Tail seat type unmanned aerial vehicle vector rotor and variable phase wing cooperative control method

Abstract

The application discloses a tail seat type unmanned aerial vehicle vector rotor and variable phase wing cooperative control method, and relates to the technical field of unmanned aerial vehicle control. The application is aimed at solving the problem of insufficient anti-interference performance of the existing tail seat type unmanned aerial vehicle in the vertical take-off and landing and transition stage. The application comprises the following steps: S1, real-time attitude direction cosine matrix conversion; S2, attitude error Lie algebra mapping expression; S3, linear feedback control based on particle swarm optimization; S4, multi-axis integral term data feature extraction; S5, integral driven multi-actuator dynamic weight distribution; and S6, each actuator control signal instruction mapping. The application is aimed at realizing multi-actuator cooperative control through an integral driven weight distribution strategy under the condition of the existing sensing system technology, so that the anti-interference performance of the tail seat type unmanned aerial vehicle in the hovering and transition flight stage is significantly improved. The application is used for attitude anti-interference control of the tail seat type unmanned aerial vehicle in the vertical take-off and landing stage and the transition flight stage.

Description

Technical Field

[0001] This invention relates to a control method for an aircraft, specifically a method for coordinated control of a vector rotor and a variable-phase wing of a tail-seat unmanned aerial vehicle (UAV), belonging to the field of UAV control technology. Background Technology

[0002] Hybrid-driven vertical takeoff and landing (VTOL) drones possess multiple flight modes, combining the advantages of various drone types to better meet the increasingly complex and diverse needs of civilian missions. However, a perfectly compatible configuration scheme does not yet exist. Among them, tail-seat drones are known for their lack of redundant drive resources, theoretically representing the configuration with the optimal drive efficiency. However, they also face issues such as poor anti-interference performance during vertical takeoff and landing and transition phases, as well as insufficient flight reliability.

[0003] Existing research has focused on enhancing the anti-jamming performance of tail-mounted unmanned aerial vehicles (UAVs). Folding the wings along their span has proven effective in mitigating interference during hovering. However, the high aerodynamic torque generated by high-speed cruise flight can limit the strength of the driving torque and hinge mechanisms responsible for wing folding and unfolding. Furthermore, the transition phase of the aircraft becomes more complex, making accurate modeling and control difficult, thus increasing the control complexity during this phase. Bionic wing-like wing folding mechanisms can also reduce the frontal area during takeoff and landing, but overly complex wings are difficult to lightweight. Moreover, similar solutions only enhance stability during hovering control, failing to provide stability gains for the more crucial transition phase.

[0004] In fact, for tail-seat drones, hovering is only for a short time, used for takeoff and landing at low altitudes. More crash risks occur during the transition phase when gusts of wind are present. Especially for tail-seat drones that use only two rotors to maintain their aerodynamic shape, the control effect of aerodynamic control surfaces is limited at low airspeeds. During landing, the airflow near the aerodynamic control surfaces decreases or even reverses, making their anti-disturbance performance significantly insufficient.

[0005] Tail-mounted UAVs incorporating vector rotors have the ability to dynamically adjust the direction of propeller thrust, enabling them to maintain sufficient control torque even at low airspeeds, thus allowing for more stable takeoff, landing, and transition maneuvers. However, adverse aerodynamic effects still exist, and maintaining the aircraft's attitude by increasing control torque is not efficient, and under strong disturbances, highly coupled control torques may cause loss of control.

[0006] In summary, existing tail-mounted UAVs suffer from insufficient anti-interference performance during vertical takeoff and landing and the transition phase. Summary of the Invention

[0007] The purpose of this invention is to address the problem of insufficient anti-interference performance of existing tail-seat unmanned aerial vehicles (UAVs) during vertical takeoff and landing and transition phases. Furthermore, it provides a method for coordinated control of a vector rotor and a variable-phase wing in a tail-seat UAV.

[0008] The technical solution of this invention is: a method for coordinated control of a vector rotor and a variable-phase wing of a tail-seat unmanned aerial vehicle, comprising the following steps:

[0009] Step S1: Direction cosine matrix transformation of real-time attitude:

[0010] Convert sensor data from tail-mounted drones into rotation matrices And based on the body angular velocity Update real-time attitude;

[0011] Step S2: Lie algebraic mapping representation of attitude error:

[0012] Calculate the desired pose matrix With the actual attitude matrix error matrix It is then transformed into a Lie algebraic form of the attitude error vector through a logarithmic mapping. ; Step S3: Linear feedback control based on particle swarm optimization:

[0013] attitude error vector and body angular velocity As input, the desired control torque is output through a linear feedback controller, and the proportional, integral, and derivative parameters of the linear feedback controller are tuned using a particle swarm optimization algorithm. Step S4: Feature extraction of multi-axis integral term data:

[0014] Extract the integral terms of the linear feedback controller along the y-axis and z-axis of the body coordinate system, and calculate the proportion of the integral terms. , and the ratio coefficient between the two ; Step S5: Integral-driven dynamic weight allocation for multiple actuators:

[0015] According to the proportion of the integral items , Based on the comparison with a preset threshold, the control weights of the vector rotor mechanism and the variable phase wing mechanism are dynamically allocated, and the desired wing phase angle is generated. ; Step S6: Mapping of control signals and commands for each actuator:

[0016] Based on the desired control torque and desired wing phase angle The thrust of the left propeller was calculated. Right propeller thrust Left propeller tilt angle Right propeller tilt angle And mapped to propeller speed command .

