A flight performance and multi-modal data fusion flight potential evaluation method

Abstract

This invention provides a method for flight potential assessment that integrates flight performance and multimodal data, belonging to the field of flight potential assessment technology. The invention first constructs a standardized simulated flight assessment task framework, decomposes the scenario, and standardizes the steps. Flight parameters and multimodal physiological and psychological data are collected and sliced ​​according to step alignment. Flight quality parameters and physiological and psychological parameters are defined, and a mapping between standardized operation steps and capability elements is established to form capability representations. Flight performance is stratified and scored to obtain completion and quality scores. Flight performance features are constructed and physiological and psychological features are extracted. A quantitative correlation is established using actual flight mission performance as the core criterion, and fusion weights are determined to synthesize capability element scores. Statistical norms are established based on test data from the subject group, and the assessment results are output. This invention improves the structure and objectivity of the assessment.

Description

Technical Field

[0001] This invention relates to the field of flight potential assessment technology, and in particular to a flight potential assessment method that integrates flight performance and multimodal data. Background Technology

[0002] In the field of flight selection and training assessment, existing technologies typically use simulated flight missions or standard training courses as a vehicle to evaluate the flight performance of test takers. In practice, evaluations are mostly based on flight parameter records, mission completion, and instructor observations. This type of approach can reflect the test takers' operational results and overall performance under specific courses, and has certain application value for training management and phased screening. However, when it is necessary to further determine the potential differences in test takers' abilities, existing assessment results often lean more towards outcome representation, and the revelation of the composition of abilities, ability shortcomings, and their causes during the operational process remains limited.

[0003] With the development of flight simulation systems, flight data recording methods, and physiological and psychological signal acquisition technologies, flight assessment technology is gradually showing a trend towards refinement and data-driven approaches. On the one hand, the assessment objects are no longer limited to the final score or a single flight indicator, but are beginning to focus on the stage performance, operational quality, and response characteristics during the flight process. On the other hand, physiological and psychological data such as eye movement, electroencephalography (EEG), electrodermal data (EDG), electromyography (EMG), and heart rate variability are being introduced into the assessment scenario to assist in analyzing changes in the subject's state in areas such as attention allocation, cognitive load, stress response, and operational stability. Correspondingly, assessment methods have also gradually evolved from single-indicator scoring to multi-source information joint analysis to improve the interpretability and discriminative power of the assessment results.

[0004] However, based on the actual use of existing technologies, several technical issues still exist that affect the accuracy of assessments and the stability of applications. First, continuous flight process data and multimodal biopsychological data often lack a unified segmentation and alignment mechanism corresponding to standardized operational procedures, making it difficult to effectively compare and integrate data from different sources under the same operational semantics. Second, existing scoring methods typically focus on whether the task was completed or the overall performance level, failing to adequately characterize the correspondence between specific operational steps and capability elements, making it difficult to form clear capability attribution results. Third, even with the introduction of multi-source data, many schemes remain at the level of simple aggregation or experience-based weighting, lacking quantitative correlation and fusion rules based on actual flight mission performance, and also lacking a statistical reference basis for similar selection targets. Therefore, there is still room for improvement in the comparability, interpretability, and stability of assessment results. Summary of the Invention

[0005] To overcome the shortcomings of existing technologies, the purpose of this invention is to provide a flight potential assessment method that integrates flight performance and multimodal data. By standardizing steps, aligning and slicing steps, using dual-channel quantitative correlation fusion, and statistical norm localization, the method improves the structure, objectivity, and interpretability of flight potential assessment.

[0006] To achieve the above objectives, the present invention provides the following solution: A method for flight potential assessment that integrates flight performance and multimodal data includes: Under the guidance of flight experts, a standardized simulated flight assessment task framework was constructed, and the standardized simulated flight assessment task framework was decomposed into scenarios and standardized in steps to obtain a standardized operation step sequence. Flight parameters and multimodal physiological and psychological data were collected from the subjects when they performed the standardized simulated flight assessment task framework. Based on the standardized operation step sequence, refined data slices were performed to align the steps, resulting in step-level data slices corresponding to each standardized operation step in the standardized operation step sequence. For each standardized operation step in the sequence of standardized operation steps, corresponding flight quality parameters and bio-psychological parameters are defined, a mapping relationship between each standardized operation step and capability elements is established, and a capability representation vector corresponding to each standardized operation step is formed based on the flight quality parameters and the bio-psychological parameters. The flight performance of the subjects is stratified and scored based on the step-level data slices to obtain a completion score and a quality score; wherein, the quality score is determined based on the comparison results of the flight parameters with the ideal trajectory or standard parameters; Flight performance characteristics are constructed based on the flight parameters, the completion score, and the quality score, and physiological and psychological characteristics are extracted from the multimodal physiological and psychological data. Using actual flight mission performance as the core criterion, quantitative correlations are established between the flight performance characteristics and the core criterion, as well as between the physiological and psychological characteristics and the core criterion. Based on the quantitative correlation, the fusion weight of the dual-channel indicators is determined, and the integrated modeling and score synthesis of each capability element are performed according to the fusion weight of the dual-channel indicators to obtain the capability element score. Statistical norms for the aforementioned ability elements are established based on test data from actual subject groups under the same assessment scenario, and flight potential assessment results are output based on these statistical norms.

[0007] The present invention discloses the following beneficial effects: (1) This invention addresses the problem of prioritizing results over process in existing flight assessments. First, it constructs a standardized simulated flight assessment task framework and decomposes the scenarios and standardizes the steps, transforming the originally continuous and complex flight task process into a standardized sequence of operational steps. This process ensures that the assessment object no longer rests on overall scores or experience-based judgments, but rather falls to a corresponding and comparable step level, providing a unified task semantic basis for subsequent scoring, attribution, and fusion modeling, thereby improving the structure and verifiability of the assessment process.

[0008] (2) This invention further establishes a correspondence between data from different sources under the same standardized operating procedure by performing refined data slicing of flight parameters and multimodal physiological and psychological data through step alignment. Compared with the problems of inconsistent time axes of multi-source data and difficulty in fusion analysis under the same operational semantics in the prior art, this technical solution can significantly improve the data alignment quality, reduce the analysis bias caused by sampling time sequence misalignment, and thus improve the reliability of subsequent feature construction and capability assessment.

[0009] (3) This invention also establishes a mapping relationship between standardized operating procedures and capability elements, and combines flight quality parameters and biopsychological parameters to form a capability representation vector, enabling the evaluation results to be further refined from whether the task was completed and the level of performance to the specific capability element performance. This technical solution is conducive to enhancing the capability attribution ability and interpretability of the evaluation results, and can more clearly distinguish the differences in control ability, cognitive state-related abilities, etc., overcoming the problem that the scoring results in the prior art are difficult to locate capability shortcomings.

[0010] (4) Based on the hierarchical scoring, this invention incorporates completion score and quality score into the construction of flight performance characteristics, and establishes quantitative correlations between flight performance characteristics, physiological and psychological characteristics and core criteria, and then determines the fusion weights of dual-channel indicators for integrated modeling and score synthesis. This technical solution avoids the shortcomings of the existing technology of simply superimposing or assigning weights based on experience for multi-source indicators, and makes the fusion weights have a criterion basis, thereby improving the objectivity, discrimination and stability of capability element scores.

