Coaxial contra-rotating double-ring duct vector propulsion system and fault-tolerant control method thereof

Abstract

The invention relates to the technical field of aircraft power systems, and discloses a coaxial contra-rotating double-ring duct vector propulsion system and a fault-tolerant control method thereof, and the coaxial contra-rotating double-ring duct vector propulsion system comprises a fuselage platform, an outer ring main lift propulsion group, an inner ring attitude control propulsion group, inter-ring Coanda diversion bridges and a distributed flight control system. The outer-ring main lift propelling groups are arranged in the circumferential direction of the fuselage, and the inner-ring attitude control propelling groups are arranged in the inner-side gap in a staggered manner; and a telescopic three-degree-of-freedom thrust vectoring nozzle mechanism is integrated at the tail part of the inner ring duct unit. When the outer ring unit breaks down, the control system drives the spray pipes of the adjacent inner ring unit to stretch out and deflect, the nozzles of the spray pipes are attached to the surface of the flow guide bridge, and an ejection flow field pointing to the lower portion of the fault unit is established through the wall attachment effect. Through an inner and outer double-ring nested architecture and a flow field reconstruction technology, active fault tolerance is realized on the premise that redundant weight is not increased, compensation thrust is amplified by utilizing fluid entrainment, and the problem that thrust stages are not matched after a main lift unit fails is effectively solved.

Description

Technical Field

[0001] This invention relates to the field of aircraft propulsion system technology, specifically to a coaxial counter-rotating dual-ring ducted vector propulsion system and its fault-tolerant control method. Background Technology

[0002] Distributed electric propulsion technology has been widely used in the design of vertical takeoff and landing (VTOL) aircraft due to its flexible layout and high redundancy. However, when faced with the failure of the main lift unit, traditional fault-tolerance strategies often rely on simple hardware redundancy, i.e., increasing the number of backup power units or increasing the power design margin of a single power unit to address potential failure risks. This passive redundancy method increases the weight of the airframe structure and introduces weightlessness loads in non-operating states, thereby reducing the overall thrust-to-weight ratio and payload efficiency of the aircraft.

[0003] Furthermore, existing aircraft typically arrange the main lift and attitude control systems independently. When the large main lift unit fails, the small attitude unit, mainly used for fine-tuning, struggles to provide sufficient direct lift compensation. Limited by space constraints, it also struggles to effectively balance the asymmetric torques caused by the failure, easily leading to attitude instability and crashes. More critically, current technology lacks a mechanism to establish fluid coupling between heterogeneous propulsion units, making it impossible to guide the jets of adjacent healthy units for targeted thrust compensation and flow field reconstruction at the moment of failure. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a coaxial counter-rotating dual-ring ducted vector propulsion system and its fault-tolerant control method, which solves the problems of low payload efficiency due to reliance on hardware redundancy and difficulty in efficient thrust compensation between heterogeneous propulsion units in existing distributed propulsion aircraft after the failure of the main lift unit.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solution: a coaxial counter-rotating double-ring duct vector propulsion system and its fault-tolerant control method.

[0006] The first aspect of this invention provides a coaxial counter-rotating dual-ring ducted vector propulsion system, comprising a fuselage platform with a ring truss structure, an outer ring main lift propulsion group, an inner ring attitude control propulsion group, an inter-ring Coanda guide bridge, and a distributed flight control system. The outer and inner ring ducted units are staggered in the circumferential direction, with the outer ring low-speed rotor and the inner ring high-speed rotor rotating in opposite directions, respectively providing low-frequency high thrust and high-frequency attitude torque. A three-degree-of-freedom vector nozzle mechanism is integrated at the tail of the inner ring ducted unit. This mechanism consists of a dual-axis orthogonal bidirectional joint assembly, a coaxial telescopic sleeve assembly driven by a linear drive unit, and a flexible sealing compensation component, enabling yaw, pitch, and stepless linear telescopic movements relative to the axis. The inter-ring Coanda guide bridge is fixedly connected between the outer and inner rings, and its working surface is a convex curved surface with a radius of curvature varying along the airflow direction. Under normal operating conditions, the coaxial telescopic sleeve assembly is in the retracted position, and a physical safety gap is maintained between the nozzle and the guide bridge, allowing the airflow to be directly discharged into the atmosphere. Only when the sleeve extends to its maximum stroke and the yaw angle meets the tangential condition, the nozzle end face adheres to the boundary layer intake zone of the guide bridge, and the ejector flow field from the inner ring to the bottom of the outer ring is established by utilizing the wall attachment effect.

[0007] The second aspect of this invention provides a fault-tolerant control method for the aforementioned coaxial counter-rotating dual-ring ducted vector propulsion system, comprising: a state monitoring module acquiring outer ring operating data; a fault diagnosis module calculating the rotational speed observation residual using a power system state observer, and determining and locking the outer ring unit experiencing thrust loss fault when the residual times out or current backflow is detected to zero; a topology reconstruction module using modulo operation to search for adjacent inner ring ducted units on the counterclockwise and clockwise sides of the fault unit as rescue units; a distributed flight control system driving the three-degree-of-freedom vector nozzle mechanisms of the two rescue units to extend to their maximum stroke and deflect the target coupling angle, so that the jet converges tangentially on the surface of the inter-ring Coanda guide bridge below the fault unit; and a control allocation solution module reorganizing the control allocation matrix, isolating the weight of the fault unit, and calculating the compensation control command of the rescue unit based on the ejector gain model, thereby achieving thrust compensation and attitude reconstruction by establishing a fluid coupling path.

[0008] This invention provides a coaxial counter-rotating dual-ring ducted vector propulsion system and its fault-tolerant control method. It has the following beneficial effects:

[0009] 1. This invention utilizes a nested inner and outer double-ring layout and an inter-ring Coanda guide bridge design to establish a cross-ring fluid path under fault conditions using a retractable vector nozzle. The jets from adjacent inner rings adhere to the wall and flow towards the fault area, amplifying thrust through fluid entrainment, thus achieving efficient active thrust compensation without increasing additional hardware redundancy weight.

[0010] 2. This invention utilizes a coaxial telescopic sleeve assembly to achieve physical configuration switching. During normal flight, the nozzle retracts to maintain a safe clearance and avoid airflow interference; it only extends to engage with the guide bridge and trigger the wall attachment effect in case of a malfunction, effectively balancing the aerodynamic efficiency of conventional cruise with the flow field reconstruction requirements in emergency situations.

[0011] 3. This invention induces ambient air to participate in the flow through the Coanda effect, enabling the small-sized high-frequency inner ring unit to generate equivalent lift sufficient to compensate for the failure of the large-sized outer ring unit. Combined with optimized control allocation, this improves the survivability of the flight platform. Attached Figure Description

[0012] Figure 1 This is a block diagram of the overall architecture of the fault-tolerant control system of the present invention;

[0013] Figure 2 This is a flowchart illustrating the fault detection and topology reconfiguration logic of the present invention.

[0014] Figure 3 This is a comparison diagram of the thrust response during the fault reconstruction stage of the present invention;

[0015] Figure 4 This is a simulation diagram of the height holding capability under a single point of failure according to the present invention;

[0016] Figure 5 The timing logic diagram for the reconfiguration of the control system of this invention is shown.

[0017] Among them, 10. Fuselage platform; 20. Outer ring main lift propulsion group; 30. Inner ring attitude control propulsion group; 40. Inter-ring Coanda guide bridge; 50. Distributed flight control system; 21. Outer ring duct unit; 22. Low-speed rotor; 23. High power density drive motor; 31. Inner ring duct unit; 32. High-speed rotor; 33. High power density drive motor; 34. Three-degree-of-freedom vector nozzle mechanism; 51. Condition monitoring module; 52. Fault diagnosis module; 53. Topology reconstruction module; 54. Control allocation calculation module. Detailed Implementation

[0018] The technical solutions in 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.

[0019] Please see the appendix Figure 1 and attached Figure 2 This invention provides a coaxial counter-rotating double-ring duct vector propulsion system and its fault-tolerant control method.

[0020] The main hardware components of the system include a fuselage platform 10, an outer ring main lift propulsion group 20, an inner ring attitude control propulsion group 30, an inter-ring Coanda guide bridge 40, and a distributed flight control system 50. In this embodiment, the fuselage platform 10 adopts a one-piece ring truss structure made of carbon fiber composite material to ensure structural rigidity under high-frequency vibration disturbances caused by the double-ring counter-rotation.

