Aircraft autonomous recovery method and aircraft facing a dynamic landing platform

Abstract

A kind of aircraft autonomous recovery method for dynamic landing platform, comprising: obtaining the first motion state information of aircraft in navigation coordinate system, and multi-source relative observation information, and estimating the second motion state information of dynamic landing platform;The dynamic pile coordinate system with the origin as the landing reference point and the coordinate axis fixed to the dynamic landing platform is constructed;Multi-source relative observation information is time-delayed alignment and motion compensation and is fused in stages, and then the first motion state information and the second motion state information are combined to determine the third motion state information of aircraft in dynamic pile coordinate system;According to the second motion state information, the predicted touchdown point of landing reference point at the time of touchdown is predicted, and the relative guidance instruction is generated in combination with the third motion state information;Relative guidance instruction is executed;When relative error, the maneuver of dynamic landing platform or observation integrity does not satisfy safety envelope, it is forbidden to continue to reduce relative height, and fly or abort final descent is executed.

Description

Technical Field

[0001] This invention relates to the field of aircraft recovery technology, and in particular to an autonomous aircraft recovery method and aircraft for dynamic landing platforms. Background Technology

[0002] With the widespread application of drones in logistics delivery, marine monitoring, emergency rescue, and other fields, the need for autonomous takeoff and recovery of drones on dynamic landing platforms (such as vehicles and ship decks) is becoming increasingly urgent. Compared to static platform landing, dynamic landing platforms have time-varying six degrees of freedom motion states (including translational velocity, acceleration, and attitude changes such as roll, pitch, and heave), which places higher demands on the relative navigation accuracy and dynamic recovery control strategies of drones.

[0003] In existing technologies, for autonomous landing of UAVs on dynamic landing platforms, there are mainly several technical approaches, including visual guidance landing, shipboard deck motion compensation, and ultra-wideband (UWB) / relative positioning enhancement. Visual guidance schemes mostly rely on airborne cameras to track cooperative targets and adjust the trajectory by estimating relative coordinates and velocity; shipboard landing schemes focus on introducing deck motion models into the guidance loop to improve the landing trajectory through prediction and compensation; relative positioning schemes such as UWB aim to improve the accuracy of close-range relative position calculation by using radio frequency ranging.

[0004] The above-mentioned technical approaches generally suffer from the following drawbacks:

[0005] Lack of predictive guidance for ground contact: Existing solutions mostly focus on tracking the target's current position or relative velocity, without explicitly estimating the platform's linear acceleration and attitude angular velocity, and lack explicit prediction of the landing point for ground contact. When the platform is in non-uniform motion such as acceleration, deceleration, or turning, control lag or overshoot is obvious, and the relative error amplifies rapidly with platform maneuvering.

[0006] Weak multi-source observation fusion and motion compensation capabilities: RTK relative positioning, UWB ranging, and visual pose information are often used independently without motion compensation fusion in a unified dynamic coordinate system. When switching sensors or when the platform attitude changes drastically, relative trajectory jumps are easily introduced, affecting the continuity and stability of control.

[0007] Incomplete modeling of platform motion constraints: Existing solutions do not systematically incorporate the heave / roll / pitch of shipborne platforms or the path curvature of vehicle-mounted platforms into relative height and attitude alignment constraints. The logic for determining the ground contact window is too simplistic and cannot meet the precise alignment requirements under complex sea or road conditions.

[0008] Lack of safety decision-making and failure response strategies: Existing solutions often lack a phased state machine design for dynamic recovery and fail to establish a reproducible safety envelope and a forced go-around interlock mechanism. In dangerous situations such as loss of observation, exceeding relative error limits, or sudden platform maneuvers, UAVs often lack effective go-around, realignment, or abort landing strategies, leaving them at risk of continuing to descend, resulting in insufficient safety redundancy.

[0009] In conclusion, the above-mentioned technical approaches are insufficient to achieve high-precision and robust autonomous recovery of UAVs on dynamic landing platforms. Summary of the Invention

[0010] This invention provides an autonomous recovery method and aircraft for dynamic landing platforms, enabling high-precision and robust autonomous recovery of aircraft such as UAVs on dynamic landing platforms.

[0011] Firstly, a method for autonomous recovery of aircraft for dynamic landing platforms is provided, including:

[0012] The system acquires first motion state information of the aircraft in the navigation coordinate system and multi-source relative observation information of the dynamic landing platform relative to the aircraft, and estimates second motion state information of the dynamic landing platform based on the first motion state information and the multi-source relative observation information. The second motion state information includes at least the position and velocity of the dynamic landing platform.