[0011] (5) Finally, this invention establishes statistical norms based on test data of actual subject groups under the same assessment scenario, and outputs flight potential assessment results based on statistical norms. Compared with the assessment method that only outputs absolute scores, this technical solution provides the assessment results with a group reference basis, which can improve the comparability and interpretability of the assessment results, facilitate standardized application in selection scenarios, and improve the practical value of the assessment conclusions. Attached Figure Description

[0012] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0013] Figure 1 This is a flowchart of a method provided in an embodiment of the present invention. Detailed Implementation

[0014] 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.

[0015] The purpose of this invention is to provide a flight potential assessment method that integrates flight performance and multimodal data. By establishing a standardized step-level data foundation and introducing criterion-driven dual-channel fusion and norm reference, the reliability, discriminability, and comparability of flight potential assessment results are enhanced.

[0016] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0017] Figure 1 The method flowchart provided in the embodiments of the present invention is as follows: Figure 1 As shown, this invention provides a method for flight potential assessment by fusing flight performance and multimodal data, comprising: Step 100: Under the guidance of flight experts, construct a standardized simulated flight assessment task framework, and decompose the standardized simulated flight assessment task framework into scenarios and standardize the steps to obtain a standardized operation step sequence. Step 200: Collect flight parameters and multimodal physiological and psychological data of the subjects when they perform the standardized simulated flight assessment task framework, and perform refined data slicing based on the standardized operation step sequence to obtain the step-level data slices corresponding to each standardized operation step in the standardized operation step sequence. Step 300: For each standardized operation step in the standardized operation step sequence, define the corresponding flight quality parameters and bio-psychological parameters, establish the mapping relationship between each standardized operation step and capability elements, and form the capability representation vector corresponding to each standardized operation step based on the flight quality parameters and bio-psychological parameters. Step 400: Based on step-level data slices, the flight performance of the subjects is stratified and scored to obtain a completion score and a quality score; the quality score is determined based on the comparison results of flight parameters with ideal trajectory or standard parameters. Step 500: Construct flight performance characteristics based on flight parameters, completion score, and quality score, and extract physiological and psychological characteristics from multimodal physiological and psychological data; Step 600: Using actual flight mission performance as the core criterion, establish quantitative correlations between flight performance characteristics and core criteria, as well as between physiological and psychological characteristics and core criteria; Step 700: Determine the fusion weight of the dual-channel indicators based on the quantitative correlation, and perform integrated modeling and score synthesis for each capability element according to the fusion weight of the dual-channel indicators to obtain the capability element score; Step 800: Establish statistical norms for capability elements based on test data of actual subject groups under the same assessment scenario, and output flight potential assessment results based on statistical norms.

[0018] Specifically, this embodiment provides a flight potential assessment method that integrates flight performance and multimodal data fusion. The preprocessing part includes steps 100 and 200. Step 100 is used to construct a standardized simulated flight assessment task framework under the guidance of flight experts, and to decompose the standardized simulated flight assessment task framework into scenarios and standardize the steps to obtain a standardized operation step sequence. Step 200 is used to collect flight parameters and multimodal physiological and psychological data when the subject executes the standardized simulated flight assessment task framework, and to perform refined data slicing based on the standardized operation step sequence to obtain step-level data slices corresponding to each standardized operation step. By first forming a standardized operation step sequence and then performing step alignment and slicing processing on multi-source time-series data, this embodiment enables subsequent capability mapping, hierarchical scoring, and fusion modeling to be based on a unified step semantics.

[0019] In step 100, this embodiment first determines the assessment scenarios for flight potential evaluation under the guidance of flight experts. These assessment scenarios include practice, five-way flight, lock-on launch, bird strike incidents, and complex weather flight. After determining the assessment scenarios, this embodiment decomposes each assessment scenario into scenario slices corresponding to each scenario. Then, it standardizes the steps of each scenario slice to form standardized operation steps corresponding to each assessment scenario. Finally, it arranges the standardized operation steps according to the task execution order of each assessment scenario to obtain a standardized operation step sequence. The output of step 100 includes at least the step name and execution order of each standardized operation step, used for subsequent step alignment and data slicing.

[0020] As shown in Table 1, the scenario decomposition and step standardization in this embodiment are performed separately according to the assessment scenarios to ensure that each standardized operation step has clear task semantics, operation objectives, and step boundaries. Specifically, the practice is scenario-decomposed into level flight practice, ascent practice one, ascent practice two, turn practice one, turn practice two, altitude hold and orientation practice one, and altitude hold and orientation practice two; the five-way flight is scenario-decomposed into pre-turn one, pre-turn two, pre-turn three, pre-turn four, glide, and landing; the lock-on launch and bird strike situations are standardized into standardized operation steps corresponding to identification, action execution, and stability recovery; the complex weather flight is standardized into standardized operation steps corresponding to scenario identification, action execution, scenario re-identification, and action adjustment. To facilitate implementation by those skilled in the art, this embodiment sets a step description item for each standardized operation step. The step description item includes at least the step name, operation objective, step start determination basis, and step end determination basis to support subsequent step alignment processing.

[0021] Table 1. Examples of assessment scenarios and scenario slices In this embodiment, the standardized operation step sequence formed after step 100 serves as the task semantic basis for subsequent data processing and capability assessment. This embodiment organizes the output of step 100 into a structured task description result, which includes at least an assessment scenario identifier, a scenario slice identifier, a standardized operation step name, step sequence information, and corresponding step description items. Furthermore, this embodiment can establish step index information based on the standardized operation step sequence, ensuring that each standardized operation step has a unique corresponding position identifier in subsequent data processing. Through the above organization method, this embodiment achieves a stable connection between standardized operation steps and subsequent parameter information processing, providing a unified reference for step-level data slicing, parameter organization, and capability mapping.

[0022] In step 200, this embodiment synchronously collects flight parameters and multimodal physiological and psychological data while the subject performs a standardized simulated flight assessment task framework, thus forming multi-source time-series data. Synchronous collection here means that this embodiment uses a unified time reference to record the flight parameter acquisition channel and the multimodal physiological and psychological data acquisition channel, thereby establishing a temporal correspondence between changes in flight state, control-related changes, and physiological and psychological state changes during the same task process. After collection is completed, this embodiment organizes the flight parameters and multimodal physiological and psychological data in a unified chronological order, and retains task process information related to the standardized operation step sequence, forming multi-source time-series data for step alignment processing.

[0023] In this embodiment, after constructing the multi-source time-series data, step alignment is performed on the multi-source time-series data based on the standardized operation step sequence obtained in step 100. Specifically, this embodiment determines the data range corresponding to each standardized operation step in the multi-source time-series data according to the step start and step end determination criteria corresponding to each standardized operation step. After determining the data range, the multi-source time-series data after step alignment is further refined into data slices according to the operation semantic units corresponding to each standardized operation step, resulting in step-level data slices corresponding to each standardized operation step in the standardized operation step sequence. The output of step 200 includes at least each step-level data slice and its corresponding step identification information, enabling subsequent steps to perform corresponding analysis, scoring, and capability mapping of flight parameters and multimodal physiological and psychological data at the standardized operation step level.

[0024] In this embodiment, the scenario decomposition and step standardization in step 100 are implemented using task semantic rules determined by flight experts. Specifically, for each standardized operation step, this embodiment predefines a step description item, which includes at least a step name, operation objective, step start determination criterion, and step end determination criterion. The step start determination criterion and the step end determination criterion are used for step alignment of multi-source time-series data in subsequent step 200. The step start determination criterion and the step end determination criterion can originate from one or more of the following: task script event records, simulator event records, and flight parameter state changes. When multiple determination criteria exist, this embodiment can determine the step boundaries according to a preset priority order to ensure that the standardized operation step sequence has consistent step division rules under different subjects and different assessment rounds.