[0017] Furthermore, in step S1:

[0018] The Euler angle rotation order of the direction cosine matrix is ​​defined as ZYX, i.e. ;

[0019] in: Here is the direction cosine matrix. It refers to the yaw angle of a tail-mounted drone during the hovering phase or the roll angle during the cruise phase. It is the pitch angle of a tail-mounted drone. It refers to the roll angle of a tail-mounted drone during the hovering phase or the yaw angle during the cruise phase.

[0020] Furthermore, in step S1:

[0021] The real-time rotation attitude matrix is ​​updated based on the differential equation of the rotation matrix and the exponential mapping relationship. ,Right now ;

[0022] in: Operators represent antisymmetric mappings, which are The inverse mapping of the operator, It is the form of angular velocity expressed in the body coordinate system. For the current moment, This represents the sampling time interval.

[0023] Furthermore, in step S1:

[0024] The exponential mapping of a matrix can be expressed by the Rodriguez formula as follows:

[0025] ;

[0026] in: It is a 3×3 identity matrix. angular velocity The length of the module.

[0027] Furthermore, in step S2:

[0028] The attitude error is expressed in matrix form as follows: ,in: This is the transpose of the desired attitude matrix;

[0029] The expression for the logarithmic mapping is: ,in: For the matrix logarithmic mapping operator, It is a mapping operator from an antisymmetric matrix to a vector.

[0030] Furthermore, in step S2:

[0031] The logarithmic mapping of a matrix, according to the Rodrigues inverse formula, is as follows:

[0032] ;

[0033] in, Let be the rotation angle, which can be expressed using the trace of the matrix and the inverse cosine formula:

[0034] ;

[0035] in, For matrix The traces.

[0036] Further, in step S3: the linear feedback controller is implemented based on the attitude error in Lie algebra form and the angular velocity data in the body coordinate system, and the parameters are adjusted according to the dynamic simulation model using a particle swarm optimization algorithm, based on the rotation angle. The fitness function is calculated using the norm of the axis angle, and the design metrics include: steady-state performance, global integral error, settling time, settling time monotonicity, and rise time. Since the axis angle norm does not have the concept of overshoot, the settling time monotonicity is included in the fitness function.

[0037] Further, in step S4: extract the output data of the linear feedback controller in the body coordinate system along the y-axis and z-axis from step S3, and calculate the proportion of the integral term. and Simultaneously calculate the proportionality coefficient between the two. ;

[0038]

[0039] ;

[0040] Where i represents the coordinate axis in the body coordinate system, which can be either the y-axis or the z-axis; , , These represent the proportional, integral, and derivative outputs of the linear feedback controller along the i-axis.

[0041] Furthermore, in step S5:

[0042] When the absolute value of the integral term is less than 0.2, it is assumed that the long-term continuous wind disturbance is small, and only the vector rotor control is activated, without the need for the variable phase wing to participate.

[0043] When the absolute value of the integral term's proportion is greater than 0.2, the auxiliary control of the variable-phase wing mechanism is activated, using a smooth cubic Bézier curve to map the respective weight values. This piecewise function can be expressed as:

[0044]

[0045]

[0046] Among them: when express When, the corresponding weighting coefficient ,when express When, the corresponding weighting coefficient ;

[0047] Then, according to right and The redistributed weights can be expressed as:

[0048]

[0049]

[0050] in:

[0051]

[0052]

[0053] The desired wing phase angle is expressed as:

[0054]

[0055] in: The gain coefficient is set to 0.785 to achieve a balance between the frontal and lateral projections of the aircraft, while avoiding insufficient lift during the transition phase.

[0056] Furthermore, in step S6, an analytical dynamic model of the vector rotor mechanism is established using the single rigid body modeling method, and the control variable is defined as:

[0057]

[0058] in: This is the thrust of the left propeller. This is the thrust of the right propeller. This refers to the tilt angle of the left propeller. The tilt angle of the right propeller;

[0059] The control force and torque generated by the vector rotor mechanism are expressed as follows:

[0060]

[0061]

[0062] in: and It is the lever arm length of the vector rotor mechanism in a tail-mounted UAV;

[0063] Based on its inverse mapping relationship, the expression form of each desired signal is obtained as follows:

[0064]

[0065]

[0066] The mapping relationship between propeller thrust and rotational speed is as follows:

[0067]

[0068] in: air density, The diameter of the propeller. This is the thrust coefficient of the propeller.

[0069] Compared with the prior art, the present invention has the following advantages:

[0070] 1. This invention indirectly measures external wind disturbance by extracting the proportion of the integral term of the PID controller. It does not require additional sensors such as anemometers and pitot tubes, and can achieve disturbance rejection control by relying only on the existing sensors of the flight control system, thus reducing system complexity and hardware costs.