[0013] A dynamic pile coordinate system is constructed with the origin as the landing reference point and the coordinate axes fixed to the dynamic landing platform, wherein the landing reference point is the predetermined contact area reference point of the aircraft on the dynamic landing platform;

[0014] With the support of the second motion state information, the multi-source relative observation information is time-delay aligned and motion-compensated and fused to obtain fused relative observation information; based on the fused relative observation information, the first motion state information and the second motion state information, the third motion state information of the aircraft in the dynamic pile coordinate system is determined, and the third motion state information includes at least relative position and relative velocity.

[0015] The predicted touchdown point of the landing reference point is predicted based on the second motion state information, and a relative guidance command is generated based on the predicted touchdown point and the third motion state information. The relative guidance command enables autonomous recovery to proceed in stages of pursuit approach, relative alignment, synchronous accompaniment, and predicted final landing.

[0016] Execute the relative guidance command.

[0017] In one possible implementation, the safety envelope is checked during the recovery process; if the relative error between the aircraft and the dynamic landing platform, or the maneuverability or observation integrity of the dynamic landing platform, does not meet the safety envelope, further descent of relative altitude is prohibited, and a go-around or abort of the final landing is performed.

[0018] In one possible implementation, the second motion state information may also include any one or more of the acceleration, angular velocity, attitude, and uncertainty of the dynamic landing platform.

[0019] In one possible implementation, the step of predicting the landing reference point based on the second motion state information to predict the landing time includes:

[0020] The estimated time to ground contact is determined based on relative altitude and expected descent rate;

[0021] Based on the position, velocity, and selectable acceleration of the dynamic landing platform, the landing reference point position corresponding to the expected touchdown time is extrapolated to obtain the predicted touchdown point; when acceleration is unavailable or the uncertainty is too large, it degenerates into uniform extrapolation based on position and velocity.

[0022] In one possible implementation, the criteria for the safety envelope include: the relative error between the aircraft and the dynamic landing platform, the maneuverability of the dynamic landing platform, and the observation integrity of the dynamic landing platform; wherein, when any criterion is violated and the duration exceeds a set threshold, the prohibition of further descent of relative altitude is executed, and a go-around or abort of the final landing is performed, and the aircraft returns to the synchronous accompaniment altitude and realigns for re-entry.

[0023] In one possible implementation, the relative error includes: horizontal position error and / or horizontal velocity error.

[0024] In one possible implementation, the maneuver of the dynamic landing platform includes: the absolute value of the acceleration and / or the absolute value of the angular velocity of the dynamic landing platform.

[0025] In one possible implementation, the observation integrity of the dynamic landing platform includes: visual lockout rate and / or ultra-wideband outage rate.

[0026] In one possible implementation, the multi-source relative observation information includes any one or more of real-time dynamic differential RTK relative observations, UWB ranging, visual observations, and altimeter observations.

[0027] In a second aspect, an aircraft is provided, including a memory and a processor, wherein the memory stores executable code, and the processor executes the executable code to implement the method described in the first aspect.

[0028] The solution provided by this invention focuses on relative state and platform motion estimation, adapting to the real dynamic recovery requirements of dynamically landing platforms. Furthermore, predictive touchdown guidance reduces lag and overshoot on the maneuvering platform, ensuring the landing point tracks the target and the touchdown time are consistent. This solution reduces erroneous handovers and trajectory jumps in dynamic scenarios through multi-source relative fusion including motion compensation and time delay alignment. Additionally, it enhances engineerable safety and reproducible failure strategies through a staged state machine and a safety envelope-based forced go-around interlock. Attached Figure Description

[0029] Figure 1 A dynamic platform recycling flowchart provided by the present invention; Figure 2 Another dynamic platform recycling flowchart provided by the present invention; Figure 3 A flowchart illustrating an autonomous recovery method for an aircraft with a dynamic landing platform, provided by the present invention. Figure 4 A comparison chart of landing success rate, CEP (Circular Error Probable) at bottom, and number of go-arounds; Figure 5 This is a diagram illustrating the relative horizontal error and relative height of a baseline and the present invention. Figure 6 A diagram illustrating the relative horizontal error and relative height compared to another baseline and the present invention; Figure 7A This is a diagram illustrating the horizontal trajectory and relative height of the present invention at a constant speed in a vehicle. Figure 7B This is a diagram illustrating the relative horizontal error and landing error of the present invention when traveling at a constant speed on a vehicle. Figure 8A This is a diagram illustrating the horizontal trajectory and relative height of the present invention during emergency braking in a vehicle. Figure 8B This diagram illustrates the relative horizontal error and landing error of the present invention during emergency braking of a vehicle. Figure 9A The diagram illustrates the horizontal trajectory and relative altitude of the invention under moderate sea conditions on a ship. Figure 9B The diagram illustrates the relative horizontal error and landing error of the present invention under moderate sea conditions on a ship. Figure 10A The diagram illustrates the horizontal trajectory and relative height of the invention during vehicle acceleration. Figure 10B The diagram illustrates the relative horizontal error and landing error of the present invention during vehicle acceleration. Detailed Implementation

[0030] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this specification, and not all embodiments. Based on the embodiments in this specification, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this specification.