[0025] In this embodiment, the step alignment in step 200 is based on the step start determination criteria and step end determination criteria predefined in step 100. The data range corresponding to each standardized operation step is determined in the multi-source time-series data, and a step-level data slice is formed within the data range. For the i-th standardized operation step in the standardized operation step sequence, this embodiment can determine the step boundary and generate a step-level data slice in the following manner: in, For the first The start time of a standardized operating procedure; For the first The end point of a standardized operating procedure; For the first A standardized operating procedure at any time The start determination function, when time... The value is 1 when the standard operating procedure is satisfied, and 0 otherwise. For the first A standardized operating procedure at any time The termination determination function, when time... The value is 1 if the standard operating procedure is met and the step completion criteria are met; otherwise, the value is 0. For the first Each standardized operation step corresponds to a step-level data slice; For a moment Multi-source time-series data; Time under a unified time base; These are the sequence numbers for standardized operating procedures.

[0026] As an example, the function of the above formula is to locate the data range of each standardized operation step on the time axis of multi-source time series data according to the predefined step boundary determination rules, and to form step-level data slices that correspond one-to-one with the standardized operation steps, so as to support subsequent step-level parameter organization, scoring and capability mapping processing.

[0027] Furthermore, in this embodiment, when a standardized operation step has multiple start-up or end-of-step determination criteria, the start-up and end-of-step determination functions can be determined according to a preset priority order. When a determination criterion is missing, the remaining valid determination criteria can be used to determine the step boundary, or the corresponding step-level data slice can be marked as a boundary data slice to be confirmed, to ensure the continuity of step alignment processing and the traceability of the data processing process. By combining the above-mentioned rule-based implementation with formulaic boundary positioning, this embodiment can achieve a stable correspondence between standardized operation step sequences and multi-source time-series data without relying on complex manual parameter tuning.

[0028] Specifically, in this embodiment, step 300, based on the step-level data slices obtained in step 200, defines corresponding flight quality parameters and biopsychological parameters for each standardized operation step in the standardized operation step sequence, establishes a mapping relationship between each standardized operation step and capability elements, and forms a capability representation vector corresponding to each standardized operation step based on the flight quality parameters and biopsychological parameters. Here, capability elements refer to assessment objects used to characterize different dimensions of flight potential, including comprehension and acceptance ability, adaptability, control ability, spatial orientation ability, situational awareness ability, attention quality, and emotional stability. Their function is to serve as target objects for organizing, scoring, and interpreting step-level parameters. The mapping relationship refers to the correspondence between standardized operation steps and capability elements. Its function is to limit which capability elements a certain standardized operation step is mainly used to characterize, and to provide a rule basis for subsequent parameter grouping and capability representation vector formation. The input of step 300 is the step-level data slices corresponding to each standardized operation step, and the output is the capability representation vector corresponding to each standardized operation step.

[0029] In this embodiment, flight quality parameters and biopsychological parameters are defined for each standardized operation step in the standardized operation step sequence. The flight quality parameters characterize the flight performance quality corresponding to that standardized operation step, while the biopsychological parameters characterize the physiological and psychological state changes corresponding to that standardized operation step. These biopsychological parameters are derived from multimodal physiological and psychological data in the step-level data slices corresponding to that standardized operation step. Because the flight quality parameters and biopsychological parameters differ in dimensions, numerical range, and fluctuation scale, this embodiment first performs homoscalation processing before forming the capability representation vector. Homoscalation processing here refers to converting parameters with different dimensions into comparable dimensionless representation values, ensuring that the flight quality parameters and biopsychological parameters can be jointly represented within the same standardized operation step. This embodiment can perform homoscalation processing in the following manner: in, For the first In the first standardized operating procedure Homoscaling results of each flight quality parameter; For the first In the first standardized operating procedure The original value of each flight quality parameter or the parameter value calculated from step-level data slices; In order to be with the first The first standardized operating procedure and the first The reference sample set corresponding to each flight quality parameter refers to the historical sample set of the same parameter collected under the same assessment scenario and the same standardized operating procedure. Its function is to provide a robust statistical benchmark for the parameter. Calculated for the median; Calculation of median absolute deviation; To prevent stable terms with a denominator of zero, a very small positive number can be used, for example... For the first In the first standardized operating procedure Homoscaling results of individual psychological parameters; For the first In the first standardized operating procedure The original values ​​of individual psychological parameters or parameter values ​​calculated from step-level data slices; In order to be with the first The first standardized operating procedure and the first A set of reference samples corresponding to individual psychological parameters; To prevent stable terms with a denominator of zero, a very small positive number can be used, such as 0.001; These are standardized operation step numbers; For flight quality parameter serial numbers; The above formula serves to eliminate the differences in the dimensions of different parameters, enabling flight quality parameters and biopsychological parameters to be jointly analyzed at the same standardized operating procedure level.

[0030] In this embodiment, for comprehension and acceptance ability, it is preferable to establish a mapping relationship between practice scenarios and standardized operational steps related to level flight, ascent, descent, and turning, and comprehension and acceptance ability. Flight quality parameters and bio-psychological parameters characterizing comprehension and acceptance ability are defined within these standardized operational steps. Specifically, flight quality parameters include: whether the action operation is accurate, whether there are any actions reversing, reflecting mastery; whether there is a refueling action in slice one of the practice task slices; whether the flight quality data of slices one, two, and four are good; and the improvement rate of flight quality data in slice three relative to slice two, and slice five relative to slice four. Corresponding bio-psychological parameters include: changes in the theta band power of the prefrontal cortex in EEG parameters after the instruction is presented, used to help explain cognitive processing speed and working memory load; and the activation level and dynamic changes of dorsolateral prefrontal cortex oxyhemoglobin concentration in functional near-infrared spectroscopy parameters during the task comprehension stage, used to help explain the efficiency of higher cognitive resource retrieval. This embodiment aggregates the above flight quality parameters and bio-psychological parameters at the corresponding standardized operational step level and uses them as step-level parameter inputs corresponding to comprehension and acceptance ability.

[0031] In this embodiment, for adaptability, it is preferable to establish a mapping relationship between the identification steps and action execution steps in bird strike incidents and lock-on launch scenarios and adaptability. Specifically, flight quality parameters include: the response time from slice one to slice two (from incident identification to the first action execution), and the accuracy rate of action execution in slice two; wherein, the response time is calculated from the time difference of the corresponding event in the step-level data slice, and the action execution accuracy rate is obtained by comparing the action execution result with the corresponding operation target of the standardized operation step. The corresponding biopsychological parameters include: skin conductance response amplitude and rise time in the skin conductance parameters, used to help explain the intensity and speed of physiological stress arousal at the moment of incident; changes in high-frequency components and changes in the low-frequency to high-frequency ratio in the heart rate variability parameters, used to help explain the balance and recovery elasticity of the autonomic nervous system; and changes in the activity of the prefrontal cortex and cingulate cortex related frequency bands in the electroencephalogram parameters, used to help explain the cognitive control and conflict monitoring process. This embodiment achieves step-level characterization of adaptability in incident scenarios through the above mapping relationship.

