Four-rotor aircraft data driving control method capable of breaking through bandwidth constraint under hostile attack

Abstract

The invention discloses a four-rotor aircraft data driving control method breaking through bandwidth constraint under hostile attack. The method specifically comprises the following steps: establishing a kinetic model of a four-rotor aircraft; converting the obtained kinetic model into an equivalent dynamic linearization equation by using a dynamic linearization method; designing a criterion function, and designing a data driving control method of the four-rotor aircraft; an event triggering mechanism is introduced, denial of service attack is considered, and then the final four-rotor aircraft safety event triggering data driving control method is obtained through arrangement of the obtained control method. According to the method, model-free adaptive control is applied to attitude control of the four-rotor aircraft, the advantages of PID are combined, malicious denial of service attacks are considered, an event triggering mechanism is introduced, the good flight quality of the aircraft is guaranteed, system bandwidth resources can be saved, and the working time of the aircraft is prolonged.

Description

Technical Field

[0001] This invention belongs to the field of automatic control of unmanned aerial vehicles, and particularly relates to a data-driven control method for quadcopter aircraft that overcomes bandwidth constraints under malicious attacks. Background Technology

[0002] Unmanned Aerial Vehicles (UAVs) are automated devices that do not require human piloting. They are remotely monitored and controlled by a backend computer and can automatically adjust or fly according to their onboard flight systems. In recent years, research on UAVs has become an unavoidable hot topic in both military and civilian fields, especially the research on micro-UAVs. Various types of aircraft are already in production and use on the market, including fixed-rotor aircraft, helicopters, and rotary-wing UAVs. Multi-rotor aircraft, with their simple mechanical structure and convenient takeoff and landing, are widely used in disaster relief search and rescue, personnel rescue, traffic monitoring, news gathering, target tracking, terrain modeling, agricultural surveying, and animal and plant protection. Multi-rotor aircraft can perform dangerous tasks in low-light and harsh environments.

[0003] Flight control technology for quadcopter UAVs is one of the key research areas in UAV development. It uses direct torque to achieve 6-DOF control, exhibiting characteristics of multivariability, nonlinearity, strong coupling, and sensitivity to disturbances. Furthermore, during flight, the aircraft is simultaneously subjected to multiple physical effects and is easily affected by external environmental disturbances such as atmospheric flow. Model accuracy and sensor precision also impact controller performance, making the design of flight control systems extremely difficult. Attitude stabilization control, the foundation of autonomous navigation for micro-flying platforms, has a crucial impact on the aircraft's flight characteristics; therefore, attitude control is the key to the entire flight control system. The most widely used control algorithm in aircraft attitude control is the PID (Proportion-Integral-Derivative) controller, which is simple and easy to implement. However, its control performance is still unsatisfactory when the controlled object has nonlinear, strongly coupled, and time-varying characteristics. Modern control theory places very high demands on the mathematical models of control systems; however, accurate modeling is very difficult, even impossible, for some control systems. Many existing mathematical models of control systems are not accurate system models, exhibiting inaccuracies in structure and parameters. Model-Free Adaptive Controller (MFAC) is a data-driven controller that does not depend on the system model and has excellent control performance for control systems of quadcopters with nonlinear, strongly coupled, and time-varying parameters. Summary of the Invention

[0004] To address the aforementioned issues in attitude control of quadcopters, and considering the problems of large model disturbances and susceptibility to external interference during quadcopter operation, this invention provides a data-driven control method for quadcopters that overcomes bandwidth constraints under malicious attacks.

[0005] The present invention provides a data-driven control method for a quadcopter that overcomes bandwidth constraints under malicious attacks, comprising the following steps:

[0006] Step 1: Establish a dynamic model of the quadcopter.

[0007] Step 2: Transform the dynamic model obtained in Step 1 into an equivalent dynamic linearized equation using the dynamic linearization method.

[0008] Step 3: Design the criterion function and design the data-driven control method for the quadcopter.

[0009] Step 4: Introduce an event triggering mechanism and consider denial-of-service attacks, and then use the control method obtained in Step 3 to obtain the final quadcopter safety event triggering data-driven control method.

[0010] Furthermore, step 1 specifically involves:

[0011] To facilitate the kinematic analysis and mathematical modeling of quadcopter aircraft, the following assumptions are made:

[0012] (1) A quadcopter is a uniformly symmetrical rigid body.