[0021] The outer ring main lift propulsion group 20 is arranged along the outer circumference of the fuselage platform 10, including... Each outer ring duct unit 21 is equipped with a large-diameter low-speed rotor 22.

[0022] Inner ring attitude control propulsion group 30 includes An inner ring duct unit 31 is installed inside the radial gap between each sector of the outer ring duct unit 21, forming a staggered nested layout. The inner ring duct unit 31 is equipped with a small-diameter high-speed rotor 32 with a rated speed range of 5000 to 8000 rpm, which is driven by a high-power-density drive motor 33 and rotates clockwise.

[0023] To achieve active flow field control, each inner ring duct unit 31 integrates a three-degree-of-freedom vector nozzle mechanism 34 at its tail. Unlike conventional single-axis nozzles, this mechanism employs a composite structure of ball joints and sliding cylinders: the base portion drives the universal ball joint via two sets of orthogonally arranged aircraft linkages to achieve pitch angle control. With yaw angle Adjustable within ±30°; the nozzle body is designed as a double-nested telescopic sleeve, driven by an embedded miniature linear screw motor, enabling continuous axial length adjustment. Its travel range covers to (e.g., ±100mm).

[0024] The inter-annular Coanda guide bridge 40 is a key static component connecting the aerodynamic characteristics of the inner and outer annulus rings. Geometrically, the inlet edge of the inter-annular Coanda guide bridge 40 is precisely located above the maximum elongation of the three-degree-of-freedom vector nozzle mechanism 34. The nozzle trajectory tangent at that time. This design establishes the physical logic of coupling only during extension: when the nozzle retracts, the airflow is directly injected into the atmosphere; only when the nozzle extends and deflects will the high-speed jet tangentially adhere to the surface of the guide bridge and be guided to the region below the outer ring.

[0025] The distributed flight control system 50, acting as the nerve center of the system, connects to the aforementioned execution units via a high-speed industrial fieldbus. The system's internal logic is divided into a status monitoring module 51, a fault diagnosis module 52, a topology reconfiguration module 53, and a control allocation and calculation module 54.

[0026] Based on the above hardware architecture, the system operates according to the following rigorous logic:

[0027] During system initialization and normal flight phases, the control allocation and calculation module 54, based on the preset control performance matrix, instructs the outer ring main lift propulsion group 20 to bear approximately 80% to 90% of the total aircraft weight, while simultaneously instructing the inner ring attitude control propulsion group 30 to maintain nozzle retraction ( In this state, the remaining gravity and attitude torque are balanced only by differential speed or micro-amplitude vector deflection.

[0028] During system operation, the fault diagnosis module 52 performs model-based fault detection in real time. Specifically, this module internally runs a power system state observer, which compares the actual motor speed in real time. Theoretical speed calculated based on voltage / current If the residuals of the two In the time window The internal temperature continuously exceeds the preset safety threshold. Or reverse current When the value becomes zero, the fault diagnosis module 52 determines that the corresponding numbered outer ring duct unit 21 has experienced an irreversible thrust loss fault and outputs the fault index ID.

[0029] Once a fault trigger signal is detected, the topology reconfiguration module 53 immediately takes over control and, based on a pre-stored geometric adjacency matrix, quickly locks onto the two inner ring duct units 31 that are spatially closest to the faulty unit as flow field repair actuators. Subsequently, the system executes two levels of reconfiguration commands in parallel: at the physical level, it instructs the three-degree-of-freedom vector nozzle mechanisms 34 of these two units to extend at full speed to... And deflect outwards to the predetermined Kornda angle. The system forcibly establishes an aerodynamic path from the inner ring to the outer ring fault zone. At the algorithm layer, the control allocation solution module 54 removes the column vectors of the faulty units and introduces a new control allocation law based on ejector gain correction to calculate the compensation thrust of the remaining healthy units and the ejector speed of the repair units, thereby completing the control strategy switch from force balance to flow field balance in milliseconds.

[0030] The core power components are mainly composed of a ducted rectifier housing, a stator support assembly, a high power density drive unit, and an axial fan rotor.

[0031] The ducted fairing housing, serving as both the airflow channel and structural framework, is integrally molded from composite materials. The inlet lip of the ducted fairing housing is designed as a convergent surface based on a hyperelliptic curve, with a contraction ratio (the ratio of inlet area to fan disk area) set between 1.15 and 1.25. This geometric feature is designed based on the fluid continuity equation.

[0032] The stator support assembly is radially arranged inside the duct, connecting the inner wall of the duct to the central motor mounting base. In this embodiment, the stator support assembly includes 7 to 11 stationary blades with an S-shaped reverse camber airfoil design. The leading edge mounting angle of this airfoil matches the absolute velocity vector angle of the airflow at the fan rotor outlet, used to convert the tangential component of the rotating airflow into axial velocity.

[0033] The high-power-density drive unit, located within the central motor mounting base, is the core unit for generating rotational torque. To meet the compact axial dimension constraints of the dual-ring duct system (thickness must be less than 40% of the duct chord length), this embodiment prioritizes the use of a coreless axial flux permanent magnet synchronous motor. The motor's rotor disk employs a dual-rotor-to-single-stator topology, and the magnetic pole array uses neodymium iron boron permanent magnets arranged in a Halebeck array to achieve a sinusoidal magnetic field focusing effect on one side. The stator disk uses slotless aluminum wire manganese windings cast with high-strength epoxy resin, eliminating cogging torque and reducing high-frequency eddy current losses. The motor provides real-time feedback of the rotor electrical angle via a built-in Hall sensor or high-precision magnetic encoder, and, in conjunction with a vector control driver, achieves millisecond-level response to speed and torque.

[0034] The axial fan rotor is directly fixed to the output shaft of the motor. For the outer ring main lift unit, the fan rotor is configured with a 5-blade structure with a large chord length and low density. The blade torsional pattern follows a free vortex design to ensure a uniform distribution of axial induced velocity from the blade root to the blade tip. For the inner ring attitude control unit, the fan rotor is configured with a 9-blade structure with a small chord length and high density to reduce rotational inertia and improve the system's acceleration and deceleration characteristics. The radial clearance between the blade tip and the inner wall of the duct... Based on the strict control of thermomechanical coupling deformation, its value follows... ,in The gap value is set to balance volumetric efficiency with the risk of mechanical interference, which is the diameter of the fan.

[0035] Based on the momentum blade element theory, there is a clear physical mapping between the aerodynamic force generated by the propulsion unit and its rotational speed. Under the condition of axial inflow neglecting crosswind interference, the first... The thrust generated by each propulsion unit and resistance It follows the following dynamic equations:

[0036] ;

[0037] ;

[0038] in: and The first The aerodynamic thrust and aerodynamic anti-torsion generated by each propulsion unit; This refers to the local atmospheric density. The diameter of the fan; This refers to the mechanical angular velocity of the motor. The thrust coefficient is a dimensionless coefficient, and for the ducted rotor in this embodiment, its typical value range is [value range missing]. ; It is a dimensionless torsional resistance coefficient, and its typical value range is... .

[0039] The above coefficients and It is not a constant, but rather related to the intake ratio. Strongly correlated nonlinear functions.

[0040] As the execution end point for realizing flow field topology switching, the three-degree-of-freedom vector nozzle mechanism 34 is configured to independently adjust the axial length of the jet stream. Pitch and yaw angles and yaw angle .

[0041] In this embodiment, the three-degree-of-freedom vector nozzle mechanism 34 mainly consists of a biaxial orthogonal bidirectional joint assembly, a coaxial telescopic sleeve assembly, and a flexible sealing compensation component. The biaxial orthogonal bidirectional joint assembly serves as the motion base, including a fixed support rigidly connected to the inner ring duct unit 31 housing, a yaw outer frame rotatable about a vertical axis, and a pitch inner frame nested within the yaw outer frame and rotatable about a horizontal axis. The yaw outer frame and the pitch inner frame are respectively linked by two sets of high-temperature, high-torque servo motors via push-pull linkages. This mechanism eliminates the kinematic transmission chain of yaw followed by pitch, thereby eliminating motion singularities within the ±30° conical working space.

