A multi-node distributed aircraft attitude cooperative control method

Abstract

The present application relates to the technical field of aircraft attitude cooperative control, and discloses a kind of multi-node distributed aircraft attitude cooperative control method, comprising: obtaining the real-time attitude data of each node in aircraft cluster and establishing communication topology structure, the attitude momentum deviation between adjacent nodes is calculated to build momentum coordination constraint field;Extract the output torque of each node in torque execution unit and angular acceleration feedback data, calculate real-time response ratio and determine real-time aerodynamic stiffness eigenvalue;Introduce virtual rack bounce compensation mechanism, adjust the control gain coefficient of momentum coordination constraint field according to real-time aerodynamic stiffness eigenvalue, solve correction torque vector and superimpose correction to basic attitude control quantity, the present application establishes the adaptive mapping of environmental load and logical constraint, eliminates the phase lag accumulation generated by nonlinear aerodynamic load, realizes the high-precision cooperation of multiple aircraft nodes in non-contact state.

Description

Technical Field

[0001] This invention relates to a multi-node distributed collaborative control method for aircraft attitude, belonging to the field of collaborative control technology for aircraft attitude. Background Technology

[0002] Current attitude coordination in multi-node distributed aircraft formations typically relies on wireless communication links between nodes. Attitude synchronization is achieved through the execution of state consensus algorithms. Under normal operating conditions, each node maintains a relatively stable flight configuration by periodically exchanging attitude parameters and control commands. Since the nodes are in a non-contact state in physical space, attitude coordination is essentially a transmission of logical information. Under high dynamic load conditions, this coupling mechanism based on logical information lags behind real-time disturbances generated by the physical environment. Measurement deviations generated by sensors and time delay jitter in wireless communication channels cause unavoidable phase delays in information interaction between nodes. When a specific node is subjected to sudden torque interference, due to the lack of immediacy of physical constraints in logical connections, attitude deviations accumulate within the formation, resulting in wave-like response lag in the system.

[0003] Besides physical lag at the hardware level, existing control frameworks also have shortcomings in meeting the high-precision synchronization requirements of extreme operating conditions. For example, Chinese invention patent CN112363535B discloses a distributed cooperative control method for leader-follower multi-aircraft, which introduces a dynamic triggering mechanism to reduce communication frequency and alleviate bandwidth burden. The core logic is based on a state-following mode based on an ideal dynamic model. Such methods focus on the convergence of logic-level parameters and do not fully consider the constraints of surface energy fluctuations of torque actuators and real-time aerodynamic stiffness characteristic value jumps on the physical execution links. When encountering nonlinear aerodynamic load impacts, it is difficult to solve the physical load mismatch problem caused by non-contact state logic lag. To alleviate such response lag, the industry usually tries to increase communication bandwidth or introduce feedforward prediction algorithms. However, increasing the communication frequency is limited by the physical capacity of the airborne wireless channel, and the execution efficiency of complex prediction algorithms on low-power microprocessors is limited. The resulting computation time will introduce additional phase lag. In addition, existing control methods usually use preset static gain parameters, which cannot adapt to dynamically changing aerodynamic loads, causing the actuator to frequently saturate commands when dealing with nonlinear interference.

[0004] Therefore, the technical problem to be solved by this invention is how to construct a momentum deviation stress balance model between adjacent nodes, convert the parameter alignment at the logical level into energy state adjustment at the physical level, eliminate the phase hysteresis caused by communication delay, and improve the coordination accuracy of heterogeneous nodes. Summary of the Invention

[0005] To address the problems mentioned in the background art, the technical solution of the present invention is as follows: A multi-node distributed aircraft attitude cooperative control method, comprising the following steps:

[0006] Step S101: The state perception unit acquires the real-time attitude data of each aircraft node in the multi-aircraft cluster and establishes the communication topology between each aircraft node.

[0007] Step S102: Calculate the attitude momentum deviation between adjacent aircraft nodes based on the real-time attitude data of each aircraft node, and establish the feedback gain correlation between each aircraft node based on the attitude momentum deviation to construct a momentum coordination constraint field that maps attitude disturbances.

[0008] Step S103: Extract the output torque of the actuator of the torque actuator unit in each aircraft node and the angular acceleration feedback data of the corresponding aircraft node in real time. By calculating the real-time response ratio between the output torque of the actuator and the angular acceleration feedback data, determine the real-time aerodynamic stiffness characteristic value that characterizes the influence of the current atmospheric load on the aircraft body.

[0009] Step S104: Introduce a virtual machine frame bounce compensation mechanism. Based on the real-time aerodynamic stiffness characteristic value, synchronously adjust the control gain coefficient of the momentum coordination constraint field so that the momentum coordination constraint field exhibits physical anti-disturbance characteristics that match the current atmospheric load level. Based on the product of the control gain coefficient and the attitude momentum deviation, calculate the correction torque vector used to offset phase hysteresis.

