Flight cabin remote real machine control method based on multi-source perception fusion

Abstract

This invention provides a remote flight control method for flight cockpits based on multi-source perception fusion, relating to the field of remote flight cockpit control. The method includes collecting cockpit control status, multi-aircraft flight status, and environmental disturbance status; generating a collaborative flight fusion state set after temporal correlation; recording the action propagation timing relationship between different aircraft; analyzing the multi-aircraft action timing correlation characteristics; verifying environmental disturbance matching relationships; separating environmental disturbance-driven actions from actions inherited by neighboring aircraft; verifying the action inheritance state of the target aircraft; hierarchically suppressing inherited actions by neighboring aircraft while retaining environmental disturbance-driven actions; generating flight correction results and integrating them into a formation control correction set; issuing remote control commands to the target aircraft; mapping the feedback status to the simulated flight cockpit; adjusting the intensity of neighboring aircraft action suppression; and outputting stable remote control results for multiple aircraft. This achieves precise separation of action sources under multi-aircraft collaboration, improving the stability of remote flight control and formation safety.

Description

Technical Field

[0001] This invention belongs to the field of remote control of flight cockpits, and more specifically, relates to a method for remote control of flight cockpits based on multi-source perception fusion. Background Technology

[0002] In multi-aircraft collaborative flight operations, formation aircraft often need to follow the actions of the lead aircraft to complete synchronous attitude and heading adjustments, which are widely used in low-altitude inspection, regional mapping, emergency rescue and other fields.

[0003] Existing remote flight control technologies mostly focus on building virtual-real synchronous control mechanisms for single aircraft, only considering the correspondence between cockpit commands and the state of a single aircraft, without taking into account the motion propagation characteristics in multi-aircraft formations. Short-term attitude corrections made by the navigator when encountering local crosswinds, sudden turbulence, or obstacle avoidance can be misinterpreted by subsequent aircraft as normal formation commands and synchronously inherited, leading to the amplification of local deviations within the formation. Simultaneously, the flight simulator cockpit can only reflect the state of a single target aircraft; operators cannot distinguish whether the target aircraft's actions originate from active cockpit control, real-world environmental disturbances, or inherited actions from neighboring aircraft. This easily leads to virtual-real control mismatch issues, potentially causing formation collisions, formation breakdowns, and other safety accidents in severe cases. Existing technologies cannot effectively solve the problem of misleading action inheritance from neighboring aircraft in multi-aircraft collaborative scenarios.

[0004] Therefore, it is necessary to provide a remote real-world control method for flight cockpits based on multi-source perception fusion to solve the above-mentioned technical problems. Summary of the Invention

[0005] To address the shortcomings of existing technologies, the present invention aims to overcome the aforementioned deficiencies and propose a remote real-world control method for flight cockpits based on multi-source perception fusion.

[0006] The present invention adopts the following technical solution.

[0007] This invention discloses a remote real-world control method for a flight cockpit based on multi-source sensing fusion, the method comprising: The system synchronously collects cockpit control status, multi-aircraft flight status, and environmental disturbance status. After unified time correlation, it generates a collaborative flight fusion status set, records the action propagation timing relationship between different aircraft, and outputs the action propagation correlation sequence. Based on the action propagation association sequence, the temporal association characteristics of multi-machine actions are analyzed, the environmental disturbance matching relationship is verified, the environmental disturbance-driven action and the neighbor machine inherited action are separated, and a set of neighbor machine action constraint relationships is generated. Based on the set of neighboring aircraft action constraint relationships, the action inheritance state of the target aircraft is verified, the inherited actions of neighboring aircraft are suppressed in stages while retaining the environmental disturbance-driven actions, and after generating independent flight correction results, they are integrated into a formation control correction set. Based on the formation control correction set, remote control commands are sent to the target aircraft and feedback from the actual aircraft is collected. The feedback status is mapped to the flight simulator cockpit, the intensity of suppression of adjacent aircraft actions is dynamically adjusted, and the results of stable remote control of multiple aircraft are output.

[0008] Preferably, the synchronous acquisition of cockpit control status, multi-aircraft flight status, and environmental disturbance status is used to generate a cooperative flight fusion status set after unified time correlation, and the action propagation timing relationship between different aircraft is recorded, outputting the action propagation correlation sequence, including: By simulating the control stick displacement sensor, throttle opening detector, rudder angle detector and cockpit attitude feedback detector in the flight cockpit, the operator's control action data is collected synchronously. After filtering out invalid instantaneous actions caused by sensor noise, a cockpit control state sequence is generated. Based on the cockpit control state sequence, all acquisition times are extracted as the synchronous acquisition reference times for multi-aircraft flight states. Synchronous acquisition commands are issued to the target aircraft and all nearby cooperating aircraft to synchronously acquire the basic flight parameters and formation relative position parameters of each aircraft and record the action timing marks to generate a multi-aircraft flight state sequence. Based on the multi-aircraft flight state sequence, the three-dimensional spatial position and precise time information of each aircraft at each synchronous acquisition moment are extracted. The local airflow changes, spatial obstacle distribution and environmental disturbance state in the flight area are sensed synchronously. After calculating the environmental disturbance intensity, an environmental disturbance state sequence is generated. The cockpit control state sequence, multi-aircraft flight state sequence, and environmental disturbance state sequence are resampled to the same main time axis. The adjacent time interpolation method is used to compensate for the time deviation caused by the difference in the acquisition cycle. After calculating the correlation strength of the fused state, a set of cooperative flight fused states is generated. Based on the fusion set of cooperative flight states, the sequential relationships of cockpit control actions, target aircraft attitude changes, navigator aircraft attitude changes, neighboring cooperative aircraft follow-up changes, and environmental disturbance changes are recorded in chronological order. Suspected neighbor aircraft action inheritance segments, active control action segments, and environmental disturbance-driven action segments are marked, and the action propagation correlation sequence is output.

[0009] Preferably, the step of simultaneously collecting the operator's control action data through the control stick displacement sensor, throttle opening detector, rudder angle detector, and cockpit attitude feedback detector in the simulated flight cockpit, filtering out invalid instantaneous actions caused by sensor noise, and generating a cockpit control state sequence includes: According to the preset acquisition cycle, the joystick displacement sensor, throttle opening detector, foot rudder angle detector and cockpit attitude feedback detector in the simulated flight cockpit are activated simultaneously to collect the raw cockpit control data of various sensors and perform preprocessing to obtain a standardized cockpit control data set. Based on a standardized cockpit control data set, the intensity of cockpit control changes within multiple consecutive acquisition cycles is calculated to distinguish between the operator's active control actions and invalid instantaneous actions caused by sensor noise, and the operator's control action data is filtered out. Key control parameters are extracted from the operator's control action data, including the lateral displacement of the joystick, the longitudinal displacement of the joystick, the throttle opening, the heading adjustment angle, the pitch control angle, and the roll control angle. The extracted key control parameters are bound to their corresponding acquisition times and sorted according to chronological order to generate a cockpit control state sequence.

[0010] Preferably, based on the cockpit control state sequence, all acquisition times are extracted as the synchronous acquisition reference times for multi-aircraft flight states. Synchronous acquisition commands are issued to the target aircraft and all nearby cooperating aircraft to synchronously acquire the basic flight parameters and formation relative position parameters of each aircraft and record the action timing markers, generating a multi-aircraft flight state sequence, including: Extract all acquisition times from the cockpit control state sequence and determine them as the synchronous acquisition reference times for multi-aircraft flight states. Send synchronous acquisition commands to the target aircraft and all nearby cooperating aircraft and output the set of synchronous acquisition reference times. Based on the set of synchronous acquisition reference times, the pitch angle, roll angle, yaw angle, spatial altitude, latitude and longitude horizontal position and ground flight speed of the target aircraft and each neighboring cooperative aircraft are synchronously acquired at each synchronous acquisition reference time, and the set of basic flight parameters is output. Based on the set of basic flight parameters, the straight-line formation spacing between the target aircraft and each neighboring cooperative aircraft, as well as the relative azimuth angle parameter of each neighboring cooperative aircraft relative to the target aircraft, are calculated, and the set of formation relative position parameters is output. Based on the set of relative position parameters of the formation, the attitude and velocity changes of each aircraft are monitored in real time, the time of occurrence of each attitude and velocity change of the target aircraft is recorded, and the follow-up time of the same change of each neighboring cooperative aircraft is recorded, and the action timing mark set is output. Based on the action timing mark set, the intensity of formation change is calculated, and all collected basic flight parameters and formation relative position parameters are bound to the corresponding synchronous acquisition reference time to generate a multi-aircraft flight state sequence.