[0042] The coaxial telescopic sleeve assembly is installed in the center hole of the pitch inner frame and swings synchronously with the universal joint. This assembly consists of a guide outer cylinder and a sliding inner cylinder. Three precision linear guide rails are symmetrically arranged on the inner wall of the guide outer cylinder, and the sliding inner cylinder is embedded in the guide rails via a slider to restrict its circumferential rotational freedom. The axial feed power of the sliding inner cylinder comes from a miniature trapezoidal lead screw assembly integrated into the cylinder wall interlayer, and the lead screw is driven by a hollow cup motor. Under control commands, the sliding inner cylinder can perform stepless linear telescopic movement relative to the guide outer cylinder.

[0043] To ensure airtightness under complex motion conditions, a high-temperature resistant flexible sealing structure is provided between the fixed support and the yaw frame, as well as at each movable joint.

[0044] The key geometric feature of this mechanism lies in its spatial fit between its maximum elongation state and the interring Coanda guide bridge 40. Specifically, the system sets a physical threshold. When the sliding inner cylinder is in the retracted position ( When the distance between the nozzle end face and the leading edge of the guide bridge is greater than 100,000, the distance between the nozzle end face and the leading edge of the guide bridge is greater than 100,000. The airflow is freely ejected; only when the sliding inner cylinder is driven to its maximum stroke ( When the yaw angle meets the tangential condition, the nozzle end face physically extends to the guide bridge boundary layer suction zone, thereby triggering the wall attachment effect.

[0045] Based on the above institutional model, the first... A kinematic model of a vector nozzle with an equal number of propulsion units. The normalized extension / retraction control variable of the nozzle is defined as... Then the current physical effective length of the nozzle The calculation is as follows:

[0046] ;

[0047] in: This is the base length when the nozzle is fully retracted; This is the maximum length when the nozzle is fully extended; It is a dimensionless travel factor, obtained by normalizing the feedback value from the displacement sensor.

[0048] Thrust vector directional control is achieved through coordinate transformation. Based on the aforementioned yaw-around... Axis, back pitch The mechanical rotation sequence of the shaft (assuming the initial axial direction of the nozzle is along) The actual coordinate system definition depends on the fuselage mounting orientation. Adjustments are made here using the nozzle's local coordinate system. (Description) The thrust direction vector at the nozzle exit. With nominal thrust The relationship is determined by the rotation matrix. describe:

[0049] ;

[0050] ;

[0051] in: The yaw angle is the range of values ​​constrained by mechanical limits. ; The pitch angle is also constrained within a certain range. .

[0052] This kinematic model is embedded in the underlying drive reconfiguration of the distributed flight control system 50. During flow field reconfiguration, the controller substitutes the target vector into the inverse solution algorithm of the aforementioned matrix. Since the deflection angle is limited to a small range of ±30°, the cosine term... Furthermore, it is far from the 90° singularity, thus ensuring the numerical stability and uniqueness of the inverse solution.

[0053] The interring Coanda guide bridge 40, as a static aerodynamic component connecting the outer ring main lift propulsion group 20 and the inner ring attitude control propulsion group 30, is essentially a fixed airfoil structure with asymmetric curvature characteristics, rigidly mounted on the radial load-bearing spokes connecting the inner ring.

[0054] In this embodiment, the main structure of the interring Coanda guide bridge 40 is manufactured using a honeycomb sandwich process with high-modulus carbon fiber reinforced epoxy resin composite material. This material selection aims to balance high structural proportional stiffness with long-term dimensional stability of the aerodynamic surface. To suppress edge effects caused by three-dimensional flow, the guide bridge has integrally formed aerodynamic end plates on both sides of its airflow direction. The height of these end plates is set to 1.5 to 2 times the characteristic thickness to create a two-dimensional flow field environment and prevent external airflow from laterally entraining and damaging the low-pressure zone attached to the wall.

[0055] From the perspective of cross-sectional geometry, the upper surface profile (i.e., working surface) of the inter-ring Coanda guide bridge 40 is divided into three functional sections along the airflow direction: jet capture zone, pressure recovery deflection zone, and wake shear fusion zone.

[0056] The jet-capturing zone is located at the inner leading edge of the guide bridge, and its geometry is characterized by a smooth, obtuse angle with a small radius of curvature (e.g., 5 mm to 10 mm). In terms of spatial topology, the tangent direction of this region has a predetermined geometric angle of attack with the theoretical jet centerline when the inner ring vector nozzle is at its maximum extension position. (range of values) This small positive angle of attack design, based on the boundary layer suction principle, helps to quickly establish a negative pressure gradient at the moment of jet contact, preventing the airflow from impacting the leading edge and generating shock waves or separation bubbles. The pressure recovery deflection zone constitutes the main working surface of the guide bridge, and its surface profile follows a logarithmic spiral or a piecewise variable curvature arc, with a curvature radius... The curvature gradually increases along the flow direction. This decreasing curvature design aims to match the physical characteristic of jet velocity decaying along the flow path and maintain a stable radial pressure gradient. To ensure that the jet does not undergo centrifugal separation at high speeds, the local curvature design of the guide bridge surface must satisfy the following Coanda stability criterion:

[0057] ;

[0058] in: The characteristic thickness of the airflow at the outlet of the inner ring nozzle (height for rectangular nozzles, diameter for circular nozzles). For the wall surface at the corner The local radius of curvature at that point; The average velocity of the jet; is the jet momentum coefficient, which characterizes the ability of a jet to maintain adhesion to the wall primarily through momentum rather than viscosity; The separation critical coefficient is used for engineering simplification. For the subsonic cold jet involved in this embodiment (Mach number...),... To reserve a safety margin, set When the ratio exceeds this range, the centrifugal force of the airflow will overcome the radial pressure gradient, leading to wall adhesion failure.

[0059] The wake shear fusion zone is located at the outer rear edge of the guide bridge, and its geometric end extends to below the outlet plane of the outer ring culvert unit 21, with the tangential direction perpendicular downward.

[0060] The spatial clearance control between the inter-annular Kornda guide bridge 40 and the three-degree-of-freedom vector nozzle mechanism 34 is the physical basis for achieving mode switching. The Euclidean distance from the center of the inner annular nozzle exit plane to the jet capture zone of the guide bridge is defined as the coupling clearance. Considering mechanical tolerances and thermal deformation, the dynamic variation model of this gap is as follows:

[0061] ;

[0062] in: The nozzle is in a fully retracted standby state. The physical interval distance at that time must satisfy the following conditions: To prevent the jet from accidentally adhering to the wall when not in operation; This is the effective mechanical extension / retraction stroke of the nozzle; This is the normalized scaling control quantity; This refers to the spatial angular deviation between the nozzle axis and the radial connecting line of the guide bridge.

[0063] When the system executes a reconfiguration command, the coupling gap... Compressed to the critical capture threshold Within (usually set) Within this micro-gap, the entrainment effect of the high-speed jet will rapidly evacuate the air between the nozzle and the guide bridge, forming a low-pressure closed loop region, forcing the jet to abruptly change from the free jet mode to the attached wall mode within milliseconds.

[0064] The ejector aerodynamic mode is the core fault-tolerant mode of the system when the outer ring main lift unit fails. Its physical essence is to use the high-speed jet in the inner ring as an active power source, and drive the stagnant air mass in the outer ring channel through the physical mediation of the Kornda guide bridge, thereby constructing a fluid duct fan effect.

[0065] In this embodiment, when the flow field reconstruction command is executed, the variable-configuration vector nozzle of the inner ring attitude control unit extends axially to its maximum stroke position, while simultaneously performing yaw deflection to adjust the jet injection angle. At this time, the nozzle exit plane spatially approaches the jet trapping zone of the Coanda guide bridge, and the physical gap between them is compressed to the critical coupling threshold. Under this geometric constraint, the high-speed jet ejected from the inner ring is restricted by the wall and cannot maintain the development of a free boundary layer. It rapidly establishes a radial negative pressure gradient on the convex surface of the guide bridge, forcing the streamlines to undergo a nearly 90° wall-attached deflection in accordance with the curvature of the guide bridge.

