A sliding mode adaptive control method for quadrotor UAV with multi-source uncertainty

Abstract

The present invention discloses a sliding mode adaptive control method for a quadrotor unmanned aerial vehicle (UAV) under multi-source uncertainty. The method comprises: combining the UAV's own resistance, external unknown disturbances, and unknown functions in the system model to calculate the total disturbance experienced by the UAV; constructing a motion model of the UAV based on the effects of actuator failures and the total disturbance; obtaining the expected value of the UAV's motion target; converting the expected value of the motion target into a smoothing process to calculate the expected value of the system; establishing a state-space representation of the UAV system; designing a hyper-distorted extended state observer to observe unknown disturbances in the system; utilizing adaptive rate design to obtain an estimated value of the fault in the system model; and calculating the control parameters of the system based on the estimated value. The extended state observer and sliding mode adaptive controller are designed based on the characteristics of each disturbance, addressing the problem of insufficient disturbance rejection.

Description

Technical Field

[0001] The present invention relates to the field of unmanned aerial vehicle (UAV) control technology, and in particular to a multi-source uncertainty sliding mode adaptive control method for a quad-rotor UAV. Background Art

[0002] In recent years, drones have gradually replaced manned aircraft in various scenarios due to their unique advantages. They can perform various tasks in environments with unknown risks and inaccessible to humans. Compared with fixed-wing drones, rotary-wing drones can take off and land vertically, are maneuverable and flexible in flight, are easy to maintain, and are small in size for easy portability and use. They have broad development prospects and are currently widely used in payload transportation, urban monitoring, environmental monitoring, precision agriculture and other fields.

[0003] A quadrotor drone is a typical highly nonlinear, strongly coupled, multivariable underactuated system, and is extremely susceptible to various uncertain factors, such as time-varying wind disturbances, model uncertainty, actuator failures, input saturation, low state detection accuracy, and input delays. These factors increase the difficulty of controlling the drone in stable flight.

[0004] In existing technologies, traditional linear control methods and more complex nonlinear control strategies are commonly used to address quadrotor stability and trajectory tracking issues. The most common linear control method, PID, is flexible, easy to implement, and has low reliance on the controlled object model. Therefore, it is widely used in industrial control systems. Among nonlinear control methods, feedback linearization is a common technique for quadrotor control. This method transforms the nonlinear model of the quadrotor into an equivalent linear model by designing a feedforward control. The controller is then designed using linear system controller design methods. Ultimately, the nonlinear feedforward control and linear controller are combined to form a complete nonlinear controller. Many nonlinear control methods have been applied to quadrotors, such as sliding mode control, backstepping control, and predictive control. However, nonlinear control methods cannot effectively suppress unknown disturbances in the presence of multiple sources of uncertainty, and they cannot specifically address actuator failures in quadrotor drones. Summary of the Invention

[0005] An embodiment of the present invention provides a multi-source uncertain downward mode adaptive control method for a quadrotor unmanned aerial vehicle (UAV) to solve the technical problems in the prior art that a quadrotor unmanned aerial vehicle (UAV) cannot effectively suppress unknown disturbances and cannot specifically solve the actuator failure problems of a quadrotor unmanned aerial vehicle (UAV).

[0006] An embodiment of the present invention provides a method for adaptively controlling a quadrotor unmanned aerial vehicle (UAV) with multi-source uncertainty sliding modes, comprising:

[0007] The total disturbance to the quadrotor is calculated by combining the quadrotor's own resistance, external unknown disturbances, and unknown functions in the system model.