[0011] Preferably, the step of analyzing the temporal correlation characteristics of multi-machine actions based on the action propagation association sequence, verifying the environmental disturbance matching relationship, separating the environmental disturbance-driven action from the neighboring machine inherited action, and generating a set of neighboring machine action constraint relationships includes: Based on the motion propagation association sequence, parameters such as the time of motion occurrence, direction of motion change, attitude change amplitude, and velocity change amplitude of the navigator, target, and neighboring cooperative aircraft are extracted. The multi-aircraft motion timing difference coefficient is calculated, and the multi-aircraft motion timing association result is generated. Based on the multi-aircraft action timing correlation results, starting from the time when the lead aircraft's action occurs, a preset time window that allows action inheritance is set. The system detects whether each neighboring cooperative aircraft exhibits the same type and direction of offset action within the preset time window, calculates the action inheritance correlation strength, and generates the action inheritance correlation results. Based on the action inheritance association results, the environmental disturbance intensity of the area where the lead aircraft and the neighboring cooperative aircraft are located at the time of the corresponding action is checked, the local difference coefficient of environmental disturbance is calculated, synchronous actions caused by the joint effect of the environment are excluded, and candidate results of neighboring aircraft action propagation are generated. Based on the candidate results of neighboring machine action propagation, the boundary between environmental disturbance-driven actions and neighboring machine inherited actions is delineated by combining the action inheritance association strength and the local difference coefficient of environmental disturbance. Mixed segments that simultaneously contain environmental disturbance-driven actions and neighboring machine inherited actions are segmented and labeled according to the order of action, generating action source separation sets. Based on the action source separation set, for each flight segment marked as an inherited action of a neighboring aircraft, the corresponding constraint level and constraint content are determined, and a set of neighboring aircraft action constraint relationships is generated.

[0012] Preferably, the step of establishing a multi-aircraft action timing correlation result, starting from the time of the lead aircraft's action, sets a preset time window that allows action inheritance, detects whether neighboring cooperative aircraft exhibit the same type and direction of offset actions within the preset time window, calculates the action inheritance correlation strength, and generates action inheritance correlation results, including: From the multi-aircraft action timing correlation results, extract the parameters of the navigator's action occurrence time, action type, action change direction and action change amplitude, and output the navigator's action parameter set; Based on the set of action parameters of the navigator aircraft, starting from the moment when the navigator aircraft's action occurs, a preset time window for allowing action inheritance is set, and the action inheritance detection window parameters are output. Based on the action inheritance detection window parameters, each neighboring cooperative aircraft is detected to determine whether it exhibits the same type and direction of offset action as the lead aircraft within a preset time window, and the neighboring aircraft action matching detection results are output. Based on the neighboring machine action matching detection results, the number of effective collections within multiple consecutive collection cycles is counted, the continuity and stability of neighboring machine action inheritance are calculated, and the action inheritance stability parameter is output. Based on the motion inheritance stability parameter, the motion change amplitude of the navigator and the neighboring cooperative aircraft are compared, the degree of matching of motion change amplitude is calculated, and random motions with excessively large differences in motion change amplitude are eliminated. The motion change amplitude matching result is then output. Based on the action sequence difference, action inheritance stability parameter and action change amplitude matching results, the action inheritance association strength is calculated, high-association action inheritance segments are marked and the action inheritance association results are output.

[0013] Preferably, the step of verifying the target aircraft's action inheritance state based on the set of neighboring aircraft action constraints, hierarchically suppressing neighboring aircraft inherited actions while retaining environmental disturbance-driven actions, and generating independent flight correction results, is then integrated into a formation control correction set, including: Based on the set of neighboring aircraft action constraint relationships, the parameters of the source aircraft of the inherited action, the type of inherited action, the direction of inheritance, the constraint level and the constraint period of the target aircraft are extracted. The current attitude adjustment, heading correction and speed following direction of the target aircraft are compared. Actions triggered by active cockpit control and driven by environmental disturbances are excluded. The action inheritance level of the target aircraft is determined and the target inheritance state determination result is output. Based on the target inheritance state determination results, a graded suppression strategy for attitude, heading and speed following quantities is formulated according to the action inheritance level and the corresponding constraint level. Suppression constraints are only applied to the action types marked as inherited by neighboring aircraft, and the inherited action suppression control results are generated. Based on the inherited action suppression control results, the corresponding environmental disturbance state sequence is checked back, the retention ratio of environmental disturbance-driven actions is calculated, the anti-disturbance correction actions driven by environmental disturbances are retained, and the repeated correction actions formed by the propagation of neighboring machines are attenuated proportionally to generate action diversion correction results. Based on the action diversion correction results, combined with the current formation space structure and aircraft safety distance requirements, the independent flight correction results that do not depend on the local offset actions of neighboring aircraft are recalculated. The flight stability control requirements of the target aircraft and the formation maintenance requirements are merged to generate a formation control correction set.

[0014] Preferably, based on the target inheritance state determination result, a graded suppression strategy for attitude, heading, and speed following quantities is formulated according to the action inheritance level and the corresponding constraint level. Suppression constraints are only applied to action types marked as inherited by neighboring aircraft, generating inherited action suppression control results, including: Based on the target inheritance state determination result, extract the target aircraft's action inheritance level, neighbor aircraft's action constraint level, inherited action type, and inheritance duration, and output the set of inherited action feature parameters. Based on the set of inherited action feature parameters, the inheritance level, constraint level and inheritance duration are comprehensively considered to dynamically calculate the inheritance action retention coefficient, determine the proportion of inherited actions that can be retained by neighboring machines, and output the inheritance action retention coefficient. Based on the inherited action retention coefficient, the attitude adjustment amount marked as roll inheritance or pitch inheritance is proportionally reduced, and the minimum safe attitude adjustment amount required to maintain flight stability is superimposed to obtain the suppressed attitude adjustment amount, and the attitude suppression correction result is output. Based on the inheritance action retention coefficient, the heading correction amount marked as yaw inheritance is proportionally reduced, and the minimum safe heading correction amount required to maintain heading stability is superimposed to obtain the suppressed heading correction amount, and the heading suppression correction result is output. Based on the inherited action retention coefficient, the speed following amount marked as speed following inheritance is proportionally reduced, and the minimum safe speed following amount required to maintain the basic formation spacing is superimposed to obtain the suppressed speed following amount, and the speed suppression correction result is output. Integrate the attitude suppression correction results, heading suppression correction results, and velocity suppression correction results to generate inherited motion suppression control results.

[0015] Preferably, the step of issuing remote control commands to the target aircraft and collecting real-aircraft feedback based on the formation control correction set, mapping the feedback state to the simulated flight cockpit, dynamically adjusting the suppression intensity of adjacent aircraft actions, and outputting multi-aircraft stable remote control results includes: Based on the formation control correction set, the final attitude correction, final heading correction, final speed correction, neighboring aircraft action suppression flag, environmental disturbance-driven action retention flag, and control correction effective period are extracted, converted into remote control correction commands that can be recognized by the target aircraft and issued. Simultaneously, pitch angle, roll angle, yaw angle, ground flight speed, spatial position, and straight formation distance with neighboring cooperative aircraft are collected after the actual aircraft is executed, and the actual aircraft execution feedback state set is output. Based on the actual aircraft execution feedback state set, the actual attitude state, heading state, speed state and formation space state of the target aircraft are mapped to the visual feedback system, attitude feedback system and formation space feedback system of the simulated flight cockpit, and the adjacent aircraft action suppression mark is displayed synchronously, and the cockpit synchronous feedback state set is output. Based on the cockpit synchronous feedback state set, the changes in lateral distance, longitudinal distance, altitude difference, relative speed and relative heading between the target aircraft and the neighboring cooperative aircraft are continuously monitored to determine whether the local offset action has a trend of continuous propagation or gradual convergence, and the spatial offset trend detection results are output. Based on the spatial offset trend detection results, when the local offset action is detected to be continuously expanding, the inherited action retention coefficient is dynamically reduced; when the offset action is detected to gradually converge to the formation reference range, the inherited action retention coefficient is gradually increased, and the updated neighbor action suppression strength result is output. Based on the updated results of the suppression intensity of neighboring aircraft actions, the latest actual aircraft execution feedback data, cockpit synchronization feedback status and suppression intensity parameters are integrated to generate a new set of cooperative flight fusion states and output multi-aircraft stable remote control results.

[0016] A remote flight cockpit control system based on multi-source sensing fusion, the system comprising: The data acquisition module is used to synchronously acquire cockpit control status, multi-aircraft flight status and environmental disturbance status, generate a collaborative flight fusion status set after unified time correlation, record the action propagation timing relationship between different aircraft, and output the action propagation correlation sequence. The separation module is used to analyze the temporal correlation characteristics of multi-machine actions based on the action propagation association sequence, verify the environmental disturbance matching relationship, separate the environmental disturbance driven actions and the neighboring machine inherited actions, and generate a set of neighboring machine action constraint relationships. The correction module is used to verify the action inheritance state of the target aircraft based on the set of neighboring aircraft action constraint relationships, suppress the inherited actions of neighboring aircraft in stages and retain the environmental disturbance-driven actions, and then integrate the generated independent flight correction results into a formation control correction set. The feedback module is used to issue remote control commands to the target aircraft and collect feedback from the actual aircraft based on the formation control correction set. It maps the feedback status to the flight simulator cockpit, dynamically adjusts the suppression intensity of adjacent aircraft actions, and outputs the stable remote control results of multiple aircraft.