[0071] 2. This invention sets an integral term percentage threshold as the actuator switching boundary. Under light wind disturbance, only the vector rotor provides control torque, while under heavy wind disturbance, the variable phase wing auxiliary control is activated. This achieves dynamic adjustment of actuator participation based on wind disturbance intensity, taking into account control requirements under different operating conditions.

[0072] 3. This invention uses a smooth weight mapping function to map the proportion of the integral term to the actuator weight coefficient, which ensures the smoothness of the change of the wing phase angle with the integral term and avoids attitude jitter caused by sudden intervention or withdrawal of the actuator.

[0073] 4. This invention achieves decoupled control of the pitch and yaw channels by calculating the proportional coefficients of the integral terms of the y-axis and z-axis and performing cross-weighting, thus avoiding the risk of coupling runaway caused by the superposition of multi-axis torques.

[0074] 5. This invention overcomes the gimbal lock problem of traditional Euler angles under large-angle maneuvers by transforming the attitude to a three-dimensional special orthogonal group and using the Rodriguez formula for exponential and logarithmic mapping, thus ensuring the attitude calculation accuracy under large-angle maneuvers. Attached Figure Description

[0075] Figure 1 This is a flowchart illustrating the method for coordinated control of vector rotor and variable phase wing in a tail-mounted unmanned aerial vehicle provided by the present invention.

[0076] Figure 2 This is a schematic diagram of the overall structure of an aircraft, representing an embodiment of the vector rotor and variable-phase wing cooperative control method for a tail-seat unmanned aerial vehicle (UAV) provided by the present invention; wherein... Defined as a body coordinate system, with its origin at... With the center of gravity of the whole machine coincide, The axis points towards the front of the fuselage. Pointing to the right side of the fuselage, Pointing downwards from the fuselage.

[0077] Figure 3 This is a schematic diagram of the driving degrees of freedom of the vector rotor mechanism in an embodiment of the coordinated control method for the vector rotor and variable-phase wing of a tail-mounted unmanned aerial vehicle provided by the present invention; wherein... Defined as the i-th vector rotor coordinate system, taking the left-side vector rotor mechanism as an example, its origin is... The point where the propeller's axis of rotation intersects with the axis of rotation of the vector mechanism. The axis coincides with the rotation axis of the vector mechanism. The shaft coincides with the propeller's shaft. The axis is determined by the right-hand rule. shaft and The angle between the shafts is defined as the rotor tilt angle. Its positive direction is around The axis is in the counterclockwise direction.

[0078] Figure 4 This is a schematic diagram of the driving degrees of freedom of the variable-phase wing mechanism in an embodiment of the vector rotor and variable-phase wing cooperative control method for a tail-mounted unmanned aerial vehicle provided by the present invention; wherein... Defined as the i-th vector rotor coordinate system, taking the right wing as an example, its origin is... This is the intersection of the wing's rotation axis and the leading edge axis. The shaft coincides with the leading edge shaft. The shaft coincides with the wing's rotation axis. The axis is determined by the right-hand rule. shaft and The angle between the axes is defined as the wing phase angle. Its positive direction is around The axis is in the counterclockwise direction.

[0079] Figure 5 A schematic diagram of the dominant control torque mechanism of the vector rotor mechanism in an embodiment of the coordinated control method of vector rotor and variable phase wing of a tail-mounted UAV provided by the present invention; Figure 5 (a) Demonstrates the mechanism by which the vector rotor generates rolling torque using differential thrust control of the propeller; Figure 5 (b) Demonstrates the mechanism by which the vector rotor section generates pitching moment, at which point the thrust vectors of the two propellers tilt in the same direction; Figure 5 (c) demonstrates the mechanism by which the vector rotor section generates yaw moment, at which point the thrust vectors of the two propellers tilt in opposite directions.

[0080] Figure 6 This is a schematic diagram of the auxiliary control torque mechanism of the variable phase wing mechanism in an embodiment of the vector rotor and variable phase wing cooperative control method for a tail-mounted UAV provided by the present invention. Figure 6 (a) demonstrates the mechanism by which the variable-phase wing section generates pitch assist torque, with the phase angles of the two wings changing symmetrically. Figure 6 (b) demonstrates the mechanism by which the variable phase wing section generates a roll assist torque, in which the phase angles of the two wings change asymmetrically. Detailed Implementation

[0081] Specific implementation method one: Combining Figure 1 This embodiment describes a method for coordinated control of a tail-mounted unmanned aerial vehicle's vector rotor and variable-phase wing, comprising the following steps:

[0082] Step S1: Direction cosine matrix transformation of real-time attitude:

[0083] Convert sensor data from tail-mounted drones into rotation matrices And based on the body angular velocity Update real-time attitude;

[0084] Step S2: Lie algebraic mapping representation of attitude error:

[0085] Calculate the desired pose matrix With the actual attitude matrix error matrix It is then transformed into a Lie algebraic form of the attitude error vector through a logarithmic mapping. ; Step S3: Linear feedback control based on particle swarm optimization:

[0086] attitude error vector and body angular velocity As input, the desired control torque is output through a linear feedback controller, and the proportional, integral, and derivative parameters of the linear feedback controller are tuned using a particle swarm optimization algorithm. Step S4: Feature extraction of multi-axis integral term data:

[0087] Extract the integral terms of the linear feedback controller along the y-axis and z-axis of the body coordinate system, and calculate the proportion of the integral terms. , and the ratio coefficient between the two ; Step S5: Integral-driven dynamic weight allocation for multiple actuators:

[0088] According to the proportion of the integral items , Based on the comparison with a preset threshold, the control weights of the vector rotor mechanism and the variable phase wing mechanism are dynamically allocated, and the desired wing phase angle is generated. ; Step S6: Mapping of control signals and commands for each actuator:

[0089] Based on the desired control torque and desired wing phase angle The thrust of the left propeller was calculated. Right propeller thrust Left propeller tilt angle Right propeller tilt angle And mapped to propeller speed command .

[0090] This invention combines vector rotor control technology with variable phase wing technology conveniently and efficiently through an integral-driven multi-actuator dynamic weight allocation scheme based on a three-dimensional special orthogonal group design. It does not require the introduction of additional sensors and can achieve coordinated control between the two using only the most basic inertial measurement unit and optical flow meter data, which significantly improves the wind resistance performance of tail-seat UAVs during vertical take-off and landing and transition flight phases.

[0091] Specific Implementation Method Two: Combining Figure 1 This embodiment describes step S1:

[0092] The Euler angle rotation order of the direction cosine matrix is ​​defined as ZYX, i.e. ;

[0093] in: Here is the direction cosine matrix. It refers to the yaw angle of a tail-mounted drone during the hovering phase or the roll angle during the cruise phase. It is the pitch angle of a tail-mounted drone. It refers to the roll angle of a tail-mounted drone during the hovering phase or the yaw angle during the cruise phase.

[0094] In step S1 of this embodiment, since the attitude expression of Euler angles has an inherent gimbal lock problem, it is difficult to accurately describe the attitude of the tail-mounted UAV when it performs large-angle pitch axis maneuvers during the transition phase. Therefore, it is first necessary to transform it into a three-dimensional special orthogonal group to obtain an accurate expression without global singularities. Subsequent control laws are designed on this smooth manifold structure.

[0095] This configuration eliminates the gimbal lock problem that exists in traditional Euler angle descriptions when the pitch angle is close to 90°, ensuring the continuity and accuracy of attitude calculation for tail-mounted UAVs when performing large-angle pitch maneuvers during the transition from vertical takeoff to horizontal cruise, and providing singularity-free attitude feedback for subsequent control laws.

[0096] Specific implementation method three: Combining Figure 1 This embodiment describes step S1:

[0097] The real-time rotation attitude matrix is ​​updated based on the differential equation of the rotation matrix and the exponential mapping relationship. ,Right now ;

[0098] in: Operators represent antisymmetric mappings, which are Inverse mapping of the operator, It is the form of angular velocity expressed in the body coordinate system. For the current moment, This represents the sampling time interval.

[0099] This configuration facilitates the conversion of the integral of the angular velocity measured by the gyroscope into the increment of the rotation matrix, thereby recursively updating the current attitude on a special three-dimensional orthogonal manifold group, ensuring the accuracy and continuity of attitude calculation.

[0100] Specific implementation method four: Combination Figure 1 This embodiment describes step S1:

[0101] The exponential mapping of a matrix can be expressed by the Rodriguez formula as follows:

[0102] ;

[0103] in: It is a 3×3 identity matrix. angular velocity The length of the module.

[0104] This configuration facilitates the expansion of the exponential mapping of the antisymmetric matrix corresponding to the angular velocity into a matrix polynomial, thereby converting the angular velocity increment measured by the gyroscope into a rotation matrix and realizing the recursive update of the attitude.

[0105] Specific Implementation Method Five: Combining Figure 1 This embodiment describes step S2:

[0106] The attitude error is expressed in matrix form as follows: ,in: This is the transpose of the desired attitude matrix;

[0107] The expression for the logarithmic mapping is: ,in: For the matrix logarithmic mapping operator, It is a mapping operator from an antisymmetric matrix to a vector.

[0108] This setup facilitates the calculation of the rotational error between the desired and actual attitudes, and converts it from matrix form to Lie algebra form as an error vector, which can then be input into the linear feedback controller for subsequent control.

[0109] Specific Implementation Method Six: Combination Figure 1 This embodiment describes step S2:

[0110] The logarithmic mapping of a matrix, according to the Rodrigues inverse formula, is as follows:

[0111] ;

[0112] in, Let be the rotation angle, which can be expressed using the trace of the matrix and the inverse cosine formula:

[0113] ;

[0114] in, For matrix The traces.

[0115] This setup facilitates the extraction of the equivalent rotation axis and rotation angle from the rotation error matrix, transforming the abstract matrix error into a three-axis error vector usable by the controller.

