 0.75-point dashed lines, and the transparency is increased to 20% to reduce the visual burden. This design conforms to the similarity principle and proximity principle of Gestalt psychology to ensure that high-priority information is prominently readable and low-priority information does not interfere with the main task.

[0196] For the specific feature element b11, the design feature deconstruction optimization is as follows: The color and proportion of the digital indication are the first level, with a stroke width-font proportion of 1:8 and a transparency of 100%; the form, dynamic characteristics, and color of the symbol identification are also the first level, with the form proportion unchanged, the dynamic characteristic set to a 3-second display time, and the transparency of 100%; the proportion and color of the text belong to the second level, with a transparency of 50% and a stroke width-font proportion of 1:10; the color of the graphic image is the first level, and the feature line belongs to the third level. Therefore, the graphic image proportion remains unchanged, the transparency is 100%, and the feature line is set to 0.75 points.

Claims

1. A method for deconstructing and optimizing the design features of an aircraft cockpit, characterized in that, It includes the following steps: S1: Divide the aircraft cockpit human-computer interaction system into multiple feature domains and feature elements; S2: Collect the eye movement data of pilots regarding the feature domains and the subjective evaluation data of the feedback information of the feature elements during flight missions; S3: Based on the eye movement data and subjective evaluation data, calculate the visible information amount of the feature elements, and generate a feature element visible information amount data set; S4: Decompose the design features of the aircraft cockpit, including the system layer, attribute layer, and feature layer, and further construct a clustering sample of the system layer-attribute layer-feature layer; S5: Calculate the average weight of each level in the clustering sample; S6: Based on the average weight of each level of the clustering sample, calculate the perception amount of each design feature of the clustering sample, so as to generate an aircraft cockpit feature perception data set; S7: Based on the feature element visible information amount data set and the aircraft cockpit feature perception data set, construct a VH-DF mapping model between the visible information amount of the feature elements and the perception amount of the design features; S8: Based on the mapping results of the VH-DF mapping model, establish a design feature decomposition and optimization paradigm.

2. The method for deconstructing and optimizing the design features of an aircraft cockpit according to claim 1, wherein The method in S1 for dividing the aircraft cockpit human-computer interaction system into multiple feature domains and feature elements is as follows: Divide the display system, control system, and warning system in the aircraft cockpit human-computer interaction system into multiple feature domains respectively. Each feature domain contains several feature elements. The feature element is the specific component element in the corresponding feature domain and is the main source for pilots to obtain information. The feature domain matrix is expressed as: Each feature domain FAs(i) contains a number of feature elements FEs(i k ) = {i1, i2, …, i k ,... i K}, where K is the number of feature elements in the feature domain FAs(i).

3. The aircraft cockpit design feature deconstruction and optimization method according to claim 1, wherein, The method in S2 for collecting the eye movement data of pilots regarding the feature domains and the subjective evaluation data of the feedback information of the feature elements during flight missions is as follows: (1) Collect the visual attention of pilots during flight missions through an eye tracker, including the number of fixation points, fixation duration, and fixation area; conduct data verification on the collected eye movement data, including data cleaning, time series alignment, and outlier removal; (2) Collect the subjective evaluation data of the feedback information of the feature elements from pilots through a subjective questionnaire survey method, including importance and salience. The importance refers to the role and contribution of the information feedback by a certain feature element during the pilot's mission execution process, and the salience is the prominence of the information feedback by a certain feature element in the pilot's perception, that is, whether a certain feature element has attracted the pilot's high attention or concern during the perception process.

4. The aircraft cockpit design feature deconstruction and optimization method according to claim 3, characterized in that The calculation method of the visible information amount of the feature elements in S3 is as follows: where Nf(i) is the number of fixation points of the pilot on the feature domain FAs(i) within the time period [0, t B , and VHW is the visual channel bandwidth, expressed as: where Afd(i) is the average fixation duration of the pilot on the feature domain FAs(i), and Ip(i k ) is the information perception degree of the pilot on the feature element FEs(i k ). The calculation method of the information perception degree is as follows: Ip(i k ) = αIi(i k ) + βFs(i k ) where, Ii (i k ) represents the salience of the feature element, Fs(i k ) represents the importance of the feature element, and α, β represent weights; H t is the fixation transition entropy of the feature domain FAs(i), expressed as: H t = -∑ i∈I P i ∑ j∈J p ij log2(p ij ), where, P i is the fixation probability of the pilot on the feature domain FAs(i) during the entire flight mission, and p ij is the fixation transfer probability between FAs(i) and FAs(j), which are respectively expressed as: Where n i is the number of times the pilot gazes at the feature domain FAs(i), I is the total number of feature domains being gazed at, and is the total number of gazes at all the feature domains being gazed at; n ij is the number of gaze transitions between FAs(i) and FAs(j), and is the total number of gaze transitions among all the feature domains being gazed at.

