Mpc-based tail stand vertical take-off uav control method

Abstract

The present application relates to the field of unmanned aerial vehicle intelligence, and specifically provides a tail stand type vertical take-off unmanned aerial vehicle control method based on MPC, comprising the following steps: S1, constructing the kinematic model and the dynamic model of the tail stand type vertical take-off unmanned aerial vehicle under the inertial coordinate system and the body coordinate system, S2, designing the reference attitude of the conversion stage based on the kinematic model and the dynamic model, S3, taking the reference attitude of the conversion stage as the expected attitude value of the prediction stage of the unmanned aerial vehicle to establish the prediction equation of the tail stand type vertical take-off unmanned aerial vehicle, S4, performing rolling optimization on the prediction model according to the constraint condition of the conversion stage, S5, performing feedback correction on the tail stand type vertical take-off unmanned aerial vehicle according to the result of the rolling optimization, the control of the transition stage conversion of the tail stand type vertical take-off unmanned aerial vehicle based on MPC proposed by the present application can be converted more quickly and has smaller height error than the traditional tail stand type vertical take-off unmanned aerial vehicle.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of intelligent vertical take-off and landing unmanned aerial vehicle, and particularly relates to a tail seat type vertical take-off and landing unmanned aerial vehicle control method based on MPC. BACKGROUND

[0002] At present, unmanned aerial vehicles are widely used in daily life and production. At present, unmanned aerial vehicles can be roughly divided into two categories: rotary wing unmanned aerial vehicles and fixed wing unmanned aerial vehicles. Rotary wing unmanned aerial vehicles can realize vertical take-off and landing in a small space, but have the disadvantages of short endurance, slow flight speed, etc. Fixed wing unmanned aerial vehicles have strong endurance and fast flight speed, but are easily limited by the environment when taking off.

[0003] The tail seat type unmanned aerial vehicle is a fixed wing vertical take-off unmanned aerial vehicle, which has the advantages of simple structure, high cruising speed and less system weight, and has a wide application and development prospect. The tail seat type unmanned aerial vehicle mainly includes five flight states. The body is vertically taken off by the rotor tension, reaches the specified height, enters the hovering state, is inclined by 90 degrees to enter the fixed wing state, lands, is inclined by 90 degrees to enter the hovering state, and finally vertically lands on the ground in the rotor mode. The conversion from the rotor to the straight fixed wing stage is called the front conversion stage, and the conversion from the fixed wing to the rotor is called the rear conversion stage. The front conversion stage and the rear conversion stage are collectively referred to as the transition stage.

[0004] At present, the transition stage is the most difficult flight state for the tail seat type unmanned aerial vehicle control. In the front conversion stage, the flight lift is composed of the rotor tension and the wing aerodynamic force, so that the unmanned aerial vehicle has the problems of large height error, long time consumption and instability in the transition stage. The current control method of the transition stage mainly focuses on accelerating the conversion speed, but the problem of large height error has not been fundamentally solved. SUMMARY

[0005] To solve the above problems, the present application provides a tail seat type vertical take-off and landing unmanned aerial vehicle control method based on MPC. The implementation of the method includes the following steps:

[0006] S1, a kinematic model and a dynamic model of the tail seat type vertical take-off and landing unmanned aerial vehicle in the inertial coordinate system and the body coordinate system are constructed;

[0007] The kinematic model includes a kinematic equation;

[0008] The kinematic equation is as follows:

[0009]

[0010] Wherein, F is the total external force, m is the body weight, u is the flight speed, L is the momentum moment, and M is the total moment;

[0011] The dynamic model equation is as follows:

[0012]

[0013] where (a x , a y , a z ), (v x , v y , v z ) and (ω x , ω y , ω z ) are the acceleration, velocity and angular velocity in the body frame respectively, α is the angle of attack of the UAV, (I x , I y , I z ) are the moments of inertia of the body frame, and are the derivatives of ω x , ω y and ω z , and are the second derivatives of x, y and z in the earth frame, M xc , M yc and M zc are the components of the rotor aerodynamic moment M Tc in the x, y and z directions in the earth frame, M xg , M yg and M zg are the components of the gyro M Tg in the x, y and z directions in the earth frame, M r is the propeller reaction moment;

[0014] S2, a reference attitude in the transition phase is designed based on the kinematic model and the dynamic model;

[0015] The transition phase includes a rotor phase and an aerodynamic phase;

[0016] The transition phase includes a rotor phase and an aerodynamic phase;

[0017] The reference state formula of the rotor phase is as follows:

[0018]

[0019] where, is the desired pitch angle, is the height error, F Z1 is the vertical component of the rotor tension;