[0008] Construct a motion model of a quadrotor UAV based on the effects of actuator failure and total disturbance;

[0009] Obtaining an expected value of a motion target of the quadrotor drone;

[0010] Convert the expected value of the moving target into a smoothing process and calculate the expected value of the system;

[0011] Establish the state space expression of the quadrotor UAV system;

[0012] Design a super-distorted expansion state observer to observe unknown disturbances in the system;

[0013] By using the design of adaptive rate, the estimated value of the fault in the system model is obtained, and the control parameters of the system are calculated based on the estimated value;

[0014] The total disturbance to which the quadrotor drone is subjected is calculated by combining the quadrotor drone's own resistance, external unknown disturbances, and unknown functions in the system model, including:

[0015] Establish a dynamic model of a quadrotor drone;

[0016] Based on the quadrotor UAV dynamics model, a quadrotor UAV dynamics model with multi-source uncertainty Lagrangian form is established;

[0017] The total disturbance to which the quadrotor drone is subjected is calculated by combining the quadrotor drone's own resistance, external unknown disturbances, and unknown functions in the system model, including:

[0018] Establish a dynamic model of a quadrotor drone;

[0019] Based on the quadrotor UAV dynamics model, a quadrotor UAV dynamics model with multi-source uncertainty Lagrangian form is established;

[0020] The method of constructing a motion model of a quadrotor drone based on the influence of actuator failure and total disturbance includes:

[0021] Modeling multiplicative or additive uncertainty in actuator output based on the effects of common actuator faults;

[0022] Construct a motion model of a quadrotor drone based on a multiplicative or additive uncertainty model and a total disturbance;

[0023] The quadrotor UAV dynamics model includes:

[0024] ,

[0025] ,

[0026] ,

[0027] in, is the position of the quadrotor drone, are the roll angle, pitch angle and yaw angle of the quadrotor drone, are the speeds of the quadrotor drone on the x-axis, y-axis, and z-axis respectively. are the roll angular velocity, pitch angular velocity and yaw angular velocity of the quadrotor drone, is the total weight of the quadrotor drone, is the acceleration due to gravity, is the control quantity of the quadrotor drone, where is the position loop control quantity of the quadrotor drone, is the attitude loop control value of the quadrotor drone, is the propeller's moment of inertia, For the The speed of the propeller, is the drag coefficient, is the resulting force vector relative to the Earth-fixed frame, is the velocity vector relative to the Earth-fixed frame, is the acceleration vector relative to the Earth-fixed frame, is the resultant moment vector, is the angular velocity vector relative to the fixed coordinate system of the drone, is the angular acceleration vector relative to the fixed coordinate system of the drone, is the diagonal inertia matrix;

[0028] When the axis of the fuselage fixed frame coincides with the main inertial axis of the quadrotor drone, is a static transformation matrix from the quadcopter frame to the earth-fixed frame:

[0029] ,

[0030] The dynamic transformation matrix is ​​established through the relationship between Euler angular rate and angular velocity:

[0031] in:

[0032] ,

[0033] in, For the conversion matrix, set the change of roll angle and pitch angle to be less than the set change threshold so that Close to the identity matrix, using Euler angular rate instead of angular velocity, we get:

[0034] ,

[0035] in, , For quadrotor drones Axial acceleration, The acceleration in the axial direction and The acceleration in the axial direction, are the roll acceleration, pitch acceleration, and yaw acceleration of the quadrotor drone;

[0036] The quadrotor UAV dynamics model based on the quadrotor UAV dynamics model is established in a multi-source uncertainty Lagrangian form, including:

[0037] The unknown external disturbance to the quadrotor UAV system is expressed as: , where D1, D2…D6 correspond to the external disturbances of the six system states of the quadrotor drone;

[0038] Combining the quadrotor's own resistance, the external unknown disturbance D, and the unknown function in the system model, the total disturbance to the quadrotor is expressed as: ,in:

[0039] ,

[0040] Assumption: For the total disturbance of the quadrotor, is continuously differentiable, and is bounded, that is ,in is a positive constant;

[0041] Accordingly, the multiplication or addition uncertainty model in the actuator output is established according to the influence of common actuator faults, including:

[0042] Create the following model:

[0043] ,

[0044] in In order to eliminate the influence of external uncertainty on the position loop control of the quadrotor drone, It is the attitude control quantity of the quadrotor drone without the influence of external uncertainty, where is the roll angle direction control quantity, is the pitch angle control quantity, is the yaw angle direction control quantity, Indicates the multiplicative loss of position loop thrust caused by the fault, It represents the multiplicative loss of torque in the roll angle direction caused by the fault, It represents the multiplicative loss of the pitch angle torque caused by the fault, represents the multiplicative loss of torque in the yaw angle direction, The additive parameter representing the position loop deviation caused by the fault, It is an additive parameter used to represent the pitch angle deviation caused by the fault. It is an additive parameter used to represent the pitch angle deviation caused by the fault. It is an additive parameter used to represent the yaw angle deviation caused by the fault. The fault parameter and The real-time value of is unknown, and τ is the attitude loop control variable;