[0116] Specific implementation method seven: Combining Figure 1 In this embodiment, step S3 involves the linear feedback controller being implemented based on attitude error in Lie algebra form and angular velocity data in the body coordinate system. The parameters are adjusted using a particle swarm optimization algorithm based on the dynamic simulation model, and the rotation angle is considered. The fitness function is calculated using the norm of the axis angle, and the design metrics include: steady-state performance, global integral error, settling time, settling time monotonicity, and rise time. Since the axis angle norm does not have the concept of overshoot, the settling time monotonicity is included in the fitness function.

[0117] This setup utilizes particle swarm optimization to automatically tune the controller parameters and designs a fitness function tailored to the characteristics of rotation angles on a special three-dimensional orthogonal group, ensuring that the attitude response is both fast and smooth.

[0118] Since both the attitude error in Lie algebra form and the body angular velocity data are in the form of third-order vectors, there are as many as nine control parameters. These parameters can be quickly tuned using a particle swarm optimization algorithm based on the dynamic simulation model. Among them, the control parameters are determined based on the axis angle. The fitness function is calculated using the norm of the axis angle. Since the axis angle norm does not have an overshoot concept, the weights of the fitness function are calculated differently from those of a conventional controller, and are designed as steady-state performance, global integral error, settling time, settling time monotonicity, and rise time. Specifically, settling time monotonicity mainly measures the proportion of non-decreasing points of the axis angle error and the maximum non-decreasing amplitude within the settling time, thereby evaluating response smoothness.

[0119] Specific implementation method eight: Combination Figure 1 In this embodiment, step S4 involves extracting the output data of the linear feedback controller along the y-axis and z-axis of the body coordinate system from step S3, and calculating the proportion of the integral term. and Simultaneously calculate the proportionality coefficient between the two. ;

[0120]

[0121] ;

[0122] Where i represents the coordinate axis in the body coordinate system, which can be either the y-axis or the z-axis; , , These represent the proportional, integral, and derivative outputs of the linear feedback controller along the i-axis.

[0123] This setup facilitates the extraction of the proportion of the controller integral term in the total output, which is used to quantify the degree of control effort required for each axis to resist continuous wind disturbance, and to calculate their proportional relationship to determine the distribution characteristics of wind disturbance along the main control axis.

[0124] This implementation quantifies the wind disturbance intensity: the magnitude of the integral term reflects the accumulated control action of the controller to eliminate steady-state errors. In attitude control, the proportional term responds to instantaneous deviations, the derivative term suppresses trend changes, and the magnitude of the integral term directly reflects the "effort" expended to resist continuous external forces (such as constant crosswinds). This is achieved by calculating the proportion of the integral term. and This transforms the abstract concept of "wind disturbance magnitude" into concrete numerical indicators, providing a quantitative basis for subsequent decision-making.

[0125] This implementation method can identify the direction of wind disturbance: by extracting the proportion of the integral terms on the y-axis (pitch) and z-axis (yaw / roll), the main direction of the wind disturbance can be determined. For example, A larger value indicates that the pitch direction is subject to continuous wind disturbance. A larger value indicates a disturbance in the yaw or roll direction. This information is used to subsequently assign actuator weights to different axes.

[0126] This implementation method calculates a proportionality coefficient. = / This quantitatively describes the proportional relationship between the degrees of disturbance in the two axes. When When it is close to 0, it means that the wind disturbance is mainly concentrated on the pitch axis. A larger value indicates that both axes are disturbed simultaneously, or that the yaw axis is more strongly disturbed. This coefficient is subsequently used for decoupling control and weight redistribution to prevent coupling runaway caused by the superposition of multi-axis torques.

[0127] Therefore, step S4 extracts the integral term features to transform the internal state of the controller into a quantitative description of the intensity, direction and axial correlation of external wind disturbance, providing a decision basis for the integral-driven dynamic weight allocation in step S5.

[0128] Specific Implementation Method Nine: Combining Figure 5 and Figure 6 This embodiment describes step S5:

[0129] When the absolute value of the integral term is less than 0.2, it is assumed that the long-term continuous wind disturbance is small, and only the vector rotor control is activated, without the need for the variable phase wing to participate.

[0130] When the absolute value of the integral term's proportion is greater than 0.2, the auxiliary control of the variable-phase wing mechanism is activated, using a smooth cubic Bézier curve to map the respective weight values. This piecewise function can be expressed as:

[0131]

[0132]

[0133] Among them: when express When, the corresponding weighting coefficient ,when express When, the corresponding weighting coefficient ;

[0134] Then, according to right and The redistributed weights can be expressed as:

[0135]

[0136]

[0137] in:

[0138]

[0139]

[0140] The desired wing phase angle is expressed as:

[0141]

[0142] in: The gain coefficient is set to 0.785 to achieve a balance between the frontal and lateral projections of the aircraft, while avoiding insufficient lift during the transition phase.

[0143] In this embodiment, 0.2 is set as the threshold for the integral term percentage. When the integral term percentage is less than or equal to 0.2, only the vector rotor provides control torque; when the integral term percentage is greater than 0.2, variable-phase wing auxiliary control is activated. This threshold is used to determine whether the wing needs to be activated for disturbance rejection.