[0017] The beneficial effects of the present invention are as follows: Compared with the prior art, the present invention has the following advantages: (1) This invention synchronously collects the cockpit control status, multi-aircraft flight status and environmental disturbance status, generates a set of cooperative flight fusion status after unified time correlation, records the action propagation timing relationship between different aircraft, and outputs the action propagation correlation sequence; based on the action propagation correlation sequence, analyzes the multi-aircraft action timing correlation characteristics, verifies the environmental disturbance matching relationship, separates the environmental disturbance-driven action and the neighboring aircraft inherited action, and generates a set of neighboring aircraft action constraint relationships; based on the set of neighboring aircraft action constraint relationships, verifies the action inheritance status of the target aircraft, suppresses the neighboring aircraft inherited action in a graded manner and retains the environmental disturbance-driven action, generates independent flight correction results and integrates them into a formation control correction set; based on the formation control correction set, issues remote control commands to the target aircraft and collects actual aircraft feedback, maps the feedback status to the simulated flight cockpit, dynamically adjusts the neighboring aircraft action suppression intensity, and outputs stable remote control results for multiple aircraft, thereby achieving accurate separation of action sources under multi-aircraft cooperation, graded suppression of neighboring aircraft inherited actions, retention of environmental disturbance resistance actions, blocking offset amplification, and improving the stability of remote actual aircraft control and formation safety.

[0018] (2) This invention achieves precise separation of action sources in multi-aircraft collaborative scenarios through multi-source perception fusion. It synchronously collects cockpit control, multi-aircraft flight, and environmental disturbance states and unifies time correlation to construct an action propagation correlation sequence, effectively distinguishing between environmental disturbance-driven actions and actions inherited from neighboring aircraft, thus fundamentally solving the problem of offset amplification caused by misleading action inheritance from neighboring aircraft. This invention adopts a hierarchical suppression mechanism, setting three constraint levels (low, medium, and high) according to the action inheritance intensity, and dynamically adjusting the suppression ratio of attitude, heading, and speed following quantities. This not only blocks the continuous propagation of local offset actions but also fully preserves the anti-disturbance correction actions and normal formation collaboration capabilities in the real environment. This invention maps the real aircraft execution state to the simulated cockpit in real time through virtual-real synchronous closed-loop feedback and dynamically adjusts the suppression intensity according to the spatial offset trend, forming a full-process closed-loop control. This invention significantly improves the stability and safety of remote real aircraft control in multi-aircraft formations under complex environments and expands the application scope of the flight simulator cockpit in multi-aircraft collaborative operation scenarios. Attached Figure Description

[0019] Figure 1 A flowchart of a remote real-world control method for a flight cockpit based on multi-source perception fusion provided in an embodiment of the present invention; Figure 2 This is a system block diagram of a remote flight cockpit control system based on multi-source perception fusion, provided as an embodiment of the present invention. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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.

[0021] like Figure 1 The diagram shown is a flowchart of a remote real-world control method for a flight cockpit based on multi-source perception fusion, provided in an embodiment of the present invention. Figure 1 The execution entity of the method shown can be a software and / or hardware device. The execution entity of this application can include, but is not limited to, at least one of the following: user equipment, network equipment, etc. User equipment can include, but is not limited to, computers, smartphones, personal digital assistants (PDAs), and the aforementioned electronic devices. Network equipment can include, but is not limited to, a single network server, a server group consisting of multiple network servers, or a cloud based on cloud computing consisting of a large number of computers or network servers. Cloud computing is a type of distributed computing, consisting of a super virtual computer composed of a group of loosely coupled computers. This embodiment does not limit this. Steps S1 to S4 are detailed as follows: S1 synchronously collects cockpit control status, multi-aircraft flight status and environmental disturbance status, generates a collaborative flight fusion status set after unified time correlation, records the action propagation timing relationship between different aircraft, and outputs the action propagation correlation sequence.

[0022] The synchronous acquisition of cockpit control status, multi-aircraft flight status, and environmental disturbance status is used to generate a cooperative flight fusion status set after unified time correlation. The action propagation timing relationship between different aircraft is recorded, and the action propagation correlation sequence is output, including: By simulating the control stick displacement sensor, throttle opening detector, rudder angle detector and cockpit attitude feedback detector in the flight cockpit, the operator's control action data is collected synchronously. After filtering out invalid instantaneous actions caused by sensor noise, a cockpit control state sequence is generated. Based on the cockpit control state sequence, all acquisition times are extracted as the synchronous acquisition reference times for multi-aircraft flight states. Synchronous acquisition commands are issued to the target aircraft and all nearby cooperating aircraft to synchronously acquire the basic flight parameters and formation relative position parameters of each aircraft and record the action timing marks to generate a multi-aircraft flight state sequence. Based on the multi-aircraft flight state sequence, the three-dimensional spatial position and precise time information of each aircraft at each synchronous acquisition moment are extracted. The local airflow changes, spatial obstacle distribution and environmental disturbance state in the flight area are sensed synchronously. After calculating the environmental disturbance intensity, an environmental disturbance state sequence is generated. The cockpit control state sequence, multi-aircraft flight state sequence, and environmental disturbance state sequence are resampled to the same main time axis. The adjacent time interpolation method is used to compensate for the time deviation caused by the difference in the acquisition cycle. After calculating the correlation strength of the fused state, a set of cooperative flight fused states is generated. Based on the fusion set of cooperative flight states, the sequential relationships of cockpit control actions, target aircraft attitude changes, navigator aircraft attitude changes, neighboring cooperative aircraft follow-up changes, and environmental disturbance changes are recorded in chronological order. Suspected neighbor aircraft action inheritance segments, active control action segments, and environmental disturbance-driven action segments are marked, and the action propagation correlation sequence is output.

[0023] Based on the joystick displacement sensor, throttle opening detector, rudder angle detector, and cockpit attitude feedback detector built into the flight simulator cockpit, the operator's real-time control actions are synchronously collected to obtain raw data such as joystick displacement, thrust opening, heading, and attitude adjustment. By filtering out invalid instantaneous actions caused by sensor noise, effective control information is extracted and organized into a time sequence to form a standardized cockpit control state sequence.

[0024] Using the acquisition time contained in the cockpit control state sequence as a unified reference time, a synchronous acquisition command is issued to the target aircraft and all nearby cooperating aircraft to synchronously acquire basic flight parameters such as pitch angle, roll angle, yaw angle, spatial position, and flight speed of each aircraft, as well as formation relative position parameters such as formation spacing and relative azimuth. At the same time, the timing marks of the target aircraft's actions and the following times of nearby cooperating aircraft are recorded to construct a multi-aircraft flight state sequence and realize the time-series synchronous acquisition of multi-aircraft flight data.

[0025] The three-dimensional spatial position and precise timestamp of each aircraft at the reference time are extracted from the multi-aircraft flight state sequence. Through airborne wind speed detection, air pressure detection, obstacle avoidance radar and visual perception devices, local airflow changes, spatial obstacle distribution and regional disturbance changes in the flight area are simultaneously perceived. The environmental disturbance intensity index is quantitatively calculated to generate an environmental disturbance state sequence that can characterize the external disturbance characteristics.

[0026] The cockpit control state sequence, multi-aircraft flight state sequence, and environmental disturbance state sequence are uniformly resampled to a time reference with the cockpit acquisition time as the main time axis. For the time deviation caused by the difference in acquisition cycle, the nearest time interpolation method is used for compensation and calibration. The correlation strength between the fusion states of multi-source data is calculated to form a time-aligned and dimension-unified collaborative flight fusion state set, thereby realizing the effective fusion of multi-source sensing data.

[0027] Based on the fusion set of cooperative flight states, the temporal sequence of cockpit active control actions, target aircraft attitude changes, navigator aircraft attitude changes, neighboring cooperative aircraft follow-up changes, and environmental disturbance changes is sorted out along the time dimension. Based on the timing, direction and intensity of the actions, suspected neighbor aircraft action inheritance segments, cockpit active control action segments and environmental disturbance driven action segments are marked, and action propagation association sequences that can characterize the causal relationship of actions are output.