[0010] Step S105: The basic attitude control quantity is superimposed and corrected using the corrected torque vector to generate a cooperative control signal, and the cooperative control signal is output to the power system to offset the phase hysteresis accumulation caused by nonlinear aerodynamic loads, thereby realizing the adaptive synchronization of physical loads and control logic of the aircraft cluster in a non-contact state.

[0011] Preferably, step S102 further includes the following steps: step S201, monitoring the real-time ratio of the output torque of the actuator of each aircraft node to the angular acceleration feedback data, and identifying the dynamic response deviation generated between different aircraft nodes; step S202, within the framework of the momentum coordination constraint field, adjusting the coupling weight of each aircraft node in the momentum coordination constraint field according to the dynamic response deviation, and reducing the coupling gain of nodes with dynamic response deviations greater than a preset threshold to reduce the internal friction torque between aircraft nodes and correct the cooperative control signal.

[0012] Preferably, step S103 further includes the following steps: step S301, extracting current pulsation characteristic signals from the torque execution units of each aircraft node, and obtaining non-stationary characteristic components in the current pulsation characteristic signals; step S302, establishing a mapping model between the non-stationary characteristic components and the environmental load, using the mapping model to predict the disturbance intensity applied to the aircraft by the environmental load, and constructing a feedforward compensation command to correct the torque vector accordingly.

[0013] Preferably, in step S104, the calculation of the compensation strength of the momentum compatibility constraint field follows the following formula: ,in, This represents the compensation strength value of the momentum-coordinated constraint field. The control gain coefficient after real-time correction for the virtual machine rack bounce compensation mechanism. This represents the attitude momentum deviation between adjacent spacecraft nodes.

[0014] Preferably, in step S101, the communication topology adopts a dynamic directed graph model based on link availability, and the communication topology is dynamically updated as the relative displacement between aircraft nodes changes.

[0015] Preferably, in step S104, the real-time aerodynamic stiffness characteristic value reflects the gradient of the drag torque generated by the external atmospheric load on the body as the attitude angle changes, and the control gain coefficient is synchronously increased as the real-time aerodynamic stiffness characteristic value increases.

[0016] Preferably, the method further includes the following steps: Step S701, when the communication link between adjacent aircraft nodes is detected to be disconnected, the disturbed node extracts historical cooperative motion features and generates a spatiotemporal cooperative trajectory fingerprint; Step S702, the autonomous prediction mode is activated, and the disturbed node predicts the virtual tension parameters in the momentum coordination constraint field based on the spatiotemporal cooperative trajectory fingerprint in order to maintain the logical pose of the disturbed node in the communication topology.

[0017] Preferably, the method further includes the following steps: Step S801, after the communication link is restored, the prediction parameters and real-time acquired data in the autonomous prediction mode are smoothed by the attenuation operator in order to eliminate the dynamic step during the control mode switching process.

[0018] Preferably, in step S105, the basic attitude control quantity is generated by a control law based on a consensus protocol, and the modified torque vector acts as a feedforward component on the power distribution link of the attitude control loop.

[0019] Preferably, the output frequency of the cooperative control signal is not less than 0.5 times the sampling frequency of the state perception unit, and the cooperative control signal drives the torque execution unit of each aircraft node in the form of a pulse width modulation signal.

[0020] Compared with the prior art, the beneficial effects of the present invention are:

[0021] 1. In the cooperative attitude control of aircraft, by mapping the angular displacement gradient and angular momentum difference between adjacent nodes as cooperative coupling tension parameters, a virtual stress field simulating physical entity constraints is constructed. This changes the traditional distributed system's synchronization method that relies solely on logical information exchange. When facing external disturbances, the aircraft formation can achieve automatic energy state balance through the momentum deviation stress model. This allows local disturbances to be canceled out by virtual tension before displacement deviation occurs, eliminating wave-like phase hysteresis caused by discrete communication delays and improving attitude synchronization accuracy in high dynamic environments.

[0022] 2. A virtual looper energy storage buffer mechanism is introduced to solve the command mismatch problem of the actuator at the physical saturation boundary. When the control torque reaches the preset threshold, the system extracts the residual attitude deviation and converts it into virtual looper height. The instantaneous spatial dimension deviation is converted into logical delay storage in the time domain, avoiding frequent switching of the actuator in the high-frequency saturation range. While protecting the life of the hardware structure, it ensures the integrity of the formation topology under extreme conditions and realizes the dynamic alignment of control commands and physical response bandwidth.

[0023] 3. By using execution sensitivity to inversely obtain the virtual execution stiffness coefficient of each node, asymmetric load distribution among heterogeneous nodes is realized. By monitoring the ratio between command torque and angular acceleration feedback value, the dynamic performance deviation of different nodes caused by aging or environmental differences is identified in real time. Then, the weight of each node in the virtual tension field is dynamically adjusted. This adaptive adjustment method reduces the internal friction torque caused by inconsistent response between nodes, optimizes the overall energy utilization efficiency of the cluster, and solves the problem of formation center of gravity drift caused by individual differences. Attached Figure Description

[0024] Figure 1 This is a flowchart of the adaptive attitude control of an aircraft that integrates aerodynamic stiffness sensing and momentum coordination according to the present invention.

