Pilot wind shear handling procedure compliance analysis system and method

Abstract

The application belongs to the technical field of data analysis, and discloses a pilot wind shear disposal procedure compliance analysis system and method. The system comprises a data acquisition module, a storage module, a processing module and a report generation module. The processing module identifies non-compliant behavior through a finite state machine program atlas, then starts a counterfactual deduction module to generate a virtual optimal trajectory, and calculates a flight resilience index based on comparative analysis. The method comprises the following steps: pre-building a finite state machine atlas; acquiring QAR data; comparing and identifying non-compliant behavior; when non-compliant behavior is identified, performing counterfactual deduction to generate a virtual trajectory; then calculating a flight resilience index; and finally generating an analysis report. The application realizes counterfactual deduction of flight operation and deepens the application value of QAR data.

Description

Technical Field

[0001] This invention belongs to the field of data analysis technology, specifically relating to a pilot wind shear handling procedure compliance analysis system and method. Background Technology

[0002] Analyzing pilots' adherence to Standard Operating Procedures (SOPs) is central to flight quality monitoring. Current technologies primarily focus on error detection, identifying discrepancies between operations and procedures through parameter thresholds or simple process comparisons. These methods suffer from two major limitations: 1) Poor debriefing effectiveness: they only inform pilots what they did wrong, failing to visually demonstrate the correct procedures and the differences in consequences, leading to low training efficiency; 2) Limited evaluation dimensions: they focus solely on deviations from procedures, neglecting the pilot's resilience—a more core competency—demonstrating their ability to recover after deviations. Summary of the Invention

[0003] The present invention aims to at least partially solve the aforementioned technical problems. Therefore, the objective of the present invention is to provide a system and method for analyzing pilot wind shear handling procedure compliance.

[0004] To achieve the above objectives, this invention provides a pilot wind shear handling procedure compliance analysis system. The system includes a data acquisition module, a storage module, a processing module, and a report generation module. The processing module identifies non-compliance behaviors using a finite state machine (FSM) procedure graph, then initiates a counterfactual inference module to generate a virtual optimal trajectory, and calculates flight resilience indices based on comparative analysis.

[0005] The data acquisition module is used to acquire QAR data containing wind shear events; the storage module is used to store the QAR data and has a built-in finite state machine program diagram modeled from the standard operating procedure.

[0006] The processing module is connected to the storage module and is used to: identify non-compliance behaviors in the pilot's actual operation by comparing the QAR data with the finite state machine program diagram; when a non-compliance behavior is identified, start a counterfactual inference module based on a generative model to generate and output a virtual QAR data trajectory simulating the optimal operation for the non-compliance behavior; and calculate and output a flight resilience quantification index characterizing the pilot's ability to restore the aircraft to a safe state after deviating from the standard operating procedure based on the analysis of the deviation recovery process between the actual QAR trajectory and the virtual QAR data trajectory.

[0007] The report generation module is used to generate analytical reports that include non-compliance behavior, counterfactual comparisons, and flight resilience indicators.

[0008] Furthermore, the counterfactual reasoning module is constructed using a generative adversarial network or a variational autoencoder. The generative adversarial network is trained with a large amount of historical QAR data to learn a flight physics model.

[0009] Furthermore, the specific method by which the counterfactual deduction module generates virtual QAR data trajectories is as follows: the control parameters corresponding to non-compliance behaviors in the actual QAR data are replaced with preset optimal control parameters, which are used as input conditions for the generative model to generate subsequent aircraft state parameter trajectories.

[0010] Furthermore, the calculation of the flight resilience quantification index takes into account at least one or more of the following factors: the time required to recover from deviation from standard operating procedures to the safe state domain, the maximum deviation of the aircraft state parameters, or the total control input of the pilot during the recovery process.

[0011] Furthermore, the safety state domain is a multidimensional spatial region defined by at least one key parameter selected from the aircraft's altitude, speed, and attitude.