[0028] The system synchronously collects the operator's control action data through joystick displacement sensors, throttle opening detectors, rudder angle detectors, and cockpit attitude feedback detectors within the simulated flight cockpit. After filtering out invalid instantaneous actions caused by sensor noise, it generates a cockpit control state sequence, including: According to the preset acquisition cycle, the joystick displacement sensor, throttle opening detector, foot rudder angle detector and cockpit attitude feedback detector in the simulated flight cockpit are activated simultaneously to collect the raw cockpit control data of various sensors and perform preprocessing to obtain a standardized cockpit control data set. Based on a standardized cockpit control data set, the intensity of cockpit control changes within multiple consecutive acquisition cycles is calculated to distinguish between the operator's active control actions and invalid instantaneous actions caused by sensor noise, and the operator's control action data is filtered out. Key control parameters are extracted from the operator's control action data, including the lateral displacement of the joystick, the longitudinal displacement of the joystick, the throttle opening, the heading adjustment angle, the pitch control angle, and the roll control angle. The extracted key control parameters are bound to their corresponding acquisition times and sorted according to chronological order to generate a cockpit control state sequence.

[0029] Based on a preset acquisition cycle, the system synchronously triggers the joystick displacement sensor, throttle opening detector, rudder angle detector, and cockpit attitude feedback detector configured in the simulated flight cockpit, acquiring raw cockpit control data output from each sensor unit in parallel. The raw data undergoes signal preprocessing, including noise reduction, amplification, analog-to-digital conversion, and format normalization, eliminating abnormal data caused by non-control factors such as signal distortion and electromagnetic interference. This generates a standardized cockpit control data set with time-aligned and dimensionally unified data, ensuring the reliability and consistency of the basic data.

[0030] Based on a standardized cockpit control data set, sensor data from multiple consecutive acquisition cycles are selected to calculate the intensity of cockpit control changes, quantifying the magnitude of change in control actions per unit time. Effective active control actions are distinguished from invalid instantaneous actions caused by sensor noise and mechanical vibration based on an intensity threshold. Low-intensity, non-continuous non-control action data are eliminated, accurately selecting effective control action data that truly reflects the operator's intent and avoiding interference from invalid signals in subsequent fusion analysis.

[0031] From the selected effective control action data, key control parameters representing the control intention are extracted, including the lateral displacement of the control stick, the longitudinal displacement of the control stick, the throttle opening, the heading adjustment angle, the pitch control angle, and the roll control angle. This comprehensively covers multi-dimensional control information such as displacement, thrust, heading, and attitude, and fully reproduces the control action characteristics of the cockpit.

[0032] Each extracted key control parameter is bound to its corresponding acquisition time, establishing a parameter-time correlation mapping. The bound data is then sorted and structured according to the chronological order of acquisition times, forming a cockpit control state sequence that is time-continuous, parameter-complete, and clearly labeled. This provides a unified time reference and cockpit control reference for the subsequent synchronous acquisition of multi-aircraft flight status and environmental disturbance status, ensuring the temporal consistency and logical correlation of multi-source sensing data fusion.

[0033] Based on the cockpit control state sequence, all acquisition times are extracted as the synchronous acquisition reference times for multi-aircraft flight states. Synchronous acquisition commands are issued to the target aircraft and all nearby cooperating aircraft to synchronously acquire the basic flight parameters and formation relative position parameters of each aircraft, and record the action timing markers, generating a multi-aircraft flight state sequence, including: Extract all acquisition times from the cockpit control state sequence and determine them as the synchronous acquisition reference times for multi-aircraft flight states. Send synchronous acquisition commands to the target aircraft and all nearby cooperating aircraft and output the set of synchronous acquisition reference times. Based on the set of synchronous acquisition reference times, the pitch angle, roll angle, yaw angle, spatial altitude, latitude and longitude horizontal position and ground flight speed of the target aircraft and each neighboring cooperative aircraft are synchronously acquired at each synchronous acquisition reference time, and the set of basic flight parameters is output. Based on the set of basic flight parameters, the straight-line formation spacing between the target aircraft and each neighboring cooperative aircraft, as well as the relative azimuth angle parameter of each neighboring cooperative aircraft relative to the target aircraft, are calculated, and the set of formation relative position parameters is output. Based on the set of relative position parameters of the formation, the attitude and velocity changes of each aircraft are monitored in real time, the time of occurrence of each attitude and velocity change of the target aircraft is recorded, and the follow-up time of the same change of each neighboring cooperative aircraft is recorded, and the action timing mark set is output. Based on the action timing mark set, the intensity of formation change is calculated, and all collected basic flight parameters and formation relative position parameters are bound to the corresponding synchronous acquisition reference time to generate a multi-aircraft flight state sequence.

[0034] All acquisition moments are extracted from the cockpit control state sequence and determined as the synchronous acquisition reference moments for multi-aircraft flight states, thereby constructing a unified time reference for multi-aircraft data acquisition. Synchronous acquisition commands are issued to the target aircraft and all nearby cooperating aircraft, outputting a set of synchronous acquisition reference moments containing all reference moments, ensuring that the multi-aircraft data acquisition timing is completely aligned with the cockpit control timing, and eliminating timing deviations caused by asynchronous acquisition.

[0035] Based on the synchronous acquisition reference time set, the basic flight parameters of the target aircraft and each neighboring cooperative aircraft are synchronously acquired at each reference time, including pitch angle, roll angle, yaw angle, spatial altitude, latitude and longitude horizontal position and ground flight speed, comprehensively covering basic flight dimensions such as attitude, position and speed, and outputting a standardized basic flight parameter set to accurately characterize the real-time flight status of each aircraft.

[0036] Based on the set of basic flight parameters, the linear formation spacing between the target aircraft and each neighboring cooperating aircraft is calculated using a spatial distance algorithm. At the same time, the relative azimuth angles of each neighboring cooperating aircraft relative to the target aircraft are solved, the spatial layout characteristics of the formation are quantified, and the set of relative position parameters of the formation is output, which fully reflects the spatial structural relationship of the formation.

[0037] Based on the set of relative position parameters of the formation, the attitude and velocity changes of each aircraft are monitored in real time. The timing of each attitude and velocity change of the target aircraft is accurately recorded, and the timing of the follow-up attitude and velocity changes of the neighboring cooperative aircraft is recorded synchronously. The temporal correspondence between the occurrence of the action and the follow-up is constructed, and a set of action timing markers is output.

[0038] Based on the action timing mark set, the intensity of formation change is quantitatively calculated to characterize the degree of dynamic change in formation. The basic flight parameters, formation relative position parameters and corresponding synchronous acquisition reference times are bound one by one to complete the temporal integration of flight data, position data and timing marks, and generate a multi-aircraft flight state sequence that is temporally continuous, parameter-complete and clearly marked.

[0039] S2, based on the action propagation association sequence, analyzes the temporal association characteristics of multi-machine actions, verifies the environmental disturbance matching relationship, separates the environmental disturbance-driven actions from the neighboring machine inherited actions, and generates a set of neighboring machine action constraint relationships.

[0040] The process involves analyzing the temporal correlation characteristics of multi-machine actions based on action propagation association sequences, verifying environmental disturbance matching relationships, separating environmental disturbance-driven actions from neighboring machine inherited actions, and generating a set of neighboring machine action constraint relationships, including: Based on the motion propagation association sequence, parameters such as the time of motion occurrence, direction of motion change, attitude change amplitude, and velocity change amplitude of the navigator, target, and neighboring cooperative aircraft are extracted. The multi-aircraft motion timing difference coefficient is calculated, and the multi-aircraft motion timing association result is generated. Based on the multi-aircraft action timing correlation results, starting from the time when the lead aircraft's action occurs, a preset time window that allows action inheritance is set. The system detects whether each neighboring cooperative aircraft exhibits the same type and direction of offset action within the preset time window, calculates the action inheritance correlation strength, and generates the action inheritance correlation results. Based on the action inheritance association results, the environmental disturbance intensity of the area where the lead aircraft and the neighboring cooperative aircraft are located at the time of the corresponding action is checked, the local difference coefficient of environmental disturbance is calculated, synchronous actions caused by the joint effect of the environment are excluded, and candidate results of neighboring aircraft action propagation are generated. Based on the candidate results of neighboring machine action propagation, the boundary between environmental disturbance-driven actions and neighboring machine inherited actions is delineated by combining the action inheritance association strength and the local difference coefficient of environmental disturbance. Mixed segments that simultaneously contain environmental disturbance-driven actions and neighboring machine inherited actions are segmented and labeled according to the order of action, generating action source separation sets. Based on the action source separation set, for each flight segment marked as an inherited action of a neighboring aircraft, the corresponding constraint level and constraint content are determined, and a set of neighboring aircraft action constraint relationships is generated.

[0041] Based on the motion propagation association sequence, the timing of motion occurrence, direction of motion change, attitude change amplitude, and velocity change amplitude of the navigator, target, and neighboring cooperative aircraft are extracted. By normalizing the time difference, attitude difference, and velocity difference, the multi-aircraft motion time sequence difference coefficient is calculated to quantitatively characterize the temporal synchronization and consistency of changes in the multi-aircraft motion occurrence, and to generate multi-aircraft motion time sequence association results.