[0144] A cubic Bézier curve is used to map the proportion of the integral term to a weighting coefficient, so that the wing phase angle remains smooth as the integral term changes, avoiding attitude jitter caused by sudden wing movements.

[0145] In addition, according to the proportionality coefficient The weight coefficients are redistributed to adjust the weight distribution relationship when both axes are disturbed simultaneously.

[0146] In addition, the gain coefficient Setting it to 0.785 keeps the wing phase angle adjustment range between 0 and 45°, preventing excessive wing rotation from causing lift loss.

[0147] Specific Implementation Method Ten: Combining Figure 3 and Figure 5This embodiment describes a step S6 in which an analytical dynamic model of the vector rotor mechanism is established using a single rigid body modeling method, and the control variable is defined as follows:

[0148]

[0149] in: This is the thrust of the left propeller. This is the thrust of the right propeller. This refers to the tilt angle of the left propeller. The tilt angle of the right propeller;

[0150] The control force and torque generated by the vector rotor mechanism are expressed as follows:

[0151]

[0152]

[0153] in: and It is the lever arm length of the vector rotor mechanism in a tail-mounted UAV;

[0154] Based on its inverse mapping relationship, the expression form of each desired signal is obtained as follows:

[0155]

[0156]

[0157] The mapping relationship between propeller thrust and rotational speed is as follows:

[0158]

[0159] in: air density, The diameter of the propeller. This is the thrust coefficient of the propeller.

[0160] In this implementation, the desired control torque and desired wing phase angle output by the upper controller are calculated as the thrust of the left propeller. Right propeller thrust Left propeller tilt angle Right propeller tilt angle And further mapped to propeller speed command. This enables the effective conversion of control commands into physical execution signals.

[0161] Combination Figures 1 to 6 Explanation of the implementation method of the present invention:

[0162] Figure 1A flowchart illustrating the method for coordinated control of a vector rotor and variable-phase wing in a tail-mounted unmanned aerial vehicle (UAV) provided by the present invention includes the following steps:

[0163] S1, Direction cosine matrix transformation of real-time attitude;

[0164] If the rotation order of Euler angles is defined as ZYX, then the real-time attitude matrix can be expressed as:

[0165]

[0166] in Here is the direction cosine matrix. It refers to the yaw angle of a tail-mounted drone during the hovering phase or the roll angle during the cruise phase. It is the pitch angle of a tail-mounted drone. It refers to the roll angle of a tail-mounted drone during the hovering phase or the yaw angle during the cruise phase.

[0167] The real-time rotation attitude matrix is ​​updated based on the differential equation of the rotation matrix and the exponential mapping relationship. :

[0168]

[0169]

[0170] In the formula, Operators represent antisymmetric mappings, which are The inverse mapping of the operator, It is the form of angular velocity expressed in the body coordinate system.

[0171] According to Rodriguez's formula, the exponential mapping of a matrix can be expressed as:

[0172]

[0173] S2, Lie algebraic mapping representation of attitude error;

[0174] The rotation error is represented in matrix form as follows:

[0175]

[0176] in This is the transpose of the desired pose matrix.

[0177] The rotation error matrix can be mapped to the Lie algebra using a logarithmic mapping:

[0178]

[0179] in For the matrix logarithmic mapping operator, It is a mapping operator from an antisymmetric matrix to a vector.

[0180] According to the Rodriguez inverse formula, the logarithmic mapping of a matrix can be expressed as:

[0181]

[0182] in The rotation angle can be expressed using the trace of the matrix and the inverse cosine formula:

[0183]

[0184] S3. Linear feedback control based on particle swarm optimization;

[0185] There are various types of linear feedback controllers for the attitude control of tail-mounted UAVs, including but not limited to PID controllers and backstepping controllers. All of these take the attitude error in Lie algebra form and the angular velocity data in the body coordinate system as input, and then output the desired control torque. In this embodiment, a PID controller is used as an example. The control parameters are quickly tuned using a particle swarm optimization algorithm based on the dynamic simulation model. The specific implementation process is as follows:

[0186] First, considering the control characteristics of the hovering and transitional flight phases of the tail-mounted UAV, the core parameters of the PSO algorithm are set, including the number of particles, maximum number of iterations, inertia weight, individual learning factor, and population learning factor. Then, the upper limit of the optimization interval for the nine control parameters corresponding to the third-order vector in the Lie algebra is set. To accelerate the algorithm's convergence speed, a particle position initialization method using initial values ​​plus random perturbations is adopted.

[0187] Then, the fitness function is designed. In the manifold structure characteristics of the three-dimensional special orthogonal group, the nine control parameters should be adjusted synchronously according to the axis angle. The fitness function is calculated using the norm of the axis angle. Since the norm of the axis angle does not have the concept of overshoot, the calculation indices of the weights of the fitness function are different from those of conventional controllers, and are designed as steady-state performance, global integral error, settling time, settling time monotonicity, and rise time.