[0032] In this embodiment, regarding control capability, it is preferable to establish a mapping relationship between the standardized operation steps in the pentagonal flight scenario and control capability, and define the degree of fit between the flight trajectory and the standard trajectory and the smoothness of attitude adjustment as flight quality parameters. The corresponding biopsychological parameters include: surface electromyographic activation timing, synergistic contraction rate, and muscle fatigue-related indicators of the joystick-related and pedal-related muscle groups in electromyographic parameters, used to help explain motor coordination and operational stability; and sensorimotor cortex-related rhythmic changes in electroencephalogram parameters, used to help explain the preparation and execution process of fine motor movements. Regarding spatial orientation capability, this embodiment preferably establishes a mapping relationship between the steps related to maintaining spatial position and stabilizing turning attitude during five-sided flight, as well as the steps related to altitude-holding orientation practice, and spatial orientation capability. The spatial position maintenance status of each slice, turning attitude stability, track deviation, heading deviation, correctness of correction direction, track deviation and heading deviation during altitude-holding orientation practice, and whether the landing gear is correctly deployed before the third turn are defined as flight quality parameters. The corresponding biopsychological parameters include fixation point, saccade path and dwell time in eye movement parameters. The analysis focuses on the scanning mode, frequency and rationality of spatial reference instruments such as attitude indicator, horizon indicator and heading indicator to help explain the visual spatial information integration strategy.

[0033] In this embodiment, for situational awareness capabilities, it is preferable to establish a mapping relationship between the standardized operational steps related to scene recognition, action execution, scene re-recognition, and action adjustment in complex weather flight scenarios and situational awareness capabilities, and define the response time and accuracy of handling different information as flight quality parameters. The corresponding biopsychological parameters include: eye movement parameters such as fixation point distribution and saccade sequence, used to characterize the systematic scanning pattern and attention allocation efficiency of multiple key information sources (including PFD, ND, engine parameters, and external environmental information); EEG parameters such as changes in parietal-occipital alpha band activity and whole-brain functional connectivity characteristics, used to help explain the process of constructing and maintaining situational awareness; and EEG parameters such as fluctuations in skin conductance levels at key decision points, used to help explain emotional arousal and decision-making pressure under high cognitive load. Regarding attention quality, this embodiment preferably establishes a mapping relationship between the standardized operation steps of continuous monitoring in the practice scenario and the five-sided flight scenario and attention quality. The flight data maintenance status under each slice and whether the correction of different deviations in altitude-fixed orientation practice is timely are defined as flight quality parameters. The corresponding biopsychological parameters include eye movement parameters such as fixation stability, blink frequency and pupil diameter changes. At the beginning of the locked launch slice, the duration from the appearance of the target to the first fixation point falling on the target area, the fixation time, the subsequent scanning pattern and pupil diameter changes are monitored to characterize the attention maintenance level, vigilance changes and cognitive load changes during the process from the appearance of the target to the formation of intentional attention and during continuous monitoring.

[0034] In this embodiment, for emotional stability, it is preferred to establish a mapping relationship between the lock-on launch, bird strike emergency, and multiple slices throughout the mission and emotional stability. Operational volatility and frequency of abnormal operations under multiple slices are defined as flight quality parameters. Corresponding biopsychological parameters include the baseline value and fluctuation amplitude of skin conductance in the skin conductance parameters, used to characterize the overall emotional arousal level and volatility across different slices of the mission, lock-on launch, and bird strike emergency, as well as the time-domain and frequency-domain trends of heart rate variability parameters, used to help explain autonomic nervous system regulation ability and long-term stress adaptation. For cases where the same standardized operation step can correspond to multiple capability elements, this embodiment establishes a mapping relationship between the standardized operation step and each capability element, and can perform consistency verification on the pre-established mapping relationship. Here, consistency verification refers to using the current subject's flight quality parameters and biopsychological parameters under the standardized operation step to determine whether the pre-established mapping relationship is supported by data. Its function is to output a consistency marker or activation marker without changing the mapping relationship pre-established by the flight expert. This embodiment can use the following methods for consistency verification and activation determination: in, For the first The standardized operating procedure for the first The mapping consistency check value of each capability element; For the capability element serial number; To achieve the first step, under the pre-defined mapping relationship set by flight experts... The first standardized operating step is used to characterize the... A set of flight quality parameter numbers for each capability element; To achieve the first step, under the pre-defined mapping relationship set by flight experts... The first standardized operating step is used to characterize the... A set of biopsychological parameter numbers for each ability element; and These represent the number of elements in the corresponding set; For the first The threshold for determining the mapping consistency of each capability element; For the first A set of mapping consistency verification values ​​for each capability element in the reference sample; This is an indicator function; it takes the value 1 when the condition is true, and 0 otherwise. For the first The standardized operating procedure for the first An activation flag is assigned to each capability element, indicating whether the standardized operational step provides effective representational support for the corresponding capability element under the current subject data. For example, when the task identification step of locking launch corresponds to strain capability and the response time, skin conductance response, and heart rate variability of the step all show significant changes. Increase and more likely to .

[0035] In this embodiment, when forming the capability representation vector corresponding to each standardized operation step based on the flight quality parameters and biopsychological parameters, the parameters are first scaled to the same size, then grouped and aggregated according to the mapping relationship, and finally concatenated to form capability element-level sub-vectors to obtain the step-level capability representation vector. Here, the capability representation vector refers to a structured data representation at the standardized operation step level, consisting of flight quality parameter representation values, biopsychological parameter representation values, and their corresponding relationship markers. Its function is to serve as input for hierarchical scoring, quantitative correlation, and fusion modeling in subsequent steps 400 to 700. This embodiment can form the capability representation vector in the following manner: in, For the first The standardized operating procedure is for the first The flight quality parameter characterization values ​​of each capability element; For the first The standardized operating procedure is for the first The biopsychological parameter representation values ​​of each ability element; For the first The standardized operating procedure is for the first The cross-channel consistency characterization value of each capability element is used to characterize the degree of consistency between the flight quality parameter characterization value and the biopsychological parameter characterization value. The closer the value is to 1, the higher the degree of consistency. To prevent stable terms with a denominator of zero, a very small positive number can be used; For the first The standardized operating procedure is for the first Sub-vectors of each capability element; In order to be with the first A set of sequence numbers representing the capability elements that establish mapping relationships through standardized operating steps; For the first Capability representation vector corresponding to each standardized operation step; symbol This indicates that vectors are concatenated in the order of capability elements. To ensure a clear correspondence between input data, processing procedures, and output results, in step 300 of this embodiment, standardized operation steps are used as units. Based on pre-established mapping relationships, the flight quality parameters and biopsychological parameters in the step-level data slices are organized and calculated, and then a capability representation vector is output. When this embodiment uses an algorithm for consistency verification, the algorithm input is the flight quality parameters and biopsychological parameters in the step-level data slice corresponding to the standardized operation step, and the algorithm output is the consistency verification result and activation flag. The algorithm output is only used to characterize the degree of support of the current data for the preset mapping relationship and does not replace the mapping relationship pre-established by flight experts.

[0036] Further, in this embodiment, step 400 is used to perform stratified scoring of the subject's flight performance based on step-level data slices, obtaining a completion score and a quality score. Here, stratified scoring refers to a scoring method that divides flight performance into two levels: completion evaluation and quality evaluation. Its function is to simultaneously reflect whether the subject has achieved the basic operational goals of each standardized operation step, as well as the operational precision and proficiency after achievement. The completion score is used to characterize task completion, serving as a benchmark indicator of competence. The quality score is used to characterize the deviation level of flight parameters from the ideal trajectory or standard parameters, the smoothness of operation, and response efficiency, thereby continuously distinguishing between good and bad performance. The inputs to step 400 are step-level data slices, the flight parameters in the step-level data slices, and the ideal trajectory or standard parameters corresponding to each standardized operation step. The outputs are the completion score, the quality score, and the quality dimension indicators used to generate the quality score.