[0013] (2) The origin of the inertial coordinate system E is located at the same position as the geometric center and the center of mass of the aircraft.

[0014] (3) The drag and gravity experienced by a quadcopter are not affected by the flight altitude and remain constant.

[0015] (4) The thrust of a quadcopter in all directions is directly proportional to the square of the thruster speed.

[0016] Establish the coordinate system for the quadcopter: the body coordinate system is as follows The ground coordinate system is The pitch angle, roll angle, and yaw angle are respectively .

[0017] Taking the center of gravity of the quadcopter as the origin, and with the center of gravity and the lifting surface of the motor on the same plane, the system model is as follows:

[0018]

[0019] in, These represent the position and velocity of the aircraft in the inertial coordinate system, respectively. ; This is the transformation matrix from the body coordinate system to the ground coordinate system; For the mass of the aircraft; The vector of the net external force acting on the aircraft body; The Euler angles of the aircraft; Angular velocity in the body coordinate system; This is the transformation matrix from the triaxial angular velocities around the body axis to Euler angles; The inertial matrix of the aircraft; For aircraft control torque.

[0020] Define the system input as:

[0021]

[0022] In the formula, , , , The lift from each of the four rotors; The lift coefficient, The drag coefficient, For the rotational speed of each rotor, For height control input, For roll control input, For pitch control input, This is the input for yaw control.

[0023] Considering air resistance, the mathematical model of the quadcopter can be obtained from the above equation:

[0024]

[0025] In the formula: , , These are the drag coefficients in each direction; ; , , These are the moments of inertia of the aircraft along its three axes in the body coordinate system. This is the distance between the center of the aircraft fuselage and the center of the propeller. This is the moment of inertia of the motor.

[0026] A quadcopter consists of 3 position variables. and 3 attitude variables The system comprises a six-degree-of-freedom flight system; the quadcopter flight control system is divided into a position loop and an attitude loop, forming a total of four channels: altitude... Channel, yaw angle passage, location With pitch angle Cascaded channels, location With roll angle Cascaded channels; where the height The channel is directly controlled by the quantity Independent control, while yaw angle passage, location With pitch angle Cascaded channels and positions With roll angle The cascade channels are controlled by three sets of cascade controllers.

[0027] The quadcopter attitude loop cascade control consists of two control loops: an outer loop and an inner loop. The outer loop comprises three proportional controllers, with the desired attitude angle as the input. and current attitude angle The output is the attitude angular velocity. The desired input of the inner-loop model-free adaptive controller is the control output of the outer loop. and current attitude angular velocity The output of the cascade attitude controller is Ultimately, combining height Channel output Generate quadcopter input It stabilizes and adjusts the flight attitude of the quadcopter.

[0028] Furthermore, step 2 specifically involves:

[0029] Attitude control employs a cascaded data-driven control strategy, with the inner loop using a model-free adaptive control algorithm and the outer loop using proportional control. For the quadcopter control system, three independent angular velocity control loops are used, each dynamically linearized.

[0030] The pitch channel is linearized as follows:

[0031] definition A vector consisting of control inputs within a time window; where This represents the window duration.

[0032] Therefore, the pitch channel dynamics model is transformed into the following equivalent partial scheme linearized data model:

[0033]

[0034] in, The pitch angular velocity, , For time-varying parameters, It is bounded; ,and This is the input for the pitch angle channel control.

[0035] The roll angle and yaw angle control algorithms are structurally identical to the pitch angle control algorithm.

[0036] Furthermore, step 3 specifically involves:

[0037] In the pitch channel control algorithm, the control input criterion function is designed as follows:

[0038]

[0039] in, This is the weighting factor, and it should be a positive number.

[0040] Minimizing the above criterion function and setting it to 0, we get:

[0041]

[0042] Among them, step size factor , The desired pitch angular velocity.

[0043] Estimation algorithm:

[0044] First, design the estimation criterion function:

[0045]

[0046] in, As a weighting factor; for The estimated value.

[0047] Minimizing the above equation yields The estimation algorithm is as follows:

[0048]

[0049] in, It is the step size factor. yes The estimated value.

[0050] Combining the control algorithm and parameter estimation algorithm described above, the inner-loop data-driven control scheme is presented as follows:

[0051]

[0052]

[0053]

[0054] in, The introduction of the reset algorithm gives the estimation algorithm a stronger tracking capability; for The initial value; It is a sufficiently small positive number.

[0055] The roll angle and yaw angle control algorithms are structurally identical to the pitch angle control algorithm.