[0025] Figure 2 This is a diagram of the closed-loop interaction architecture between the momentum coordination constraint field and the spacecraft node, which integrates a virtual compensation mechanism according to the present invention. Detailed Implementation

[0026] The present invention will be further described in detail below with reference to the accompanying drawings, but the embodiments of the present invention are not limited thereto; the following embodiments are intended to explain the present invention and are not intended to limit the scope of protection of the present invention.

[0027] This invention provides a multi-node distributed aircraft attitude cooperative control method, based on a momentum coordination mechanism simulating a physical continuum. It acquires the motion parameters of each node in the aircraft cluster through a state-sensing unit, establishing a dynamic communication topology between nodes. Based on this, it constructs a momentum coordination constraint field to characterize the strength of logical constraints by calculating the attitude momentum deviation between adjacent nodes. Furthermore, it utilizes a virtual machine frame bounce compensation mechanism to sense the aerodynamic impedance generated by atmospheric environmental loads, dynamically adjusting the control gain. Finally, it achieves closed-loop synchronous adjustment of physical loads and control logic through current feedback from the actuators. During distributed formation flight, the aircraft nodes are in a non-contact state in space, and attitude coordination relies on discrete wireless communication links, causing the transmission of logical information to lag behind real-time disturbances generated by the physical environment. To address this obstacle, this invention's solution, in its execution steps... At that time, the inertial measurement unit and global positioning system onboard the aircraft are used as state awareness units to... The system collects real-time attitude data such as Euler angles, angular velocities, and triaxial accelerations from each spacecraft node at a sampling frequency. Based on the relative distance between nodes and the signal link quality, the system establishes a communication topology based on a dynamic directed graph model, and maintains and updates it in real time as the displacement between nodes changes, providing a definite information flow reference for momentum coordination. Addressing the response hysteresis caused by the lack of immediate physical constraints on logical connections, the present invention's solution... The paper introduces a construction procedure for a momentum coordination constraint field; extracts the attitude angle vector and angular momentum vector of adjacent spacecraft nodes in the same inertial coordinate system, and calculates the difference between them, i.e., the attitude momentum deviation; defines the angular displacement gradient between adjacent nodes as a virtual stretching amount, and uses a preset mapping function to convert the virtual stretching amount into a cooperative coupling tension parameter, which is used to simulate the stress state between physical entities; when the attitude deviation between adjacent nodes increases, the cooperative coupling tension parameter increases synchronously, generating a virtual restoring torque that returns the nodes to a synchronized state. This procedure converts discrete logical alignment into a continuous energy state balance process, reducing the cumulative phase error caused by communication delay.

[0028] To address the challenge of nonlinear aerodynamic loads changing drastically with environmental variations, the present invention provides a solution in the following steps: The system configures a solution procedure for real-time aerodynamic stiffness characteristic values; it monitors the output parameters of the aircraft's torque actuator, extracts the motor drive current signal, and converts it into the actuator output torque. Simultaneously, it acquires angular acceleration feedback data measured by the gyroscope. In the collection Previously, an adaptive variable window long cutoff frequency filtering procedure was performed on the angular rate signal acquired by the sensor to eliminate high-frequency signal oscillations introduced by random noise from the low-cost inertial measurement unit. The system then adjusted the current sampling frequency accordingly. The ratio of the noise power spectral density to the smoothing window length is adjusted to map the first-order difference of the filtered angular velocity as follows: Using the aircraft's pre-set physical rotational inertia matrix The environmental resistance offset is corrected, and the final system is based on the formula. To obtain real-time aerodynamic stiffness characteristic values ​​that characterize the impact of current atmospheric environmental loads on the aircraft. ,in, This represents the real-time aerodynamic stiffness characteristic value, in units of... , The torque output by the actuator, in units of , This is angular acceleration feedback data, in units of , Sampling frequency, in units of , This is the physical moment of inertia matrix, in units of... To balance the trade-off between system response speed and stability, the present invention provides the following steps: The system implements a virtual machine frame bounce compensation mechanism; and establishes real-time aerodynamic stiffness characteristic values. With control gain coefficient The monotonic mapping relationship between them enables dynamic correction of the momentum coordination constraint field compensation strength; when real-time aerodynamic stiffness eigenvalues ​​are detected... When the gain increases, the system synchronously increases the control gain coefficient. The system calculates the compensation strength value of the momentum-coordinated constraint field. The calculation formula is as follows: ,in, This represents the compensation strength value of the momentum-coordinated constraint field, in units of... ; To control the gain coefficient; The attitude momentum deviation between adjacent aircraft nodes is used to calculate the corrective torque vector used to offset phase hysteresis. This process uses the instantaneous feedback of aerodynamic impedance to adjust the stiffness of the virtual connectors, so that the entire formation exhibits the characteristics of a flexible continuum with environmental awareness.