[0020] The reference state formula of the aerodynamic phase is as follows:

[0021]

[0022] where, F maxFor the maximum pull of the rotor, k(θ) is a time-varying quantity with respect to θ, and F is the minimum pull required for the current moment to balance the gravity;

[0023] S3, taking the reference attitude of the conversion stage as the expected attitude value of the prediction stage of the unmanned aerial vehicle, a prediction equation of the tail seat type vertical take-off unmanned aerial vehicle is established;

[0024] The prediction equation is as follows:

[0025]

[0026] The sampling period is set to T=0.02s, and the discrete equation of formula (5) is as follows:

[0027]

[0028] Wherein, Q(t) is a state observation matrix, is an expected state observation matrix, is a derivative of the state observation matrix, is a derivative of the state observation matrix, A(t) is a corresponding coefficient matrix, B(t) is a corresponding coefficient matrix, V(t) is a control input, is an expected control quantity;

[0029] S4, the prediction model is rolling optimized according to the constraint condition of the conversion stage;

[0030]

[0031]

[0032] Wherein, R is a control input error penalty matrix, ε is a relaxation factor, and k(θ) is a control rate coefficient;

[0033] S5, the tail seat type vertical take-off unmanned aerial vehicle is feedback corrected according to the result of the rolling optimization.

[0034] Preferably, the reference state of the aerodynamic stage

[0035]

[0036] Wherein, β is a fixed conversion rate, and k is a conversion coefficient;

[0037] Preferably, in step S3, the state observation matrix is selected as follows:

[0038]

[0039] Preferably, in step S5, the N cThe optimization result obtained by the inner solution is transmitted to the flight control through the MAVLink protocol, the flight control issues instructions to the actuator in the Gazebo virtual environment, trajectory tracking and attitude optimization are realized, and the position and attitude information is transmitted to the flight control to realize real-time state feedback.

[0040] Preferably, the tail stand type vertical take-off unmanned aerial vehicle is a four-rotor tail stand type vertical take-off unmanned aerial vehicle.

[0041] The present application can achieve the following technical effects:

[0042] The transition phase of the tail stand type vertical take-off unmanned aerial vehicle control method based on MPC provided by the present application can predict the position and attitude of the vertical take-off unmanned aerial vehicle compared with the traditional control method, optimize the predicted result, correct according to the optimization result, realize faster conversion and have smaller height error. BRIEF DESCRIPTION OF DRAWINGS

[0043] Figure 1 is a flowchart of the tail stand type vertical take-off unmanned aerial vehicle control method based on MPC provided by the embodiment of the present application;

[0044] Figure 2 is a model prediction and control input calculation flowchart of the tail stand type vertical take-off unmanned aerial vehicle control method based on MPC provided by the embodiment of the present application;

[0045] Figure 3 is a flight mode conversion schematic diagram of the tail stand type vertical take-off unmanned aerial vehicle control method based on MPC provided by the embodiment of the present application;

[0046] Figure 4 is a force diagram of the front and rear conversion phase of the tail stand type vertical take-off unmanned aerial vehicle control method based on MPC provided by the embodiment of the present application;

[0047] Figure 5 is a height change diagram of the front and rear conversion phase of the tail stand type vertical take-off unmanned aerial vehicle control method based on MPC provided by the embodiment of the present application. DETAILED DESCRIPTION

[0048] Hereinafter, embodiments of the present application will be described with reference to the accompanying drawings. In the following description, the same modules are denoted by the same reference numerals. In the case of the same reference numerals, their names and functions are also the same. Therefore, detailed descriptions thereof will not be repeated.

[0049] In order to make the purpose, technical scheme and advantages of the present application more clear, the present application will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and do not constitute a limitation on the present application.

[0050] Figure 1 A flowchart of a tail-sitter vertical take-off unmanned aerial vehicle control method based on MPC (Model Predictive Control) provided by an embodiment of the present application is shown.

[0051] As shown in the figure, the tail-sitter vertical take-off unmanned aerial vehicle control method based on MPC provided by the embodiment of the present application comprises the following steps: Figure 1

[0052] S1, constructing a kinematic model and a dynamic model of the tail-sitter vertical take-off unmanned aerial vehicle in an inertial coordinate system and a body coordinate system;

[0053] The inertial coordinate system is {I}={x i , y i , z i}, the x i -axis points to the north, the y i -axis points to the east, and the z i -axis points to the center of the earth; the body coordinate system is {B}={x b , y b , z b}, the body coordinate system complies with the right-hand rule, the x b -axis points to the direction in which the head of the unmanned aerial vehicle advances, the y b -axis points to the right side of the unmanned aerial vehicle, and the z b -axis direction is determined by the right-hand rule and points to the center of the earth; the rotation matrix formula of the body coordinate system to the terrestrial coordinate system is as follows:

[0054]

[0055] wherein, θ is a roll angle, is a pitch angle, and is a yaw angle;

[0056] The kinematic model comprises a kinematic equation;

[0057] The kinematic equation is as follows:

[0058]

[0059] wherein, F is a total sum of external forces, m is the weight of the body, u is the flight speed, L is the moment of momentum, and M is the total moment;

[0060] The total sum of external forces equation is as follows:

[0061] F=F G +F T +F L +F D (12);

[0062] wherein, F G is the gravity of the body, F​T F is the thrust generated by the four rotors L F is the lift D F is the drag

[0063] The body gravity in the body coordinate system is as follows:

[0064]

[0065] Where F Gb is the body gravity in the body coordinate system

[0066] The thrust generated by the four rotors is as follows:

[0067]

[0068] Where C T is the rotor tension coefficient, p is the air density, w is the rotor speed, and D is the rotor diameter

[0069] The sum of the pitching moment, lift and drag of the wing is as follows:

[0070]

[0071] Where p is the air density, S is the wing area, C L and C D are the lift and drag coefficients respectively

[0072] The total moment M is composed of the gyroscopic moment M Tg , the rotor aerodynamic moment M Tc , and the wing aerodynamic moment M A

[0073] The dynamic model equation is as follows:

[0074]

[0075] Where (a x , a y , a z ), (v x , v y , v z ) and (w x , w y , w z ) are the acceleration, speed and angular velocity in the inertial coordinate system respectively, a is the angle of attack of the unmanned aerial vehicle, is the roll angle, (I x , I y , I z ) is the body coordinate axis moment of inertia, and are w​x , ω y , and the derivative of ω z , , and are the second derivatives of x, y, and z in the geodetic coordinate system, M xc , M yc , and M zc are the aerodynamic moments M Tc in the x, y, and z directions in the geodetic coordinate system, M xg , M yg , and M zg are the gyroscopic moments M Tg in the x, y, and z directions in the geodetic coordinate system, M r is the propeller reaction moment;

[0076] S2, design conversion phase reference attitude;

[0077] In the tail stand vertical take-off unmanned aerial vehicle fixed wing mode and rotor mode conversion phase, the height of the unmanned aerial vehicle is controlled by the rotor and the fixed wing, and the control height and forward flight speed are controlled by the pitch angle. The vertical component F z1 of the rotor tension gradually decreases to 0 with the decrease of the pitch angle in the front conversion phase, and the vertical component F z2 of the aerodynamic lift gradually increases to balance the gravity. The rear conversion phase is opposite to this.

[0078] Take the pitch angle from 0 in the rotor mode to is the rotor phase, to 90° is the aerodynamic phase. The influence of F z2 is ignored in the rotor phase, the reference state is the expected pitch angle , the height error , and the control constraint factor is F z1 > mg. The reference state in the aerodynamic phase is the height error , and the control input constraint factor is the angle of attack constraint: α0< α < α M ;

[0079] Wherein, α0is the zero lift angle of attack, and α M is the critical stall angle of attack;

[0080] At the same time, the rotor tension should satisfy:

[0081] mg·k(φ) < F < 65%F max ;

[0082] Wherein, F max is the maximum rotor tension, is a time-varying quantity related to θ, and is the minimum tension required to balance the gravity at the current moment;

[0083] Transition phase state transition phase attitude model as follows:

[0084] Transition phase includes rotor phase and aerodynamic phase;

[0085] Rotor phase reference state formula as follows:

[0086]

[0087] Wherein, is the desired pitch angle, is the height error, F Z1 is the vertical component of rotor pull force;

[0088] Aerodynamic phase reference state formula as follows:

[0089]

[0090] Wherein, F max is the maximum rotor pull force, k (θ) is a time-varying quantity about θ, F is the minimum pull force required to balance the gravity at the current time;

[0091] Wherein, β is the fixed transition rate, k is the transition coefficient;

[0092] S3, with the reference attitude of the transition phase as the expected attitude value of the prediction phase of the unmanned aerial vehicle, the prediction equation of the tail seat vertical take-off unmanned aerial vehicle is established;

[0093] Select state observation matrix

[0094] The prediction equation is as follows:

[0095]

[0096] The sampling period is set to T = 0.02s, and the discrete equation of equation (5) is as follows:

[0097]