5. The method for deconstructing and optimizing the design features of an aircraft cockpit according to claim 1, characterized in that The method in S4 for decomposing the design features of the aircraft cockpit, including the system layer, attribute layer, and feature layer, and further constructing a clustering sample of the system layer-attribute layer-feature layer is as follows: (1) The display system, control system, and warning system in the aircraft cockpit human-machine interaction system are regarded as the system layer; the attribute layer further decomposes the macro design elements of each feature element in the system layer into more specific design elements, including digital indication, text, symbols, graphics and images, swing controllers, slide controllers, press controllers, rotary controllers, main visual warning lights, process status, warning flashes, and warning indications; the feature layer further refines the visual features of each design element on the basis of the attribute layer, including color, form, size, proportion, dynamic characteristics, feature lines, brightness, contrast, transparency, texture, thickness, angle, height, shadow depth, motion trajectory, form change, contour line, edge angle, brightness change, reflectivity, smoothness, distortion, and spacing. (2) Pair the vocabulary in the feature layer, the vocabulary in the attribute layer, and the pictures in the system layer to form multiple "system-attribute-feature" combinations, invite multiple experts to evaluate the relevance of various combinations, conduct a credibility assessment on the evaluation results, and select the combination with a high credibility assessment as the clustering sample.

6. The method for deconstructing and optimizing the design features of an aircraft cockpit according to claim 1, characterized in that, (2) The method for calculating the average weight of each level in the clustering sample described in S5 is as follows: (1) First, construct a relative importance scoring matrix to evaluate the relative importance between elements within each level in the clustering sample and conduct a consistency test. The scoring matrix is expressed as: where the factor in the matrix represents the relative importance between two elements, and z represents the number of elements within the level in the clustering sample. (2)Normalize the evaluation matrix for the system layer, attribute layer, and feature layer with the best consistency test results. Divide each element in the matrix by the sum of the corresponding column, and then calculate the weight of each factor in the normalized scoring matrix. Add up the weights of each level and take the average to obtain the average weights of the system layer, attribute layer, and feature layer respectively.

7. The method for deconstructing and optimizing the design features of an aircraft cockpit according to claim 6, characterized in that (5) The method for calculating the perceived quantity of each design feature in the clustering sample described in S6 is as follows: where i', j', and k' are the indices of the design features of the system level, attribute level, and feature level respectively, and R xy is the association relationship between the design feature x and the design feature y of the attribute level, and R yg is the association relationship between the design feature y of the attribute level and the design feature g of the feature level; when the design feature y of the attribute level belongs to the design feature x of the system level, R xy = 1, otherwise R xy = 0; When the design feature g of the feature layer belongs to the design feature y of the attribute layer, R yg = 1, otherwise R yg = 0.

8. The aircraft cockpit design feature deconstruction and optimization method according to claim 1, characterized in that (6) The VH-DF mapping model described in S7 includes data type t, data objects, and the mapping relationship m and mapping intensity s between data objects. The construction method is as follows: (1) First, construct the model views View-VH and View-DF, which contain the data objects as the feature element visual information dataset and the aircraft cockpit feature perception dataset respectively; let the data object set in the View-VH view be D = {d1, d2, d3…d n’}, where each data object represents a specific data type observable in the View-VH view; let the data object set in the View-DF view be F = {f1, f2, f3…f m'}, where each data object corresponds to another data representation form in the View-DF view; use t1d1 and t1d1 to represent the projection expressions of the data object d1 of data type t1 and the data object d2 of data type t1 in the views View-VH and View-DF respectively; (2)t i d j Mapping relationship m with t i' d j' judgment: When i = i', m = 1, a dominant mapping is formed, and the mapping relationship is m1. When i ≠ i', m = 0, a recessive mapping is formed, and the mapping relationship is m0. (3) Use the cosine distance as a similarity measurement tool to calculate the mapping strength s between data objects d i and f j as follows: Among them, respectively represent the feature vectors of the data objects d i and f i and the norms of the vectors are respectively. The similarity value ranges from [0, 1]. The closer it is to 1, the higher the similarity between the two. ​ Based on the calculation of the mapping intensity between all data objects, the VH-DF mapping matrix S is obtained, which is expressed as: Among them, s ij represents the mapping strength between the data object d i and f i Under the premise that the mapping relationship is m1, if s ij ≠0, it means that there is an effective mapping relationship between d i and f i When the mapping relationship is m0, the values are all invalid mappings.

9. The method for deconstructing and optimizing the design features of an aircraft cockpit according to claim 8, wherein (10) The method for establishing an optimization paradigm for the design feature deconstruction based on the mapping result of the VH-DF mapping model is as follows: According to the effective mapping relationship and mapping intensity between the visible information quantity of the feature element and the perceived quantity of the design feature revealed by the VH-DF mapping model, the feature elements in the aircraft cockpit are hierarchically divided, and optimized according to the mapping intensity and the level where each feature element is located. Specifically: For display-type feature elements, adjust the color, transparency, and appearance time according to the mapping intensity to achieve the distinction between primary and secondary. For feature elements with a high mapping intensity and high priority, maintain prominent visual features to attract the attention of pilots. For feature elements with a lower mapping intensity and secondary level, gradually weaken their visual features. Avoid interference. For rotary-type and swing-type feature elements, adjust the color and contour change according to the mapping intensity and the priority of the feature element to ensure that pilots can quickly identify and operate key controllers. For warning-type feature elements, adjust the color, dynamic appearance time, and font proportion according to the mapping intensity and the priority of the feature element to optimize the pilot's perception and response ability to abnormal information.

Images