[0042] Based on the multi-aircraft action timing correlation results, the time when the lead aircraft's action occurs is taken as the starting point. A preset time window is set as the action inheritance judgment interval. Each aircraft is checked to see whether the neighboring cooperative aircraft has the same type and direction of offset action as the lead aircraft within the window. The action inheritance correlation strength is calculated by combining the number of effective matches, timing differences, and amplitude matching degree. The credibility of neighboring aircraft action inheritance is quantitatively characterized, action inheritance correlation results are generated, and suspected neighboring aircraft action inheritance segments are screened.

[0043] Based on the action inheritance association results, the environmental disturbance intensity of the area where the lead aircraft and the neighboring cooperative aircraft are located at the time of the action is checked. The local difference coefficient of environmental disturbance is calculated by the disturbance intensity ratio, the distance difference between the disturbance center and the disturbance action timing difference. Local disturbances and global disturbances are distinguished, synchronous actions caused by multiple aircraft under common environmental disturbances are excluded, candidate results of neighboring aircraft action propagation are generated, and interference terms caused by common environmental effects are eliminated.

[0044] Based on the candidate results of neighboring machine action propagation, the action inheritance association strength and the local difference coefficient of environmental disturbance are integrated to delineate the boundary between environmental disturbance-driven actions and neighboring machine inherited actions. For mixed action segments that contain both types of causes, they are segmented and marked according to the time sequence of environmental disturbance acting first and neighboring machine inheritance following, clarifying the source of each action segment, generating a separate set of action sources, and achieving accurate tracing of the cause of actions.

[0045] Based on the separation set of action sources, for each flight segment of inherited actions from neighboring aircraft, low, medium, and high constraint levels and corresponding constraint contents in attitude, heading, and speed dimensions are determined according to the inheritance intensity, duration, and proportion of environmental influence, thereby generating a set of neighboring aircraft action constraint relationships.

[0046] By using the above methods, we can accurately distinguish between actions driven by environmental disturbances and actions inherited from neighboring aircraft, forming a hierarchical constraint basis. This provides a constraint benchmark for subsequent formation control corrections, preventing the risk of amplified actions inherited from neighboring aircraft from the root.

[0047] The method based on the multi-aircraft action timing correlation results, starting from the time of the lead aircraft's action, sets a preset time window that allows action inheritance, detects whether each neighboring cooperative aircraft exhibits the same type and direction of offset action within the preset time window, calculates the action inheritance correlation strength, and generates action inheritance correlation results, including: From the multi-aircraft action timing correlation results, extract the parameters of the navigator's action occurrence time, action type, action change direction and action change amplitude, and output the navigator's action parameter set; Based on the set of action parameters of the navigator aircraft, starting from the moment when the navigator aircraft's action occurs, a preset time window for allowing action inheritance is set, and the action inheritance detection window parameters are output. Based on the action inheritance detection window parameters, each neighboring cooperative aircraft is detected to determine whether it exhibits the same type and direction of offset action as the lead aircraft within a preset time window, and the neighboring aircraft action matching detection results are output. Based on the neighboring machine action matching detection results, the number of effective collections within multiple consecutive collection cycles is counted, the continuity and stability of neighboring machine action inheritance are calculated, and the action inheritance stability parameter is output. Based on the motion inheritance stability parameter, the motion change amplitude of the navigator and the neighboring cooperative aircraft are compared, the degree of matching of motion change amplitude is calculated, and random motions with excessively large differences in motion change amplitude are eliminated. The motion change amplitude matching result is then output. Based on the action sequence difference, action inheritance stability parameter and action change amplitude matching results, the action inheritance association strength is calculated, high-association action inheritance segments are marked and the action inheritance association results are output.

[0048] From the multi-aircraft action timing correlation results, key parameters such as the time of occurrence of the navigator's action, action type, direction of action change, and magnitude of action change are accurately extracted to construct a set of navigator action parameters, which fully characterizes the navigator's role as the source of the action.

[0049] Based on the set of action parameters of the navigator aircraft, taking the moment when the navigator aircraft's action occurs as the starting point, a preset time window is set to allow neighboring aircraft to inherit the action, the effective time interval for action inheritance judgment is determined, and the action inheritance detection window parameters are output. The judgment range of inherited actions is limited from the temporal dimension to avoid misjudgment caused by temporal misalignment.

[0050] Based on the action inheritance detection window parameters, action matching detection is performed on each neighboring cooperative aircraft to determine whether it has an offset action of the same type and direction as the lead aircraft within a preset time window. The neighboring aircraft action matching detection results are output, and suspected inherited action segments that meet the time sequence and type characteristics are preliminarily screened.

[0051] Based on the neighboring machine action matching detection results, the number of valid collections that meet the matching conditions within multiple consecutive collection cycles is counted, the continuity and stability of neighboring machine action inheritance are calculated, and action inheritance stability parameters are generated to eliminate unstable interference such as single random matching and short-term abnormal fluctuations, thus ensuring the reliability of inherited action recognition.

[0052] Based on the motion inheritance stability parameter, the motion change amplitude of the lead aircraft and the neighboring cooperative aircraft are compared, the degree of matching of motion change amplitude is quantitatively calculated, and random motions with excessively large amplitude differences and lack of correlation are eliminated. The motion change amplitude matching result is output, and invalid matching items are further filtered from the dimension of motion intensity.

[0053] By combining the differences in action timing, the stability parameters of action inheritance, and the matching results of action change amplitude, the action inheritance correlation strength is calculated by weighting, the confidence of neighboring machine action inheritance is quantified, and action inheritance segments with high correlation are marked, and the action inheritance correlation results are output.

[0054] The above methods can be used to quantitatively assess the credibility of neighboring machine action inheritance, accurately identify real inheritance behavior, eliminate accidental interference actions, and generate quantifiable and traceable action inheritance association results.

[0055] S3, based on the set of neighboring aircraft action constraint relationships, verifies the action inheritance state of the target aircraft, suppresses the inherited actions of neighboring aircraft in stages while retaining the environmental disturbance-driven actions, and integrates the generated independent flight correction results into a formation control correction set.

[0056] The set of neighboring aircraft action constraint relationships is used to verify the action inheritance state of the target aircraft, suppress inherited actions of neighboring aircraft in a graded manner while retaining environmental disturbance-driven actions, and generate independent flight correction results, which are then integrated into a formation control correction set, including: Based on the set of neighboring aircraft action constraint relationships, the parameters of the source aircraft of the inherited action, the type of inherited action, the direction of inheritance, the constraint level and the constraint period of the target aircraft are extracted. The current attitude adjustment, heading correction and speed following direction of the target aircraft are compared. Actions triggered by active cockpit control and driven by environmental disturbances are excluded. The action inheritance level of the target aircraft is determined and the target inheritance state determination result is output. Based on the target inheritance state determination results, a graded suppression strategy for attitude, heading and speed following quantities is formulated according to the action inheritance level and the corresponding constraint level. Suppression constraints are only applied to the action types marked as inherited by neighboring aircraft, and the inherited action suppression control results are generated. Based on the inherited action suppression control results, the corresponding environmental disturbance state sequence is checked back, the retention ratio of environmental disturbance-driven actions is calculated, the anti-disturbance correction actions driven by environmental disturbances are retained, and the repeated correction actions formed by the propagation of neighboring machines are attenuated proportionally to generate action diversion correction results. Based on the action diversion correction results, combined with the current formation space structure and aircraft safety distance requirements, the independent flight correction results that do not depend on the local offset actions of neighboring aircraft are recalculated. The flight stability control requirements of the target aircraft and the formation maintenance requirements are merged to generate a formation control correction set.

[0057] From the set of neighboring aircraft action constraints, key parameters such as the source aircraft of the inherited action, the type of inherited action, the direction of inheritance, the constraint level, and the constraint period are extracted for the target aircraft. These parameters are compared with the target aircraft's current attitude adjustment, heading correction, and speed following direction. Effective actions triggered by active cockpit control or directly driven by environmental disturbances are excluded. Only passive actions caused by neighboring aircraft inheritance are assessed for their level, and the target inheritance status determination result is output to clarify the degree to which the target aircraft is affected by the actions of neighboring aircraft.

[0058] Based on the target inheritance state determination results, and combined with the action inheritance level and preset constraint level, a graded suppression strategy for attitude, heading, and speed following quantities is formulated. The suppression operation only applies to the action types marked as inherited by neighboring aircraft, avoiding false suppression of normal control and disturbance rejection actions, generating inherited action suppression control results, and achieving differentiated and precise suppression constraints.

[0059] Based on the inherited action suppression control results, the environmental disturbance state sequence for the corresponding time period is reviewed, and the retention ratio of environmental disturbance-driven actions is quantitatively calculated. Anti-disturbance correction actions triggered by environmental disturbances are fully retained to ensure the aircraft's autonomous stability in response to external disturbances. Repeated correction actions propagated from neighboring aircraft are proportionally attenuated to weaken ineffective following effects, generating action diversion correction results to separate effective and ineffective actions.