[0029] Real-time aerodynamic stiffness eigenvalues With control gain coefficient The monotonic mapping relationship between them was determined through an offline calibration program. The wind tunnel test environment simulated the airflow impact at different Reynolds numbers, and the output torque of the actuator was recorded. With angular acceleration feedback data The steady-state ratio is determined, and the maximum feedback gain that maintains the system's attitude stability is measured. Based on this, a system is constructed... A control gain lookup table for index values ​​is stored in the onboard processor's non-volatile memory. This lookup table has 21 preset index points, with an index step size of 5 Nm / radian per square second, covering an aerodynamic stiffness range from 0 to 100 units. During real-time retrieval, if the currently calculated characteristic value falls between two index points, the system extracts the gain value corresponding to the lower index and the gain value corresponding to the higher index, calculates the linear percentage of the characteristic value within the step size range, and then performs a weighted average of the two gain values ​​to calculate the current control gain coefficient. This ensures that gain adjustment remains continuous during stiffness jumps. In real-time flight, the system, based on the logic determined in steps S101 to S105, utilizes the control gain calculated in step S103... Retrieve from the lookup table and output the corresponding value using linear interpolation. To adjust the virtual work stiffness of the momentum-coordinated constraint field in real time, so that the control commands are dynamically adapted to the aerodynamic load intensity, wherein, This represents the real-time aerodynamic stiffness characteristic value. To control the gain coefficient, To output torque to the actuator, This is angular acceleration feedback data.

[0030] In scenarios where the actuator faces physical output limits that may lead to command saturation, this invention introduces a virtual looper energy storage buffer procedure. During the generation of the final coordinated control signal, the system executes a linear mapping procedure from torque command to PWM duty cycle and acquires the feedback current signals of each aircraft node in real time. To determine whether the actuator has entered the physical saturation region; if the system detects... To achieve the maximum stall current of the propulsion motor And the duration exceeds During each sampling period, the virtual looper energy storage buffer procedure is automatically activated, which locks the current output duty cycle and reduces the control gain coefficient. In this manner, the torque increment exceeding the physical output limit is stored in a virtual loop register composed of first-order low-pass inertial elements, and mapped to a virtual loop height parameter to achieve stress release; After falling back to the linear operating range, the system smoothly releases the virtual loop height parameter according to a preset exponential decay function, thus smoothly eliminating the accumulated deviation. This is a feedback current signal, in units of... , This is to control the gain coefficient.

[0031] The virtual looper energy storage buffer procedure is implemented by integrating a first-order low-pass inertial element integral register into the control law. The system sets the upper limit of the linear operating range of the actuator to the maximum stall current of the propulsion motor. Feedback current signal Upon reaching the upper limit, the system locks the output duty cycle and stops increasing the physical execution torque. Torque command deviations exceeding the physical limit are input to the integration register and converted into virtual loop height parameters for logical storage, pending... Once the torque drops below the upper limit, the system releases the accumulated deviation in the integral register according to a preset exponential decay function, and smoothly adds the torque increment to the cooperative control signal. The decay constant of the exponential decay function is set to the peak response time of the power system. This eliminates the accumulated error of command saturation and reduces the structural impact caused by dynamic step jumps. For feedback current signal; to address the interference of surface energy state fluctuations in the propulsion system on attitude stability, the present invention provides the following steps: to The system implements a feedforward compensation procedure based on current ripple characteristics. It extracts the current ripple characteristic signal of the motor from the ESC unit and uses a Fast Fourier Transform (FFT)-based method to obtain the non-stationary characteristic components in the current signal that are inconsistent with the reference power consumption. At the hardware implementation level, the system sets the current sampling frequency of the ESC unit to 1000Hz, continuously capturing 1024 sampling points each time to construct an analysis frame, which is then stored in a circular buffer. The overlap rate of the sliding window is set to 50%, meaning that the Fourier transform logic is triggered once every 512 new data points received in the buffer. The system focuses on extracting frequency ripple characteristics. The sum of energy amplitudes within a 20Hz range around the center frequencies of the third and fifth harmonics is defined as the non-stationary characteristic intensity of the current pulsation characteristic signal. This serves as the original input for subsequent disturbance intensity determination. Using a pre-calibrated dynamic transfer function model, the slope of the variation of the non-stationary characteristic component is mapped to the predicted disturbance intensity value exerted on the body by the environmental load. Based on this, a feedforward compensation command is constructed. The torque vector is pre-corrected before the attitude deviation is formed, and the disturbance is absorbed by utilizing the leading characteristics of energy overflow in the power system, thus shortening the system's recovery time under pulsed airflow interference.

