Pilot multi-task processing efficiency evaluation method under complex situation

Abstract

The invention belongs to the technical field of aviation safety management and pilot efficiency evaluation, relates to a pilot multi-task processing efficiency evaluation method under a complex situation, and aims to solve the problem that a traditional method does not comprehensively consider a real flight environment and has a blank in the field of multi-task processing efficiency evaluation. The method comprises the steps that flight parameter data, eye movement data and task data of a pilot during flight are acquired and preprocessed; screening and calculating from three dimensions of attention distribution, work memory and task switching and burst task processing efficiency to obtain a key parameter data set; performing flight stage division based on flight parameter data in combination with the flight stage transfer atlas and the weighted multi-dimensional matching distance; presetting a weight matrix according to cognitive requirements of different flight stages; and matching a preset weight matrix according to a flight stage division result, and carrying out weighted fusion on each key parameter value to obtain a final performance evaluation value. In combination with multi-source data analysis, efficient evaluation of pilot multi-task processing efficiency is realized.

Description

Technical Field

[0001] The present invention belongs to the technical fields of aviation safety management and pilot performance evaluation, and specifically relates to a method for evaluating pilot multi-tasking performance in complex situations. Background Art

[0002] In the aviation field, scientific evaluation of pilot task processing efficiency is a core component of ensuring flight safety and optimizing cockpit operational processes. Currently, relevant patents exist related to pilot workload assessment: for example, the patent "System and Method for Pilot Workload Assessment Using Multi-Source Data" (CN118691142A) utilizes real-time physiological monitoring data to identify pilots' cognitive load capacity, analyze their attention allocation and decision-making abilities, and then assess the cognitive load status during real-time tasks, outputting workload assessment results. The patent "Online Assessment Method for Pilot Task Load Based on Eye Movement and Physiological Data" (CN113229791A) collects eye movement and physiological data, classifies the data using a support vector machine hierarchical combination model, and constructs a five-category online assessment model for task load, which calculates and provides feedback on task load levels in real time.

[0003] However, existing technologies have limitations. First, their assessment dimensions are relatively narrow, focusing primarily on pilot skill or workload during a single task. These assessments fail to fully encompass the complex multitasking scenarios encountered in actual flight, such as flight status monitoring, responding to control commands, and handling equipment failures. Furthermore, they lack a systematic assessment of multitasking performance. Second, while existing technologies incorporate physiological data to categorize task load, determining whether a pilot is at low, relatively low, medium, relatively high, or high task load levels, these assessments focus solely on workload intensity and fail to delve into the assessment of pilot performance in multitasking scenarios (e.g., task prioritization, parallel processing efficiency, and error rate control). In summary, existing technologies fail to comprehensively assess pilots' multitasking performance and fail to truly reflect their comprehensive task handling capabilities in complex flight scenarios. Summary of the Invention

[0004] To address the aforementioned issues in the prior art, namely, the inability of the prior art to fully reflect a pilot's actual task processing efficiency in complex scenarios involving multiple tasks due to a single evaluation dimension and the lack of multi-tasking performance evaluation, thereby affecting flight safety and efficiency, the present invention, in its first aspect, proposes a method for evaluating a pilot's multi-tasking performance in complex scenarios. The method comprises the following steps: Acquiring and preprocessing the pilot's flight parameter data, eye movement data, and task data during the flight to obtain preprocessed data; the preprocessing includes time synchronization, data cleaning, segmentation labeling, and region of interest division; the time synchronization includes using a sliding time window and cognitive alignment synchronization method to achieve time synchronization of multi-source data; Extracting and quantifying features of the eye movement data and task data in the pre-processed data from three dimensions: attention allocation, working memory and task switching, and sudden task processing efficiency, to obtain key parameter values, and thus obtain a key parameter data set; According to the different cognitive requirements of each flight phase, a weight matrix is ​​preset using the analytic hierarchy process and applied to the key parameter data set; Combining the flight phase transfer map and weighted multi-dimensional matching distance, the flight phase division is performed using the pre-processed flight parameter data to obtain the flight phase division results. The preset weight matrix is ​​matched according to the flight phase division result, and the key parameter values ​​are combined and weighted fusion is performed to obtain the final multi-task processing performance evaluation value.

[0005] In some preferred embodiments, the eye movement data includes fixation time, scan frequency, and pupil diameter; The task data includes burst task processing data, multi-task parallel processing data, task switching data, and task priority tags; The burst task processing data includes first fixation time, attention recovery time, switching delay, and error rate.

[0006] In some preferred embodiments, the segmentation labeling includes labeling by task stage and labeling by subtask; The region of interest is divided based on the cockpit functional areas, including the primary flight display, navigation display, and radio control panel.

[0007] In some preferred embodiments, the quantitative data set includes attention allocation parameters, working memory and task switching parameters, and burst task processing efficiency parameters; The attention allocation parameters include the normalized regional fixation time ratio, normalized saccade frequency, pupil diameter change, normalized first fixation time, and attention recovery rate; The working memory and task switching parameters include switching delay, error rate change, and parallel task interference; The burst task processing performance parameters include normalized response time and decision parameter values.