[0066] The technological advantage of the ejector aerodynamic mode lies in its thrust multiplication characteristic. When the attached-wall jet passes the trailing edge of the guide bridge and enters the wake region below the outer ring channel, a strong turbulent shear layer is formed between the high-speed jet core region and the surrounding low-speed ambient fluid. Based on the principles of fluid viscosity and momentum exchange, this shear layer continuously entrains surrounding still air into the flow, causing the jet cross-section to expand along the flow path and increasing the total exhaust mass flow rate. This momentum transfer process not only changes the velocity distribution of the flow field but also induces additional pressure thrust on the surface of the guide bridge and the inner wall of the outer ring duct.

[0067] Based on the above physical mechanism, the first [unit / item] in ejection mode... The total equivalent repair force vector generated by a rescue unit with an equal number of propulsion units Instead of following the simple law of reaction force, it adopts a composite model that is modified by introducing an ejection gain coefficient:

[0068] ;

[0069] in: This represents the total effective force vector, which includes both direct thrust and ejected additional thrust. The mass flow rate of the active jet ejected by the inner ring rescue unit is determined by the motor speed and the cross-sectional area of ​​the inner ring duct. The average velocity vector of the active jet at the trailing edge is mainly directed vertically downwards; is a dimensionless ejector gain coefficient, characterizing the additional thrust gain obtained by the system through entrainment of the surrounding fluid.

[0070] In this embodiment, It is not a fixed constant, but rather serves as a geometric characteristic parameter of the guide bridge. The function. Integrating the main jet outlet area Effective mixing area with the outer ring ejected region The ratio (i.e., the area ratio) Based on the momentum conservation theory and the CFD calibration results of this embodiment, The range of values ​​for the engineering parameters is set as follows:

[0071] ;

[0072] The lower limit of this value range corresponds to the simple wall-attached flow condition under static atmospheric conditions, while the upper limit corresponds to the optimal diffuser configuration condition generated by the outer ring duct structure. To ensure the conservatism and safety of the calculation model, in the thrust distribution solution of the flight control system, a value is typically taken as... This means that flow field reconstruction can bring at least 20% additional thrust gain, which does not require additional electrical power and is purely due to the improved energy conversion efficiency of the fluid dynamic structure.

[0073] In addition, to maintain a stable entrainment effect, the Reynolds number of the inner ring jet is... It needs to be maintained at the critical value The above measures ensure that the attached flow remains in a turbulent state, thereby overcoming the adverse pressure gradient through high-energy turbulent micro-elements and preventing premature separation of the airflow at the large curvature turning point of the guide bridge. This hydrodynamic constraint directly defines the minimum speed limit of the inner loop motor in rescue mode. That is, when performing a reconfiguration task, even if the attitude control requirements are low, the inner loop unit must be maintained above the base idle speed to maintain the connectivity of the hydro-aerodynamic bridge.

[0074] The ejector aerodynamic mode is the core working mode of this system, which utilizes the fluid physics properties to achieve thrust enhancement and fault tolerance.

[0075] In this embodiment, when the variable configuration vector nozzle of the inner ring attitude control unit performs a full extension action, i.e., the normalized telescopic control quantity... At a speed of 1.0, the nozzle exit plane spatially crosses the physical isolation zone in independent mode, approaching the leading edge notch of the interring Coanda guide bridge 40. At this point, the physical gap between the nozzle lip and the guide bridge surface is compressed to the critical coupling threshold. (Preferred setting is 3mm to 5mm). Under this geometric constraint, the high-speed jet ejected from the inner ring is restricted by the single-sided wall, and cannot maintain the axisymmetric entrainment boundary condition of the free jet, resulting in the pressure on the lower side of the jet beam being lower than the pressure on the upper side of the environment.

[0076] Based on the Coanda effect in fluid dynamics, the aforementioned lateral pressure gradient provides the centripetal force required to maintain the curved flow, forcing the high-speed fluid to overcome inertia and continuously deflect along the convex surface of the guide bridge. This physical process transforms the momentum flow that was originally jetted horizontally or obliquely into a vertically downward induced flow without any mechanical moving parts.

[0077] The technological advantage of the ejector aerodynamic mode lies in its thrust multiplication characteristic, which is physically derived from the viscous entrainment of fluids. When the high-energy wall-attached flow passes the trailing edge of the guide bridge and enters the low-energy wake region below the outer ring channel, a strong turbulent shear layer is formed between the high-speed mainstream and the surrounding stationary fluid. Based on the principle of momentum transfer, fluid particles within the shear layer continuously transfer kinetic energy to the outer low-speed fluid, thereby entraining a large amount of ambient air and accelerating it downwards with the mainstream. This process not only increases the total mass flow rate of the working fluid participating in the work but also induces additional pressure thrust components on the suction surface of the guide bridge and the inner wall of the outer ring duct.

[0078] To quantify this complex fluid dynamics process, this embodiment establishes an equivalent thrust model based on momentum conservation. (The text then abruptly shifts to a seemingly unrelated topic: "In ejection mode, and the...") The total effective thrust vector generated by an inner-ring rescue unit with an equal number of propulsion units is... Expressed as:

[0079] ;

[0080] in: This represents the total reaction force vector acting on the machine body after flow field reconstruction. This vector is a combination of the direct thrust generated by the motor and the additional thrust induced by the fluid. The mass flow rate of the active jet ejected by the inner ring propulsion unit depends on the integral of the effective intake cross-sectional area of ​​the inner ring duct and the exhaust velocity of the fan. The average velocity vector of the active jet as it leaves the trailing edge of the guide bridge, with its direction mainly downward along the vertical direction of the fuselage; is a dimensionless ejector gain coefficient, characterizing the additional thrust gained by the system through the fluid entrainment mechanism.

[0081] It should be noted that, It is not a fixed constant; its value depends on the dimensionless geometric characteristic parameters of the guide bridge. In this embodiment, Area of ​​the outer ring culvert being ejected Same inner ring jet outlet area The ratio (i.e., the area ratio) The relationship is non-linear and positively correlated. Based on the geometric configuration of this embodiment (wherein...) Furthermore, the ratio of the guide bridge chord length to the nozzle height ), The range of values ​​for the project is defined as follows: This means that without increasing the electrical power input of the inner ring motor, the system can obtain a passive thrust gain of 15% to 40% simply by changing the flow field topology, thereby effectively compensating for the lift gap after the failure of the outer ring main unit.

[0082] Furthermore, to maintain the stability of the wall-attached flow and prevent boundary layer separation at the large curvature transition section of the guide bridge, the Reynolds number of the inner ring jet is... It needs to be maintained at the critical value The above. This hydrodynamic constraint requires that the inner loop unit's rotor speed must be maintained at the minimum sustaining speed during rescue operations. Above this, to ensure that the boundary layer is in a fully developed turbulent state, the high energy transport characteristics of turbulent micro-clusters are used to resist the adverse pressure gradient.

[0083] In this embodiment, the dynamic description of the system is based on the body coordinate system fixed to the physical center of mass of the body. Based on this. To accurately describe the distribution along the circumference of the annular fuselage... The force state of the propulsion unit (comprising the outer ring main lift unit and the inner ring attitude control unit) is defined as follows: Local coordinate system of each propulsion unit Its origin Anchored at the geometric center of the channel. For any... Each propulsion unit has a control input vector. It includes three key physical degrees of freedom: the mechanical angular velocity of the drive motor. And the pitch angle of the vector nozzle relative to the duct axis. and yaw angle .

[0084] Based on blade element momentum theory, the aerodynamic force generated by the propulsion unit essentially depends on the relative velocity and angle of attack of the airflow cut by the profiler blades. Under high Reynolds number conditions that neglect airflow distortion, the initial thrust amplitude along the duct axis exhibits a non-linear proportional relationship with the square of the motor speed. However, the specific projection of this initial thrust in the airframe coordinate system is constrained by the real-time pointing of the vector nozzle. Considering the motion transmission logic of the mechanical structure, this embodiment uses a chain-like rotational transformation to describe the spatial mapping of the thrust vector: first, the thrust vector is rotated to the duct local coordinate system via pitch and yaw matrices, and then rotated to the airframe coordinate system via the installation azimuth matrix.