[0037] In this embodiment, the achievement of basic operational goals by subjects in each standardized operation step is determined based on step-level data slices, resulting in a completion score. The achievement of basic operational goals here refers to whether the necessary operations corresponding to a standardized operation step are completed as required, whether the preset state requirements of the standardized operation step are met, and whether there are any obvious errors that conflict with the goals of the standardized operation step. Its function is to provide a completion assessment for the standardized operation step. This embodiment can calculate the completion score in the following manner: in, For the first The completion determination value of each standardized operation step; For the first A set of basic operational objectives that were not achieved in a standardized operational step, wherein the elements of the set are derived from event records, status records or flight parameter determination results in the data slice at that step level; To complete the score; This refers to the total number of standardized operational steps involved in the scoring within the framework of this standardized simulated flight assessment task. For example, if an assessment contains 20 standardized operational steps, then this can be taken as... ;symbol The meaning is the same as the definition in step 300 above. The purpose of the above formula is to uniformly convert the completion judgment of each standardized operation step into a summable benchmark score, so as to form a baseline indicator reflecting the completion of the task.

[0038] In this embodiment, the flight parameters in the step-level data slices are compared with the ideal trajectory or standard parameters to obtain the comparison results. Here, the ideal trajectory or standard parameters refer to a pre-defined reference flight trajectory, reference state curve, or target parameter range for each standardized operation step, serving as a comparison benchmark for quality evaluation. When a standardized operation step's primary objective is trajectory tracking, the ideal trajectory is preferred for comparison; when a standardized operation step's primary objective is state maintenance or action achievement, the standard parameters are preferred for comparison. This embodiment can first construct a time-by-time normalized comparison result: in, For the first In the first standardized operating procedure The flight parameter at the ... Normalized comparison results at each sampling time; For the first step-level data slice The flight parameter at the ... The actual value at each sampling time; In order to be with the first The ideal trajectory value or standard parameter value corresponding to each standardized operation step, when compared using standard parameters. It can be a constant value or a piecewise constant value; For the first In the first standardized operating procedure The robust scaling parameters of each flight parameter in the reference sample are derived from the statistical results of the flight parameters corresponding to historical samples under the same assessment scenario and the same standardized operating procedures. To prevent stable terms with a denominator of zero, a very small positive number can be used, such as 0.001; For flight parameter serial numbers; This refers to the sampling time sequence within the data slice at this step level. The purpose of the above formula is to unify the deviations of different flight parameters to a comparable scale, in order to support the subsequent calculation of quality dimension indicators.

[0039] In this embodiment, at least one quality dimension indicator among deviation value, operational smoothness, and response efficiency is calculated based on the comparison results. Here, quality dimension indicators refer to indicators used to characterize different aspects of operational quality, further calculated from the comparison results or step-level data slices. Their function is to break down the quality evaluation into multiple interpretable dimensions. This embodiment preferably calculates deviation value, operational smoothness, and response efficiency simultaneously, and can adopt the following method: in, For the first The deviation value index of each standardized operation step is used to characterize the overall deviation level of the actual flight parameters relative to the ideal trajectory or standard parameters in that step. For the first The original value of the smoothness index of the standardized operation steps is used to characterize the smoothness of the changes in flight parameters in the step. The second-order difference component is used to characterize the severity of parameter fluctuations. The larger the value, the less smooth the operation. For the first The raw value of the response efficiency index for each standardized operation step is used to characterize the efficiency level of the key action or state response in that step relative to the reference response time. The larger the value, the slower the response. For the first Each standardized operation step corresponds to a set of sampling times for a step-level data slice; This is the set of sampling times from which second-order difference calculations can be performed; The first A set of flight parameter numbers used to calculate deviation values, operational smoothness, and response efficiency in a standardized operating procedure is derived from the scoring rule definition of that standardized operating procedure. For the first In the first standardized operating procedure Robust scaling parameters of the second-order difference components of each flight parameter in the reference sample; For the first In the first standardized operating procedure The response time of a response event from the moment it is triggered to the moment the target state is reached; This is the reference response time for the corresponding response event in the reference sample; and To prevent stable terms with a denominator of zero, a very small positive number can be used, such as 0.001; (Signature) The meaning is the same as the definition in step 300 above. The purpose of the above formula is to further convert the flight parameter comparison results into interpretable quality dimension indicators.

[0040] In this embodiment, quality scores are generated based on quality dimension indicators. To facilitate the comprehensive analysis of different quality dimensions, this embodiment first converts the original indicators corresponding to deviation values, operational smoothness, and response efficiency into unidirectional scoring values, and then determines the weights of each quality dimension based on its discriminative power in the reference sample. The scientific weights referred to here are the quality dimension weights automatically determined by the statistical results of the reference sample; their function is to reduce the influence of manually assigned weights on the scoring results. This embodiment can adopt the following method: in, , , The first Each standardized operating step corresponds to a score for deviation value, operation smoothness, and response efficiency. The larger the value, the better the quality. For the first In the first standardized operating procedure The ability to differentiate across quality dimensions; and The first In the first standardized operating procedure The quality dimension is a set of indicator samples in the high-performance reference sample group and the low-performance reference sample group, which can be divided according to historical assessment results or training evaluation results. For the first In the first standardized operating procedure The set of indicator samples for each quality dimension in all reference samples; For the first In the first standardized operating procedure Weights for each quality dimension; For the first The set of quality dimension indices for standardized operating steps involved in the quality score calculation; for example, when using deviation value, operational smoothness, and response efficiency simultaneously, one could take... ; For the first Step-level quality score for each standardized operating procedure; This is the quality score for this assessment; and To prevent stable terms with a denominator of zero, a very small positive number can be used, such as 0.001; (Signature) The meaning is the same as the definition in step 300 above. The function of the above formula is to automatically determine the weights of the quality dimensions and generate quality scores while preserving the interpretability of the quality dimensions, and then... Constrain the step-level quality score to ensure that standardized operation steps that fail to achieve the basic operation objective do not generate undue gains in the quality score.

[0041] In this embodiment, step 400 uses explicit judgment rules and explicit calculation formulas to process the flight parameters in the step-level data slice. The correspondence between the input, processing, and output is clear. The input includes the step-level data slice, flight parameters, and ideal trajectory or standard parameters. The intermediate results include at least the completion judgment value, time-by-time normalized comparison results, and quality dimension indicators. The output is the completion score. and quality When a standardized operating procedure defines only some quality dimension indicators such as deviation value, operational smoothness, and response efficiency, this embodiment calculates the corresponding unidirectional score value and quality dimension weight only for the defined and calculable quality dimensions, and generates a step-level quality score for the standardized operating procedure after normalizing the weights of the participating dimensions. Through the above-mentioned hierarchical scoring process, this embodiment retains the benchmark requirements for basic operation compliance while continuously distinguishing high-precision and high-efficiency operational performance, thereby meeting the comprehensive evaluation needs for safety baseline and fine control performance in flight selection and training scenarios.