[0008] In some preferred embodiments, the standardized regional gaze time percentage is obtained by calculating the ratio of the total gaze time of the pilot in the key instruments, communication interface, and navigation screen areas to the total mission duration to obtain the regional gaze time percentage, and normalizing the regional gaze time percentage to obtain the standardized regional gaze time percentage; The standardized scan frequency is obtained by calculating the ratio of the number of gaze shifts per unit time to the total time to obtain the scan frequency, and normalizing the scan frequency based on preset lower and upper limits of the scan frequency to obtain the standardized scan frequency; The pupil diameter change is obtained by calculating the difference between the mean pupil diameter during the burst task and the mean pupil diameter during the regular task, and calculating the ratio of the difference to the standard deviation of the pupil diameter during the regular task to obtain the pupil diameter change; The standardized first fixation time is obtained by calculating the difference between the time node when the pilot fixates on the emergency task-related area and the time node when the emergency task starts, and normalizing the difference to obtain the standardized first fixation time; The attention recovery rate is obtained by calculating the difference between the time node when the current task ends and the time node when the sudden task ends; and calculating the ratio of the fixation duration of returning to the original task to the difference to obtain the attention recovery rate.

[0009] In some preferred embodiments, the switching delay is obtained by: when the first operation response time of task A in the dual-task environment is shorter than the average response time when only task A is executed, directly assigning a value of 1; Otherwise, calculate the relative difference ratio between the first operation response time of Task A in the dual-task environment and the average response time when only Task A is performed; The error rate change is obtained by: obtaining the error rate change through the relative change rate of the error rate of task A in the dual-task environment and the error rate of task A in the single-task environment; The parallel task interference degree is obtained by taking the completion quality of task A in the single-task state as a benchmark and calculating the deviation degree of the completion quality of task A in the dual-task state.

[0010] In some preferred embodiments, the standardized post-response time is obtained by calculating the actual time difference from the time node when the emergency occurs to the time node when the pilot starts to respond to the operation, and dividing the actual time difference by the maximum reaction time for normalization to obtain the standardized post-response time.

[0011] In some preferred embodiments, the time synchronization is implemented by using a sliding time window and cognitive alignment synchronization method to achieve time synchronization of multi-source data, specifically including the following steps: Step 11, obtaining the pilot's eye movement data sequence, mission data sequence and their corresponding timestamp sequence during the flight mission; Step 12, preprocessing the timestamps of the eye movement data and the task data to obtain a corrected timestamp sequence; the preprocessing includes linear time alignment, abnormal timestamp detection, and missing value processing; Step 13, based on the corrected timestamp sequence and the set time window, perform coarse synchronization alignment of the eye movement and task data in a sliding time window to construct a set of synchronized sample pairs; Step 14: Based on the cognitive alignment feature fusion of the adaptive optimal transmission, semantic-level fine alignment and feature fusion are performed on the set of synchronized sample pairs to generate a fused feature sequence as the time-synchronized data.

[0012] In some preferred embodiments, the flight phase transition map and the weighted multidimensional matching distance are combined to perform flight phase division on the flight parameter data in the preprocessed data to obtain the flight phase division result, and the method is as follows: S1, obtain a set of flight situation segments and a set of vertical motion situation of a preset flight phase, extract and weight the multi-dimensional features of the situation segments, and construct a set of situation segment feature vectors and a set of preset phases with boundary constraints; S2, based on the preset stage set, constructs a flight stage transition graph with stages as nodes and reasonable transfer relationships as edges; S3, based on the flight phase transition map, integrating the feature differences between the situation segment and the preset phase and the out-of-bounds penalty, and calculating the weighted multi-dimensional matching distance; S4, mapping each situation segment into a center state point sequence in a two-dimensional space based on the weighted multi-dimensional matching distance and trajectory data; Calculating the normalized curvature by calculating the ratio of the vertical distance from the midpoint of the central state point sequence to the line segment to the line segment length; The normalized curvature is combined with the dynamic penalty term and the structural consistency penalty coefficient to construct a structure-sensitive matching function, and the structure-sensitive matching is performed to generate a comprehensive matching score; S5, generating an original matching confidence based on the comprehensive matching score, fusing the original matching confidence, the context consistency score, and the historical situation segment matching average confidence to obtain a fused confidence; S6, selecting the best flight phase that matches the current situation segment according to the fused confidence level; the flight phase includes a take-off phase, a climb phase, a cruise phase, a descent phase, an approach phase, and a landing phase.

[0013] In some preferred embodiments, the final multi-task processing performance evaluation value is obtained after the weighted fusion, and the calculation method is as follows: ; ; ; ; Where, is the final multi-tasking performance evaluation value, 、 、 They are attention allocation value, working memory and task switching value, and sudden task processing efficiency value. 、 、 They are attention allocation parameters, working memory and task switching parameters, and burst task processing efficiency parameters. The parameter in The weight coefficient corresponding to each flight phase, 、 、 They are attention allocation parameters, working memory and task switching parameters, and burst task processing efficiency parameters. The value of the parameter.