[0085] In summary, the first physical process The three-dimensional thrust vector generated by each propulsion unit in the body coordinate system The model is as follows:

[0086] ;

[0087] in: The principal dynamic vector mapped to the body coordinate system, its The axial component provides lift. The axis component provides the translational driving force; For the first The inherent installation azimuth angle of each propulsion unit on the fuselage circumference, for The uniformly distributed configuration, whose value is... ; Representing respectively axis, axis, The standard basic rotation matrix of the axis, this multiplication structure clearly defines the spatial transformation order of yaw, pitch, and positioning, ensuring the uniqueness of vector control solution; The comprehensive aerodynamic thrust coefficient, which is determined by static thrust bench test, characterizes the efficiency of a specific duct configuration in converting mechanical energy into aerodynamic thrust. The motor speed is limited by the maximum current characteristics of the driver, and the physical constraint range is... (For example ).

[0088] Besides the force vector required for linear motion, the torque required to maintain flight attitude stability is another core control variable. This includes the total torque generated by the propulsion unit on the airframe mass. It is not a single source, but a composite physical quantity resulting from the superposition of thrust eccentricity and aerodynamic anti-torsion effects. Especially in the double-ring coaxial counter-rotating layout employed in this invention, the characteristic of adjacent units rotating in opposite directions (i.e., (Alternating positive and negative), pure yaw torque can be generated by adjusting the speed difference without tilting the fuselage.

[0089] Based on the principles of rigid body dynamics, the torque generation model is expressed as follows:

[0090] ;

[0091] in: To point from the body's center of mass to the first The rigid position vector of the duct center is equal to the number of propulsion units. This vector is precisely determined by the CAD geometric parameters of the fuselage structure and determines the length of the control lever arm. Represents the vector cross product operation. The physical meaning of this operation is to actively change the distance between the thrust line of action and the center of mass by using the deflection of the vector nozzle, thereby achieving agile control of pitch and roll torque. The aerodynamic drag torque coefficient characterizes the torque required for a fan to overcome air viscous resistance. Similarly, it is calibrated experimentally; As the rotation direction identification factor, it is specified that the counterclockwise rotation unit takes Rotate the unit clockwise to take This design allows the total counter-torque of the system to be naturally balanced when it is suspended.

[0092] The above dynamic model constructs the control quantity With state quantity The deterministic causal chain between them. In subsequent fault-tolerant control steps, this model exists as the basic operator regardless of whether the system is in independent mode or ejection mode; the only difference is that when entering ejection mode, the above-mentioned... Based on this, the ejector gain term calculated in the aforementioned embodiments is superimposed to correct the thrust loss.

[0093] In this embodiment, when the inner ring unit performs a fully extended action ( After triggering the Coanda effect, the resulting aerodynamic thrust no longer follows the momentum theorem for isolated systems. At this point, the high-speed jet, acting as a momentum transfer source, continuously entrains surrounding still air through the viscous effect of the turbulent shear layer, increasing the total mass flow rate of the working fluid participating in the work. To transform this complex hydrodynamic gain into an algebraic constraint executable by the control algorithm, a nonlinear correction factor is introduced on top of the basic thrust model.

[0094] Based on the principles of momentum conservation and energy conversion, the ejection mode and the first Rescue units of equal number of propulsion units act vertically on the aircraft. Actual effective lift of the shaft Expressed as:

[0095] ;

[0096] in: This represents the actual lift component obtained by the system after flow field reconstruction. This value is a key input for controlling the calculation of vertical acceleration in the distribution matrix. The basic aerodynamic thrust coefficient of the inner ring ducted fan is obtained by mean calibration from static bench experiments under undisturbed flow field conditions. This refers to the real-time mechanical angular velocity of the drive motor. The entrainer gain coefficient is a dimensionless scalar that characterizes the fluid entrainment efficiency. In this embodiment, this coefficient is not a fixed constant, but is determined by the geometry of the guide bridge. Specifically, Area of ​​the outer ring being ejected Same inner ring jet outlet area The area ratios are positively correlated. Based on the geometric constraints set in this embodiment ( ), The range of values ​​for the project is limited to The specific values ​​are obtained by consulting the aerodynamic database pre-stored in the flight control computer.

[0097] The flow field reconstruction not only enhances the thrust amplitude but also alters the spatial distribution characteristics of aerodynamic forces. In ejector mode, since most of the lift is generated by the induced pressure on the surface of the guide bridge and the inner wall of the outer duct, the effective pressure center of the propulsion unit shifts radially outward. This shift in physical position means that, with the same thrust magnitude, the system can generate a larger attitude recovery torque.

[0098] To accurately describe this torque-enhancing effect, and in conjunction with the first The total torque generated by an equal number of rescue units on the body's center of mass Revised to:

[0099] ;

[0100] in: The vector of the repair force acting vertically, i.e. ; The vector of gyroscopic torque and counter-torque generated by the rotation of the motor and fan; This is the equivalent aerodynamic point position vector in ejection mode. In this embodiment, the magnitude of this vector is... No longer equal to the physical installation radius of the inner ring motor Instead, it is corrected to point towards the center of fluid pressure between the inner and outer rings. Its value can be estimated in engineering using the following formula: ,in Pressure distribution weighting coefficient (range of values) This revised lever arm length directly improves the control efficiency of the rescue unit in the pitch and roll channels, enabling the system to balance the unbalanced torque caused by the failure of the outer loop at lower speeds.

[0101] In this embodiment, when the system self-test confirms that the entire machine... When all propulsion units are in a healthy state, the control logic defaults to maintaining independent aerodynamic mode. At this time, the normalized scaling ratio command for all inner-loop attitude units is forcibly locked to zero. The Korndaer guide bridge is inactive. In this configuration, the total number of controllable input variables of the system (including 12 motor speeds and 24 nozzle deflection angles) is much greater than the control requirements of the four degrees of freedom of the body motion, which mathematically constitutes a redundant system with infinitely many solutions.

[0102] To address this redundancy allocation problem, this embodiment does not employ a simple pseudo-inverse matrix method, but instead introduces an optimization control allocation strategy based on quadratic programming. The core of this strategy lies in constructing a cost function that incorporates both energy and mechanical constraints. By finding the optimal operating point for the entire system, it achieves command tracking while simultaneously balancing mechanical wear equilibrium and maximizing aerodynamic efficiency.

[0103] The flight control computer first generates a total vertical lift, including the desired lift, based on the position and attitude errors fed back by the navigation system. and triaxial torque Virtual control command vector Subsequently, the system constructs and solves an optimization objective function of the following form. :

[0104] ;

[0105] in: Let be the system control input vector to be solved, and its dimension is . It includes the motor thrust of each unit (proportional to the square of the rotational speed) and the pitch and yaw angles of the nozzle; The desired deviation state vector of the actuator is usually set to the corresponding nozzle deflection angle of 0 degrees. This is intended to guide the system to prioritize the use of thrust differential rather than frequent mechanical deflection to generate torque, thereby extending the mechanical life of the vector mechanism. To achieve dynamic control efficiency, considering the Jacobian matrix of the vector with respect to the control input, the system calculates partial derivatives in real time based on the current operating point state within each control cycle. This is used to update the matrix elements, thereby locally linearizing the nonlinear model; The virtual instruction tracking weight matrix is ​​typically set to a large diagonal matrix (e.g., ...). (on a scale of magnitude) to ensure that the optimizer prioritizes eliminating attitude tracking errors as its highest priority task; Assigning weight matrices to the actuators is a key parameter for achieving hierarchical control in this embodiment. Based on the physical characteristics of the dual-loop architecture, the outer loop unit set... The corresponding weighting coefficients are set to be less than the inner loop unit set. Weighting coefficients ( This asymmetric weighting logic forces the algorithm to automatically favor the use of outer loop units with large diameters and high lift coefficients when allocating steady-state lift loads; while when dealing with high-frequency attitude disturbances, it utilizes inner loop units with higher weight penalties but smaller moments of inertia for a faster response.

[0106] When solving the above objective function, the hard constraints of the physical system must be strictly adhered to. This embodiment defines the physical upper and lower limits of the motor speed. , and the mechanical stroke limit of the vector nozzle Inequality constraints are incorporated into the solver. For such convex optimization problems, the effective set method or the primal-dual interior-point method can be used for iterative solutions. To prevent matrix singularities from causing solution divergence, this embodiment also introduces a damping term into the objective function. (in For small positive real numbers, such as This is to ensure the stability of numerical calculations.