[0098] Wherein, Q (t) is the state observation matrix, is the expected state observation matrix, is the derivative of the state observation matrix, is the expected derivative of the state observation matrix, A (t) is the corresponding coefficient matrix, B (t) is the corresponding coefficient matrix, V (t) is the control input, is the expected control quantity;

[0099] Take the state prediction variable as The attitude of the vertical take-off unmanned aerial vehicle is estimated, and a discretized prediction equation is obtained as follows:

[0100] The prediction equation includes a state vector and an attitude control vector in a prediction time domain;

[0101] The prediction time domain is taken as N p , the state vector X p The equation is as follows:

[0102] X p_Np ={X(k+1|k), X(k+2|k), …, X(k+N x -1|k), X(k+N p |k)} (16);

[0103] X p (k) = P(k|k)·X(k) + H(k|k)V(k) (17);

[0104] Y p_Np ={Y(k+1|k), Y(k+2|k), …, Y(k+N x -1|k), Y(k+N p |k)} (18);

[0105] Y p (k) = ψ(k|k)X(k) + Θ(k|k){V p (k|k) - V p (k-1|k)} (19);

[0106] X p_Np is a state expression of a time sequence of {k+1, k+N p}, X p (k) represents a state expression of X p at each time point, Y p_Np is an output expression of a time sequence of {k+1, k+N p}, Y p (k) represents a predicted output Y p of the prediction time domain controller at each time point;

[0107] P(k|k) is a prediction time domain conversion matrix of the state X(k), and the expression is as follows:

[0108]

[0109] H(k|k) is a prediction time domain conversion matrix of the control V(k), and the expression is as follows:

[0110]

[0111] Ψ(k|k) is the output Y p (k) of the prediction time domain X(k) conversion matrix, the expression is:

[0112]

[0113] Θ(k|k) is the output Y p (k) of the prediction time domain control V(k) conversion matrix, the expression is:

[0114]

[0115] Wherein, I4 is a unit matrix;

[0116] Taking the control time domain as N c , the attitude control vector equation is as follows:

[0117]

[0118]

[0119] Wherein, V(k) is the control quantity, is the attitude control input in the prediction time domain;

[0120] S4, according to the constraint condition of the conversion stage, the prediction model is rolled and optimized;

[0121]

[0122]

[0123] Wherein, formula (7) optimizes the position of the vertical take-off unmanned aerial vehicle, and formula (8) optimizes the attitude of the vertical take-off unmanned aerial vehicle;

[0124] Wherein, Q is the state observation matrix, R is the control input error penalty matrix, ε is the relaxation factor, k(θ) is the control rate coefficient, in the conversion stage, the unmanned aerial vehicle is prone to oscillation phenomenon, if the conversion is too slow, the problem of falling high in the former conversion stage and climbing in the latter conversion stage increases, and the forward flight distance of the conversion stage is lengthened, taking the control input error as the penalty factor can improve the conversion rate and shorten the mode conversion time of the unmanned aerial vehicle;

[0125] S5, according to the optimization result, the tail seat type vertical take-off unmanned aerial vehicle is feedback corrected, based on the MAVROS function in ROS (Robot Operating System, robot operating system), the every N cThe optimization result obtained by the inner solution is transmitted to the flight control through the MAVLink protocol, the flight control issues instructions to the actuator in the Gazebo virtual environment to realize trajectory tracking and attitude optimization, and meanwhile, the position and attitude information is transmitted to the flight control to realize real-time state feedback.

[0126] Figure 2 A model prediction and control input calculation process of the tail stand vertical take-off unmanned aerial vehicle control method based on MPC provided by the embodiment of the application is shown.

[0127] Figure 3 A flight mode conversion schematic diagram of the tail stand vertical take-off unmanned aerial vehicle control method based on MPC is shown, and the flight state of the tail stand unmanned aerial vehicle mainly includes five states, the body is vertically taken off by the rotor pull, enters the hovering state after reaching the specified height, the body is tilted by 90 degrees to enter the fixed wing state, the cruising mode is started, when landing, the body is tilted by -90 degrees to enter the hovering state, and finally vertically lands on the ground in the rotor mode. The conversion from the rotor to the fixed wing stage is called the front conversion stage, the conversion from the fixed wing stage to the rotor stage is called the rear conversion stage, and the two stages are collectively called the transition stage.

[0128] Figure 4 A force diagram of the front and rear conversion stages of the tail stand vertical take-off unmanned aerial vehicle control method based on MPC is shown, in the fixed wing mode and the rotor mode conversion stage of the tail stand vertical take-off unmanned aerial vehicle, the height of the unmanned aerial vehicle is controlled by the rotor and the fixed wing, and the control height and the forward flight speed are controlled by the pitch angle α. The vertical component F Z1 In the front conversion stage, the vertical component F Z2 of the aerodynamic lift gradually increases from 0 to balance the gravity, and the rear conversion stage is opposite to this.