[0014] Beneficial effects of the present invention: 1) Existing techniques for analyzing pilot status are often limited to single-dimensional maneuvering skill performance or isolated inferences of cognitive load. Such simplified models fail to fully reflect the complexity of real-world flight environments. This invention continues the principle of quantitative analysis of eye movement and task data, further utilizing rigorous correlation modeling to objectively characterize the comprehensive effectiveness of pilots' multitasking performance in highly dynamic environments. Specifically, this performance is reflected in their ability to effectively allocate attention, rapidly switch task focus, accurately execute critical operations, and maintain situational awareness within the constraints of limited cognitive resources. This design overcomes the limitations of traditional approaches that rely solely on maneuvering skills or a single load metric. The resulting characterization more realistically reflects the core competencies of pilots in actual flight, which require them to simultaneously manage the multiple cognitive and operational challenges of driving, monitoring, communicating, and making decisions. This approach is highly consistent with the actual demands of multitasking in real-world flight scenarios. 2) The nature of tasks, peak cognitive load, operational precision requirements, and risk levels faced by pilots in different flight phases, such as takeoff, climb, cruise, approach, and landing, vary significantly. Based on this, the present invention continues the idea of ​​dynamic division of flight phases and further constructs an adaptive weighting mechanism: by finely identifying the characteristics of each phase and adjusting the evaluation weights in a targeted manner to adapt to the characteristics of the phase tasks, it can measure performance fluctuations under high load and changing scenarios in complex real-world flight scenarios (such as approach in severe weather, emergency fault handling, and operation in high-density airspace), providing a highly scenario-based, objective basis for targeted training planning and capacity building, and solving the problem that existing technologies are difficult to comprehensively measure pilots' ability to cope with complex scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Other features, objects and advantages of the present application will become more apparent upon reading the detailed description of non-limiting embodiments made with reference to the following drawings: Figure 1 The present invention is a flowchart of the steps of a method for evaluating the multi-tasking performance of pilots in complex situations.

[0016] Figure 2 The present invention is a multi-tasking efficiency evaluation model diagram of a pilot multi-tasking efficiency evaluation method under complex situations. DETAILED DESCRIPTION

[0017] The present application will be further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are intended only to illustrate the relevant invention and are not intended to limit the invention. It should also be noted that, for ease of description, only portions relevant to the relevant invention are shown in the accompanying drawings.

[0018] It should be noted that, in the absence of conflict, the embodiments and features of the embodiments in this application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0019] In order to more clearly illustrate the pilot multi-tasking performance evaluation method under a complex situation of the present invention, the following is combined with Figures 1 to 2 Each step in the embodiment of the present invention is described in detail.