[0107] Finally, the optimal control quantity obtained by the solution is... The signals are discretized into low-level PWM drive signals and CAN bus angle commands, which are sent to the electronic speed controller and servo controller respectively to drive the coordinated action of each propulsion unit. This model-based allocation method ensures that the system can perform flight missions with optimal aerodynamic efficiency across the entire flight envelope.

[0108] In this embodiment, to accurately identify the health status of the propulsion unit in a complex flight environment, a set of parallel-running state observers reside within the flight control computer. For each node in the outer ring main lift unit set O, the system no longer relies solely on a single error code fed back by the motor driver, but instead employs a model reference-based residual evaluation technique. The core logic of this technique lies in comparing the measured responses of physical sensors with the theoretical predictions of the dynamic model, thereby separating out systematic deviations caused by physical failures.

[0109] Definition and the first The number of outer ring propulsion units is equal to the number of propulsion units. )exist Normalized fault characteristic residual at time step for:

[0110] ;

[0111] in: The real-time mechanical angular velocity of the motor is acquired by a Hall sensor or photoelectric encoder; The desired speed command issued by the controller; This is a first-order inertial element operator (implemented in the discrete domain via difference equations), where This is the motor time constant (calibrated to 30ms~50ms in this embodiment). The purpose of introducing this step is to perform dynamic lag processing on the command signal to ensure consistency with the physical motor dynamic characteristics before calculating the residual, thereby eliminating spurious dynamic errors caused by sudden changes in the command. The normalization factor is the maximum rated speed of the motor.

[0112] To avoid misjudgments caused by gusts of wind or communication glitches, the system does not directly utilize instantaneous residuals. Instead of triggering a decision, a sliding time window mechanism is used. Only when the residual amplitude is continuously within the decision window... (For example, exceeding the preset safety threshold for 100ms, or approximately 40 control cycles) At that time, the fault flag bit It was only recently located. Among them, The value range is set to 0.10 to 0.15. The threshold is set based on the fact that it covers the upper limit of speed deviation caused by normal pneumatic load fluctuations.

[0113] Once confirmed, the When an outer ring unit experiences an irreversible thrust loss fault, the control logic immediately freezes the control output of that unit and enters the topology indexing stage. Based on the dual-ring coaxial staggered layout adopted in this invention, the only units physically capable of fluid repair of the faulty unit via the Coanda guide bridge are the two inner ring units that are geometrically adjacent to each other.

[0114] In this embodiment, the system employs a strict alternating odd-even numbering rule: outer ring cells occupy odd-numbered index positions, and inner ring cells occupy even-numbered index positions. To quickly and unambiguously locate physical neighbors at the software level, the system utilizes the cyclic symmetry of the circular array to construct an index mapping operator based on modulo operations. This operator can automatically handle the boundary conditions connecting the first and last elements from number 12 to number 1, without requiring the writing of complex conditional branch statements.

[0115] The formula for calculating the index of spatially adjacent rescue units is as follows:

[0116] The rescue unit number located on the counter-clockwise side (left adjacent) of the faulty unit:

[0117] ;

[0118] The rescue unit number located clockwise (right adjacent) to the faulty unit:

[0119] ;

[0120] in: The total number of propulsion units; The faulty outer loop unit number determined through the aforementioned steps (must be an odd number); Number the inner ring rescue unit located clockwise from the fault unit; The inner ring rescue unit is numbered on the counterclockwise side of the fault unit.

[0121] For example, when detected When unit number fails, the formula is calculated as follows: Unit number (right neighbor) and Unit number (left neighbor). These two index values ​​will serve as key addressing pointers, directly passed to subsequent configuration reconfiguration modules to activate the nozzle extension / retraction actuators and vector deflection servos of specific units, thereby establishing a hydro-aerodynamic link across the rings.

[0122] In this embodiment, once the fault index module locks the inner ring rescue unit number located on both sides of the fault sector... and The flight control system immediately triggered a configuration lock command for these two specific units. This command has the highest priority control authority, enabling it to override conventional attitude stabilization loops and force the underlying mechanical servo mechanisms to execute preset flow field reconfiguration actions.

[0123] The physical reconfiguration process consists of two synchronous mechanical processes: nozzle axial extension and vector orientation deflection. For the nozzle axial extension motion, the controller directs the control to the unit... and The telescopic mechanism driver sends a stepped control signal to normalize the telescopic ratio. The default contraction position (0) is set to the maximum extension position (1). In response to this command, a linear actuation component (such as a miniature trapezoidal screw or electric push rod) integrated into the outer wall of the duct drives the cylindrical vector nozzle to slide outward along the fuselage axis. During this sliding process, the nozzle maintains a zero yaw angle to pass through the radial channel reserved inside the guide bridge until the physical exit plane of the nozzle completely crosses the geometric leading edge of the guide bridge. Subsequently, the controller drives the yaw servo mechanism to deflect the nozzle to one side of the guide bridge, thereby achieving seamless contact between the nozzle and the guide surface of the Coanda guide bridge. This sliding process continues until the physical exit plane of the nozzle completely crosses the upper edge of the inlet lip of the outer duct and achieves a geometrically tight fit with the guide surface of the Coanda guide bridge. This physical contact is a prerequisite for achieving the Coanda effect, and its technical purpose is to eliminate the geometric step between the jet exit and the guide surface, preventing early turbulent separation or momentum dissipation of the high-speed airflow before wall attachment. To ensure the inner ring jet accurately injects into the area below the faulty unit along the tangential direction of the guide bridge, the yaw angle of the rescue unit must be adjusted for vector-directed deflection maneuvers. Apply precise guidance constraints. Based on the body coordinate system defined in this embodiment (with counterclockwise rotation defined as the positive direction of the yaw angle), and considering that the inner ring unit is distributed between the outer ring units, the rescue unit located on the counterclockwise side (left adjacent) of the fault point... It needs to be turned clockwise, and the rescue unit is located on the clockwise side (right adjacent) of the fault point. It needs to be deflected counterclockwise to form a converging flow field pointing towards the center of the fault.

[0124] Accordingly, the system applies the following mirror deflection angle constraint to the rescue unit:

[0125] ;

[0126] in: These are the yaw angle commands for the left and right rescue units, respectively. This is the preset optimal fluid coupling angle. In this embodiment, this angle is not arbitrarily selected, but rather a geometric constant obtained through wind tunnel testing, and its value typically ranges from [value missing]. The physical basis for determining this angle is to make the tangential angle between the jet core region and the leading edge of the guide bridge zero, thereby maximizing the momentum retention rate of the attached flow; The pitch angle command is forcibly locked at this stage. The technical purpose of this constraint is to suppress the escape of the vertical component of the jet, ensuring that all momentum participates in the horizontal Coanda entrainment process.

[0127] From the mathematical implementation perspective of the control allocation algorithm, the execution of this step signifies a dimensionality reduction in the system's optimization solution space. In the reconfiguration mode, the rescue unit no longer participates in the torque allocation of the entire system as a free control variable, but is transformed into known state parameters. Mathematically, the system constructs the following set of equality constraints. :

[0128] ;

[0129] in, This indicates that the system is in reconfiguration mode, and the control input vector... Forced structured constraints; Indicates the first The control input vector of each propulsion unit; Indicates the time when the motor is The rotational speed; This indicates the offset of coupling terms that are manually introduced or retained in locked mode.

[0130] Although the nozzle configuration of the rescue unit is mechanically locked, the rotation speed of its motor is... In subsequent steps, it can still be retained as a scalar variable to be optimized, or it can be directly set as the saturation speed depending on the severity of the fault. This control strategy of hard constraint of configuration and soft adjustment of energy ensures the determinism of the flow field topology in physical terms, so that subsequent algorithms can directly call the ejector gain model to calculate the thrust compensation without having to deal with the uncertainty of the flow field structure in real time, thereby reducing the computational load under fault conditions.

[0131] In this embodiment, with the completion of the physical reconfiguration, the multiple propulsion units of the entire machine are logically divided into three mutually exclusive functional subsets: the outer ring unit set that stopped due to a fault. (Its output is forcibly set to zero), a set of inner-loop rescue units performing fluid repair while in a position-locked state. and the remaining healthy unit set that maintains the standard vector maneuver mode. Since the nozzle attitude of the rescue unit has been mechanically locked and the aerodynamic forces have generated Coanda gains, the original control performance matrix is ​​no longer applicable, and a reconfigurable control model incorporating fluid coupling characteristics must be constructed.