[0060] Based on the action diversion correction results, and combined with the current formation space structure and aircraft safety distance requirements, independent flight correction results that do not depend on the local offset actions of neighboring aircraft are recalculated. The target aircraft's own flight stability control requirements are integrated with the formation formation maintenance requirements to form the final formation control correction set. This provides accurate and reliable control basis for subsequent remote control command issuance, enabling stable remote real-world control in multi-aircraft collaborative scenarios.

[0061] The above methods can accurately identify the action inheritance state of the target aircraft, implement graded suppression of the inherited actions of neighboring aircraft, and fully preserve the environmental disturbance-driven actions. Formation stability control can be achieved through independent flight correction, and the step-by-step amplification of the local deviation actions of neighboring aircraft can be blocked from the control link, thus ensuring the safety and stability of remote real-aircraft control.

[0062] Based on the target inheritance state determination result, and according to the action inheritance level and corresponding constraint level, a graded suppression strategy for attitude, heading, and speed following quantities is formulated. Suppression constraints are only applied to action types marked as inherited by neighboring aircraft, generating inherited action suppression control results, including: Based on the target inheritance state determination result, extract the target aircraft's action inheritance level, neighbor aircraft's action constraint level, inherited action type, and inheritance duration, and output the set of inherited action feature parameters. Based on the set of inherited action feature parameters, the inheritance level, constraint level and inheritance duration are comprehensively considered to dynamically calculate the inheritance action retention coefficient, determine the proportion of inherited actions that can be retained by neighboring machines, and output the inheritance action retention coefficient. Based on the inherited action retention coefficient, the attitude adjustment amount marked as roll inheritance or pitch inheritance is proportionally reduced, and the minimum safe attitude adjustment amount required to maintain flight stability is superimposed to obtain the suppressed attitude adjustment amount, and the attitude suppression correction result is output. Based on the inheritance action retention coefficient, the heading correction amount marked as yaw inheritance is proportionally reduced, and the minimum safe heading correction amount required to maintain heading stability is superimposed to obtain the suppressed heading correction amount, and the heading suppression correction result is output. Based on the inherited action retention coefficient, the speed following amount marked as speed following inheritance is proportionally reduced, and the minimum safe speed following amount required to maintain the basic formation spacing is superimposed to obtain the suppressed speed following amount, and the speed suppression correction result is output. Integrate the attitude suppression correction results, heading suppression correction results, and velocity suppression correction results to generate inherited motion suppression control results.

[0063] Based on the target inheritance state determination results, the action inheritance level of the target aircraft, the action constraint level of the neighboring aircraft, the inherited action type and the inheritance duration are extracted to construct a set of inherited action feature parameters, which fully characterize the intensity, constraint requirements, action dimension and duration of the inherited action.

[0064] Based on the set of inherited action feature parameters, the inherited action retention coefficient is dynamically calculated by combining the action inheritance level, the neighboring machine action constraint level, and the inheritance duration. This quantitatively determines the proportion of inherited actions that can be retained by the neighboring machine. The coefficient is adaptively adjusted according to the inheritance intensity, constraint level, and duration to achieve fine matching of the suppression force and output the inherited action retention coefficient.

[0065] Based on the inheritance action retention coefficient, the attitude adjustment amount marked as roll inheritance or pitch inheritance is proportionally reduced according to the retention coefficient, and then superimposed with the minimum safe attitude adjustment amount required to maintain flight stability to obtain the suppressed attitude adjustment amount. The attitude suppression correction result is output, which takes into account both suppression effect and attitude stability.

[0066] Based on the inheritance action retention coefficient, the heading correction amount marked as yaw inheritance is proportionally reduced, and the minimum safe heading correction amount required to maintain heading stability is superimposed to obtain the suppressed heading correction amount. The heading suppression correction result is output to avoid excessive amplification of heading follow.

[0067] Based on the inherited action retention coefficient, the speed following amount marked as speed following inheritance is proportionally reduced, and the minimum safe speed following amount required to maintain the basic formation spacing is superimposed to obtain the suppressed speed following amount. The speed suppression correction result is output to balance the formation maintenance and suppression requirements.

[0068] By integrating attitude suppression correction results, heading suppression correction results, and speed suppression correction results, a complete and unified inherited action suppression control result is formed, providing precise control input for subsequent action diversion correction and formation control, and realizing hierarchical and controllable suppression of inherited actions of neighboring aircraft.

[0069] The above methods can be used to implement graded and precise suppression of three types of inherited actions: attitude, heading, and speed. At the same time, minimum safety control quantities can be superimposed to ensure basic flight stability while blocking ineffective following, and generate inherited action suppression control results that can be directly used in the control link.

[0070] S4, based on the formation control correction set, sends remote control commands to the target aircraft and collects feedback from the actual aircraft. It maps the feedback status to the flight simulator cockpit, dynamically adjusts the intensity of suppression of adjacent aircraft actions, and outputs stable remote control results for multiple aircraft.

[0071] The formation control correction set is used to issue remote control commands to the target aircraft and collect feedback from the actual aircraft. The feedback state is mapped to the simulated flight cockpit, dynamically adjusting the suppression intensity of adjacent aircraft actions, and outputting stable remote control results for multiple aircraft, including: Based on the formation control correction set, the final attitude correction, final heading correction, final speed correction, neighboring aircraft action suppression flag, environmental disturbance-driven action retention flag, and control correction effective period are extracted, converted into remote control correction commands that can be recognized by the target aircraft and issued. Simultaneously, pitch angle, roll angle, yaw angle, ground flight speed, spatial position, and straight formation distance with neighboring cooperative aircraft are collected after the actual aircraft is executed, and the actual aircraft execution feedback state set is output. Based on the actual aircraft execution feedback state set, the actual attitude state, heading state, speed state and formation space state of the target aircraft are mapped to the visual feedback system, attitude feedback system and formation space feedback system of the simulated flight cockpit, and the adjacent aircraft action suppression mark is displayed synchronously, and the cockpit synchronous feedback state set is output. Based on the cockpit synchronous feedback state set, the changes in lateral distance, longitudinal distance, altitude difference, relative speed and relative heading between the target aircraft and the neighboring cooperative aircraft are continuously monitored to determine whether the local offset action has a trend of continuous propagation or gradual convergence, and the spatial offset trend detection results are output. Based on the spatial offset trend detection results, when the local offset action is detected to be continuously expanding, the inherited action retention coefficient is dynamically reduced; when the offset action is detected to gradually converge to the formation reference range, the inherited action retention coefficient is gradually increased, and the updated neighbor action suppression strength result is output. Based on the updated results of the suppression intensity of neighboring aircraft actions, the latest actual aircraft execution feedback data, cockpit synchronization feedback status and suppression intensity parameters are integrated to generate a new set of cooperative flight fusion states and output multi-aircraft stable remote control results.

[0072] Based on the formation control correction set, the final attitude correction, final heading correction, final speed correction, adjacent aircraft action suppression flags, environmental disturbance-driven action retention flags, and control correction effective periods are extracted. These control parameters are converted into remote control correction commands recognizable by the target aircraft and issued for execution. Simultaneously, pitch angle, roll angle, yaw angle, ground speed, spatial position, and straight-line formation distance with neighboring cooperating aircraft are collected after the actual aircraft's execution. A set of actual aircraft execution feedback states is output to accurately characterize the actual aircraft control execution effect.

[0073] Based on the actual aircraft's execution feedback state set, the target aircraft's actual attitude, heading, velocity, and formation space states are mapped to the visual feedback system, attitude feedback system, and formation space feedback system of the flight simulator cockpit, respectively. Simultaneously, adjacent aircraft action suppression markers are displayed on the cockpit interface, allowing the operator to monitor the actual aircraft's operation and suppression status in real time. The system outputs a synchronized cockpit feedback state set, achieving real-time synchronization between virtual and real states.

[0074] Based on the cockpit synchronous feedback state set, the system continuously monitors the changes in lateral spacing, longitudinal spacing, altitude difference, relative speed, and relative heading between the target aircraft and nearby cooperating aircraft. It quantitatively analyzes the spatial offset, rate of change, and duration, determines whether the local offset action continues to propagate or gradually converges, and outputs the spatial offset trend detection results.

[0075] Based on the spatial offset trend detection results, the suppression intensity is dynamically and adaptively adjusted: when the local offset action continues to expand, the inherited action retention coefficient is dynamically reduced to enhance the suppression of neighboring machine actions; when the offset action gradually converges to the formation reference range, the inherited action retention coefficient is gradually increased to restore normal formation coordination capability, and the updated neighboring machine action suppression intensity result is output to achieve dynamic optimization of the suppression strategy.