[0032] The mapping model between non-stationary characteristic components and environmental loads is composed of a pre-calibrated polynomial coefficient matrix. Phase current signals from the torque actuator under different turbulence intensities are collected from the electronic control unit (ECU). The distribution characteristics of incoherent harmonic energy related to the motor's fundamental frequency are extracted using Fast Fourier Transform (FFT). A functional relationship is established between the harmonic energy amplitude and the external disturbance torque. The coefficients within the model are determined using a least-squares fitting algorithm. In step S302, the system extracts current pulsation characteristic signals in real time and inputs them into the model. The system calculates the predicted value of the equivalent disturbance intensity exerted by the environmental load on the machine body. This predicted value serves as the feedforward term correction torque vector, and active torque counterbalancing is performed before the attitude angle undergoes physical deflection. Considering the possibility of instantaneous communication link disconnection due to complex electromagnetic environments, the present invention's solution in step... to The invention establishes a spatiotemporal cooperative trajectory fingerprint and autonomous prediction procedure. When the monitoring unit determines that the communication link has been logically interrupted through packet loss rate statistics, the disturbed node activates a virtual memory mode to extract historical cooperative motion features such as the attitude change rate and first derivative trend of adjacent nodes within a preset time window before the interruption, and constructs a spatiotemporal cooperative trajectory fingerprint. The disturbed node executes a polynomial extrapolation algorithm based on the spatiotemporal cooperative trajectory fingerprint to generate predicted tension parameters. After the communication link is restored, the system eliminates the residual between the predicted value and the real-time acquired value through a linear attenuation operator to ensure the continuity of torque during the control mode switching process and to ensure the morphological integrity of the formation under the link failure condition. The steps of this invention are as follows: In this process, the generated corrected torque vector is superimposed on the basic attitude control quantity generated by the consensus protocol to generate the final cooperative control signal; the cooperative control signal drives the actuator in the form of a pulse width modulation signal; during this process, the system acquires the feedback current signal of the actuator. The system utilizes feedback current signals to evaluate the execution effect of dynamic torque compensation vectors and corrects the internal weight parameters of stress compensation models based on execution deviations. This constructs a closed loop from environmental perception and logical mapping to physical execution, enabling high-precision attitude coordination among multiple aircraft nodes in a non-contact state.

[0033] Example 1: At an altitude of [missing information - likely an altitude range], swarm aircraft And the flight speed is During high-altitude cruise operations, the formation encountered peak wind speeds of Sudden nonlinear airflow disturbances caused the navigator node to... Internal generation The pitch angle deviation, due to drastic changes in physical load, causes phase lag in the logical synchronization commands between the slave and lead aircraft nodes; to address synchronization failures caused by instantaneous load impacts, the state perception unit uses... The sampling frequency is used to obtain the pitch angle displacement gradient of the navigator node and input it into the momentum coordination constraint field. Logical deviations are converted into tension feedback simulating physical connections by calculating the cooperative coupling tension parameters. Simultaneously, the torque execution unit inside each node monitors the motor output current in real time and uses the formula... Calculate the real-time aerodynamic stiffness eigenvalue ,in, This represents the real-time aerodynamic stiffness characteristic value, in units of... , The torque output by the actuator, in units of , The angular acceleration feedback data is obtained from the gyroscope onboard the aircraft, and the unit is angular acceleration. .

[0034] when As the virtual machine frame rises with increasing atmospheric disturbance, the bounce compensation mechanism is based on... Adjust the control gain coefficient in real time according to the changing trend. This aligns the virtual connection stiffness between nodes with the current aerodynamic impedance strength. When the actuator torque output approaches the linear margin threshold, the virtual loop energy storage buffer procedure stores the residual attitude deviation in a virtual loop register composed of first-order low-pass inertial elements, mapping it to virtual loop height parameters to achieve stress release. By closed-loop coupling of the information flow reference provided by the state perception unit and the physical feedback provided by the torque execution unit, the system reconstructs the attitude parameter alignment into a physical-level momentum balance adjustment, enabling the aircraft formation to respond to disturbances. The pitch angle deviation was converged, and the attitude synchronization accuracy of the entire formation was maintained at [value missing]. Within.

[0035] Example 2: A distributed flight simulation platform was used to perform a six-degree-of-freedom full-physical-quantity coupled experiment to verify the performance of the aircraft swarm when encountering high altitudes. And the Reynolds number is The adjustment accuracy under high-altitude flow field conditions; the flow field environment is simulated by solving the unsteady Navier-Stokes equations, and the data sampling frequency is set to [value missing]. That is, selecting the sampling period for The signal-to-noise ratio superimposed in the test signal is Gaussian white noise is used to simulate industrial electromagnetic interference, and a root mean square value is introduced. The random wind field disturbance is used as an environmental excitation to construct a complete verification chain from the input of the original physical quantity to the output of the control command. Four sets of comparative samples are set to observe the impact of different control mechanisms on the system performance. The sample of this invention adopts a complete method including momentum coordination constraint field, virtual machine frame bounce compensation mechanism and virtual loop energy storage buffer procedure. The first control group adopts the traditional control method based on the consensus protocol. The second control group removes the virtual machine frame bounce compensation mechanism based on the sample of this invention. The third control group removes the virtual loop energy storage buffer procedure. The specific experimental test data are recorded in Table 1.