[0045] Accordingly, the motion model of the constructed quadrotor drone is as follows:

[0046]

[0047] Furthermore, the conversion of the expected value of the moving target into a smoothing process and calculation of the expected value of the system include:

[0048] According to the desired yaw angle , determine the new x-axis, y-axis and z-axis control inputs;

[0049] Determine the state space of the quadrotor based on the new x-axis, y-axis, and z-axis control inputs to determine the desired trajectory;

[0050] A desired vector is defined using a desired roll angle, a desired pitch angle, and a desired trajectory;

[0051] The expected vector is input into a nonlinear tracking differentiator to obtain a smoothed system expected value.

[0052] Furthermore, the design of the super-distorted expansion state observer is used to observe unknown disturbances in the system, including:

[0053] Expanding the state space expression of the quadrotor UAV system;

[0054] By using the designed super-distorted extended state observer, adjusting the adjustable gain of the extended state observer, the estimated error system of the extended state observer is obtained, and the unknown disturbance in the observation system is calculated.

[0055] Furthermore, the design of the adaptive rate is used to obtain an estimated value of the fault in the system model, and the control parameters of the system are calculated based on the estimated value, including:

[0056] The unknown faults and corresponding parameters in the system are obtained by the adaptive law, and the control quantities of the position loop and the attitude loop of the quadrotor UAV are expressed by the designed global fast terminal sliding surface according to the position faults and corresponding parameters in the system.

[0057] Furthermore, the desired yaw angle , determine the new x-axis, y-axis and z-axis control inputs, including:

[0058] According to the received yaw angle expected value and the desired target point, respectively define the quadrotor drone model axis, Axis and Axis control amount for:

[0059] ,

[0060] According to the characteristics of the control quantity of the actuator error with unknown parameters, new axis, Axis and The axis control inputs are and , and meet the following conditions:

[0061] ,

[0062] The state space expression of the quadrotor drone is determined based on the new x-axis, y-axis and z-axis control inputs:

[0063] ,

[0064] in, , ,

[0065] , , , ;

[0066] , ,

[0067] ,

[0068] ,

[0069] , The intermediate amount calculated by the above formula is, is the system state vector composed of the six-dimensional state of the UAV system, is the derivative of the system state vector composed of the six-dimensional state of the UAV system, is the derivative of the UAV system state vector, It is the second-order derivative of the system state vector composed of the six-dimensional state of the UAV system, expressed as the speed and angular velocity of the system, is the vector of the drone's gravity conversion, is the vector transformed from the multiplicative coefficient of the actuator fault, is the matrix of actuator fault additive coefficient transformation, and is the attitude loop control variable of the quadrotor drone, where is the control quantity of the roll angle direction of the quadrotor drone, is the control quantity in the pitch angle direction, is the control quantity in the yaw angle direction;

[0070] Determining the state space of the quadrotor drone based on the new x-axis, y-axis, and z-axis control inputs and determining the desired trajectory includes:

[0071] Determine the desired roll angle based on the new x-, y-, and z-axis control inputs and pitch angle , according to the desired roll angle and pitch angle Determine the desired trajectory:

[0072] ,

[0073] Defining a desired vector using a desired roll angle, a desired pitch angle, and a desired trajectory includes:

[0074] Generate the desired vector using the desired trajectory and the desired attitude angle:

[0075] ,

[0076] The nonlinear tracking differentiator comprises:

[0077] ,

[0078] Among them, the nonlinear function The definition is as follows:

[0079] ,

[0080] in, is the required deflection angle The tracking signal, yes The differential signal of and are the speed factor and filter factor of the tracking differentiator, , is a nonlinear function The intermediate parameters of , used to calculate nonlinear functions.