[0056] Furthermore, step 4 specifically involves:

[0057] Consider a denial-of-service attack. This represents the attacker's dormant period; the attacker's attack count and total attack time must meet the following conditions:

[0058]

[0059] in, These represent the number of attacks and the total duration of the attacks, respectively. They represent two constants respectively; it is easy to obtain ; These represent two nonlinear functions related to the number of attacks and the duration of the attack, respectively.

[0060] To make full use of bandwidth resources, an event triggering mechanism is introduced.

[0061] The outer loop triggering condition is designed as follows:

[0062]

[0063] in, They are a constant scalar and a positive definite weighted matrix, respectively, and .

[0064] The outer ring safety event-triggered data-driven controller is designed as follows:

[0065]

[0066] in, , , .

[0067] The triggering conditions for the inner ring event are designed as follows:

[0068]

[0069] in, They are a constant scalar and a positive definite weighted matrix, respectively, and .

[0070] The inner-loop security event-triggered data-driven controller is designed as follows:

[0071]

[0072]

[0073]

[0074] in, , , ; ; ; , ; ; ; ; This represents the moment the event was triggered.

[0075] Furthermore, at the pitch angle In a proportional-model-free adaptive cascade controller, pitch angle The desired signal for the proportional controller is subtracted from the actual attitude of the current aircraft to obtain the current pitch angle error. This error is then processed by the proportional controller and output. The desired signal for the next-stage pitch angular velocity controller is obtained after passing through the model-free adaptive controller. As the control input signal for a quadcopter.

[0076] Roll angle In a proportional-model-free adaptive cascade controller, the roll angle The desired signal for the proportional controller is subtracted from the actual attitude of the aircraft to obtain the current roll angle. The error is processed by a proportional controller, and the output is... As the desired signal for the next-stage roll rate controller, after passing through the model-free adaptive controller, we obtain... As the control input signal for a quadcopter.

[0077] Yaw angle In a proportional-model-free adaptive cascade controller, yaw angle The desired signal for the proportional controller is subtracted from the actual attitude of the current aircraft to obtain the current yaw angle error. This error is then processed by the proportional controller and output. As the desired signal for the next-stage yaw rate controller, after passing through the model-free adaptive controller, we obtain... As the control input signal for a quadcopter.

[0078] Furthermore, when using voltage to drive the motor, The value is taken as the voltage value; when using PWM for motor drive... Different PWM pulse widths can be selected.

[0079] The beneficial technical effects of this invention are as follows:

[0080] The quadrotor proportional-model-free adaptive anti-interference attitude cascade controller designed in this invention has a simple structure. It does not require establishing a precise mathematical model of the system; the corresponding control scheme can be implemented based solely on the system's input and output data. This eliminates the dependence of controller design on the mathematical model of the controlled system and other related theoretical challenges. It is not only applicable to quadrotor systems, but also, with appropriate modifications, can be applied to various unmanned aerial vehicle (UAV) systems. This is thanks to the model-independent nature of MFAC and the extended state observer. Furthermore, the introduction of an event-triggered mechanism eliminates the fixed-time mechanism of transmitting sampling signals at each sampling moment, reducing redundant signal updates and transmissions, improving the effective utilization of the system's limited bandwidth resources, and enhancing the aircraft's operating efficiency. Attached Figure Description

[0081] Figure 1 This is a schematic diagram of the aircraft's body coordinate system and the geodetic coordinate system.

[0082] Figure 2 This is a block diagram of the quadcopter control structure of the present invention.

[0083] Figure 3 This is a hardware structure diagram of the control system for an example.

[0084] Figure 4 This is a flowchart illustrating the implementation of attitude cascade control in this invention.

[0085] Figure 5 The image shows a 3D simulation experiment of a quadcopter as an example.

[0086] Figure 6 The pitch angle of the embodiment A curve comparing the expected value with the actual value.

[0087] Figure 7 Roll angle for the example A curve comparing the expected value with the actual value.

[0088] Figure 8 Yaw angle for the embodiment A curve comparing the expected value with the actual value.

[0089] Figure 9 The pitch angle of the embodiment A diagram showing the triggering time and triggering interval of the channel.

[0090] Figure 10 Roll angle for the example A diagram showing the triggering time and triggering interval of the channel.