[0036] Table 1: Comparison of attitude synchronization performance between the present invention sample group and each control group under different disturbance intensities

[0037]

[0038] Referring to the measurement results in Table 1, when the disturbance intensity is... to During this period, the root mean square value of the attitude synchronization deviation of the sample group of the present invention remained at [value missing]. The following; among them, in Under the disturbance conditions, the real-time aerodynamic stiffness characteristic value of the sample group of this invention is calculated. for The control gain coefficient is dynamically increased based on this value. This aligns the virtual stiffness of the momentum-coordinated constraint field with the aerodynamic impedance; in contrast, control group two, lacking an adaptive compensation mechanism for aerodynamic impedance, exhibits a synchronization deviation that increases to [missing information]. This confirms the role of the virtual machine frame bounce compensation mechanism in improving attitude stability when dealing with nonlinear aerodynamic load jumps; when the disturbance intensity experienced by the prototype of this invention is further increased to At that time, the system observed that the root mean square value of the attitude synchronization deviation increased to The adjustment response time has also been extended to The physical factor contributing to this performance change lies in the output torque of the actuator. Reached The physical saturation extreme value caused the residual attitude deviation stored in the virtual loop register to exceed the buffer limit; the experiment introduced [something] into the communication link. The instantaneous interruption condition verified the reliability of the spatiotemporal cooperative trajectory fingerprint and autonomous prediction procedure. Measurement data showed that the disturbed node utilized the preceding sequence during the link loss. Extrapolating the first derivative of the attitude angle at each sampling point for prediction, the pose drift is only... This indicates that the momentum coordination constraint field eliminates phase hysteresis by reconstructing discrete logic into physical-level energy state regulation.

[0039] Example 3: This example combines Figures 1 to 2 This paper describes a multi-node distributed cooperative attitude control method for aircraft, such as... Figure 1 As shown, in step S101, the state perception unit acquires the real-time attitude data of each aircraft node in the multi-aircraft cluster and establishes a dynamic communication topology based on link availability. In step S102, the attitude momentum deviation between adjacent nodes is calculated, feedback gain correlation is established, and a momentum coordination constraint field that can map attitude disturbances is constructed accordingly. Then, in step S103, the output torque and angular acceleration feedback data of the actuator are extracted, the real-time response ratio is calculated, and the real-time aerodynamic stiffness characteristic value representing the influence of the current atmospheric load is determined. After that, in step S104, a virtual machine frame bounce compensation mechanism is introduced, the control gain coefficient is adjusted according to the real-time aerodynamic stiffness characteristic value, and the correction torque vector used to offset phase hysteresis is calculated. Finally, in step S105, the correction torque vector is used to superimpose and correct the basic attitude control quantity, generate a cooperative control signal output to the power system, and realize physical load adaptive synchronization in non-contact state.

[0040] like Figure 2As shown, the system uses a momentum coordination constraint field located in the logical domain as its core, serving as a virtual tension and energy balance center. Internally, it integrates a virtual machine frame bounce compensation module responsible for dynamic control gain adjustment, and a virtual loop energy storage buffer module for physical saturation elimination. Its core architecture forms a closed-loop interaction with peripheral aircraft nodes. Aircraft node A is equipped with a state awareness unit including IMU / GPS and a torque execution unit including a motor / ESC, and deploys a spatiotemporal cooperative trajectory fingerprint algorithm. Aircraft node B has the same hardware configuration and is equipped with aerodynamic stiffness calculation capabilities. Aircraft node C has the same hardware configuration and is equipped with feedforward compensation commands. During operation, the system feeds back attitude momentum deviation and real-time aerodynamic stiffness characteristics as input parameters to the momentum coordination constraint field. The calculated corrected torque vector acts on aircraft node C to resist nonlinear disturbances and physical impacts caused by atmospheric loads.

[0041] Example 4: When performing formation cruise missions in complex urban conditions, the aircraft swarm faces the tube effect disturbance caused by tall buildings and the physical obstruction of radio signals by buildings; when the lead aircraft node is in When subjected to shear winds with frequently changing directions, the state perception unit observes a nonlinear transition in the angular displacement gradient between the navigator node and the slave node; the physical output boundary of the actuator is determined through a pre-executed offline calibration program, and the stall torque of the propulsion motor under maximum operating voltage is measured on the test bench. , and select of As the linear margin threshold of the actuator; the system at discrete time steps The incremental momentum coordination constraint algorithm is then executed; this algorithm is used to determine the attitude momentum deviation at the current moment. Update the compensation strength value of the momentum compatibility constraint field The specific calculation formula is as follows: ,in: For the first The compensation intensity value of the momentum-coordinated constraint field for each sampling period, in units of ; The compensation intensity value for the previous sampling period, in units of ; To control the gain coefficient; This refers to the attitude momentum deviation between adjacent spacecraft nodes collected at the current moment. Through this incremental update procedure, the system achieves continuous and smooth adjustment of the virtual stress state, avoiding abrupt changes in logical commands under large disturbances, and linking the adjustment commands with the physical aerodynamic stiffness characteristic values. Dynamic matching accuracy control within The calculation error is within the acceptable range.