[0012] To achieve the above objectives, the present invention also provides a method for analyzing pilot compliance with wind shear handling procedures. This method includes the following steps:

[0013] The standard operating procedure is pre-modeled as a finite state machine program diagram;

[0014] Collect QAR data containing wind shear events;

[0015] By comparing the QAR data with the finite state machine program diagram, non-compliant behaviors of the program are identified;

[0016] When non-compliance behavior is identified, a counterfactual inference module based on a generative model is used to generate and output a virtual QAR data trajectory that simulates the optimal operation.

[0017] Based on the analysis of the deviation recovery process between the actual QAR trajectory and the virtual QAR trajectory, a quantitative index of flight resilience is calculated and output.

[0018] Generate an analysis report that includes non-compliance behavior, counterfactual inferences, and flight resilience indicators.

[0019] Furthermore, the counterfactual reasoning module is constructed using a generative adversarial network or a variational autoencoder, which is trained on a large amount of historical QAR data to learn a flight physics model.

[0020] Furthermore, the steps for generating virtual QAR data trajectories specifically include: replacing the control parameters corresponding to non-compliance behaviors in the actual QAR data with preset optimal control parameters, using them as input conditions for the generative model, and generating subsequent aircraft state parameter trajectories.

[0021] Furthermore, the calculation of the flight resilience quantification index takes into account at least one or more of the following factors: the time required to recover from deviation from standard operating procedures to the safe state domain, the maximum deviation of the aircraft state parameters, or the total control input of the pilot during the recovery process.

[0022] Furthermore, the safety state domain is a multidimensional spatial region defined by at least one key parameter selected from the aircraft's altitude, speed, and attitude.

[0023] The beneficial effects of this invention are as follows:

[0024] This invention enables counterfactual simulation of flight operations. The system visualizes flight outcomes under optimal operation, providing pilots with comparisons, debriefings, and training effects far exceeding traditional methods. This invention expands the evaluation focus from avoiding mistakes to effectively correcting them. By quantitatively assessing pilots' recovery efficiency and control level after deviations, it can more scientifically and comprehensively identify high-level pilots.

[0025] In this invention, the accurate error detection of FSM, the intuitive teaching of generative AI, and the in-depth evaluation of resilience theory work together to form a complete closed loop from problem discovery to solution demonstration and capability evaluation, which greatly enhances the analytical value of QAR data. Attached Figure Description

[0026] Figure 1 This is a system structure block diagram of an embodiment of the present invention.

[0027] Figure 2 This is a flowchart of a method according to an embodiment of the present invention. Detailed Implementation

[0028] The technical solutions of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0029] It should be understood that, and also noted, in the embodiments, the functions / actions may appear in a different order than those shown in the figures. For example, depending on the functions / actions involved, they may actually be performed substantially concurrently, or sometimes the two figures shown consecutively may be performed in reverse order.

[0030] like Figure 1 and Figure 2 As shown, the pilot wind shear handling procedure compliance analysis system in this embodiment mainly includes a data acquisition module, a storage module, a processing module, and a report generation module.

[0031] The core processing flow of the method in this embodiment is completed in the processing module, which can be logically divided into an FSM program verification module, a counterfactual deduction module, and a flight resilience assessment module. The specific implementation method is as follows:

[0032] Step 1: FSM Procedure Verification and Non-Compliance Behavior Identification

[0033] Finite State Machine Program Graph: In the storage module, the standard operating procedure for wind shear detachment is formally modeled in advance and constructed into a finite state machine program graph. This finite state machine program graph defines the correct sequence of operations (state transition paths) and the conditions that key parameters (such as pitch angle and throttle) should satisfy in each state.

[0034] Path Comparison: The FSM program verification module receives QAR data collected by the data acquisition module and maps the key parameter sequences therein to an actual state transition path. Then, this actual path is compared with the standard path in the FSM program graph.