[0091] Figure 11 Yaw angle for the embodiment A diagram showing the triggering time and triggering interval of the channel. Detailed Implementation

[0092] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0093] All flight maneuvers and missions of a quadcopter are based on the premise that the aircraft has a stable flight attitude, which shows the important role that aircraft attitude control plays in aircraft research.

[0094] This invention combines classic PID control and model-free adaptive control in a cascade control system, fully utilizing the advantages of both control algorithms. Applying the proportional-model-free adaptive cascade control method to quadcopter attitude control first considers the basic structure and dynamic characteristics of the quadcopter. The quadcopter's rotors are distributed in four directions: left front, left rear, right front, and right rear. All four rotors are on the same altitude plane, and their structures and radii are identical. For example... Figure 1 As shown, in the quadcopter, motors 1 and 3 rotate counterclockwise while motors 2 and 4 rotate clockwise. Therefore, when the aircraft is in balanced flight, the gyroscopic effect and aerodynamic torque effect are canceled out. The lift of the quadcopter is achieved by varying the different rotational speeds of the four propellers, and the quadcopter's forward, backward, left, and right movements are achieved by changes in the pitch and roll angles of the fuselage. Therefore, the quadcopter's dynamic model is divided into an attitude-position inner and outer loop control structure, and the control law is also divided into inner loop attitude control and outer loop position control based on this. The specific design principle of the control law (e.g.) Figure 2 As shown below: When the quadcopter is in flight, set the target position and target heading angle. The desired attitude angle is input from the PID controller in the outer position loop to the attitude controller in the inner loop. The inner-loop attitude is the core of this invention, and the inner-loop attitude controller is constructed as a novel proportional-model-free adaptive cascade controller. The attitude controller will take the output of the outer loop... As the controller input, combined with the current actual attitude angle and attitude angular velocity, three identical P-MFAC three-channel controllers are used to output... Combined with position Channel output The control inputs for the four propellers of the quadcopter are generated. This forms a closed-loop control system, enabling the quadcopter to maintain a good flight condition and achieve the desired external and target heading angles. Complete the flight mission.

[0095] The present invention provides a data-driven control method for a quadcopter that overcomes bandwidth constraints under malicious attacks, comprising the following steps:

[0096] Step 1: Establish a dynamic model of the quadcopter.

[0097] To facilitate the kinematic analysis and mathematical modeling of quadcopter aircraft, the following assumptions are made:

[0098] (1) A quadcopter is a uniformly symmetrical rigid body.

[0099] (2) The origin of the inertial coordinate system E is located at the same position as the geometric center and the center of mass of the aircraft.

[0100] (3) The drag and gravity experienced by a quadcopter are not affected by the flight altitude and remain constant.

[0101] (4) The thrust of a quadcopter in all directions is directly proportional to the square of the thruster speed.

[0102] Establish the coordinate system of the quadcopter, such as Figure 1 As shown, the body coordinate system is The ground coordinate system is The pitch angle, roll angle, and yaw angle are respectively .

[0103] Taking the center of gravity of the quadcopter as the origin, and with the center of gravity and the lifting surface of the motor on the same plane, the system model is as follows:

[0104]

[0105] in, These represent the position and velocity of the aircraft in the inertial coordinate system, respectively. ; This is the transformation matrix from the body coordinate system to the ground coordinate system; For the mass of the aircraft; The vector of the net external force acting on the aircraft body; The Euler angles of the aircraft; Angular velocity in the body coordinate system; This is the transformation matrix from the triaxial angular velocities around the body axis to Euler angles; The inertial matrix of the aircraft; For aircraft control torque.

[0106] Define the system input as:

[0107]

[0108] In the formula, , , , The lift from each of the four rotors; The lift coefficient, The drag coefficient, For the rotational speed of each rotor, For height control input, For roll control input, For pitch control input, This is the input for yaw control.

[0109] Considering air resistance, the mathematical model of the quadcopter can be obtained from the above equation:

[0110]

[0111] In the formula: , , These are the drag coefficients in each direction; ; , , These are the moments of inertia of the aircraft along its three axes in the body coordinate system. This is the distance between the center of the aircraft fuselage and the center of the propeller. This is the moment of inertia of the motor.

[0112] A quadcopter consists of 3 position variables. and 3 attitude variables The system comprises a six-degree-of-freedom flight system; the quadcopter flight control system is divided into a position loop and an attitude loop, forming a total of four channels: altitude... Channel, yaw angle passage, location With pitch angle Cascaded channels, location With roll angle Cascaded channels; where the height The channel is directly controlled by the quantity Independent control, while yaw angle passage, location With pitch angle Cascaded channels and positions With roll angle The cascade channels are controlled by three sets of cascade controllers.