[0042] When the formation enters a building-shaded area, causing an increase in the packet loss rate of the communication link, the monitoring unit activates the counting logic judgment mechanism; the system sets a packet loss counter threshold. for When continuous When the signal feedback for a sampling period is lost, the disturbed node is forced to switch to virtual memory mode; at this time, the disturbed node retrieves the previous sequence stored in the circular queue. The attitude angular rate data from each sampling point are used to construct a fifth-order prediction polynomial using the sliding window least squares method, thereby generating a spatiotemporal cooperative trajectory fingerprint; the prediction output term of the spatiotemporal cooperative trajectory fingerprint... The calculation process is as follows: ,in: for The tension parameters predicted at any given time; The coefficients are obtained through least squares fitting, with subscripts. Values to ; The predicted time offset is calculated from the time of loss, in units of During the prediction process, the system judges the convergence of the predicted trajectory by monitoring the second derivative of the prediction residuals. If the slope of the residual fluctuation exceeds the preset safety criterion, the system reduces... The step size enters a low-dynamic hold state to ensure that the formation maintains its geometric configuration and the overall attitude coordination deviation of the cluster remains within a certain range until the communication link is restored. .

[0043] Example 5: During the ground standby condition before takeoff of the swarm aircraft, the system initiates a deviation calibration procedure to eliminate the influence of sensor noise on the initial state of the momentum coordination constraint field; the state perception unit collects the attitude deviation of each aircraft node in a stationary state and locates the momentum balance reference point based on the angular displacement gradient between adjacent aircraft nodes; during this process, the system locks the zero position of the momentum coordination constraint field so that the output value of the cooperative coupling tension parameter when there is no external aerodynamic load is zero. Specifically, the system collects attitude data and uses the arithmetic mean as a static benchmark. If the standard deviation of the attitude pulsation of a node exceeds a certain threshold, the system will take action. Then the ESC unit corrects the current offset of the actuator, aligning the physical state of the node with the logic reference.

[0044] When the system is calibrated to map the sensitivity of the momentum-coordinated constraint field, the amplitude injected into the pitch axis of the navigator node is the output torque of the actuator. The system receives pulse signals and monitors the overall stress response of the cluster; the system relies on real-time aerodynamic stiffness characteristic values. The response curve is used to correct the internal weights of the stress compensation model, and the following phase of each slave vehicle node is calculated. If the phase hysteresis exceeds a preset threshold, the system increases the control gain coefficient in fixed steps. Until the feedback current signal The virtual stretching amount exhibits a monotonic change that is positively correlated with the amount of stretching, among which This represents the real-time aerodynamic stiffness characteristic value, in units of... , To control the gain coefficient, This is a feedback current signal, in units of... This process establishes a linear mapping relationship between the physical execution layer and the virtual stress layer to eliminate deviations in formation geometry caused by differences in individual dynamic characteristics.

[0045] Example 6: In the case of a heterogeneous formation of multiple rotorcraft performing a high-altitude canyon inspection mission, the nodes exhibit inconsistent open-loop response characteristics due to differences in airframe inertia and propulsion efficiency. When the system detects yaw drift of the lead aircraft node in the canyon's narrow wind field, the system executes a pre-response consistency calibration procedure to eliminate the adjustment hysteresis caused by differences in airframe physical properties. The steady-state output torque of the propulsion motor under different operating voltages is pre-measured using a controlled ground test environment to establish the actuator output torque. The system uses a mapping table between the control commands and the actual control commands to calculate predicted angular acceleration values ​​based on the measured three-axis rotational inertia matrix of the airframe, thus providing real-time aerodynamic stiffness characteristic values. The solution provides a definite physical quantity reference.

[0046] The system extracts the real-time response ratio of each aircraft node under different airflow velocities through frequency sweep excitation testing, and maps the statistical variance of the response deviation to the initial weight distribution of the momentum-coordinated constraint field, thereby determining the control gain coefficient. The adaptive evolution trajectory; during this process, the energy storage characteristics of the virtual loop register are calibrated, and the peak response time of the power system is selected. of The exponential decay constant of the virtual loop height parameter is used to balance the contradiction between the residual deviation elimination rate and the power consumption of the propulsion motor; when the aircraft cluster encounters an event lasting for a period of time... When subjected to external turbulent impacts, the system adjusts the virtual tension of the momentum coordination constraint field according to the parameter matrix preset in the calibration procedure, so that the feedback current signal... Closely track the rate of change of attitude momentum deviation, where, This represents the real-time aerodynamic stiffness characteristic value, in units of... , The torque output by the actuator, in units of , This is angular acceleration feedback data, in units of , This is a feedback current signal, in units of... , The peak response time of the power system, in units of Test data shows that the synchronization delay of heterogeneous formations under strong disturbance conditions is reduced to [missing information]. Within.