[0031] To facilitate understanding of the technical solution of this invention, the technical terms that may be involved in this invention will be introduced first.

[0032] Dynamic stakes and the dynamic stake coordinate system P: A dynamic stake refers to a predetermined contact area reference point (stake point, such as the center of the landing pad) on the dynamic landing platform. The dynamic stake coordinate system P is a relative reference coordinate system for landing, with the landing reference point as its origin and its coordinate axes fixed to the dynamic platform. P moves accordingly when the platform translates, turns, or changes its attitude. Relative navigation, guidance errors, and safety criteria are all uniformly expressed under this system.

[0033] relative state The relative state is the deviation of the UAV from the landing reference point (dynamic stake) under condition P, which is determined by the transformation. Obtain. At least includes relative position. With relative velocity When attitude observation and estimation are available, it is preferable to include relative attitude / heading deviation. This describes "how much the aircraft has deviated relative to the platform and how closely it is following," rather than the absolute state of the aircraft and platform within their respective geoids.

[0034] Predicted touchdown point The predicted contact point is determined at the current estimated time t, based on the platform's motion state (at least position and velocity, preferably acceleration). The extrapolated estimate of the platform's landing reference point's position in the navigation coordinate system. The remaining contact time is estimated. Second-order extrapolation is commonly used, as shown in equation (2); when acceleration is unavailable or uncertain, it degenerates into uniform extrapolation. Relative guidance uses this predicted position (which is then transformed to P) as the tracking target, rather than the currently observed snapshot position of the pad point.

[0035] Safety envelope (dynamic recovery safety envelope): The safety envelope is a set of time-varying, verifiable constraints set during autonomous approach and final descent to ensure safe ground contact or safe abort. The criteria are jointly constituted by the relative state under condition P, the platform's motion state, and the relative observation integrity, etc. (e.g., and ground-touching windows, etc.). When any constraint is broken and the duration exceeds At this time, further reduction in relative altitude is prohibited, and a forced go-around / relative climb will be triggered. Its "dynamic" nature is reflected in the fact that the thresholds and observations change with the recovery phase, relative distance, and platform maneuvering, rather than the static alignment conditions on a fixed apron.

[0036] In scenarios such as logistics rendezvous, ship support, and aircraft hangar recovery, the landing surface of an aircraft is often in motion. Although engineering can provide relative observations such as Global Navigation Satellite System (GNSS), Real-Time Kinematic (RTK), UWB, and downward-looking vision, if the "current pad coordinates" are still tracked according to the static landing logic, the platform's translation, acceleration, deceleration, turning, or heave / attitude motion caused by waves will directly become relative alignment errors, which will be amplified during the final descent, resulting in poor recovery safety.

[0037] To address this situation, the present invention provides an autonomous aircraft recovery solution for dynamic landing platforms, which can improve the safety of aircraft recovery on dynamic landing platforms. For example... Figure 1 , Figure 2 As shown, the dynamic platform recovery in this scheme can be described as a closed loop of steps S1–S5. Here, S1 is observation and platform state estimation, S2 is dynamic stake system relative navigation and multi-source fusion, S3 is predictive ground contact guidance, S4 is the state machine for the dynamic recovery phase, and S5 is the safety envelope and forced go-around.

[0038] The concept of this invention is not to make a single sensor more accurate, but to make the above decision-making process calculable, threshold-comparable, and hard-locked with flight control.

[0039] Step S1 can be executed by the observation and platform state estimation module. Specifically, in step S1, the aircraft state provided by the airframe navigation is input. (Without replacing the original body navigation filter), acquire RTK relative, UWB, visual, altimeter and its quality words / integrity markers; and estimate the motion state of the dynamic landing platform. (Position, velocity, acceleration, attitude, angular velocity, and covariance P_p). Output for use by S2–S5.