[0020] This invention proposes a method for evaluating pilot multi-tasking performance in complex situations. Figure 1 , the method comprises the following steps: Step 1: Acquire the pilot's flight parameter data, eye movement data, and task data during the flight and preprocess them to obtain preprocessed data; the preprocessing includes time synchronization, data cleaning, segmentation labeling, and region of interest division; the time synchronization includes using a sliding time window and cognitive alignment synchronization method to achieve time synchronization of multi-source data; In this embodiment, the eye movement data includes gaze time, scan frequency, pupil diameter, gaze point distribution, and scan path; the task data includes burst task processing data, multi-task parallel processing data, task switching data, task priority label, and task completion time; The burst task processing data includes first fixation time, attention recovery time, switching delay, error rate, as well as the time point of emergency triggering, burst task response time and processing results; it also includes error types, such as communication delay and navigation deviation; The time synchronization includes implementing the time synchronization of multi-source data using a sliding time window and cognitive alignment synchronization method, specifically including the following steps: Step 11: Obtain the pilot's eye movement data sequence E={e1,e2,...,eT} and mission data sequence T={t1,t2,...,tN} during the flight mission, as well as the timestamp sequence of the eye movement data. , timestamp sequence of task data ; Step 12: pre-process the timestamps of the eye movement data and task data to obtain a corrected timestamp sequence; including linear time alignment (using least squares fitting to correct the overall time offset between the eye movement and task systems), abnormal timestamp detection (setting a time difference threshold), and , if the adjacent time difference , it is marked as an outlier and skipped), missing value processing (using linear interpolation or attention mask mechanism to ignore missing samples); Step 13: Based on the corrected timestamp sequence and the set time window Perform coarse synchronization alignment of eye movement and task data in a sliding time window to construct a set of synchronized sample pairs , where K is the number of valid synchronization samples, For the Eye movement samples, For the Task samples, eye movement events Timestamp is , mission events Timestamp is , ; Implement candidate set screening based on temporal proximity to provide input for subsequent fine alignment; Step 14: Based on the cognitive alignment feature fusion of adaptive optimal transmission, the semantic level fine alignment and feature fusion are performed on the synchronized sample pair set after coarse synchronization to generate a fused feature sequence. , as the data after time synchronization, specifically: For each eye movement sample , define it and the task sample The cognitive alignment cost , integrating multi-dimensional factors to achieve the transition from "time alignment" to "cognitive consistency alignment"; ; ; ; Where, To learn the projection function, eye movements and tasks are mapped to a unified semantic space respectively; is the similarity between eye movement behavior and task semantics, is the time sensitivity coefficient, is the flight phase state corresponding to the eye movement sample, Encode the function for the mission dynamics and output the time sensitivity level of the phase, such as the approach phase 1.8, cruising stage is 0.5, is a learnable scaling parameter, Sigmoid normalization and time sensitivity coefficient setting are used to automatically increase the time alignment weight in highly dynamic phases (such as approach) to prevent mismatching due to slight time deviations. is the cognitive load coefficient, , indicating the cognitive load level of the current flight phase, The basic semantic weight (learnable or fixed) and cognitive load coefficient setting are used to automatically increase the cognitive load coefficient during high-load phases, when the cost of semantic matching errors is higher, forcing attention to focus on semantically related tasks and improving the cognitive rationality of synchronization; Based on the cognitive alignment cost, the eye movement samples are As a "supply point" of attention resources (the total amount is 1), the task sample Considered as a “demand point”, the attention allocation matrix that minimizes the total transmission cost is solved , used to achieve fine alignment; ; in, , for the The attention allocation vector of eye movement samples, Indicates that the eye movement samples The attention resources are allocated to the task sample The proportion of , indicating that the total attention resource is 1, , Indicates a task The "cognitive reception capacity" can be set based on task priority. is a set of transmission constraints that satisfies ; The Sinkhorn algorithm is used for lightweight solution, and the attention allocation formula is: , the optimal attention allocation result is obtained, which reflects the optimal attention configuration under the minimum cognitive alignment cost, realizes the identification of truly semantically relevant task events from the candidate set, and is the direct output of fine alignment; Based on the cognitive alignment cost and the optimal attention allocation result, a cognitive alignment fusion feature is generated to further encode dynamic trend information in the alignment process; ; Where, The attention weights obtained for the optimal transmission are from the vector No. elements, is the linear projection of the task features, The cost function is the task feature The gradient of , which indicates the direction in which the task representation should be adjusted to better align eye movement behavior, The trend enhancement coefficient is adaptively quantified based on the task dynamics and cognitive load of the current flight phase. For example, the alignment error (reflecting the current matching quality) is calculated based on the attention allocation Aij∗ and the cost c(ei,tj). The error is mapped to alignment instability through a function and finally multiplied by a basic scaling factor (which can be learned or fixed) to obtain the trend enhancement coefficient. When the alignment error is large, the alignment instability approaches 1, and the trend term is automatically enhanced. When the alignment is good, the trend term is weakened to avoid overfitting. This enables the fusion process to have the ability to perceive and dynamically adjust the alignment state, further improving the stability and responsiveness of fine alignment. Cognitive alignment fusion features are used as time-synchronized data for subsequent preprocessing, such as segmentation annotation and region of interest (AOI) segmentation; In this way, a progressive architecture is developed: coarse synchronization screening of spatiotemporal proximity is achieved through sliding time windows → fine alignment of semantic-level cognition is achieved based on cognitive consistency modeling → fusion feature generation. This improves the accuracy and interpretability of cross-modal data matching while ensuring efficiency, solving the problems of "rough time window synchronization", "missing semantic associations", and "poor cognitive interpretability of fusion results" in existing technologies. The segmentation marking includes marking by mission phase (such as cruise, emergency) and marking by sub-task (communication, navigation); The area of ​​interest (AOI) division is based on the cockpit functional partitioning, including the primary flight display (PFD), navigation display (ND), and radio control panel (RCP); Step 2, see Figure 2 , extracting and quantifying the eye movement data and task data in the pre-processed data from the three dimensions of attention allocation, working memory and task switching, and sudden task processing efficiency, obtaining the values ​​of each key parameter, and then obtaining a key parameter data set; In this embodiment, the key parameter data set includes attention allocation parameters, working memory and task switching parameters, and burst task processing efficiency parameters; The attention allocation parameter measures the pilot's ability to allocate limited visual resources to different tasks or information sources in a complex flight environment, including active or automatic attention to the instrument panel, navigation display, external environment, cockpit warning signals, and interactive interfaces. Its core is to dynamically balance multiple visual inputs to ensure that key information is processed first while maintaining overall situational awareness. Specifically, it includes the standardized regional fixation time percentage, standardized scan frequency, pupil diameter change, standardized first fixation time, and attention recovery rate. Together, they can comprehensively measure the rationality and effectiveness of the pilot's visual resource allocation, determine whether they can accurately capture key content in complex information, and ensure the efficiency of information acquisition during flight. The standardized regional fixation time ratio is used to quantify the pilot's attention to key areas and determine whether the pilot pays sufficient attention to the core information source. The method for obtaining the standardized regional fixation time ratio is as follows: the ratio of the pilot's total fixation time on key instruments, communication interfaces, and navigation screens to the total mission duration is calculated, and the standardized regional fixation time ratio is obtained by normalizing the regional fixation time ratio. The standardized regional fixation time ratio is specifically expressed by the following formula: ; ; Where, is the proportion of regional gaze time, The total time the pilot looks at key instruments, communication interfaces, and navigation