[0035] Identifying non-compliance behavior: If the actual path deviates from the standard path at some point, such as missing a step (e.g., not pressing the TOGA switch), incorrect sequence, or parameters not meeting the requirements (e.g., lever speed is lower than the SOP requirement), the module records the type of non-compliance behavior, the time of occurrence, and related parameters.

[0036] Step 2: Counterfactual inference based on generative models:

[0037] When the FSM program verification module identifies a non-compliance behavior, the counterfactual reasoning module is activated to perform a What-if analysis.

[0038] 2.1 Model Construction: The core of this module is a pre-trained deep generative model, such as a generative adversarial network or a variational autoencoder. By learning from massive amounts of historical QAR data (including various normal and abnormal flight scenarios), this model has mastered the dynamic response patterns of the aircraft under different control inputs, thus approximating a high-fidelity flight physics model.

[0039] 2.2 Counterfactual Condition Setting: Assume that non-compliance behavior is identified at time T1, such as "the pilot's stick pull rate is only 5 degrees / second, lower than the SOP requirement of 8 degrees / second". The module uses the actual QAR data sequence before time T1 as historical context, and then artificially constructs a counterfactual condition: at time T1, the stick pull rate is corrected to the optimal value of "8 degrees / second".

[0040] 2.3 Virtual Trajectory Generation: The module uses historical context and counterfactual conditions as input to the generative model. Based on its learned flight patterns, the model begins to generate and output a completely new virtual QAR data trajectory starting from time T1, simulating the optimal operating conditions. This virtual trajectory demonstrates how the aircraft's flight state (altitude, speed, attitude, etc.) would have evolved had the pilot operated correctly.

[0041] Step 3: Quantitative assessment of flight resilience:

[0042] After obtaining the actual trajectory and the virtual (optimal) trajectory, the flight resilience assessment module quantifies the pilot's error correction capability.

[0043] 3.1 Safe State Domain Definition: This module pre-defines a multi-dimensional safe state domain based on aircraft performance and safety regulations. The safe state domain consists of normal ranges for a series of key parameters, including but not limited to aircraft altitude, speed, attitude (e.g., roll angle, pitch angle), vertical velocity, and energy state. In this embodiment, a specific safe state domain can be defined as {altitude > 500ft, speed > Vref, |roll angle| < 15°}. The ideal trajectory corresponding to the SOP is typically located in the central region of this safe state domain.

[0044] 3.2 Deviation-Recovery Process Analysis: The module compares and analyzes the actual QAR trajectory. When non-compliance occurs, the actual trajectory may deviate from or even completely leave this safe state domain. The module will record:

[0045] Maximum deviation: The maximum difference between a key parameter (such as altitude) in the actual trajectory and the boundary of the safety domain or the virtual optimal trajectory.

[0046] Recovery time: The time taken from the moment the actual trajectory deviates from the safe zone until the pilot's corrective actions bring it fully back into the zone.

[0047] Total control inputs: The integral of the total control inputs applied by the pilot during the recovery time, such as the stick / wheel inputs, reflects the cost and smoothness of the recovery operation.

[0048] 3.3 Resilience Index Calculation: The module uses a weighted formula to calculate the final flight resilience index, for example:

[0049] Resilience index = w1 * (1 / recovery time) + w2 * (1 / maximum deviation) + w3 * (1 / total control input)

[0050] w1, w2, and w3 are weighting coefficients. The higher the index, the stronger the pilot's ability to recover the aircraft from deviations, i.e., the better the flight resilience.

[0051] Step 4: Generate an in-depth analysis report:

[0052] Finally, the report generation module integrates all the above analysis results to generate an in-depth analysis report. The report will be clearly presented graphically:

[0053] List of identified non-compliance behaviors;

[0054] The comparison graph between the actual flight trajectory and the virtual optimal trajectory derived from counterfactual reasoning allows pilots to intuitively understand the differences in consequences brought about by different operations.