[0081] Furthermore, the state space expression of the expanded quadrotor drone system is:

[0082] ,

[0083] in, , ,

[0084] The super-twisted expansion state observer is:

[0085] ,

[0086] in, is the output state of the extended state observer, is the velocity observation value of the extended state observer, is the perturbation observation value of the extended state observer, is the adjustable gain of the extended state observer, assuming that the estimated error state is and ;

[0087] The detailed form of the superwarp algorithm is:

[0088] ,

[0089] , , , , ,

[0090] The designed super-distorted extended state observer is used to adjust the adjustable gain of the extended state observer to obtain the estimated error system of the extended state observer, and the unknown disturbance existing in the observation system is calculated, which is achieved by the following method:

[0091] ,

[0092] Accordingly, the unknown faults and corresponding parameters in the system are obtained by the adaptive law, and according to the position faults in the system and the corresponding parameters, the following are included:

[0093] set up 、 、 and is the unknown parameter of the fault in the system 、 、 and The online estimation of is designed by the following adaptive law:

[0094] ,

[0095] in, is a positive constant, is the transpose of the position loop sliding mode surface quantity, It is the transpose of the attitude loop sliding mode face quantity;

[0096] The control quantities designed by using the designed global fast terminal sliding surface to express the control quantities of the position loop and the attitude loop of the quadrotor drone include:

[0097] The drone is designed with the following control quantities:

[0098] ,

[0099] in, is the control quantity of the quadrotor UAV position loop, where For quadrotor drones Control amount in the axial direction, for The control amount in the axis direction and for Control quantity in the axis direction; is the attitude loop control variable of the quadrotor drone, where is the control quantity of the roll angle direction of the quadrotor drone, is the control quantity in the pitch angle direction and is the control quantity in the yaw angle direction, and is the designed global fast terminal sliding surface, where For quadrotor drones Sliding surface in the axial direction, for Sliding surface in the axial direction, for Axial synovial surface, is the sliding surface of the quadrotor UAV in the rolling angle direction, is the sliding surface in the pitch angle direction and is the sliding membrane surface in the yaw angle direction, and are the position and velocity tracking errors. and is the tracking error of angular velocity and angular acceleration. and are greater than 0, and yes , is a strictly positive constant of the control gain of the controller, , , , ,

[0100] ,

[0101] in:

[0102] ,

[0103] ,

[0104] ,

[0105] ,

[0106] ,

[0107] ,

[0108] ,

[0109] .

[0110] The embodiment of the present invention provides a multi-source uncertain downward mode adaptive control method for a quadrotor unmanned aerial vehicle. The method combines the self-resistance of the quadrotor unmanned aerial vehicle with external unknown disturbances and unknown functions in the system model to calculate the total disturbance experienced by the quadrotor unmanned aerial vehicle; constructs a motion model of the quadrotor unmanned aerial vehicle according to the influence of actuator failure and the total disturbance; obtains the expected value of the motion target of the quadrotor unmanned aerial vehicle; converts the expected value of the motion target into a smoothing process to calculate the expected value of the system; establishes a state space expression of the quadrotor unmanned aerial vehicle system; designs a super-distorted expansion state observer for observing unknown disturbances in the system; utilizes the design of an adaptive rate to obtain an estimated value of the fault in the system model, and calculates the control parameters of the system based on the estimated value. By combining the quadrotor's inherent resistance with unknown external disturbances and unknown functions in the system model, and taking into account the effects of actuator failures, the quadrotor's motion model is modified. The control variables are optimized based on the various actuator failure parameters. The desired attitude angle is determined based on the optimized control variables, and the desired trajectory is determined based on the desired attitude angle. The trajectory vector is defined using the desired trajectory and smoothed. An optimized hyper-warped extended state observer is used to obtain the extended state observer's estimated error system, and the adaptive rate is used to design the system's internal fault parameters and obtain the quadrotor's control variables. An extended state observer and sliding mode adaptive controller are designed based on the characteristics of each disturbance, addressing the problem of insufficient disturbance rejection. BRIEF DESCRIPTION OF THE DRAWINGS