[0188] Specifically, steady-state performance is calculated based on the integral of the axis-angle norm error, the mean steady-state error, and the oscillation intensity during the last 20% of the simulation time, including a penalty for severe oscillations; global integral error is the integral value of the absolute value of the axis-angle norm error over the entire cycle; settling time is the time it takes for the axis-angle norm error to enter and remain within ±5% of the initial error range, including a maximum penalty for exceeding 90% of the total time; settling time monotonicity mainly measures the proportion of non-decreasing points and the maximum non-decreasing amplitude of the axis-angle error during the settling time; rise time is the time it takes for the axis-angle error to decrease from its initial value to 10% of the initial error.

[0189] S4. Feature extraction of multi-axis integral term data;

[0190] Extract the output data of the linear feedback controller in the body coordinate system along the y-axis and z-axis from step S3, and calculate the proportion of the integral term. and Simultaneously calculate the proportionality coefficient between the two. :

[0191]

[0192] ;

[0193] Where i represents the coordinate axis in the body coordinate system, which can be either the y-axis or the z-axis; , , These represent the proportional, integral, and derivative outputs of the linear feedback controller along the i-axis.

[0194] S5, Integral-driven dynamic weight allocation for multiple actuators;

[0195] The piecewise function for calculating the weights of each wing is as follows: when the absolute value of the integral term is less than 0.2, it is assumed that the long-term continuous wind disturbance is relatively small, and only vector rotor control is activated, without the need for the variable-phase wing to participate; when the absolute value of the integral term is greater than 0.2, auxiliary control of the variable-phase wing mechanism is activated, and a smooth cubic Bézier curve is used to map the respective weight values. This piecewise function can be expressed as:

[0196]

[0197]

[0198] Among them when express When, the corresponding weighting coefficient ,when express When, the corresponding weighting coefficient .

[0199] Then according to right and The redistributed weights can be expressed as:

[0200]

[0201]

[0202] in:

[0203]

[0204]

[0205] The desired wing phase angle is expressed as:

[0206]

[0207] in The gain coefficient is set to 0.785 to achieve a balance between the frontal and lateral projections of the aircraft, while avoiding insufficient lift during the transition phase.

[0208] S6. Control signal command mapping for each actuator;

[0209] An analytical dynamic model of the vector rotor mechanism is established using the single rigid body modeling method, and the control variables are defined as follows:

[0210]

[0211] in This is the thrust of the left propeller. This is the thrust of the right propeller. This represents the tilt angle of the left propeller. Let be the tilt angle of the right propeller. The control force and torque generated by the vector rotor mechanism can be expressed as:

[0212]

[0213]

[0214] in and It is the lever arm length of the vector rotor mechanism in a tail-mounted UAV.

[0215] Based on its inverse mapping relationship, the expression form of each desired signal can be obtained as follows:

[0216]

[0217]

[0218] The mapping relationship between propeller thrust and rotational speed is as follows:

[0219]

[0220] in air density, The diameter of the propeller. This is the thrust coefficient of the propeller.

[0221] exist Figure 2 The image shows an embodiment of a tail-seat unmanned mechanism. Figure 3This demonstrates the driving degrees of freedom of the vector rotor mechanism. Figure 4 This demonstrates the actuation degrees of freedom of the variable-phase wing mechanism. Figure 5 The principle of how a vector rotor mechanism generates the dominant control torque is described. Figure 6 The principle of the auxiliary control torque generated by the variable phase wing mechanism is described.

[0222] While the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the invention. Those skilled in the art can make other changes within the spirit of the invention and apply it to fields not mentioned in the invention. Of course, all such changes made in accordance with the spirit of the invention should be included within the scope of protection claimed by the invention.

Claims

1. A method for coordinated control of a vector rotor and a variable-phase wing in a tail-seat unmanned aerial vehicle (UAV), characterized in that: Includes the following steps: Step S1: Direction cosine matrix transformation of real-time attitude: Convert sensor data from tail-mounted drones into rotation matrices And based on the body angular velocity Update real-time attitude; Step S2: Lie algebraic mapping representation of attitude error: Calculate the desired pose matrix With the actual attitude matrix error matrix It is then transformed into a Lie algebraic form of the attitude error vector through a logarithmic mapping. ; Step S3: Linear feedback control based on particle swarm optimization: attitude error vector and body angular velocity As input, the desired control torque is output through a linear feedback controller, and the proportional, integral, and derivative parameters of the linear feedback controller are tuned using a particle swarm optimization algorithm. Step S4: Feature extraction of multi-axis integral term data: Extract the integral terms of the linear feedback controller along the y-axis and z-axis of the body coordinate system, and calculate the proportion of the integral terms. , and the ratio coefficient between the two ; Step S5: Integral-driven dynamic weight allocation for multiple actuators: According to the proportion of the integral items , Based on the comparison with a preset threshold, the control weights of the vector rotor mechanism and the variable phase wing mechanism are dynamically allocated, and the desired wing phase angle is generated. ; Step S6: Mapping of control signals and commands for each actuator: Based on the desired control torque and desired wing phase angle The thrust of the left propeller was calculated. Right propeller thrust Left propeller tilt angle Right propeller tilt angle And mapped to propeller speed command .