[0040] Step S2 can be executed by the dynamic pile system relative navigation and multi-source fusion module. Specifically, in step S2, the dynamic pile coordinate system P is maintained. With support, multi-source relative observations are transformed to P, time delay alignment and platform motion compensation are completed, and weighted or gated fusion is performed in stages according to distance / altitude; then combined with Calculate the relative state according to formula (1) (At least (Relative attitude, etc.). The output is used in steps S3 and S5.

[0041] Step S3 can be executed by the predictive ground contact guidance module. Specifically, in step S2, the module... Extrapolation yields the predicted contact point (Equations (2) and (3)) (This can be achieved using altimeters / relative height measurements), and combined with... Generate relative guidance commands. Track the platform reference point at the expected touchdown moment, not the current ground contact point. Determine whether final landing is allowed to be interlocked with steps S4 and S5.

[0042] Step S4 can be executed by the dynamic recovery phase state machine module. In step S4, the approach is followed by pursuit, relative alignment, synchronous accompaniment, predicted final landing, and touchdown completion / go-around re-entry. The entry conditions for each phase are determined by the relative quantity and threshold under P, and the relative guidance command given in step S3 is executed.

[0043] Step S5 can be executed by the safety envelope and forced go-around module. In step S5, a time-varying constraint is formed by integrating relative error, platform maneuvering, and observation integrity; if any constraint is violated and persists for more than [a certain period]... Prohibit further descent and force a go-around / relative climb, return to control step S4 to synchronize with the altitude and then realign.

[0044] Based on the foregoing content, this invention provides an autonomous recovery method for aircraft using a dynamic landing platform. This method can be executed by aircraft such as drones. The method may include... Figure 3 The steps shown.

[0045] Step 301: Obtain the first motion state information of the aircraft in the navigation coordinate system and the multi-source relative observation information of the dynamic landing platform relative to the aircraft, and estimate the second motion state information of the dynamic landing platform based on the first motion state information and the multi-source relative observation information. The second motion state information includes at least the position and velocity of the dynamic landing platform.

[0046] The dynamic landing platform can be simply referred to as a platform or a mobile platform.

[0047] In some embodiments, the multi-source relative observation information includes any one or more of real-time dynamic differential RTK relative observations, UWB ranging, visual observations, and altimeter observations.

[0048] In some embodiments, the second motion state information may further include any one or more of the acceleration, angular velocity, attitude, and uncertainty of the dynamic landing platform.

[0049] Step 301 pertains to observation and platform state estimation. During each control cycle, the state of the human-machine interface (HMI) under the local navigation system (NED) can be acquired. Multi-source observation results (RTK, UWB, visual, altimeter and quality words) for dynamic landing platforms.

[0050] Among them, extended Kalman filtering / unscented filtering and other methods are used to estimate the platform state. Observation sources can be combined according to priority (if a measurement is missing, it will be downgraded without interrupting the entire link). Observation sources are shown in Table 1.

[0051] Table 1. Sources of observation

[0052]

[0053] Step 301 can output: Platform location ,speed (Preferred acceleration) Attitude and angular velocity and covariance For use in steps 302-306. If If the value is too large, the safety envelope should be tightened or entry into the predicted final descent should be prohibited.

[0054] Step 302: Construct a dynamic pile coordinate system with the origin as the landing reference point and the coordinate axes fixed to the dynamic landing platform, wherein the landing reference point is the predetermined contact area reference point of the aircraft on the dynamic landing platform.

[0055] In step 302, dynamic pile system relative navigation and multi-source fusion are performed.

[0056] Step 302 includes defining a dynamic pile coordinate system P. Following the technical terminology described above, a reference point at the predetermined contact area (dynamic pile / pile point, such as the center of the landing pad) is used as the origin, and the coordinate axes are fixed to the platform (e.g., forward along the vehicle's nose / ship's length, starboard along the starboard side, vertically upwards), forming a relative reference coordinate system for landing. When the platform moves or turns... It then translates and rotates.

[0057] Step 302 includes relative state calculation (transformation T in equation (1)). The relative state is... The deviation of the aircraft's position, speed (and selectable attitude) relative to the landing reference point. Let the local navigation system's aircraft and platform position / speed be denoted as... , , , The rotation matrix corresponding to the platform's attitude is (Transfer the platform system vector to the navigation system). Then the relative position and relative velocity are... The following can be taken:

[0058]

[0059]

[0060] in This is the velocity term caused by the platform's rotation; if the angular velocity is not yet observable, this cross-product term can be omitted, but the prediction uncertainty will be increased. And tighten the envelope. The relative state vector is denoted as... At least includes , Preferably, relative heading / attitude is included.