[0113] like Figure 4 As shown, the quadcopter attitude loop cascade control consists of two control loops: an outer loop and an inner loop. The outer loop comprises three proportional controllers, with the desired attitude angle as the input. and current attitude angle The output is the attitude angular velocity. The desired input of the inner-loop model-free adaptive controller is the control output of the outer loop. and current attitude angular velocity The output of the cascade attitude controller is Ultimately, combining height Channel output Generate quadcopter input It stabilizes and adjusts the flight attitude of the quadcopter.

[0114] Step 2: Transform the dynamic model obtained in Step 1 into an equivalent dynamic linearized equation using the dynamic linearization method.

[0115] Attitude control employs a cascaded data-driven control strategy, with the inner loop using a model-free adaptive control algorithm and the outer loop using proportional control. For the quadcopter control system, three independent angular velocity control loops are used, each dynamically linearized.

[0116] The pitch channel is linearized as follows:

[0117] definition A vector consisting of control inputs within a time window; where This represents the window duration.

[0118] Therefore, the pitch channel dynamics model is transformed into the following equivalent partial scheme linearized data model:

[0119]

[0120] in, The pitch angular velocity, , For time-varying parameters, It is bounded; ,and This is the input for the pitch angle channel control.

[0121] The roll angle and yaw angle control algorithms are structurally identical to the pitch angle control algorithm.

[0122] Step 3: Design the criterion function and design the data-driven control method for the quadcopter.

[0123] In the pitch channel control algorithm, the control input criterion function is designed as follows:

[0124]

[0125] in, This is the weighting factor, and it should be a positive number.

[0126] Minimizing the above criterion function and setting it to 0, we get:

[0127]

[0128] Among them, step size factor , The desired pitch angular velocity.

[0129] Estimation algorithm:

[0130] First, design the estimation criterion function:

[0131]

[0132] in, As a weighting factor; for The estimated value.

[0133] Minimizing the above equation yields The estimation algorithm is as follows:

[0134]

[0135] in, It is the step size factor. yes The estimated value.

[0136] Combining the control algorithm and parameter estimation algorithm described above, the inner-loop data-driven control scheme is presented as follows:

[0137]

[0138]

[0139]

[0140] in, The introduction of the reset algorithm gives the estimation algorithm a stronger tracking capability; for The initial value; It is a sufficiently small positive number.

[0141] The roll angle and yaw angle control algorithms are structurally identical to the pitch angle control algorithm.

[0142] Step 4: Introduce an event triggering mechanism and consider denial-of-service attacks, and then use the control method obtained in Step 3 to obtain the final quadcopter safety event triggering data-driven control method.

[0143] Consider a denial-of-service attack. This represents the attacker's dormant period; the attacker's attack count and total attack time must meet the following conditions:

[0144]

[0145] in, These represent the number of attacks and the total duration of the attacks, respectively. They represent two constants respectively; it is easy to obtain ; These represent two nonlinear functions related to the number of attacks and the duration of the attack, respectively.

[0146] To make full use of bandwidth resources, an event triggering mechanism is introduced.

[0147] The outer loop triggering condition is designed as follows:

[0148]

[0149] in, They are a constant scalar and a positive definite weighted matrix, respectively, and .

[0150] The outer ring safety event-triggered data-driven controller is designed as follows:

[0151]

[0152] in, , , .

[0153] The triggering conditions for the inner ring event are designed as follows:

[0154]

[0155] in, They are a constant scalar and a positive definite weighted matrix, respectively, and .

[0156] The inner-loop security event-triggered data-driven controller is designed as follows:

[0157]

[0158]

[0159]

[0160] in, , , ; ; ; , ; ; ; ; This represents the moment the event was triggered.

[0161] Example:

[0162] The hardware structure diagram of the quadcopter designed in this invention is as follows: Figure 3As shown, the microcontroller is the core of the entire architecture; the power module provides power to the quadcopter and is crucial for its continuous flight time; the remote controller and ground station transmit control signals to the controller via a wireless transmission module, and the quadcopter measurement module transmits measurement information to the controller, which is also transmitted wirelessly to the ground station; after passing through the controller, the microcontroller transmits motor control signals to the motor drive module, driving the rotor motors to achieve the flight mission. The 3D simulation of the quadcopter is as follows... Figure 5 As shown.