[0076] Based on the updated results of the suppression intensity of neighboring aircraft actions, the latest actual aircraft execution feedback data, cockpit synchronization feedback status and suppression intensity parameters are integrated to reconstruct the temporal correlation of multi-source data, generate a new set of cooperative flight fusion states, and finally output the stable remote control results of multi-aircraft, forming a closed-loop iterative control to ensure the continuous stability and reliability of remote actual aircraft control in multi-aircraft cooperative scenarios.

[0077] like Figure 2 The diagram shown is a system block diagram of a remote flight cockpit control system based on multi-source perception fusion provided in an embodiment of the present invention. The system includes: The data acquisition module is used to synchronously acquire cockpit control status, multi-aircraft flight status and environmental disturbance status, generate a collaborative flight fusion status set after unified time correlation, record the action propagation timing relationship between different aircraft, and output the action propagation correlation sequence. The separation module is used to analyze the temporal correlation characteristics of multi-machine actions based on the action propagation association sequence, verify the environmental disturbance matching relationship, separate the environmental disturbance driven actions and the neighboring machine inherited actions, and generate a set of neighboring machine action constraint relationships. The correction module is used to verify the action inheritance state of the target aircraft based on the set of neighboring aircraft action constraint relationships, suppress the inherited actions of neighboring aircraft in stages and retain the environmental disturbance-driven actions, and then integrate the generated independent flight correction results into a formation control correction set. The feedback module is used to issue remote control commands to the target aircraft and collect feedback from the actual aircraft based on the formation control correction set. It maps the feedback status to the flight simulator cockpit, dynamically adjusts the suppression intensity of adjacent aircraft actions, and outputs the stable remote control results of multiple aircraft.

[0078] Figure 2 The apparatus of the illustrated embodiment can be used to perform corresponding actions. Figure 1 The steps in the method embodiments shown are implemented in a similar manner and have similar technical effects, and will not be repeated here.

[0079] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A remote real-world control method for a flight cockpit based on multi-source perception fusion, characterized in that, The method includes: The system synchronously collects cockpit control status, multi-aircraft flight status, and environmental disturbance status. After unified time correlation, it generates a collaborative flight fusion status set, records the action propagation timing relationship between different aircraft, and outputs the action propagation correlation sequence. Based on the action propagation association sequence, the temporal association characteristics of multi-machine actions are analyzed, the environmental disturbance matching relationship is verified, the environmental disturbance-driven action and the neighbor machine inherited action are separated, and a set of neighbor machine action constraint relationships is generated. Based on the set of neighboring aircraft action constraint relationships, the action inheritance state of the target aircraft is verified, the inherited actions of neighboring aircraft are suppressed in stages while retaining the environmental disturbance-driven actions, and after generating independent flight correction results, they are integrated into a formation control correction set. Based on the formation control correction set, remote control commands are sent to the target aircraft and feedback from the actual aircraft is collected. The feedback status is mapped to the flight simulator cockpit, the intensity of suppression of adjacent aircraft actions is dynamically adjusted, and the results of stable remote control of multiple aircraft are output.

2. The remote real-world control method for a flight cockpit based on multi-source perception fusion according to claim 1, characterized in that, The synchronous acquisition of cockpit control status, multi-aircraft flight status, and environmental disturbance status is used to generate a cooperative flight fusion status set after unified time correlation. The action propagation timing relationship between different aircraft is recorded, and the action propagation correlation sequence is output, including: By simulating the control stick displacement sensor, throttle opening detector, rudder angle detector and cockpit attitude feedback detector in the flight cockpit, the operator's control action data is collected synchronously. After filtering out invalid instantaneous actions caused by sensor noise, a cockpit control state sequence is generated. Based on the cockpit control state sequence, all acquisition times are extracted as the synchronous acquisition reference times for multi-aircraft flight states. Synchronous acquisition commands are issued to the target aircraft and all nearby cooperating aircraft to synchronously acquire the basic flight parameters and formation relative position parameters of each aircraft and record the action timing marks to generate a multi-aircraft flight state sequence. Based on the multi-aircraft flight state sequence, the three-dimensional spatial position and precise time information of each aircraft at each synchronous acquisition moment are extracted. The local airflow changes, spatial obstacle distribution and environmental disturbance state in the flight area are sensed synchronously. After calculating the environmental disturbance intensity, an environmental disturbance state sequence is generated. The cockpit control state sequence, multi-aircraft flight state sequence, and environmental disturbance state sequence are resampled to the same main time axis. The adjacent time interpolation method is used to compensate for the time deviation caused by the difference in the acquisition cycle. After calculating the correlation strength of the fused state, a set of cooperative flight fused states is generated. Based on the fusion set of cooperative flight states, the sequential relationships of cockpit control actions, target aircraft attitude changes, navigator aircraft attitude changes, neighboring cooperative aircraft follow-up changes, and environmental disturbance changes are recorded in chronological order. Suspected neighbor aircraft action inheritance segments, active control action segments, and environmental disturbance-driven action segments are marked, and the action propagation correlation sequence is output.

3. The remote real-world control method for a flight cockpit based on multi-source perception fusion according to claim 2, characterized in that, The system synchronously collects the operator's control action data through joystick displacement sensors, throttle opening detectors, rudder angle detectors, and cockpit attitude feedback detectors within the simulated flight cockpit. After filtering out invalid instantaneous actions caused by sensor noise, it generates a cockpit control state sequence, including: According to the preset acquisition cycle, the joystick displacement sensor, throttle opening detector, foot rudder angle detector and cockpit attitude feedback detector in the simulated flight cockpit are activated simultaneously to collect the raw cockpit control data of various sensors and perform preprocessing to obtain a standardized cockpit control data set. Based on a standardized cockpit control data set, the intensity of cockpit control changes within multiple consecutive acquisition cycles is calculated to distinguish between the operator's active control actions and invalid instantaneous actions caused by sensor noise, and the operator's control action data is filtered out. Key control parameters are extracted from the operator's control action data, including the lateral displacement of the joystick, the longitudinal displacement of the joystick, the throttle opening, the heading adjustment angle, the pitch control angle, and the roll control angle. The extracted key control parameters are bound to their corresponding acquisition times and sorted according to chronological order to generate a cockpit control state sequence.

4. The remote real-world control method for a flight cockpit based on multi-source perception fusion according to claim 2, characterized in that, Based on the cockpit control state sequence, all acquisition times are extracted as the synchronous acquisition reference times for multi-aircraft flight states. Synchronous acquisition commands are issued to the target aircraft and all nearby cooperating aircraft to synchronously acquire the basic flight parameters and formation relative position parameters of each aircraft, and record the action timing markers, generating a multi-aircraft flight state sequence, including: Extract all acquisition times from the cockpit control state sequence and determine them as the synchronous acquisition reference times for multi-aircraft flight states. Send synchronous acquisition commands to the target aircraft and all nearby cooperating aircraft and output the set of synchronous acquisition reference times. Based on the set of synchronous acquisition reference times, the pitch angle, roll angle, yaw angle, spatial altitude, latitude and longitude horizontal position and ground flight speed of the target aircraft and each neighboring cooperative aircraft are synchronously acquired at each synchronous acquisition reference time, and the set of basic flight parameters is output. Based on the set of basic flight parameters, the straight-line formation spacing between the target aircraft and each neighboring cooperative aircraft, as well as the relative azimuth angle parameter of each neighboring cooperative aircraft relative to the target aircraft, are calculated, and the set of formation relative position parameters is output. Based on the set of relative position parameters of the formation, the attitude and velocity changes of each aircraft are monitored in real time, the time of occurrence of each attitude and velocity change of the target aircraft is recorded, and the follow-up time of the same change of each neighboring cooperative aircraft is recorded, and the action timing mark set is output. Based on the action timing mark set, the intensity of formation change is calculated, and all collected basic flight parameters and formation relative position parameters are bound to the corresponding synchronous acquisition reference time to generate a multi-aircraft flight state sequence.

5. The remote real-world control method for a flight cockpit based on multi-source perception fusion according to claim 1, characterized in that, The process involves analyzing the temporal correlation characteristics of multi-machine actions based on action propagation association sequences, verifying environmental disturbance matching relationships, separating environmental disturbance-driven actions from neighboring machine inherited actions, and generating a set of neighboring machine action constraint relationships, including: Based on the motion propagation association sequence, parameters such as the time of motion occurrence, direction of motion change, attitude change amplitude, and velocity change amplitude of the navigator, target, and neighboring cooperative aircraft are extracted. The multi-aircraft motion timing difference coefficient is calculated, and the multi-aircraft motion timing association result is generated. Based on the multi-aircraft action timing correlation results, starting from the time when the lead aircraft's action occurs, a preset time window that allows action inheritance is set. The system detects whether each neighboring cooperative aircraft exhibits the same type and direction of offset action within the preset time window, calculates the action inheritance correlation strength, and generates the action inheritance correlation results. Based on the action inheritance association results, the environmental disturbance intensity of the area where the lead aircraft and the neighboring cooperative aircraft are located at the time of the corresponding action is checked, the local difference coefficient of environmental disturbance is calculated, synchronous actions caused by the joint effect of the environment are excluded, and candidate results of neighboring aircraft action propagation are generated. Based on the candidate results of neighboring machine action propagation, the boundary between environmental disturbance-driven actions and neighboring machine inherited actions is delineated by combining the action inheritance association strength and the local difference coefficient of environmental disturbance. Mixed segments that simultaneously contain environmental disturbance-driven actions and neighboring machine inherited actions are segmented and labeled according to the order of action, generating action source separation sets. Based on the action source separation set, for each flight segment marked as an inherited action of a neighboring aircraft, the corresponding constraint level and constraint content are determined, and a set of neighboring aircraft action constraint relationships is generated.