[0061] Step 303: With the support of the second motion state information, the multi-source relative observation information is time-delay aligned and motion-compensated and fused to obtain fused relative observation information; based on the fused relative observation information, the first motion state information and the second motion state information, the third motion state information of the aircraft in the dynamic pile coordinate system is determined, and the third motion state information includes at least relative position and relative velocity.

[0062] In step 303, multi-source observations are unified to {P} and motion-compensated fusion is performed. RTK, UWB, visual, and altimeter observations relative to the target can be transformed to {P} respectively. The relative position (and optional velocity) is observed. Unlike static landing, the following operations must be performed before fusion:

[0063] Time delay alignment: Aligning the timestamps of each sensor to the same estimated time;

[0064] Motion compensation: Extrapolate or backtrack the observation to that moment using the platform's velocity / acceleration to avoid treating the platform displacement as a relative error;

[0065] Phased weighting / gating (e.g., calibrated by platform size). As shown in Table 2, platform size can be relative ramp distance or height.

[0066] Table 2, Stage Division

[0067]

[0068] When there is a quality difference or a loss of lock, the weight of the source is reduced or the source is gated and rejected; the relative position / velocity estimate and its covariance are fused and output for use in guidance and envelope.

[0069] Step 304: Based on the second motion state information, predict the predicted touchdown point of the landing reference point at the touchdown time, and generate a relative guidance command based on the predicted touchdown point and the third motion state information. The relative guidance command enables autonomous recovery to proceed in stages of pursuit approach, relative alignment, synchronous accompaniment, and predicted final landing.

[0070] In some embodiments, the step of predicting the landing reference point based on the second motion state information to predict the landing time includes: determining the expected landing time based on the relative height and the expected descent rate; extrapolating the landing reference point position corresponding to the expected landing time based on the position, velocity, and optional acceleration of the dynamic landing platform to obtain the predicted landing point; when acceleration is unavailable or the uncertainty is too large, it degenerates into uniform extrapolation based on position and velocity.

[0071] In step 304, predictive ground contact guidance is performed. Following the above explanation of technical terminology, the predicted point of contact... This is an extrapolation estimate of the platform's landing reference point position at the expected touchdown time. Expected remaining touchdown time (illustrated, Equation (3)):

[0072]

[0073] in Relative height (take) Vertical component or height measurement). The expected rate of decline can also be obtained through trajectory optimization. .like If the observation is invalid, entry into the predicted final descent is prohibited.

[0074] Platform-side predicted touch point (Equation (2)):

[0075]

[0076] When acceleration is unobservable or When the value is too large, it degenerates into uniform extrapolation. and improve Tighten the wrapping thread.

[0077] The relative target tracked by the aircraft is Generated after transformation to P; horizontal channel elimination The vertical descent rate is controlled according to the expected rate of descent in stages, based on the difference from the predicted target. In shipboard scenarios, phase prediction of vertical waves can be performed, or touchdown can be allowed only within a low vertical velocity window; optional attitude prediction can be superimposed to ensure that the relative attitude enters the alignment window before final descent. Compared with tracking only the current pad point observation, predictive touchdown guidance ensures that the tracked target is consistent with the touchdown time, which can reduce hysteresis and overshoot caused by platform maneuvering.

[0078] Step 305: Execute the relative guidance command. In step 305, the dynamic recovery phase state machine may be executed. The various phases are shown in Table 3.

[0079] Table 3

[0080]

[0081] During the phase, transfers are only permitted in accordance with regulations; when the envelope triggers a go-around, the aircraft must retreat from the predicted final landing to synchronized accompaniment (or a higher safe altitude), and must not continue to descend relative altitude during the period of envelope violation.

[0082] During the recovery process, the safety envelope is checked; if the relative error between the aircraft and the dynamic landing platform, or the maneuverability or observation integrity of the dynamic landing platform, does not meet the safety envelope, further reduction in relative altitude is prohibited, and a go-around or abort of the final landing is performed.

[0083] In some embodiments, the criteria for the safety envelope include: the relative error between the aircraft and the dynamic landing platform, the maneuvering of the dynamic landing platform, and the observation integrity of the dynamic landing platform; wherein, when any criterion is violated and the duration exceeds a set threshold, the prohibition of further descent of relative altitude is executed, and a go-around or abort of the final landing is performed, and the aircraft returns to the synchronous accompaniment altitude and realigns for re-entry.

[0084] In one example of this embodiment, the relative error includes: horizontal position error and / or horizontal velocity error. In one example of this embodiment, the maneuvering of the dynamic landing platform includes: the absolute value of the dynamic landing platform's acceleration and / or the absolute value of its angular velocity. In one example of this embodiment, the observation integrity of the dynamic landing platform includes: visual lock-on rate and / or ultra-wideband outage rate.