[0042] In this embodiment, step 500 is used to construct flight performance features based on flight parameters, completion score, and quality score, and to extract physiological and psychological features corresponding to the ability elements from multimodal physiological and psychological data. Here, flight performance features refer to a structured feature representation composed of flight parameters and their step-level and task-level scoring results, whose function is to characterize the subject's overt flight performance within a standardized simulated flight assessment task framework. Physiological and psychological features refer to a structured feature representation extracted from multimodal physiological and psychological data and organized according to ability elements, whose function is to characterize changes in the internal physiological and psychological state related to each ability element. Multimodal physiological and psychological data includes one or more of eye-tracking data, electroencephalogram (EEG) data, electrodermal conductance data, electromyography (EMG) data, heart rate variability data, and functional near-infrared spectroscopy (FIR) data. The inputs to step 500 are step-level data slices, flight parameters, completion score, quality score, and multimodal physiological and psychological data; the outputs are flight performance features and physiological and psychological features, which are used by step 600 to establish quantitative correlations.

[0043] In this embodiment, when combining flight parameters, completion scores, and quality scores to construct flight performance features, it is preferable to first statistically summarize the flight parameters at the standardized operation step level, and then concatenate them with the step-level score and task-level score to obtain the flight performance feature vector for subsequent quantitative correlation modeling. This embodiment can adopt the following method: in, For flight performance characteristics; For the first Each standardized operating procedure corresponds to a step-level flight performance sub-feature; For the first A set of flight parameter numbers for each standardized operational step involved in the construction of flight performance characteristics, the set being determined according to the scoring rules of that standardized operational step; symbols This indicates that vectors are concatenated in a preset order; the meanings of the other symbols are the same as in step 400 above. The purpose of the above formula is to unify the step-level statistical results of flight parameters with the completion score and quality score into flight performance characteristics that can be directly used to establish quantitative correlations.

[0044] In this embodiment, when extracting physiological and psychological features corresponding to ability elements from multimodal physiological and psychological data, it is preferable to continue using the mapping relationship between standardized operation steps and ability elements in step 300, as well as the step-level ability representation results formed by physiological and psychological parameters, and to aggregate them according to ability elements to obtain ability element-level physiological and psychological features. This embodiment can adopt the following approach: in, For the first The physiological and psychological characteristics corresponding to each ability element; In order to be with the first A set of standardized operational steps for establishing mapping relationships between capability elements; , , The first one formed in step 300 respectively The standardized operating procedure is for the first The physiological and psychological parameter representation values, cross-channel consistency representation values, and activation markers for each ability element have the same meaning as in step 300 above. The function of the above formula is to convert the step-level representation results of multimodal physiological and psychological data into ability element-level physiological and psychological characteristics, thereby enabling a one-to-one correspondence between physiological and psychological characteristics and subsequent ability element scoring objects.

[0045] In this embodiment, step 600 is used to establish quantitative correlations between flight performance characteristics and core criteria, as well as between physiological and psychological characteristics and core criteria, using actual flight mission performance as the core criterion. Here, the core criterion refers to the target quantity used to characterize actual flight mission performance, serving as a unified reference for modeling the quantitative correlation between the two types of characteristics. Quantitative correlation refers to the process of establishing a calculable correspondence between characteristics and core criteria through an explicit model, its function being to obtain a measure of the explanatory or predictive ability of each channel to the core criterion. In this embodiment, the model building stage is processed on a per-subject-sample basis, introducing sample index i; for the k-th ability element, quantitative correlation models are constructed between flight performance channel characteristics and physiological and psychological channel characteristics and core criteria, respectively, for step 700 to determine the fusion weights of the dual-channel indicators.

[0046] In this embodiment, when establishing quantitative correlations between flight performance characteristics and physiological and psychological characteristics and the core criterion, an interpretable explicit linear model is preferably used, and the correlation strength index is used to characterize the degree of quantitative correlation between each channel and the core criterion. This embodiment can be implemented in the following manner: in, For the first The sample is for the first The flight performance channel characteristics of each capability element can be derived from the flight performance characteristics. According to the The feature index corresponding to each capability element is extracted; For the first The sample is for the first Physiological and psychological pathway characteristics of each ability element; and These are the core criterion estimates output by the two-channel quantization correlation model; , , , The first Each capability element corresponds to the intercept term and coefficient vector of the quantitative correlation model; , , The first The estimated vectors of flight performance channels, physiological and psychological channels, and core criterion vectors for each capability element on the sample set; Corr Calculate the correlation coefficient; , The first The quantitative correlation strength index of each capability element in the flight performance channel and the physiological and psychological channel; , To prevent stable terms with a denominator of zero, a very small positive number can be used, such as 0.001. The purpose of the above formula is to obtain a correlation strength index that can be used for subsequent weight determination while establishing quantitative correlations with the core criterion in both channels.

[0047] In this embodiment, step 700 is used to determine the fusion weights of the dual-channel indicators based on quantitative correlation. Then, based on the fusion weights of the dual-channel indicators, an integrated modeling and score synthesis are performed on each capability element to obtain the score for each capability element. Here, the flight performance channel weight and the physiological and psychological channel weight refer to the weights of the two channels, respectively. The channel-level weights of each capability element across the two channels; the dual-channel indicator fusion weight refers to the fusion weight structure formed by combining the channel-level weights with the intra-channel feature weights. Its function is to ensure that the score synthesis simultaneously reflects both the channel contribution and the intra-channel indicator contribution. This embodiment can determine the channel-level weights and the dual-channel indicator fusion weights in the following manner: in, , The first The weighting of each capability element in the flight performance channel and the weighting of the physiological and psychological channel; , , To prevent stable terms with a denominator of zero, a very small positive number can be used. , The first The set of feature numbers of each capability element involved in modeling in the flight performance channel and the physiological and psychological channel; , These are the first numbers in the coefficient vector of the corresponding quantization correlation model. The and the first One coefficient; , These are the feature weights within the channel; , These are the feature weight vectors within the channel; For the first The dual-channel index fusion weights of each capability element. The above formula serves to automatically determine the weights of the two channels based on the quantitative correlation strength, and further form a fusion weight structure that can be used for the synthesis of integrated laws and common numbers.

[0048] In this embodiment, when performing integrated modeling and score synthesis for each capability element based on the dual-channel index fusion weights, it is preferable to first convert the output of the two-channel quantitative correlation model into channel scores of the same scale, and then synthesize them according to channel-level weights to obtain the scores for each capability element. This embodiment can adopt the following method: in, , The first The low and high reference boundaries of the core criterion for each capability element are used to convert the output of the two-channel model to a unified score range. , The first The sample at the th Flight performance channel scores and physiological and psychological channel scores for each capability element; , To prevent stable terms with a denominator of zero, a very small positive number can be used, such as 0.001; This is a truncation function used to restrict the input to the range of 0 to 1; For the first The sample at the th The above formula serves to process the outputs of the two-channel model on the same scale and synthesize the scores using the core criterion as a unified reference. This ensures that the scores of each capability element are always formed around the actual flight mission performance, and that the source of the scores can be traced back to the quantitative correlation and corresponding fusion weights of the two channels.

[0049] In this embodiment, step 800 includes three processing stages: collecting test data from a real group of test subjects under the same assessment scenario; establishing statistical norms for each ability element based on the test data; and outputting flight potential assessment results based on the statistical norms. The real group of test subjects refers to a collection of actual test subjects used for similar flight selection or training assessments, serving to provide a baseline for population ability distribution consistent with actual application scenarios. The same assessment scenario refers to assessment conditions consistent with the standardized simulated flight assessment task framework used in the current test subject's assessment, and with consistent assessment scenario composition, standardized operation step sequence, and scoring rule version, ensuring statistical comparability of ability element scores among different test subjects. The inputs to step 800 include at least the scores of each ability element output in step 700, the assessment scenario identifier of the current test subject, the version of the standardized operation step sequence, the version of the scoring rule, and the personnel category identifier; the outputs of step 800 include at least the norm positioning results for each ability element and the flight potential assessment results determined based on the norm positioning results.