[0132] To map the physical flow field changes to the algebraic equations at the algorithmic level, the system first modifies the control efficiency matrix. The column vectors are reorganized and their values ​​corrected to generate a reconstructed function matrix. For the remaining set of healthy units In the rescue unit, the corresponding control column vector is no longer generated by conventional trigonometric functions, but is replaced by a constant vector based on the ejection physics model. Specifically, its thrust coefficient terms are modified as follows: (in This is the ejector gain coefficient, and its value typically ranges from [value range missing]. Between them), the torque vector is based on a fixed coupling angle. The nozzle extension position is recalculated. The technical purpose of this correction is to explicitly account for the additional lift contribution and torque center shift caused by the entrainment of surrounding air by the jet in the algorithm model, preventing control divergence due to model mismatch.

[0133] Based on the reconstructed system model, the system constructs the following reconstructing optimization allocation problem, aiming to find an optimal set of control increments to minimize instruction tracking error:

[0134] ;

[0135] in: This is the reconstructed control input vector to be solved. It should be noted that the dimension of this vector has changed from that in the normal mode. Reduced to Dimension. This is because of the remaining set of healthy units. Each unit in the system retains three degrees of freedom: rotational speed, pitch, and yaw. The inner ring rescue unit is assembled... The rescue unit in the middle retains only one degree of freedom: rotational speed (its vector angle has been fixed as a known parameter). (in the matrix) The desired force and torque command vectors generated for the upper-level guidance law; This term is introduced as the result of the previous control cycle. It aims to limit the high-frequency jumps of control commands in the time domain, thereby protecting the motor bearings and servo gears from electromechanical shocks. The Tikhonov regularization coefficient is set to a value of 1 in this embodiment. This coefficient was introduced to address the matrix rank deficiency problem caused by actuator failure, i.e., when... When approaching singularity, a damping term is introduced to ensure the stability of numerical inversion and avoid calculation errors such as division by zero.

[0136] Considering the real-time requirements of embedded systems (typically less than 2ms), this embodiment uses the damped least squares method to derive the closed-form analytical solution to the above optimization problem:

[0137] ;

[0138] in, It is the identity matrix. The calculated value is... This contains optimization instructions for all available cells, including those belonging to the remaining healthy cell set. The components are analyzed into rotational speed and vector angle commands, used to generate the main balancing torque; it belongs to the inner ring rescue unit set. The component is only interpreted as a rotation speed command, which drives the rescue unit to generate a high-energy jet to maintain the stability of the ejector flow field.

[0139] If the calculated control quantity exceeds the physical limit (e.g.) The system employs a weighted direction-maintaining saturation and cutoff strategy: prioritizing proportional reduction of the yaw moment. The contribution of strictly ensuring vertical lift With pitch / roll moment The output is matched. This degradation logic conforms to the flight safety principle of survival over heading, ensuring that the aircraft can maintain its attitude level and make a safe emergency landing under extreme failure conditions.

[0140] Specific application examples:

[0141] In this embodiment, a six-axis, twelve-rotor ring-ducted logistics UAV is selected as the physical carrier to verify the coaxial counter-rotating dual-ring ducted vector propulsion system and its fault-tolerant control method of the present invention.

[0142] System parameter configuration: The UAV fuselage adopts a carbon fiber ring truss structure, with a designed takeoff weight of 48 kg and an outer ring diameter of 1.8 meters. The specific configuration of the propulsion system is as follows:

[0143] The outer ring main lift propulsion group consists of 6 evenly distributed duct units (numbered oddly), with a maximum static thrust of 120 Newtons per unit.

[0144] Inner ring attitude control propulsion group: It consists of 6 evenly distributed duct units (numbered evenly), with a maximum static thrust of 40 Newtons per unit, located at the sector clearance of each outer ring unit.

[0145] Inter-ring Coanda guide bridge: Installed at the connection between the inner and outer rings, its geometric surface characteristics are designed to maximize the ejector gain coefficient. The calibration mean is 0.25.

[0146] Fault scenario preset: The experiment simulates a drone performing a hovering mission at a height of 50 meters. During the simulation time... At a certain time, the No. 1 main lift unit of the outer ring is set to experience a motor jamming fault, and its thrust output instantly drops from the hovering balance value to zero.

[0147] Fault-tolerant control logic calculation: Based on the control method of this invention, the system executes the following logic:

[0148] Fault Lockout: The status monitoring module confirmed that Unit 1 had failed after detecting that the speed residual error exceeded the limit for 80 milliseconds.

[0149] Neighborhood Index: The topology reconstruction module uses the ring index algorithm to determine that the spatial neighbor of cell 1 is cell 2. (counterclockwise left neighbor) and Unit 12 ( (clockwise to the right neighbor).

[0150] Command generation: The control allocation and calculation module generates a reconstruction command, requiring the nozzles of units 2 and 12 to extend and deflect towards the center of the fault sector, while increasing the rotation speed to establish an ejector flow field.

[0151] To verify the effectiveness of the aforementioned fault-tolerant control strategy, a hardware-in-the-loop simulation platform incorporating a dynamic model was built, and static bench thrust tests and dynamic flight simulation tests were conducted. The following is in conjunction with the appendix to the instruction manual. Figure 3 To be continued Figure 5 Explain the experimental results.

[0152] Bench tests were conducted to investigate the thrust generation characteristics of the inner-ring rescue unit during the reconfiguration process. The test results are attached. Figure 3 As shown.

[0153] Traditional solution (dashed line): corresponds to the conventional inner ring unit operating mode, where the nozzle remains constricted, with no ejection effect. After a step command is input, the thrust increases with the motor speed. However, due to the blade size, the steady-state thrust is saturated at 40 Newtons.

[0154] The present invention scheme (solid line): corresponds to the flow field reconstruction mode. Simultaneously with motor startup, the nozzle extends and deflects.

[0155] Data Analysis: From Appendix Figure 3 It can be seen that the thrust curve of the present invention is in the middle of the rising phase ( At a certain point (seconds), a slope change occurs, and this region is marked as the flow field reconstruction gain region. This is because the high-speed jet, attached to the Coanda guide bridge, entrains ambient air, generating additional induced lift. The final steady-state thrust reaches approximately 50 Newtons. Compared to the traditional approach, the system achieves a 25% additional thrust gain without increasing the motor's rated power. This data verifies the thrust multiplication mechanism of the ejector aerodynamic mode.

[0156] Appendix Figure 5 The timing of the response of the underlying actuators when the system dealt with the failure of Unit 1 was recorded, with the timeline taken from 4.9 seconds to 5.3 seconds.

[0157] Fault occurrence and detection ( The third sub-graph shows that the rotational speed of the faulty unit (No. 1) is... The second dropped to zero. After a short fault assessment window, the system... The physical reconstruction command is triggered within seconds.

[0158] Mechanism coordination ( ): Scalability (First sub-figure): The nozzle extension control of the rescue unit is a step signal, quickly switching from the retracted position (0) to the maximum extension position (1).

[0159] Yaw angle (Second sub-diagram): To achieve airflow convergence, the rescue unit is located clockwise from the fault point. (Solid line) deflects counterclockwise; the rescue unit is located on the counterclockwise side of the fault point. (Dashed line) Deflects clockwise. This mirror deflection ensures that the jet tangentially enters the guide bridge surface.

[0160] rotational speed (Third sub-figure): The speed command of the rescue unit (No. 2 / No. 12) is synchronously increased. Due to the aforementioned fluid gain, the motor speed does not need to reach the limit value to provide the required compensation force. The figure shows that the speed smoothly transitions to between 6000-8000 RPM.

[0161] Appendix Figure 4 The height channel response curve of the whole machine under single-point fault disturbance is shown, and the control effect of the present invention solution is compared with that of the traditional diagonal cut-off solution.

[0162] Comparison group (dashed line): A traditional diagonal shutdown strategy was adopted (i.e., if unit 1 fails, unit 7 is forcibly shut down). Due to the instantaneous loss of power from two main lift units, the aircraft still experienced a significant drop in height despite the remaining motors operating at full load. Data shows a maximum drop of approximately 3 meters, and the recovery time to the target height (50 meters) was relatively long, exhibiting overshoot oscillations, indicating a decrease in system damping ratio and deterioration in stability.