screens. is the total duration of the task, is a benchmark value, which can be determined by expert experience; The standardized post-scan frequency, the number of sight shifts per unit time, reflects the speed and rationality of the pilot's sight switching between different areas. If the scan frequency is too high, ineffective scanning may be caused by cognitive overload or unreasonable interface layout; if it is too low, key areas may be ignored. The method for obtaining the post-scan frequency is as follows: the ratio of the number of sight shifts per unit time to the total time is calculated to obtain the scan frequency, and the scan frequency is standardized based on the preset lower and upper limits of the scan frequency to obtain the standardized post-scan frequency, which is specifically expressed by the following formula: ; ; Where, is the scanning frequency, is the number of sight shifts per unit time, is the total time, They are respectively the preset lower limit and upper limit of the scanning frequency, which can be determined by expert experience; The pupil diameter change and pupil dilation reflect the intensity of cognitive load. A larger pupil diameter change indicates a higher cognitive load, which intuitively reflects the pilot's physiological reaction when a sudden mission occurs, reflects his or her alertness to sudden situations and changes in psychological load, and assists in determining his or her initial state for responding to sudden missions. The method for obtaining the pupil diameter change is as follows: the difference between the mean pupil diameter during the sudden mission and the mean pupil diameter during the routine mission is calculated, and the ratio of the difference to the standard deviation of the pupil diameter during the routine mission is calculated to obtain the pupil diameter change, which is specifically expressed by the following formula: ; Where, is the mean pupil diameter during the burst task, is the mean pupil diameter during the routine task, is the standard deviation of pupil diameter during the routine task; Emergency mission response: This refers to the pilot's ability to quickly interrupt the current mission and reallocate cognitive resources to identify the problem, make decisions, and implement countermeasures when an unexpected event (such as engine failure, system alarm, or bird strike) occurs. The core of this capability is to quickly switch attention modes (from routine monitoring to emergency response) and maintain decision-making accuracy under time pressure; The standardized first fixation time measures the pilot's reaction speed to an emergency task (for example, the time delay for the pilot to fixate on the relevant area after an emergency occurs (such as an alarm sound in the cabin)), and evaluates the pilot's ability to quickly perceive and pay attention to emergency information. The method for obtaining the standardized first fixation time is to calculate the difference between the time node when the pilot fixates on the relevant area of ​​the emergency task and the time node when the emergency task starts, and to normalize the difference to obtain the standardized first fixation time, which is specifically expressed by the following formula: ; ; Where, is the first fixation time, It is the time point for pilots to watch the relevant areas of emergency missions; The start time node of the emergency task; The attention recovery rate evaluates the pilot's efficiency in quickly returning his attention to the original task after completing the sudden task, reflecting his attention flexibility and the smoothness of task switching. The method for obtaining the attention recovery rate is as follows: the difference between the end time of the current task and the end time of the sudden task is calculated; the ratio of the fixation duration of returning to the original task to the difference is calculated to obtain the attention recovery rate, which is specifically expressed by the following formula: ; Where, The time node when the current task ends; The end time of the emergency task; is the fixation duration of returning to the original task; The working memory and task switching metric measures the additional time and increased error rate caused by the reallocation of cognitive resources when switching between multiple parallel tasks. It directly reflects the pilot's flexibility and efficiency when handling multiple tasks. The task switching cost can be broken down into the following two sub-indicators, quantified from the time and accuracy dimensions: switching delay, error rate change, and parallel task interference. The switching delay can quantify the time loss of the pilot when switching between two tasks, and intuitively reflects the smoothness of the transition from one task to another. If the value is smaller, it means that the switching time is relatively shorter and the task switching efficiency is higher. Conversely, the efficiency is lower, which helps to evaluate the pilot's time management ability in a multi-tasking environment. The method for obtaining it is: when the first operation response time of Task A in the dual-task environment is shorter than the average response time when executing Task A alone, it is directly assigned a value of 1; Otherwise, calculate the relative difference ratio between the first operation response time of Task A in the dual-task environment and the average response time when only Task A is performed; It is specifically expressed in the following formula: ; Where, is the first operation response time of task A in the dual-task environment; is the average response time when only task A is executed; The error rate change reflects the degree of impact on the pilot's task accuracy when processing multiple tasks in parallel. If the value is positive and large, it indicates that the dual-task environment has a greater impact on task accuracy and the pilot is more likely to make errors during task switching. Conversely, it indicates that the pilot can better maintain task accuracy during multi-tasking. It is an important basis for evaluating task processing quality. The error rate change is obtained by calculating the relative change rate of the error rate of task A in the dual-task environment and the error rate of task A in the single-task environment. It is specifically expressed by the following formula: ; Where, is the error rate of task A in the single-task environment; is the error rate of Task A in the dual-task environment; The parallel task interference degree effectively measures the degree of interference between multiple tasks on the completion quality of tasks when they are being carried out simultaneously. A larger value indicates a greater negative impact of dual-task parallelism on the completion quality of Task A, and a stronger interference to the pilot when handling parallel tasks. Conversely, a lower value indicates weaker interference. This provides strong support for assessing a pilot's ability to maintain task quality in a multi-tasking environment. The degree of interference is calculated by taking the completion quality of Task A in the single-task state as a benchmark and calculating the degree of deviation in the completion quality of Task A when dual-task parallelism is used. Specifically, it is expressed as follows: ; Where, is the completion quality of task A in the single-task state; The completion quality of task A when two tasks are performed in parallel; The interference degree range is [0,1], where 0 indicates no interference (dual-task performance = single-task performance) and 1 indicates complete interference (dual-task performance = 0); The emergency task handling efficiency parameters comprehensively characterize the pilot's comprehensive ability to handle emergency tasks from two key dimensions: reaction speed and decision quality. They provide an important basis for judging their handling level in emergency situations, including standardized response time and decision parameter values. The standardized response time can accurately quantify the pilot's reaction speed to an emergency situation. After standardization, the time differences caused by the characteristics of different emergency tasks can be eliminated, facilitating horizontal comparison of the emergency response capabilities of different pilots or the same pilot in different scenarios. The shorter the reaction time, the more timely the pilot responds to the emergency task, which buys more time to ensure flight safety. The method for obtaining the standardized response time is: calculating the actual time difference from the time the emergency occurs to the time the pilot's operational response begins, dividing the actual time difference by the maximum reaction time for normalization, and obtaining the standardized response time. It is specifically expressed by the following formula: ; ; Where, The time point when the pilot's operational response begins. The time point for emergency situations; The decision parameter value represents the pilot's judgment ability and the rationality of the handling strategy in emergency missions from the perspective of decision quality. It can intuitively reflect the pilot's decision-making level in a high-pressure environment. A high score indicates that the pilot can make scientific, safe and efficient decisions. It is a key parameter for measuring the pilot's core emergency handling capabilities. The method is as follows: a three-level weighting is used: if the optimal response strategy is selected, the weight coefficient is 1; if a suboptimal but safe strategy is selected, the weight coefficient is 0.5; if an incorrect or dangerous strategy is selected, the weight coefficient is 0; ; Step 3: Based on the different cognitive requirements of each flight phase, a weight matrix is ​​preset using the analytic hierarchy process and applied to the key parameter dataset; In this embodiment, taking attention allocation A as an example, the preset weight matrix is ​​shown in Table 1: Table 1:

[0021] Combining the flight phase transfer map and the weighted multidimensional matching distance, the flight parameter data in the preprocessed data are divided into flight phases to obtain a flight phase division result; S1, obtain the flight situation segment set, represented by S, , where the kth situation segment is defined as , including the multi-time data of pressure altitude A and descent rate V (N≥2), for each flight situation segment Extracting original feature vectors , They are the mean value of pressure altitude, the mean value of descent rate, the derived speed of pressure altitude, and the derived acceleration of descent rate. After normalization and weighting, a weighted feature vector is generated to obtain the feature vector set of the situation segment: , is the weight coefficient, the sum of which is 1; obtain the vertical motion situation set of the preset flight phase, represented by Z, , the jth stage Expressed as , we get a preset stage set with boundary constraints, where For the flight phase The upper and lower limits of the pressure altitude, Flight phase The upper and lower limits of the decline rate; S2, constructing a flight phase transition map based on the preset phase set Z, , where: the node set V corresponds to the preset stage set Z; the edge set E represents the reasonable transfer relationship between flight stages (used for subsequent structure-sensitive matching and global consistency optimization); S3, based on the flight phase transition map, the feature differences between the situation segment and the preset phase and the out-of-bounds penalty are integrated to calculate the weighted multi-dimensional matching distance Specifically, the weighted multidimensional matching distance function is used to calculate the k-th situation segment With the jth preset stage Weighted multidimensional matching distance : ; For the The weighted feature vector of each situation segment; For the Preset Stage Reference weighted features of (such as central value or typical value); For the situation segment Beyond the stage the extent of the pressure altitude or descent rate boundary; is the out-of-bounds penalty coefficient (learnable or fixed); S4, constructing a structure-sensitive matching function based on the weighted multi-dimensional matching distance and trajectory data to generate a comprehensive matching score, the method is as follows: Based on the weighted multi-dimensional matching distance and trajectory data, each situation segment is mapped into a center state point sequence in a two-dimensional space; Calculating the normalized curvature by calculating the ratio of the vertical distance from the midpoint of the central state point sequence to the line segment to the line segment length; The normalized curvature is combined with the dynamic penalty term and the structural consistency penalty coefficient to construct a structure-sensitive matching function, and the structure-sensitive matching is performed to generate a comprehensive matching score; The structure-sensitive matching function is: ; ; ; ; Where, For the Situational stages and phases The comprehensive matching score of is a dynamic penalty term, is the structural consistency penalty coefficient, is the sensitivity coefficient, It is a saturated nonlinear function to prevent the penalty from growing infinitely. , based on the first three points Calculate normalized curvature The larger the value is, the more likely the flight state will change dramatically in this range, and the higher the possibility of phase jump. point To line segment The vertical distance, For line segments length, each situation segment Mapped to its central state point, it forms a trajectory sequence of the aircraft in the two-dimensional situation space composed of (pressure altitude, descent rate); S5, generating an original matching confidence based on the comprehensive matching score, fusing the original matching confidence, the context consistency score, and the historical situation segment matching average confidence to obtain a fused confidence; The fusion formula is: ; Where, is the original matching confidence, Score the context consistency (calculated by the matching degree between the stage sequence and G in a sliding window with window size W ≥ 2), is the confidence level of the historical similarity segment ( Situation segments with feature similarity higher than the threshold are matched The average confidence level of ; S6, selecting the current situation segment according to the fused confidence Match the best flight phase; ; Furthermore, dynamic programming or Hidden Markov Model (HMM) can be used to analyze the stage sequence Perform global optimization to ensure that it satisfies the edge constraints E in the flight phase transition graph G (i.e., the transition relationship between adjacent phases is reasonable), thereby improving the temporal consistency of the recognition results. Supports expert feedback or cluster analysis, including online updates of the following: The feature range of the preset flight phase template Z is as follows: ; The edge set E of the flight phase transition graph G; Feature weighting , dynamic fusion weight , penalty coefficient , out-of-bounds penalty coefficient ; To adapt to new flight missions (such as special maneuvers and changes in stage characteristics under extreme weather conditions) and improve the algorithm's generalization ability; Step 4: Matching the preset weight matrix according to the flight phase division result, combining the key parameter values, and obtaining a final multi-task processing performance evaluation value after weighted fusion; In this embodiment, the final multi-task processing performance evaluation value is obtained after the weighted fusion, and the calculation method is as follows: ; ; ; ; Where, is the final multi-tasking performance evaluation value, 、 、 They are attention allocation value, working memory and task switching value, and sudden task processing efficiency value. 、 、 They are attention allocation parameters, working memory and task switching parameters, and burst task processing efficiency parameters. The parameter in The weight coefficient corresponding to each flight phase, 、 、 They are attention allocation parameters, working memory and task switching parameters, and burst task processing efficiency parameters. The value of the parameter.