[0047] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0048] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A multi-node distributed aircraft attitude cooperative control method, characterized in that, Comprise the following steps: Step S101, the state perception unit acquires the real-time attitude data of each aircraft node in the multi-aircraft cluster, and establishes the communication topology structure between each aircraft node; Step S102, according to the real-time attitude data of each aircraft node, the attitude momentum deviation between adjacent aircraft nodes is calculated, and the feedback gain correlation between each aircraft node is established according to the attitude momentum deviation, and the momentum coordination constraint field mapping the attitude disturbance is constructed; Step S103, real-time extraction of the execution mechanism output torque of the torque execution unit in each aircraft node and the angular acceleration feedback data of the corresponding aircraft node, the real-time response ratio between the execution mechanism output torque and the angular acceleration feedback data is calculated, and the real-time aerodynamic stiffness characteristic value representing the influence of the current atmospheric environment load on the body is determined; Step S104, a virtual rack bounce compensation mechanism is introduced, the control gain coefficient of the momentum coordination constraint field is adjusted synchronously according to the real-time aerodynamic stiffness characteristic value, so that the momentum coordination constraint field presents physical anti-disturbance characteristics matched with the current atmospheric environment load magnitude, and based on the product term of the control gain coefficient and the attitude momentum deviation, the correction torque vector for offsetting phase lag is calculated; Step S105, the correction torque vector is used to superimpose and correct the basic attitude control quantity, generate a cooperative control signal, and output the cooperative control signal to the power system to offset the phase lag accumulation caused by nonlinear aerodynamic load, realize the adaptive synchronization of physical load and control logic of the aircraft cluster in the non-contact state.

2. The multi-node distributed aircraft attitude cooperative control method according to claim 1, characterized in that, In step S102, further comprising the following steps: step S201, monitoring the real-time proportion of the execution mechanism output torque and the angular acceleration feedback data of each aircraft node, identifying the dynamic response deviation generated between different aircraft nodes; step S202, under the framework of the momentum coordination constraint field, the coupling weight of each aircraft node in the momentum coordination constraint field is adjusted according to the dynamic response deviation, and the coupling gain of the node whose dynamic response deviation is greater than the preset threshold is reduced to reduce the internal friction torque between the aircraft nodes and correct the cooperative control signal.

3. The multi-node distributed aircraft attitude cooperative control method of claim 1, wherein, In step S103, further comprising the following steps: step S301, extracting current pulsation characteristic signal from the torque execution unit of each aircraft node, and obtaining non-stationary characteristic component in the current pulsation characteristic signal; step S302, establishing a mapping model between the non-stationary characteristic component and the environmental load, using the mapping model to predict the disturbance intensity exerted by the environmental load on the body, and constructing a feedforward compensation instruction to correct the torque vector.

4. The multi-node distributed aircraft attitude cooperative control method of claim 1, wherein, In step S104, the compensation strength calculation of the momentum coordination constraint field follows the formula: wherein, is the compensation strength value of the momentum coordination constraint field, is the control gain coefficient of the virtual rack bounce compensation mechanism after real-time correction, is the attitude momentum deviation between adjacent aircraft nodes.

5. The multi-node distributed aircraft attitude cooperative control method of claim 1, wherein, In step S101, the communication topology structure adopts a dynamic directed graph model based on link availability, and the communication topology structure is dynamically updated with the relative displacement between the aircraft nodes.

6. The multi-node distributed aircraft attitude cooperative control method of claim 1, wherein In step S104, the real-time aerodynamic stiffness characteristic value reflects the change gradient of the resistance torque generated by the external atmospheric environment load on the body with the change of the attitude angle, and the control gain coefficient is synchronously adjusted with the increase of the real-time aerodynamic stiffness characteristic value.

7. The multi-node distributed aircraft attitude cooperative control method of claim 1, wherein, Further comprising the following steps: Step S701, when the communication link between adjacent aircraft nodes is disconnected, the disturbed node extracts historical cooperative motion characteristics to generate a space-time cooperative trajectory fingerprint; Step S702, activate the autonomous prediction mode, and the disturbed node predicts the virtual tension parameter in the momentum coordination constraint field according to the space-time collaborative trajectory fingerprint, so as to maintain the logical pose of the disturbed node in the communication topology.

8. The multi-node distributed aircraft attitude cooperative control method of claim 7, wherein, Further comprising the following steps: Step S801, after the communication link is recovered, the prediction parameters in the autonomous prediction mode and the real-time collected data are smoothed by the attenuation operator, so as to eliminate the dynamic step in the control mode switching process.

9. The multi-node distributed aircraft attitude cooperative control method of claim 1, wherein, In step S105, the basic attitude control quantity is generated by a control law based on a consistency protocol, and the correction moment vector acts as a feedforward component on the power distribution link of the attitude control loop.

10. The multi-node distributed aircraft attitude cooperative control method of claim 1, wherein, The output frequency of the collaborative control signal is not less than 0.5 of the sampling frequency of the state perception unit, and the collaborative control signal drives the torque execution unit of each aircraft node in the form of a pulse width modulation signal.

Images