[0163] This invention group (solid line): Employs a fluid ejection reconfiguration strategy. Thanks to the rapid intervention of neighboring units and the additional lift provided by the Coanda effect, the system effectively fills the thrust gap. Data shows that the maximum altitude drop after a failure is controlled within 0.5 meters, and the system can quickly converge and stabilize back to the target altitude of 50 meters.

[0164] Conclusion: Experimental data show that the system and method provided by this invention, by triggering fluid gain through changes in physical configuration, offer greater equivalent control force with the same energy consumption. Compared to traditional fault-tolerant strategies, this approach reduces altitude loss under fault conditions and improves the aircraft's survival probability and flight stability after actuator failure.

Claims

1. A coaxial counter-rotating double-ring ducted vector propulsion system, characterized in that, Includes a fuselage platform (10), on which an outer ring main lift propulsion group (20) is arranged circumferentially, and the outer ring main lift propulsion group (20) includes multiple outer ring duct units (21); An inner ring attitude control propulsion group (30) is provided in the inner gap of the outer ring main lift propulsion group (20). The inner ring attitude control propulsion group (30) includes multiple inner ring duct units (31) that are staggered with the outer ring duct unit (21). The outer ring main lift propulsion group (20) and the inner ring attitude control propulsion group (30) are fixedly connected by an inter-ring Coanda guide bridge (40); The tail of the inner ring duct unit (31) is integrated with a three-degree-of-freedom vector nozzle mechanism (34), which is capable of axial extension and deflection relative to the axis of the inner ring duct unit (31). The vector propulsion system also includes a distributed flight control system (50), which controls the extension and deflection of the three-degree-of-freedom vector nozzle mechanism (34) so ​​that the nozzle of the three-degree-of-freedom vector nozzle mechanism (34) fits against the surface of the inter-ring Coanda guide bridge (40), thereby establishing an ejector flow field pointing from the inner ring duct unit (31) to below the outer ring duct unit (21).

2. The coaxial counter-rotating double-ring ducted vector propulsion system according to claim 1, characterized in that, The three-degree-of-freedom vector nozzle mechanism (34) includes a biaxial orthogonal bidirectional joint assembly, a coaxial telescopic sleeve assembly, and a flexible sealing compensation component; The dual-axis orthogonal bidirectional joint assembly is capable of performing yaw and pitch movements; the coaxial telescopic sleeve assembly is nested inside the dual-axis orthogonal bidirectional joint assembly and is driven by a linear drive unit. The inlet edge of the inter-ring Coanda guide bridge (40) is positioned above the tangent of the nozzle trajectory when the three-degree-of-freedom vector nozzle mechanism (34) is at its maximum elongation, and the working surface of the inter-ring Coanda guide bridge (40) is a convex surface with a radius of curvature that varies along the airflow direction.

3. The coaxial counter-rotating double-ring ducted vector propulsion system according to claim 1, characterized in that, The low-speed rotor (22) in the outer ring main lift propulsion group (20) rotates in the opposite direction to the high-speed rotor (32) in the inner ring attitude control propulsion group (30); The outer ring main lift propulsion group (20) provides low-frequency high thrust, and the inner ring attitude control propulsion group (30) provides high-frequency attitude torque; The distributed flight control system (50) establishes a communication connection with the outer ring main lift propulsion group (20), the inner ring attitude control propulsion group (30), and the three-degree-of-freedom vector nozzle mechanism (34).

4. A coaxial counter-rotating double-ring ducted vector propulsion system according to claim 2, characterized in that, A physical safety clearance is maintained between the three-degree-of-freedom vector nozzle mechanism (34) and the inter-ring Coanda guide bridge (40); When the coaxial telescopic sleeve assembly is in the retracted position, the nozzle end face of the three-degree-of-freedom vector nozzle mechanism (34) is physically isolated from the inter-ring Coanda guide bridge (40), and the airflow is directly discharged into the atmosphere; Only when the coaxial telescopic sleeve assembly extends to its maximum stroke and the yaw angle meets the tangential condition, the nozzle end face extends into the boundary layer intake zone of the inter-annular Coanda guide bridge (40), triggering the wall attachment effect.

5. A coaxial counter-rotating double-ring ducted vector propulsion system according to claim 1, characterized in that, The fuselage platform (10) has a ring truss structure; The number of outer ring duct units (21) and inner ring duct units (31) are equal and they are evenly and alternately distributed along the circumference. The diameter of the low-speed rotor (22) in the outer ring duct unit (21) is larger than the diameter of the high-speed rotor (32) in the inner ring duct unit (31).

6. A fault-tolerant control method for a coaxial counter-rotating double-ring ducted vector propulsion system, characterized in that, The method applied to the coaxial counter-rotating dual-ring ducted vector propulsion system according to any one of claims 1-5 includes the following steps: The status monitoring module (51) collects the operating data of the outer ring main lift propulsion group (20) and transmits it to the fault diagnosis module (52); The fault diagnosis module (52) determines whether the outer ring duct unit (21) has a thrust loss fault based on the operating data. If it is determined that a certain outer ring duct unit (21) has a thrust loss fault, the outer ring duct unit (21) that has a thrust loss fault is marked as a fault unit and the index number of the fault unit is locked. The topology reconstruction module (53) searches for an inner ring culvert unit (31) that is spatially adjacent to the fault unit as a rescue unit based on the index number; The distributed flight control system (50) controls the three-degree-of-freedom vector nozzle mechanism (34) of the rescue unit to perform axial extension and deflection actions, and constructs aerodynamic path, fluid coupling path and fluid-aerodynamic link through the inter-ring Coanda guide bridge (40); The control allocation solution module (54) reorganizes the control allocation matrix, isolates the weights of the faulty units, and calculates the compensation control commands of the rescue units based on the ejector gain model.

7. The fault-tolerant control method for a coaxial counter-rotating double-ring ducted vector propulsion system according to claim 6, characterized in that, The specific steps by which the fault diagnosis module (52) determines whether the outer ring duct unit (21) has experienced a thrust loss fault based on the operating data are as follows: Using a dynamic system state observer, the observation residual between the actual speed and the theoretical speed of the high power density drive motor (23) is calculated in real time; When the observation residual continuously exceeds the safety allowable threshold within the judgment time window, or when the current backflow is detected to return to zero, the corresponding outer ring culvert unit (21) is judged to have an irreversible fault.

8. The fault-tolerant control method for a coaxial counter-rotating double-ring ducted vector propulsion system according to claim 6, characterized in that, The step of searching for the inner ring culvert unit (31) adjacent to the fault unit space as the rescue unit specifically includes: Based on the ring topology, the modulo operation is used to calculate the number of the first rescue unit located on the counterclockwise side of the fault unit and the number of the second rescue unit located on the clockwise side of the fault unit, respectively. The first rescue unit and the second rescue unit are the inner ring duct unit (31) for the flow field reconstruction to be performed.

9. The fault-tolerant control method for a coaxial counter-rotating double-ring ducted vector propulsion system according to claim 8, characterized in that, The specific steps of the distributed flight control system (50) controlling the three-degree-of-freedom vector nozzle mechanism (34) of the rescue unit to perform axial extension and deflection actions include: The three-degree-of-freedom vector nozzle mechanism (34) of the first rescue unit and the second rescue unit extends to the maximum stroke position; The three-degree-of-freedom vector nozzle mechanism (34) of the first rescue unit is driven to deflect the target coupling angle in a clockwise direction, and the three-degree-of-freedom vector nozzle mechanism (34) of the second rescue unit is driven to deflect the target coupling angle in a counterclockwise direction, so that the two jets converge tangentially on the surface of the inter-ring Coanda guide bridge (40) below the fault unit.

10. The fault-tolerant control method for a coaxial counter-rotating double-ring ducted vector propulsion system according to claim 6, characterized in that, In the step of calculating the compensation control command of the rescue unit based on the ejector gain model, the ejector gain model is established in the following manner: When calculating the equivalent thrust generated by the rescue unit, a dimensionless ejector gain coefficient is introduced to correct the original aerodynamic thrust; The ejector gain coefficient is positively correlated with the ratio of the ejected area of ​​the outer ring duct unit (21) to the jet outlet area of ​​the inner ring duct unit (31), and is used to characterize the additional thrust gain obtained by entraining ambient air by fluid.

Images