[0085] Each cycle is determined according to the following steps:

[0086] Calculate the relative horizontal position error Horizontal speed error (Depend on (Achieved relative to the expected target)

[0087] Check whether the absolute value of the horizontal position error is not greater than And the absolute value of the horizontal velocity error is no greater than ;

[0088] Check platform maneuverability: absolute value of acceleration absolute value of angular velocity (Sudden braking, sharp turns, large swaying, etc.);

[0089] Check observation integrity: visual lock loss, UWB interruption and If it is too large, it will be considered as damage;

[0090] Check the ground contact window (shipborne): whether the wave height / vertical speed / attitude are within the allowable window; optional to coordinate with integrity monitoring LAL / HPL thresholds;

[0091] If any one of them is destroyed, a destruction timer will be triggered; if it continues for more than [time period missing], the destruction timer will be triggered. If it still does not recover, prohibit further descent, issue a relative climb / go-around command to the flight controller, and re-align after returning to the synchronized altitude; if it recovers before the timer is full, the timer can be cleared (anti-shake).

[0092] This enables high-precision and robust autonomous recovery of drones and other aircraft on dynamic landing platforms.

[0093] The symbols related to the state and coordinate system involved in this invention are shown in Table 4.

[0094] Table 4

[0095]

[0096] The safety envelope symbols involved in this invention are shown in Table 5.

[0097] Table 5

[0098]

[0099] Next, the formula involved in this invention will be introduced.

[0100] Equation (1) defines the relative state:

[0101]

[0102] Equation (2) predicts the location of ground contact:

[0103]

[0104] Equation (3) degenerates into uniform velocity prediction when the platform acceleration is unobservable or the uncertainty is too large. and correspondingly improve Tighten the safety harness.

[0105] Estimated remaining time to ground:

[0106]

[0107] in Relative height, The expected rate of decline can also be obtained through trajectory optimization. .like If the observation is invalid, then entering the prediction of the final descent is prohibited.

[0108] Equation (4) Safety envelope determination (horizontal channel):

[0109]

[0110] Furthermore, the platform's maneuverability satisfies the absolute value of acceleration. absolute value of angular velocity The above conditions apply if the duration of the damage exceeds [a certain value]. If the conditions are still not met, a forced go-around will be triggered.

[0111] The technical parameters involved in this invention are shown in Table 6.

[0112] Table 6

[0113]

[0114] This invention also provides a method for relative navigation and autonomous recovery of unmanned aerial vehicles (UAVs) for dynamic platforms, comprising:

[0115] Observation and Platform State Estimation: Obtaining UAV State Information In addition, multi-source relative observation information of the dynamic landing platform is used to estimate the platform's motion state. This includes at least the platform location and speed;

[0116] Dynamic pile system relative navigation and multi-source fusion: Establishing a dynamic pile coordinate system P, in With support, time delay alignment and motion compensation are performed on multi-source relative observations, and the data is fused in stages, then combined with... Calculate relative state (At least relative position and relative velocity);

[0117] Predictive ground-contact guidance: based on Extrapolate to predict the touchdown point, and combine with Generate relative guidance commands;

[0118] Dynamic recovery phase state machine: autonomous recovery is advanced in stages of pursuit and approach, relative alignment, synchronous accompaniment, and predicted final descent;

[0119] Safety envelope and forced go-around: When the relative error, platform maneuvering or observation integrity does not meet the safety envelope, descent is prohibited and a go-around or termination of landing must be performed.

[0120] In some embodiments, the platform state further includes acceleration, attitude, angular velocity, and uncertainty P_p.

[0121] In some embodiments, RTK relative, UWB, and visual motion compensation fusion under P (including time delay alignment and phased weighting or gating by distance / height); specific expansion of relative state transformation T.

[0122] In some embodiments, the time to ground is predicted. Calculation and landing point prediction formulas; shipborne vertical / attitude ground contact window; vehicle-mounted non-integrity constraints or path curvature feedforward.

[0123] In some embodiments, the entry conditions and safety envelope criteria for each stage, the forced go-around interlock, and the interface with the flight control are specified.

[0124] In some embodiments, the method is coordinated with a multi-source integrity monitoring LAL / HPL threshold.

[0125] In some embodiments, the relative filter is specifically selected from EKF, UKF, factor graph, or multiple hypothesis tracking.

[0126] The solution of this invention focuses on relative state and platform motion estimation, adapting to the real dynamic recovery needs of vehicle-mounted / ship-mounted systems, rather than simplifying the dynamic platform into a slowly moving static point.