[0163] To verify the proportional-model-free adaptive cascade control scheme proposed in this invention, a MATLAB simulation experiment was conducted. The simulation results are shown in the figure. Figure 6 , Figure 7 , Figure 8 As shown in the simulation comparison curves, under the present invention, the quadcopter attitude control can be adjusted basically according to the desired attitude, and has a better control effect than the traditional P-PID attitude control scheme. This verifies the feasibility and effectiveness of the cascade control method proposed in this invention to achieve precise and high-quality flight control of quadcopter UAVs.

[0164] Figure 9 , Figure 10 , Figure 11 The trigger times and intervals for the three channels are shown respectively. The height of the lever represents the interval between two adjacent trigger times, and the position of the lever represents the trigger time. It can be seen that, while ensuring the stability of the aircraft, the event triggering mechanism proposed in this invention significantly reduces the update frequency of system signals and saves system bandwidth resources.

[0165] Finally, it should be noted that the above are preferred embodiments of the present invention. Although the present invention has been described in detail through the above preferred embodiments, those skilled in the art should understand that various changes in form and detail can be made without departing from the scope claimed by the claims of the present invention.

Claims

1. A data-driven control method for a quadcopter that overcomes bandwidth constraints under malicious attacks, characterized in that, Includes the following steps: Step 1: Establish a dynamic model of the quadcopter; Step 2: Transform the dynamic model obtained in Step 1 into an equivalent dynamic linearized equation using the dynamic linearization method; Step 3: Design the criterion function and design the data-driven control method for the quadcopter; Step 4: Introduce an event triggering mechanism and consider denial-of-service attacks, and then use the control method obtained in Step 3 to obtain the final quadcopter safety event triggering data-driven control method.

2. The data-driven control method for a quadcopter to overcome bandwidth constraints under malicious attacks as described in claim 1, characterized in that, Step 1 specifically involves: To facilitate the kinematic analysis and mathematical modeling of quadcopter aircraft, the following assumptions are made: (1) A quadcopter is a uniformly symmetrical rigid body; (2) The origin of the inertial coordinate system E is located at the same position as the geometric center and the center of mass of the aircraft; (3) The drag and gravity experienced by a quadcopter are not affected by flight altitude and remain constant. (4) The thrust of a quadcopter in all directions is directly proportional to the square of the thruster speed; Establish the coordinate system for the quadcopter: the body coordinate system is as follows The ground coordinate system is ; Pitch angle, roll angle, and yaw angle are respectively ; Taking the center of gravity of the quadcopter as the origin, and with the center of gravity and the lifting surface of the motor on the same plane, the system model is as follows: ； in, These represent the position and velocity of the aircraft in the inertial coordinate system, respectively. ; This is the transformation matrix from the body coordinate system to the ground coordinate system; For the mass of the aircraft; The vector of the net external force acting on the aircraft body; The Euler angles of the aircraft; Angular velocity in the body coordinate system; This is the transformation matrix from the triaxial angular velocities around the body axis to Euler angles; The inertial matrix of the aircraft; For aircraft control torque; Define the system input as: ； In the formula, , , , The lift generated by each of the four rotors; The lift coefficient, The drag coefficient, For the rotational speed of each rotor, For height control input, For roll control input, For pitch control input, This is the yaw control input; Considering air resistance, the mathematical model of the quadcopter can be obtained from the above equation: ； In the formula: , , These are the drag coefficients in each direction; ; , , These are the moments of inertia of the aircraft along its three axes in the body coordinate system. This is the distance between the center of the aircraft fuselage and the center of the propeller. This is the moment of inertia of the motor. A quadcopter consists of 3 position variables. and 3 attitude variables The system comprises a six-degree-of-freedom flight system; the quadcopter flight control system is divided into a position loop and an attitude loop, forming a total of four channels: altitude... Channel, yaw angle passage, location With pitch angle Cascaded channels, location With roll angle Cascaded channels; where the height The channel is directly controlled by the quantity Independent control, while yaw angle passage, location With pitch angle Cascaded channels and positions With roll angle The cascaded channels are controlled by three sets of cascade controllers respectively; The quadcopter attitude loop cascade control consists of two control loops: an outer loop and an inner loop. The outer loop comprises three proportional controllers, with the desired attitude angle as the input. and current attitude angle The output is the attitude angular velocity. The desired input of the inner-loop model-free adaptive controller is the control output of the outer loop. and current attitude angular velocity The output of the cascade attitude controller is Ultimately, combining height Channel output Generate quadcopter input It stabilizes and adjusts the flight attitude of the quadcopter.