2. The method for coordinated control of vector rotor and variable phase wing of a tail-mounted unmanned aerial vehicle according to claim 1, characterized in that: In step S1: The Euler angle rotation order of the direction cosine matrix is ​​defined as ZYX, i.e. ; in: Here is the direction cosine matrix. It refers to the yaw angle of a tail-mounted drone during the hovering phase or the roll angle during the cruise phase. It is the pitch angle of a tail-mounted drone. It refers to the roll angle of a tail-mounted drone during the hovering phase or the yaw angle during the cruise phase.

3. The method for coordinated control of vector rotor and variable phase wing of a tail-mounted unmanned aerial vehicle according to claim 2, characterized in that: In step S1: The real-time rotation attitude matrix is ​​updated based on the differential equation of the rotation matrix and the exponential mapping relationship. ,Right now ; in: Operators represent antisymmetric mappings, which are Inverse mapping of the operator, It is the form of angular velocity expressed in the body coordinate system. For the current moment, This represents the sampling time interval.

4. The method for coordinated control of vector rotor and variable phase wing of a tail-mounted unmanned aerial vehicle according to claim 3, characterized in that: In step S1: The exponential mapping of a matrix can be expressed by the Rodriguez formula as follows: ； in: It is a 3×3 identity matrix. angular velocity The length of the module.

5. The method for coordinated control of vector rotor and variable phase wing of a tail-mounted unmanned aerial vehicle according to claim 1, characterized in that: In step S2: The attitude error is expressed in matrix form as follows: ,in: This is the transpose of the desired attitude matrix; The expression for the logarithmic mapping is: ,in: For the matrix logarithmic mapping operator, It is a mapping operator from an antisymmetric matrix to a vector.

6. The method for coordinated control of vector rotor and variable-phase wing of a tail-mounted unmanned aerial vehicle according to claim 5, characterized in that: In step S2: The logarithmic mapping of a matrix, according to the Rodrigues inverse formula, is as follows: ； in, Let be the rotation angle, which can be expressed using the trace of the matrix and the inverse cosine formula: ； in, For matrix The traces.

7. The method for coordinated control of vector rotor and variable phase wing of a tail-mounted unmanned aerial vehicle according to claim 1, characterized in that: In step S3: the linear feedback controller is implemented based on the attitude error in Lie algebra form and the angular velocity data in the body coordinate system, and the parameters are adjusted according to the dynamic simulation model using a particle swarm optimization algorithm, based on the rotation angle. The fitness function is calculated using the norm of the axis angle, and the design metrics include: steady-state performance, global integral error, settling time, settling time monotonicity, and rise time. Since the axis angle norm does not have the concept of overshoot, the settling time monotonicity is included in the fitness function.

8. The method for coordinated control of vector rotor and variable phase wing of a tail-mounted unmanned aerial vehicle according to claim 1, characterized in that: In step S4: Extract the output data of the linear feedback controller in the body coordinate system along the y-axis and z-axis from step S3, and calculate the proportion of the integral term. and Simultaneously calculate the proportionality coefficient between the two. ; ； Where i represents the coordinate axis in the body coordinate system, which can be either the y-axis or the z-axis; , , These represent the proportional, integral, and derivative outputs of the linear feedback controller along the i-axis.

9. The method for coordinated control of vector rotor and variable phase wing of a tail-mounted unmanned aerial vehicle according to claim 1, characterized in that: In step S5: When the absolute value of the integral term is less than 0.2, it is assumed that the long-term continuous wind disturbance is small, and only the vector rotor control is activated, without the need for the variable phase wing to participate. When the absolute value of the integral term's proportion is greater than 0.2, the auxiliary control of the variable-phase wing mechanism is activated, using a smooth cubic Bézier curve to map the respective weight values. This piecewise function can be expressed as: Among them: when express When, the corresponding weighting coefficient ,when express When, the corresponding weighting coefficient ; Then, according to right and The redistributed weights can be expressed as: in: The desired wing phase angle is expressed as: in: The gain coefficient is set to 0.785 to achieve a balance between the frontal and lateral projections of the aircraft, while avoiding insufficient lift during the transition phase.

10. The method for coordinated control of vector rotor and variable phase wing of a tail-mounted unmanned aerial vehicle according to claim 1, characterized in that: In step S6, an analytical dynamic model of the vector rotor mechanism is established using the single rigid body modeling method, and the control variables are defined as follows: in: This is the thrust of the left propeller. This is the thrust of the right propeller. This represents the tilt angle of the left propeller. The tilt angle of the right propeller; The control force and torque generated by the vector rotor mechanism are expressed as follows: in: and It is the lever arm length of the vector rotor mechanism in a tail-mounted UAV; Based on its inverse mapping relationship, the expression form of each desired signal is obtained as follows: The mapping relationship between propeller thrust and rotational speed is as follows: in: air density, The diameter of the propeller. This is the thrust coefficient of the propeller.

Images