[0127] The solution of this invention predicts that ground contact guidance reduces hysteresis and overshoot on the maneuvering platform, and the landing point tracks the target and the moment of ground contact is consistent.

[0128] The multi-source relative fusion in the solution of this invention includes motion compensation and time delay alignment, which reduces erroneous switching and trajectory jumps in dynamic scenes.

[0129] The present invention includes a phased state machine and a safety envelope forced re-flight interlock, which improves engineerable safety and reproducible failure strategies.

[0130] The solution of this invention can be combined with methods such as multi-source integrity monitoring and continuous approach to GNSS degradation corridors to form a continuous and provably secure dynamic recovery link.

[0131] Next, the solution of this invention will be introduced in conjunction with specific application scenarios.

[0132] Example 1: Vehicle-mounted mobile platform recycling

[0133] The vehicle is equipped with a landing pad and visual beacon on its roof, and can optionally broadcast the vehicle's RTK speed. A typical recovery can be understood as follows timeline:

[0134] Pursuit and Approach: The drone, based on RTK, pursues the vehicle to the safe corridor above and behind it.

[0135] Relative alignment: UWB / visual establishment of stable relative observation, in Eliminate horizontal and heading errors;

[0136] Synchronous following: Matches vehicle horizontal speed, relative hovering / slow following; continuous monitoring. , Relative threshold;

[0137] Predicting the final descent: After the envelope is satisfied, predict the final descent point according to formulas (2) and (3), instead of focusing on the center of the current pad pixel;

[0138] Interruption of maneuver: If the vehicle accelerates or turns, causing the relative error or platform acceleration to exceed the limit and continue to do so. Immediately return to the synchronized altitude, realign, and then land;

[0139] Touchdown: Successful landing detection completes recovery.

[0140] Example 2: Shipboard Deck Recovery

[0141] Based on the horizontal link in Example 1, deck attitude and vertical motion estimation are added:

[0142] The vehicle can be used for pursuit, approach / relative alignment / synchronous accompaniment, while maintaining a greater safety margin in relative height.

[0143] Before entering the predicted final descent phase, additional checks are performed to ensure that the vertical velocity and attitude angle fall within the touchdown window.

[0144] If the window is not met, maintain synchronized altitude and wait for the next window, or go around and re-enter;

[0145] When the window and horizontal envelope are satisfied, the landing will be completed by the predicted landing point.

[0146] Example 3: Integration with multi-source sensors

[0147] For long distances, relative position is calculated using RTK differential positioning between the aircraft and the platform; for medium distances, UWB is used; and for short distances, vision is employed. The entire process is dynamic. The system is integrated, and observation compensation and time delay alignment are performed on platform motion. Compared with the baseline of "only following the current beacon position without platform velocity estimation", the landing point error and miss rate under dynamic maneuvering are significantly improved.

[0148] Next, we will introduce the simulation verification and results.

[0149] To reduce actual flight costs and support comparisons of implementation examples, a schematic-level digital simulation was conducted on the dynamic recovery closed loop: platform motion (static apron / vehicle-mounted constant speed, acceleration, emergency braking, turning / shipborne sea conditions), multi-source observation (RTK / UWB / visual, with the ability to inject lock loss), and baseline (only matching the current pad point, weak velocity matching) were compared with the scheme of this invention (platform velocity feedforward + acceleration prediction landing point + stage state machine + safety envelope). The dynamics are a first-order velocity tracking model, used for qualitative / semi-quantitative comparison, and do not replace field evaluation.

[0150] The Monte Carlo summary (20 times per condition) is shown in Table 7.

[0151] Table 7

[0152]

[0153] Note: Under onboard constant speed / acceleration injection, the baseline landing point error increases significantly or fails to meet the success threshold; the success rate and CEP of the proposed solution are significantly improved. In the emergency braking condition (D04), ground contact can be completed within the envelope by predictive feedforward (0 misses); in the acceleration condition, a safety envelope miss is triggered, followed by a second approach and ground contact (average miss 1.0).

[0154] Figure 4 It shows the landing success rate, CEP (Circular Error Probable) at bottom, and number of go-arounds.

[0155] Figure 5 The relative horizontal error and relative height are shown in the comparison between a baseline and the present invention.

[0156] Figure 6 The relative horizontal error and relative height are shown in the comparison between another baseline and the present invention.

[0157] Figure 7A The horizontal trajectory and relative height of the present invention at a constant speed on a vehicle are shown. Figure 7B The relative horizontal error and landing error of the present invention at constant speed on a vehicle are shown.