[0055] The final calculated score for flight resilience quantification index is accompanied by an evaluation comment.

[0056] Through the above-described embodiments, this invention elevates program compliance analysis from a simple right-or-wrong judgment to a level of intelligent teaching and competency assessment.

[0057] This invention is not limited to the above-described optional embodiments. Anyone can derive other various forms of products under the guidance of this invention. However, regardless of any changes made in their shape or structure, any technical solution that falls within the scope of the claims of this invention shall be protected by this invention.

Claims

1. A pilot wind shear handling procedure compliance analysis system, characterized in that, include: The data acquisition module is used to collect QAR data including wind shear events; The storage module is used to store the QAR data and has a built-in finite state machine program diagram modeled from the standard operating procedure. The processing module, connected to the storage module, is used for: By comparing the QAR data with the finite state machine program diagram, non-compliance behaviors in the pilot's actual operation can be identified. When non-compliance behavior is identified, a counterfactual inference module based on a generative model is activated. For the non-compliance behavior, a virtual QAR data trajectory simulating the optimal operation is generated and output. The specific way the counterfactual inference module generates the virtual QAR data trajectory is as follows: the control parameters corresponding to the non-compliance behavior in the actual QAR data are replaced with preset optimal control parameters, which are used as input conditions for the generative model to generate the subsequent aircraft state parameter trajectory. Based on the analysis of the deviation recovery process between the actual QAR trajectory and the virtual QAR data trajectory, a flight resilience quantification index characterizing the pilot's ability to restore the aircraft to a safe state after deviating from the standard operating procedure is calculated and output. The calculation of the flight resilience quantification index takes into account at least one or more of the following factors: the time required to recover from the deviation from the standard operating procedure to the safe state domain, the maximum deviation of the aircraft state parameters, or the total control input of the pilot during the recovery process. The report generation module is used to generate analytical reports that include non-compliance behavior, counterfactual comparisons, and flight resilience indicators.

2. The system according to claim 1, characterized in that, The counterfactual reasoning module is constructed using a generative adversarial network or a variational autoencoder. The generative adversarial network is trained on a large amount of historical QAR data to learn a flight physics model.

3. The system according to claim 1, characterized in that, The safety state domain is a multidimensional spatial region defined by at least one key parameter selected from the aircraft's altitude, speed, and attitude.

4. A method for analyzing pilot compliance with wind shear handling procedures, characterized in that, Includes the following steps: The standard operating procedure is pre-modeled as a finite state machine program diagram; Collect QAR data containing wind shear events; By comparing the QAR data with the finite state machine program diagram, non-compliant behaviors of the program are identified; When non-compliance behavior is identified, a counterfactual inference module based on a generative model is used to generate and output a virtual QAR data trajectory that simulates the optimal operation. The specific steps for generating virtual QAR data trajectories include: replacing the control parameters corresponding to non-compliance behaviors in the actual QAR data with preset optimal control parameters, using them as input conditions for the generative model, and generating subsequent aircraft state parameter trajectories. Based on the analysis of the deviation recovery process of the actual QAR trajectory and the virtual QAR trajectory, the flight resilience quantification index is calculated and output. The calculation of the flight resilience quantification index shall take into account at least one or more of the following factors: the time required to recover from deviation from standard operating procedures to the safe state domain, the maximum deviation of the aircraft state parameters, or the total control input of the pilot during the recovery process. Generate an analysis report that includes non-compliance behavior, counterfactual inferences, and flight resilience indicators.

5. The method according to claim 4, characterized in that, The counterfactual reasoning module is constructed using a generative adversarial network or variational autoencoder, which is trained on a large amount of historical QAR data to learn a flight physics model.

6. The method according to claim 4, characterized in that, The safety state domain is a multidimensional spatial region defined by at least one key parameter selected from the aircraft's altitude, speed, and attitude.

Images