[0022] In this way, by combining the three major parameters of attention allocation, working memory and task switching, and sudden task processing efficiency with their corresponding weight vectors, we first calculate the attention allocation value, working memory and task switching value, and sudden task processing efficiency value, and then integrate these three values ​​to obtain the multi-task processing efficiency evaluation value, which can comprehensively, objectively and accurately reflect the pilot's overall ability to handle multiple tasks in parallel in complex flight scenarios.

[0023] Although the various steps in the above embodiment are described in the above-mentioned order, those skilled in the art will understand that in order to achieve the effect of this embodiment, different steps do not have to be executed in such an order. They can be executed simultaneously (in parallel) or in a reverse order. These simple changes are within the scope of protection of the present invention.

[0024] Those skilled in the art should be able to appreciate that the modules and method steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two, and the programs corresponding to the software modules and method steps can be placed in random access memory (RAM), internal memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disks, removable disks, CD-ROMs, or any other form of storage medium known in the art. In order to clearly illustrate the interchangeability of electronic hardware and software, the above description has generally described the components and steps of each example according to their functions. Whether these functions are performed in electronic hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0025] The terms "first", "second", etc. are used to distinguish similar objects, rather than to describe or indicate a particular order or sequence.

[0026] The term "comprise" or any other similar term is intended to cover non-exclusive inclusion such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed or inherent to such process, method, article, or apparatus.

[0027] Thus far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art may make equivalent changes or substitutions to the relevant technical features, and the technical solutions after such changes or substitutions will fall within the scope of protection of the present invention.

Claims

1. A method for evaluating pilot multi-tasking performance in complex situations, characterized by: The method comprises the following steps: Acquiring and preprocessing the pilot's flight parameter data, eye movement data, and task data during the flight to obtain preprocessed data; the preprocessing includes time synchronization, data cleaning, segmentation labeling, and region of interest division; the time synchronization includes using a sliding time window and cognitive alignment synchronization method to achieve time synchronization of multi-source data; Extracting and quantifying features of the eye movement data and task data in the pre-processed data from three dimensions: attention allocation, working memory and task switching, and sudden task processing efficiency, to obtain key parameter values, and thus obtain a key parameter data set; According to the different cognitive requirements of each flight phase, a weight matrix is ​​preset using the analytic hierarchy process and applied to the key parameter data set; Combining the flight phase transfer map and weighted multi-dimensional matching distance, the flight phase division is performed using the pre-processed flight parameter data to obtain the flight phase division results. The preset weight matrix is ​​matched according to the flight phase division result, and the key parameter values ​​are combined and weighted fusion is performed to obtain the final multi-task processing performance evaluation value.

2. The method for evaluating pilot multi-tasking performance in complex situations according to claim 1, characterized in that: The eye movement data includes gaze time, scanning frequency, and pupil diameter; the task data includes burst task processing data, multi-task parallel processing data, task switching data, and task priority labels; The burst task processing data includes first fixation time, attention recovery time, switching delay, and error rate.

3. The method for evaluating pilot multi-tasking performance in complex situations according to claim 1, characterized in that: The segmentation marking includes marking by task stage and marking by subtask; The region of interest is divided based on the cockpit functional areas, including the primary flight display, navigation display, and radio control panel.

4. The method for evaluating pilot multi-tasking performance in complex situations according to claim 1, wherein: The key parameter data set includes attention allocation parameters, working memory and task switching parameters, and burst task processing efficiency parameters; The attention allocation parameters include the normalized regional fixation time ratio, normalized saccade frequency, pupil diameter change, normalized first fixation time, and attention recovery rate; The working memory and task switching parameters include switching delay, error rate change, and parallel task interference; The burst task processing performance parameters include normalized response time and decision parameter values.