[0158] Figure 8A The horizontal trajectory and relative height of the present invention during emergency braking of a vehicle are shown. Figure 8B The relative horizontal error and landing error of the present invention during vehicle emergency braking are shown.

[0159] Figure 9A The horizontal trajectory and relative altitude of the present invention in shipboard medium sea state are shown. Figure 9B The relative level error and landing error of the present invention in shipboard medium sea state are shown.

[0160] Figure 10A The horizontal trajectory and relative height of the present invention during vehicle acceleration are shown. Figure 10B The relative horizontal error and landing error of the present invention during vehicle acceleration are shown.

[0161] It should be understood that the descriptions such as "first" and "second" in this article are merely for the sake of simplicity in description and to distinguish similar concepts, and do not have any other limiting function.

[0162] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to interchangeably. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments.

[0163] The foregoing has described specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in a different order than that shown in the embodiments and may still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require the specific or sequential order shown to achieve the desired result. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0164] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of each example have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this invention. The software modules can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disks, removable disks, CD-ROMs, or any other form of storage medium known in the art.

[0165] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for autonomous recovery of an aircraft for a dynamic landing platform, characterized in that, include: The system acquires first motion state information of the aircraft in the navigation coordinate system and multi-source relative observation information of the dynamic landing platform relative to the aircraft, and estimates second motion state information of the dynamic landing platform based on the first motion state information and the multi-source relative observation information. The second motion state information includes at least the position and velocity of the dynamic landing platform. A dynamic pile coordinate system is constructed with the origin as the landing reference point and the coordinate axes fixed to the dynamic landing platform, wherein the landing reference point is the predetermined contact area reference point of the aircraft on the dynamic landing platform; With the support of the second motion state information, the multi-source relative observation information is time-delay aligned and motion-compensated and then fused to obtain fused relative observation information; Based on the fusion of relative observation information, the first motion state information and the second motion state information, the third motion state information of the aircraft in the dynamic pile coordinate system is determined. The third motion state information includes at least relative position and relative velocity. The predicted touchdown point of the landing reference point is predicted based on the second motion state information, and a relative guidance command is generated based on the predicted touchdown point and the third motion state information. The relative guidance command enables autonomous recovery to proceed in stages of pursuit approach, relative alignment, synchronous accompaniment, and predicted final landing. Execute the relative guidance command.

2. The method according to claim 1, characterized in that, The method further includes: During the recovery process, the safety envelope is checked; if the relative error between the aircraft and the dynamic landing platform, or the maneuverability or observation integrity of the dynamic landing platform, does not meet the safety envelope, further reduction in relative altitude is prohibited, and a go-around or abort of the final landing is performed.

3. The method according to claim 1, characterized in that, The second motion state information also includes any one or more of the acceleration, angular velocity, attitude, and uncertainty of the dynamic landing platform.

4. The method according to claim 1, characterized in that, The predicted touchdown point of the landing reference point based on the second motion state information includes: The estimated time to ground contact is determined based on relative altitude and expected descent rate; Based on the position, velocity, and selectable acceleration of the dynamic landing platform, the landing reference point position corresponding to the expected touchdown time is extrapolated to obtain the predicted touchdown point; when acceleration is unavailable or the uncertainty is too large, it degenerates into uniform extrapolation based on position and velocity.

5. The method according to claim 2, wherein the criterion for the safety envelope includes: The relative error between the aircraft and the dynamic landing platform, the maneuverability of the dynamic landing platform, and the observation integrity of the dynamic landing platform; When any criterion is violated and the duration exceeds a set threshold, the prohibition of further descent of relative altitude is executed, and a go-around or abort of final landing is performed, and the aircraft returns to the synchronous accompaniment altitude and realigns for re-entry.

6. The method according to claim 5, characterized in that, The relative error includes: horizontal position error and / or horizontal velocity error.

7. The method according to claim 5, characterized in that, The maneuvering of the dynamic landing platform includes: the absolute value of the acceleration and / or the absolute value of the angular velocity of the dynamic landing platform.

8. The method according to claim 5, characterized in that, The observation integrity of the dynamic landing platform includes: visual lockout rate and / or ultra-wideband outage rate.

9. The method according to claim 1, characterized in that, The multi-source relative observation information includes any one or more of the following: real-time dynamic differential RTK relative observation, UWB ranging, visual observation, and altimeter observation.

10. An aircraft, characterized in that, The method includes a memory and a processor, wherein the memory stores executable code, and the processor executes the executable code to implement the method of any one of claims 1-9.

Images