[0050] In this embodiment, when collecting test data from actual subject groups under the same assessment scenario, at least the following fields are collected and stored: subject identifier, test time, assessment scenario identifier, standardized operation step sequence version, scoring rule version, personnel category identifier, scores of each ability element, and necessary test validity identifier. The personnel category identifier is used to distinguish between trainee subjects, experienced pilot subjects, or other preset categories, in order to establish stratified statistical norms later. The test validity identifier is used to indicate whether the test meets data quality requirements. After collection, this embodiment performs consistency verification and cleaning processing on the test data, including at least: removing records missing scores for any key ability element, removing records where the assessment scenario identifier is inconsistent with the target norm scenario, removing records where the standardized operation step sequence version is inconsistent, removing records where the scoring rule version is inconsistent, and reviewing and removing or separately archiving abnormal score records that exceed a reasonable range. After the above processing, a valid test dataset for establishing statistical norms is obtained, thereby ensuring that subsequent norm establishment and evaluation results output are based on data with the same caliber and rules.

[0051] In this embodiment, statistical norms for each ability element are established based on the effective test dataset. The statistical norms refer to reference standards established for the distribution pattern of a certain ability element score in the actual subject population. Their function is to convert the individual subject's ability element score into its relative position and uniform scale score within the population. For each ability element, this embodiment establishes at least the following norm content: the ranking result of the ability element score, a quantile position lookup table (as shown in Table 2), a uniform scale conversion table (as shown in Table 3) or conversion parameters, and the boundary of the individual ability level interval. The quantile position lookup table is used to map the input ability element score to the relative position of the ability element within the corresponding population; the uniform scale conversion table or conversion parameters are used to convert different ability element scores to the same score scale for subsequent integration; the boundary of the individual ability level interval is used to convert the positioning result into an individual ability level. To ensure accurate norm retrieval, this embodiment binds and stores each set of statistical norms with the corresponding assessment scenario identifier, standardized operation step sequence version, scoring rule version, personnel category identifier, and norm version number, and records the sample size and sample time range used to establish the norm.

[0052] Table 2. Comparison of quantile positions Table 3 Unified Scale Conversion Table In this embodiment, when locating the scores of each ability element based on the statistical norms, the matching statistical norms are first retrieved and called according to the current subject's assessment scenario identifier, standardized operation step sequence version, scoring rule version, and personnel category identifier. If multiple available norm versions exist, the statistical norm with the valid version status and the latest update time is prioritized. Subsequently, the scores of each ability element obtained in step 700 are input item by item into the corresponding statistical norm of the ability element to obtain the quantile position, uniform scale score, and individual ability level of the ability element. Here, the quantile position refers to the relative ranking position of the ability element score in the matched population, which reflects the current subject's level relative to the same group of people; the uniform scale score refers to the result after converting the scores of different ability elements to the same scale, which provides a weighted composite input for comprehensive potential assessment; the individual ability level refers to the level result obtained based on the level interval boundary of the ability element, which supports the subsequent identification of advantageous and weak ability elements. For any capability element that fails to match a statistical norm, this embodiment outputs a norm missing flag and stops the level determination of that capability element to avoid outputting unreliable results under norm mismatch conditions.

[0053] In this embodiment, when determining the flight potential assessment result based on the positioning result, a pre-stored assessment rule table is used for processing. The assessment rule table includes at least a list of capability elements, a comprehensive weight corresponding to each capability element, a comprehensive potential level division range, a list of key capability elements, and low-level restriction conditions. Specifically, this embodiment first weights and synthesizes the uniform scale scores of each capability element according to the comprehensive weight to obtain a comprehensive potential score; then, it determines the comprehensive potential level based on the comprehensive potential level division range; and then, it verifies the comprehensive potential level based on the list of key capability elements and the low-level restriction conditions. When any key capability element level is lower than a preset lower limit, the comprehensive potential level is restricted, downgraded, or marked as having a training risk. Further, this embodiment identifies dominant and weak capability elements based on the comparison results of the quantile position of each capability element with preset high and low quantile thresholds. Capability elements above the high quantile threshold are marked as dominant capability elements, and capability elements below the low quantile threshold are marked as weak capability elements. The final flight potential assessment results should include at least the scores of each capability element, the percentile position of each capability element, the unified scale score of each capability element, the individual capability level of each capability element, the comprehensive potential score, the comprehensive potential level, and the identification results of the dominant and weak capability elements, so as to clarify the correspondence between the input source, processing process and output conclusion of the assessment results.

[0054] The beneficial effects of this invention are as follows: This invention constructs a standardized simulated flight assessment task framework under the guidance of flight experts, and decomposes and standardizes the assessment scenarios to form a standardized sequence of operational steps. Based on this standardized sequence, it performs fine-grained data slicing of flight parameters and multimodal physiological and psychological data, aligning each step. This transforms continuous, mixed, multi-source time-series data into step-level data slices that correspond one-to-one with specific operational semantics. Therefore, the evaluation process no longer focuses on the overall task result but extends to the specific performance of each standardized operational step, significantly improving the structure, comparability, and traceability of the evaluation process, and providing a unified data foundation for subsequent capability element analysis.

[0055] This invention defines flight quality parameters and biopsychological parameters for each standardized operational step, establishes a mapping relationship between standardized operational steps and capability elements, and constructs flight performance characteristics by combining completion scores and quality scores. Simultaneously, it extracts biopsychological characteristics corresponding to capability elements from multimodal biopsychological data, and establishes quantitative correlations between the two types of characteristics and the core criterion using actual flight mission performance as the core criterion. This processing method can reflect both explicit operational performance using flight parameters and internal state changes using multimodal biopsychological data, and uses the core criterion as a unified reference for dual-channel fusion, thereby improving the systematicness, robustness, and interpretability of the evaluation results, and reducing biases caused by single-indicator evaluation or judgment based solely on experience.

[0056] This invention further establishes statistical norms for each ability element based on test data from actual subject groups under the same assessment scenario. After locating the scores of each ability element based on these statistical norms, it outputs flight potential assessment results, enabling the assessment conclusions to be interpreted within the context of a real population distribution. Compared to subjectively setting fixed thresholds, this scheme can more objectively reflect the relative level of the subject within the same population group and supports the quantile positioning, level determination, and identification of dominant and weak ability elements for each ability element. This provides consistent and practically valuable data for flight selection, tiered management, and targeted training.