[0129] Figure 5 A height change diagram of the front and rear conversion stages of the tail stand vertical take-off unmanned aerial vehicle control method based on MPC is shown, and the height error of the unmanned aerial vehicle when the improved transition stage conversion is completed is reduced compared with the height error of the PID (Proportion Integration Differentiation) controller.

[0130] Although the embodiments of the application have been shown and described above, it should be understood that the above embodiments are exemplary and should not be construed as limiting the application. Those skilled in the art can make changes, modifications, replacements and variations to the above embodiments within the scope of the application.

[0131] The specific embodiments of the application described above do not constitute a limitation on the protection scope of the application. Any various other corresponding changes and modifications made according to the technical concept of the application should be included in the protection scope of the claims of the application.

Claims

1. A control method for a tail-mounted vertical takeoff and landing (VTOL) unmanned aerial vehicle (UAV) based on MPC, characterized in that, Includes the following steps: S1. Construct and establish the kinematic and dynamic models of the tail-mounted vertical take-off and landing UAV in the inertial coordinate system and the body coordinate system; The kinematic model includes kinematic equations; The kinematic equations are as follows: Where F is the sum of the net external forces, m is the weight of the aircraft, u is the flight speed, L is the angular momentum, and M is the total torque; The dynamic model equations are as follows: （2）； in, , and These represent the acceleration, velocity, and angular velocity in the inertial coordinate system, respectively. For the angle of attack of the drone flight, The moment of inertia of the machine's coordinate axes. and They are respectively , The derivative, , and Let x, y, and z be the second derivatives in the geodetic coordinate system, respectively. respectively rotor aerodynamic torque The components in the x, y, and z directions in the geodetic coordinate system. They are gyroscopes The components in the x, y, and z directions in the geodetic coordinate system. For propeller reaction torque, For lift, As resistance, The thrust generated by the four rotors is m, and the weight of the airframe is m. S2. Design the reference attitude during the conversion stage based on the kinematic and dynamic models; The conversion phase includes a rotor phase and an aerodynamic phase; The reference state formula for the rotor stage is as follows: in, For the desired pitch angle, For height error, The vertical component generated by the rotor thrust; The formula for the reference state during the aerodynamic phase is as follows: in, This is the maximum thrust of the rotor. For about The time variable is F, which is the minimum tension required to balance gravity at the current moment; S3. Using the reference attitude of the transition phase as the expected attitude value of the UAV prediction phase, establish the prediction equation for the tail-mounted vertical take-off UAV. The prediction equation is as follows: （5）； Setting the sampling period to T = 0.02s, the discretized equation of equation (5) is as follows: （6）； Where Q(t) is the state observation matrix, Let be the desired state observation matrix. (t) is the derivative of the state observation matrix. Let A(t) be the derivative of the expectation of the state observation matrix. The corresponding coefficient matrix, B(t), is The corresponding coefficient matrix, V(t), represents the control input. (t) represents the desired control quantity; S4. Perform rolling optimization of the prediction model based on the constraints of the transformation stage; （7）； （8）； Where R is the control input error penalty matrix. As a relaxation factor, To control the rate coefficient, This represents the predicted output of the predictive time-domain controller. The output expression at each time point; S5. Feedback correction is performed on the tail-mounted vertical take-off and landing UAV based on the results of the rolling optimization.

2. The control method for a tail-mounted vertical takeoff and landing UAV based on MPC as described in claim 1, characterized in that, The reference state of the aerodynamic phase = （9）; Where β is the fixed conversion rate and k is the conversion coefficient.

3. The control method for a tail-mounted vertical takeoff and landing UAV based on MPC as described in claim 1, characterized in that, In step S3, the state observation matrix is ​​selected as follows: （10）。 4. The control method for a tail-mounted vertical takeoff and landing UAV based on MPC as described in claim 1, characterized in that, In step S5, each The optimization results obtained from the internal solution are transmitted to the flight controller via the MAVLink protocol. The flight controller then issues commands to the actuators in the Gazebo virtual environment to achieve trajectory tracking and attitude optimization. Simultaneously, position and attitude information are transmitted to the flight controller for real-time status feedback. This indicates the control time domain.

5. The control method for a tail-mounted vertical takeoff and landing UAV based on MPC as described in claim 1, characterized in that, The tail-mounted vertical takeoff and landing (VTOL) drone is a quadcopter tail-mounted VTOL drone.

Images