5. The method for evaluating pilot multi-tasking performance in complex situations according to claim 4, characterized in that: The standardized regional gaze time percentage is obtained by calculating the ratio of the total gaze time of the pilot in the key instruments, communication interface, and navigation screen areas to the total mission time to obtain the regional gaze time percentage, and normalizing the regional gaze time percentage to obtain the standardized regional gaze time percentage; The standardized scan frequency is obtained by calculating the ratio of the number of gaze shifts per unit time to the total time to obtain the scan frequency, and normalizing the scan frequency based on preset lower and upper limits of the scan frequency to obtain the standardized scan frequency; The pupil diameter change is obtained by calculating the difference between the mean pupil diameter during the burst task and the mean pupil diameter during the regular task, and calculating the ratio of the difference to the standard deviation of the pupil diameter during the regular task to obtain the pupil diameter change; The standardized first fixation time is obtained by calculating the difference between the time node when the pilot fixates on the emergency task-related area and the time node when the emergency task starts, and normalizing the difference to obtain the standardized first fixation time; The attention recovery rate is obtained by calculating the difference between the time node when the current task ends and the time node when the sudden task ends; The ratio of the fixation duration of returning to the original task to the difference is calculated to obtain the attention recovery rate.

6. The method for evaluating pilot multi-tasking performance in complex situations according to claim 4, characterized in that: The switching delay is obtained by directly assigning a value of 1 when the first operation response time of task A in the dual-task environment is shorter than the average response time when only task A is executed; Otherwise, calculate the relative difference ratio between the first operation response time of Task A in the dual-task environment and the average response time when only Task A is performed; The error rate change is obtained by: obtaining the error rate change through the relative change rate of the error rate of task A in the dual-task environment and the error rate of task A in the single-task environment; The parallel task interference degree is obtained by taking the completion quality of task A in the single-task state as a benchmark and calculating the deviation degree of the completion quality of task A in the dual-task state.

7. The method for evaluating pilot multi-tasking performance in complex situations according to claim 4, characterized in that: The standardized post-response time is obtained by calculating the actual time difference between the time when the emergency occurs and the time when the pilot begins to respond to the operation, and dividing the actual time difference by the maximum response time for normalization to obtain the standardized post-response time.

8. The method for evaluating pilot multi-tasking performance in complex situations according to claim 1, characterized in that: The time synchronization includes implementing the time synchronization of multi-source data using a sliding time window and cognitive alignment synchronization method, specifically including the following steps: Step 11, obtaining the pilot's eye movement data sequence, mission data sequence and their corresponding timestamp sequence during the flight mission; Step 12, preprocessing the timestamps of the eye movement data and the task data to obtain a corrected timestamp sequence; the preprocessing includes linear time alignment, abnormal timestamp detection, and missing value processing; Step 13, based on the corrected timestamp sequence and the set time window, perform coarse synchronization alignment of the eye movement and task data in a sliding time window to construct a set of synchronized sample pairs; Step 14: Based on the cognitive alignment feature fusion of the adaptive optimal transmission, semantic-level fine alignment and feature fusion are performed on the set of synchronized sample pairs to generate a fused feature sequence as the time-synchronized data.

9. The method for evaluating pilot multi-tasking performance in complex situations according to claim 1, characterized in that: The flight phase transfer map and the weighted multi-dimensional matching distance are combined to divide the flight parameter data in the pre-processed data into flight phases to obtain the flight phase division result, and the method is as follows: S1, obtain a set of flight situation segments and a set of vertical motion situation of a preset flight phase, extract and weight the multi-dimensional features of the situation segments, and construct a set of situation segment feature vectors and a set of preset phases with boundary constraints; S2, based on the preset stage set, constructs a flight stage transition graph with stages as nodes and reasonable transfer relationships as edges; S3, based on the flight phase transition map, integrating the feature differences between the situation segment and the preset phase and the out-of-bounds penalty, and calculating the weighted multi-dimensional matching distance; S4, mapping each situation segment into a center state point sequence in a two-dimensional space based on the weighted multi-dimensional matching distance and trajectory data; Calculating the normalized curvature by calculating the ratio of the vertical distance from the midpoint of the central state point sequence to the line segment to the line segment length; The normalized curvature is combined with the dynamic penalty term and the structural consistency penalty coefficient to construct a structure-sensitive matching function, and the structure-sensitive matching is performed to generate a comprehensive matching score; S5, generating an original matching confidence based on the comprehensive matching score, fusing the original matching confidence, the context consistency score, and the historical situation segment matching average confidence to obtain a fused confidence; S6, selecting the best flight phase that matches the current situation segment according to the fused confidence level; the flight phase includes a take-off phase, a climb phase, a cruise phase, a descent phase, an approach phase, and a landing phase.

10. The method for evaluating pilot multi-tasking performance in complex situations according to claim 1, characterized in that: After the weighted fusion, the final multi-task processing performance evaluation value is obtained, and the calculation method is as follows: ； ； ； ； Where, is the final multi-tasking performance evaluation value, 、 、 They are attention allocation value, working memory and task switching value, and sudden task processing efficiency value. 、 、 They are attention allocation parameters, working memory and task switching parameters, and burst task processing efficiency parameters. The parameter in The weight coefficient corresponding to each flight phase, 、 、 They are attention allocation parameters, working memory and task switching parameters, and burst task processing efficiency parameters. The value of the parameter.

Images