6. The remote real-world control method for a flight cockpit based on multi-source perception fusion according to claim 5, characterized in that, The method based on the multi-aircraft action timing correlation results, starting from the time of the lead aircraft's action, sets a preset time window that allows action inheritance, detects whether each neighboring cooperative aircraft exhibits the same type and direction of offset action within the preset time window, calculates the action inheritance correlation strength, and generates action inheritance correlation results, including: From the multi-aircraft action timing correlation results, extract the parameters of the timing of the action occurrence, action type, direction of action change and magnitude of action change of the navigator, and output the navigator action parameter set; Based on the set of action parameters of the navigator aircraft, starting from the moment when the navigator aircraft's action occurs, a preset time window for allowing action inheritance is set, and the action inheritance detection window parameters are output. Based on the action inheritance detection window parameters, each neighboring cooperative aircraft is detected to determine whether it exhibits the same type and direction of offset action as the lead aircraft within a preset time window, and the neighboring aircraft action matching detection results are output. Based on the neighboring machine action matching detection results, the number of effective collections within multiple consecutive collection cycles is counted, the continuity and stability of neighboring machine action inheritance are calculated, and the action inheritance stability parameter is output. Based on the motion inheritance stability parameter, the motion change amplitude of the navigator and the neighboring cooperative aircraft are compared, the degree of matching of motion change amplitude is calculated, and random motions with excessively large differences in motion change amplitude are eliminated. The motion change amplitude matching result is then output. Based on the action sequence difference, action inheritance stability parameter and action change amplitude matching results, the action inheritance association strength is calculated, high-association action inheritance segments are marked and the action inheritance association results are output.

7. The remote real-world control method for a flight cockpit based on multi-source perception fusion according to claim 1, characterized in that, The set of neighboring aircraft action constraint relationships is used to verify the action inheritance state of the target aircraft, suppress inherited actions of neighboring aircraft in a graded manner while retaining environmental disturbance-driven actions, and generate independent flight correction results, which are then integrated into a formation control correction set, including: Based on the set of neighboring aircraft action constraint relationships, the parameters of the source aircraft of the inherited action, the type of inherited action, the direction of inheritance, the constraint level and the constraint period of the target aircraft are extracted. The current attitude adjustment, heading correction and speed following direction of the target aircraft are compared. Actions triggered by active cockpit control and driven by environmental disturbances are excluded. The action inheritance level of the target aircraft is determined and the target inheritance state determination result is output. Based on the target inheritance state determination results, a graded suppression strategy for attitude, heading and speed following quantities is formulated according to the action inheritance level and the corresponding constraint level. Suppression constraints are only applied to the action types marked as inherited by neighboring aircraft, and the inherited action suppression control results are generated. Based on the inherited action suppression control results, the corresponding environmental disturbance state sequence is checked back, the retention ratio of environmental disturbance-driven actions is calculated, the anti-disturbance correction actions driven by environmental disturbances are retained, and the repeated correction actions formed by the propagation of neighboring machines are attenuated proportionally to generate action diversion correction results. Based on the action diversion correction results, combined with the current formation space structure and aircraft safety distance requirements, the independent flight correction results that do not depend on the local offset actions of neighboring aircraft are recalculated. The flight stability control requirements of the target aircraft and the formation maintenance requirements are merged to generate a formation control correction set.

8. The remote real-world control method for a flight cockpit based on multi-source perception fusion according to claim 7, characterized in that, Based on the target inheritance state determination result, and according to the action inheritance level and corresponding constraint level, a graded suppression strategy for attitude, heading, and speed following quantities is formulated. Suppression constraints are only applied to action types marked as inherited by neighboring aircraft, generating inherited action suppression control results, including: Based on the target inheritance state determination result, extract the target aircraft's action inheritance level, neighbor aircraft's action constraint level, inherited action type, and inheritance duration, and output the set of inherited action feature parameters. Based on the set of inherited action feature parameters, the inheritance level, constraint level and inheritance duration are comprehensively considered to dynamically calculate the inheritance action retention coefficient, determine the proportion of inherited actions that can be retained by neighboring machines, and output the inheritance action retention coefficient. Based on the inherited action retention coefficient, the attitude adjustment amount marked as roll inheritance or pitch inheritance is proportionally reduced, and the minimum safe attitude adjustment amount required to maintain flight stability is superimposed to obtain the suppressed attitude adjustment amount, and the attitude suppression correction result is output. Based on the inheritance action retention coefficient, the heading correction amount marked as yaw inheritance is proportionally reduced, and the minimum safe heading correction amount required to maintain heading stability is superimposed to obtain the suppressed heading correction amount, and the heading suppression correction result is output. Based on the inherited action retention coefficient, the speed following amount marked as speed following inheritance is proportionally reduced, and the minimum safe speed following amount required to maintain the basic formation spacing is superimposed to obtain the suppressed speed following amount, and the speed suppression correction result is output. Integrate the attitude suppression correction results, heading suppression correction results, and velocity suppression correction results to generate inherited motion suppression control results.

9. The remote real-world control method for a flight cockpit based on multi-source perception fusion according to claim 1, characterized in that, The formation control correction set is used to issue remote control commands to the target aircraft and collect feedback from the actual aircraft. The feedback state is mapped to the simulated flight cockpit, dynamically adjusting the suppression intensity of adjacent aircraft actions, and outputting stable remote control results for multiple aircraft, including: Based on the formation control correction set, the final attitude correction, final heading correction, final speed correction, neighboring aircraft action suppression flag, environmental disturbance-driven action retention flag, and control correction effective period are extracted, converted into remote control correction commands that can be recognized by the target aircraft and issued. Simultaneously, pitch angle, roll angle, yaw angle, ground flight speed, spatial position, and straight formation distance with neighboring cooperative aircraft are collected after the actual aircraft is executed, and the actual aircraft execution feedback state set is output. Based on the actual aircraft execution feedback state set, the actual attitude state, heading state, speed state and formation space state of the target aircraft are mapped to the visual feedback system, attitude feedback system and formation space feedback system of the simulated flight cockpit, and the adjacent aircraft action suppression mark is displayed synchronously, and the cockpit synchronous feedback state set is output. Based on the cockpit synchronous feedback state set, the changes in lateral distance, longitudinal distance, altitude difference, relative speed and relative heading between the target aircraft and the neighboring cooperative aircraft are continuously monitored to determine whether the local offset action has a trend of continuous propagation or gradual convergence, and the spatial offset trend detection results are output. Based on the spatial offset trend detection results, when the local offset action is detected to be continuously expanding, the inherited action retention coefficient is dynamically reduced; when the offset action is detected to gradually converge to the formation reference range, the inherited action retention coefficient is gradually increased, and the updated neighbor action suppression strength result is output. Based on the updated results of the suppression intensity of neighboring aircraft actions, the latest actual aircraft execution feedback data, cockpit synchronization feedback status and suppression intensity parameters are integrated to generate a new set of cooperative flight fusion states and output multi-aircraft stable remote control results.

10. A remote flight cockpit control system based on multi-source sensing fusion, applied to the remote flight cockpit control method based on multi-source sensing fusion as described in any one of claims 1-9, characterized in that, The system includes: The data acquisition module is used to synchronously acquire cockpit control status, multi-aircraft flight status and environmental disturbance status, generate a collaborative flight fusion status set after unified time correlation, record the action propagation timing relationship between different aircraft, and output the action propagation correlation sequence. The separation module is used to analyze the temporal correlation characteristics of multi-machine actions based on the action propagation association sequence, verify the environmental disturbance matching relationship, separate the environmental disturbance driven actions and the neighboring machine inherited actions, and generate a set of neighboring machine action constraint relationships. The correction module is used to verify the action inheritance state of the target aircraft based on the set of neighboring aircraft action constraint relationships, suppress the inherited actions of neighboring aircraft in stages and retain the environmental disturbance-driven actions, and then integrate the generated independent flight correction results into a formation control correction set. The feedback module is used to issue remote control commands to the target aircraft and collect feedback from the actual aircraft based on the formation control correction set. It maps the feedback status to the flight simulator cockpit, dynamically adjusts the suppression intensity of adjacent aircraft actions, and outputs the stable remote control results of multiple aircraft.

Images