3. The data-driven control method for a quadcopter to overcome bandwidth constraints under malicious attacks according to claim 2, characterized in that, Step 2 specifically involves: Attitude control adopts a cascaded data-driven control strategy, with the inner loop using a model-free adaptive control algorithm and the outer loop using proportional control. For the quadcopter control system, three independent angular velocity control loops are used for dynamic linearization. The pitch channel is linearized as follows: definition A vector consisting of control inputs within a time window; where This refers to the window duration. Therefore, the pitch channel dynamics model is transformed into the following equivalent partial scheme linearized data model: ； in, The pitch angular velocity, , For time-varying parameters, It is bounded; ,and For pitch angle channel control input; The roll angle and yaw angle control algorithms are structurally identical to the pitch angle control algorithm.

4. The data-driven control method for a quadcopter to overcome bandwidth constraints under malicious attacks according to claim 3, characterized in that, Step 3 specifically involves: In the pitch channel control algorithm, the control input criterion function is designed as follows: ； in, The weighting factor is a positive number. Minimizing the above criterion function and setting it to 0, we get: ； Among them, step size factor , The desired pitch rate; Estimation algorithm: First, design the estimation criterion function: ； in, As a weighting factor; for The estimated value; Minimizing the above equation, we get The estimation algorithm is as follows: ； in, It is the step size factor. yes The estimated value; Combining the control algorithm and parameter estimation algorithm described above, the inner-loop data-driven control scheme is presented as follows: ； ； ； in, The introduction of the reset algorithm gives the estimation algorithm a stronger tracking capability; for Initial value; It is a sufficiently small positive number; The roll angle and yaw angle control algorithms are structurally identical to the pitch angle control algorithm.

5. The data-driven control method for a quadcopter to overcome bandwidth constraints under malicious attacks according to claim 4, characterized in that, Step 4 specifically involves: Consider a denial-of-service attack. This represents the attacker's dormant period; the attacker's attack count and total attack time must meet the following conditions: ； in, These represent the number of attacks and the total duration of the attacks, respectively. They represent two constants respectively; it is easy to obtain ; These represent two nonlinear functions related to the number of attacks and the duration of the attack, respectively. To fully utilize bandwidth resources, an event-triggered mechanism is introduced; The outer loop triggering condition is designed as follows: ； in, They are a constant scalar and a positive definite weighted matrix, respectively, and ; The outer ring safety event-triggered data-driven controller is designed as follows: ； in, , , ; The triggering conditions for the inner ring event are designed as follows: ； in, They are a constant scalar and a positive definite weighted matrix, respectively, and ; The inner-loop security event-triggered data-driven controller is designed as follows: ； 6. ； in, , , ; ; ; , ; ; ; ; This represents the moment the event was triggered.

7. The data-driven control method for a quadcopter to overcome bandwidth constraints under malicious attacks according to claim 2, characterized in that, At pitch angle In a proportional-model-free adaptive cascade controller, pitch angle The desired signal for the proportional controller is subtracted from the actual attitude of the current aircraft to obtain the current pitch angle error. This error is then processed by the proportional controller and output. The desired signal for the next-stage pitch angular velocity controller is obtained after passing through the model-free adaptive controller. As the control input signal for a quadcopter; Roll angle In a proportional-model-free adaptive cascade controller, the roll angle The desired signal for the proportional controller is subtracted from the actual attitude of the aircraft to obtain the current roll angle. The error is processed by a proportional controller, and the output is... As the desired signal for the next-stage roll rate controller, after passing through the model-free adaptive controller, we obtain... As the control input signal for a quadcopter; Yaw angle In a proportional-model-free adaptive cascade controller, yaw angle The desired signal for the proportional controller is subtracted from the actual attitude of the current aircraft to obtain the current yaw angle error. This error is then processed by the proportional controller and output. As the desired signal for the next-stage yaw rate controller, after passing through the model-free adaptive controller, we obtain... As the control input signal for a quadcopter.

8. The data-driven control method for a quadcopter to overcome bandwidth constraints under malicious attacks according to claim 2, characterized in that, When using voltage to drive the motor The value is taken as the voltage value; when using PWM for motor drive... Different PWM pulse widths can be selected.

Images