[0057] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0058] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. Furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for flight potential assessment that integrates flight performance and multimodal data, characterized in that, include: Under the guidance of flight experts, a standardized simulated flight assessment task framework was constructed, and the standardized simulated flight assessment task framework was decomposed into scenarios and standardized in steps to obtain a standardized operation step sequence. Flight parameters and multimodal physiological and psychological data were collected from the subjects when they performed the standardized simulated flight assessment task framework. Based on the standardized operation step sequence, refined data slices were performed to align the steps, resulting in step-level data slices corresponding to each standardized operation step in the standardized operation step sequence. For each standardized operation step in the sequence of standardized operation steps, corresponding flight quality parameters and bio-psychological parameters are defined, a mapping relationship between each standardized operation step and capability elements is established, and a capability representation vector corresponding to each standardized operation step is formed based on the flight quality parameters and the bio-psychological parameters. The flight performance of the subjects is stratified and scored based on the step-level data slices to obtain a completion score and a quality score; wherein, the quality score is determined based on the comparison results of the flight parameters with the ideal trajectory or standard parameters; Flight performance characteristics are constructed based on the flight parameters, the completion score, and the quality score, and physiological and psychological characteristics are extracted from the multimodal physiological and psychological data. Using actual flight mission performance as the core criterion, quantitative correlations are established between the flight performance characteristics and the core criterion, as well as between the physiological and psychological characteristics and the core criterion. Based on the quantitative correlation, the fusion weight of the dual-channel indicators is determined, and the integrated modeling and score synthesis of each capability element are performed according to the fusion weight of the dual-channel indicators to obtain the capability element score. Statistical norms for the aforementioned ability elements are established based on test data from actual subject groups under the same assessment scenario, and flight potential assessment results are output based on these statistical norms.

2. The flight potential assessment method based on flight performance and multimodal data fusion according to claim 1, characterized in that, Under the guidance of flight experts, a standardized simulated flight assessment task framework was constructed. This framework was then broken down into scenarios and standardized in terms of steps, resulting in a standardized sequence of operational steps, including: Under the guidance of the flight experts, assessment scenarios for flight potential evaluation were determined; these assessment scenarios included practice, five-way flight, lock-on launch, bird strike incidents, and flight in complex weather conditions. Each of the assessment scenarios is decomposed to obtain a scenario slice corresponding to each of the assessment scenarios; Standardize the steps for each of the aforementioned scenario slices to form standardized operation steps corresponding to each of the aforementioned assessment scenarios; The standardized operation steps are arranged according to the task execution order of each assessment scenario to obtain the standardized operation step sequence.

3. The flight potential assessment method based on flight performance and multimodal data fusion according to claim 2, characterized in that, Each of the aforementioned assessment scenarios is decomposed into scenario slices, including: The exercises are broken down into scenarios, resulting in level flight exercise, ascent exercise 1, ascent exercise 2, turning exercise 1, turning exercise 2, altitude hold and orientation exercise 1, and altitude hold and orientation exercise 2. The five-sided flight scenario is broken down into the following segments: before the first turn, before the second turn, before the third turn, before the fourth turn, glide, and landing. The locking and launching steps are standardized to obtain standardized operation steps corresponding to task identification, action execution, and stable recovery. The bird strike incident is standardized to obtain standardized operation steps corresponding to incident identification, action execution, and stable recovery. The steps of the complex weather flight are standardized to obtain standardized operation steps corresponding to scene recognition, action execution, scene re-recognition and action adjustment.

4. The flight potential assessment method based on flight performance and multimodal data fusion according to claim 1, characterized in that, Flight parameters and multimodal physiological and psychological data were collected from the subjects when they performed the standardized simulated flight assessment task framework. Based on the standardized operation step sequence, refined data slices were performed with step alignment to obtain step-level data slices corresponding to each standardized operation step in the standardized operation step sequence, including: Simultaneously collect flight parameters and multimodal physiological and psychological data of the subjects when they perform the standardized simulated flight assessment task framework, so that the flight parameters and the multimodal physiological and psychological data constitute multi-source time-series data; The multi-source time-series data is aligned according to the standardized operation step sequence. The multi-source time-series data after step alignment is refined into data slices according to the operation semantic units corresponding to each standardized operation step in the sequence of standardized operation steps, and the step-level data slices corresponding to each standardized operation step are output.

5. The flight potential assessment method based on flight performance and multimodal data fusion according to claim 1, characterized in that, For each standardized operation step in the sequence of standardized operation steps, corresponding flight quality parameters and biopsychological parameters are defined. A mapping relationship between each standardized operation step and capability elements is established. Based on the flight quality parameters and the biopsychological parameters, a capability representation vector corresponding to each standardized operation step is formed, including: For each standardized operation step in the sequence of standardized operation steps, flight quality parameters and bio-psychological parameters are defined respectively; Establish a mapping relationship between each of the standardized operating steps and the capability elements, so that each of the standardized operating steps corresponds to at least one capability element; the capability elements include comprehension and acceptance ability, adaptability, control ability, spatial orientation ability, situational awareness ability, attention quality, and emotional stability. Based on the flight quality parameters and bio-psychological parameters corresponding to each of the standardized operating steps, a capability representation vector corresponding to each of the standardized operating steps is formed.

6. The flight potential assessment method based on flight performance and multimodal data fusion according to claim 1, characterized in that, Based on the step-level data slices, the flight performance of the subjects is stratified and scored to obtain completion scores and quality scores, including: Based on the step-level data slices, the participants' achievement of basic operational goals in each of the standardized operational steps is determined to obtain a completion score. The flight parameters in the step-level data slices are compared with the ideal trajectory or standard parameters to obtain the comparison results. Based on the comparison results, at least one quality dimension indicator among deviation value, operational smoothness, and response efficiency is calculated. A quality score is generated based on the aforementioned quality dimension indicators.

7. The flight potential assessment method based on flight performance and multimodal data fusion according to claim 1, characterized in that, Flight performance characteristics are constructed based on the flight parameters, the completion score, and the quality score, and physiological and psychological characteristics are extracted from the multimodal physiological and psychological data, including: The flight parameters, the completion score, and the quality score are combined to construct flight performance characteristics; Extract physiological and psychological features corresponding to the ability elements from the multimodal physiological and psychological data; The multimodal physiological and psychological data includes one or more of the following: eye-tracking data, electroencephalogram (EEG) data, electrodermal conductance data, electromyography (EMG) data, heart rate variability data, and functional near-infrared spectroscopy (FIR) data.

8. The flight potential assessment method based on flight performance and multimodal data fusion according to claim 1, characterized in that, Using actual flight mission performance as the core criterion, quantitative correlations are established between the flight performance characteristics and the core criterion, as well as between the physiological and psychological characteristics and the core criterion, including: The actual flight mission performance is defined as the core criterion. A quantitative correlation is established between the flight performance characteristics and the core criteria based on the flight performance characteristics; A quantitative correlation is established between the physiological and psychological characteristics and the core criteria based on the aforementioned physiological and psychological characteristics.

9. The flight potential assessment method based on flight performance and multimodal data fusion according to claim 1, characterized in that, Based on the quantitative correlation, the fusion weights of the dual-channel indicators are determined, and the integrated modeling and score synthesis of each capability element are performed according to the fusion weights of the dual-channel indicators to obtain the capability element scores, including: Based on the quantitative correlation between the flight performance characteristics and the core criteria, the flight performance channel weights are determined; Based on the quantitative correlation between the physiological and psychological characteristics and the core criteria, the weights of the physiological and psychological channels are determined. The weights for the fusion of the dual-channel indicators are determined based on the weights of the flight performance channel and the weights of the physiological and psychological channel. Based on the weighted fusion of the dual-channel indicators, each capability element is modeled and its score is synthesized in an integrated manner to obtain the score of each capability element.

10. The flight potential assessment method based on flight performance and multimodal data fusion according to claim 1, characterized in that, Statistical norms for the aforementioned ability elements are established based on test data from actual test subjects under the same assessment scenario, and flight potential assessment results are output based on these statistical norms, including: Collect test data from actual test subjects under the same assessment scenario; Statistical norms for each of the aforementioned capability elements are established based on the test data. The scores of each capability element are located based on the statistical norms, and the flight potential assessment result is determined based on the location results.

Images