[0111] Other features, objects and advantages of the present invention will become more apparent upon reading the detailed description of non-limiting embodiments made with reference to the following drawings:

[0112] Figure 1 1 is a flow chart of a multi-source uncertainty sliding mode adaptive control method for a quadrotor UAV provided by an embodiment of the present invention;

[0113] Figure 2 The invention provides a quadrotor drone with a multi-source uncertainty sliding mode adaptive control method. Axis direction position loop control quantity diagram;

[0114] Figure 3 The invention provides a quadrotor drone with a multi-source uncertainty sliding mode adaptive control method. Axis direction position loop control quantity diagram;

[0115] Figure 4 The invention provides a quadrotor drone with a multi-source uncertainty sliding mode adaptive control method. Axis direction position loop control quantity diagram;

[0116] Figure 5 It is a graph of attitude loop control quantities in the roll angle direction of a quadrotor drone in the multi-source uncertainty sliding mode adaptive control method for a quadrotor drone provided by an embodiment of the present invention;

[0117] Figure 6 It is a graph of attitude loop control quantities in the pitch angle direction of a quadrotor drone in the multi-source uncertainty sliding mode adaptive control method for a quadrotor drone provided by an embodiment of the present invention;

[0118] Figure 7 It is a graph of attitude loop control quantities in the yaw angle direction of a quadrotor drone in the multi-source uncertainty sliding mode adaptive control method for a quadrotor drone provided by an embodiment of the present invention;

[0119] Figure 8 This is a position state curve diagram of a quadrotor drone using the quadrotor drone multi-source uncertainty downward mode adaptive control method provided by an embodiment of the present invention;

[0120] Figure 9 This is an angular state curve diagram of a quadrotor drone using the quadrotor drone multi-source uncertainty downward mode adaptive control method provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0121] The present invention will be further described in detail below with reference to the accompanying drawings and examples. It will be understood that the specific embodiments described herein are intended only to illustrate the present invention and are not intended to limit the present invention. It should also be noted that, for ease of description, the accompanying drawings only illustrate portions relevant to the present invention, not all structures.

[0122] Example 1

[0123] Figure 1 : This is a flow chart of a method for adaptively controlling a multi-source uncertain sliding mode of a quadrotor drone provided by the first embodiment of the present invention. This embodiment is applicable to the case of adaptively controlling a multi-source uncertain sliding mode of a quadrotor drone against various interferences, and specifically includes the following steps:

[0124] In step 110, the quadrotor drone's own resistance, external unknown disturbances, and unknown functions in the system model are combined to calculate the total disturbance experienced by the quadrotor drone. A motion model of the quadrotor drone is constructed based on the impact of the actuator failure and the total disturbance.

[0125] Exemplarily, it may include: establishing a quadrotor UAV dynamics model; and establishing a quadrotor UAV dynamics model in a multi-source uncertainty Lagrangian form based on the quadrotor UAV dynamics model.

[0126] First, establish the dynamic model of the quadrotor drone:

[0127] ,

[0128] , ,

[0129] in, is the position of the quadrotor drone, are the roll angle, pitch angle and yaw angle of the quadrotor drone, are the speeds of the quadrotor drone on the x-axis, y-axis, and z-axis respectively. are the roll angular velocity, pitch angular velocity and yaw angular velocity of the quadrotor drone, is the total weight of the quadrotor drone, is the acceleration due to gravity, is the position loop control quantity of the quadrotor drone, are the roll angle control value, pitch angle control value and yaw angle control value of the quadrotor drone, is the propeller's moment of inertia, For the The speed of the propeller, is the drag coefficient, is the resulting force vector relative to the Earth-fixed frame, where for The resulting force vector in the direction of the axis relative to the Earth-fixed frame, for The resulting force vector in the direction of the axis relative to the Earth-fixed frame, for The resulting force vector is oriented with respect to the Earth-fixed frame.

[0130] is the velocity vector relative to the Earth-fixed frame, is the acceleration vector relative to the Earth-fixed frame, is the resultant moment vector, where is the moment vector in the roll angle direction, is the pitch angle direction torque vector, is the torque vector in the yaw angle direction, is the angular velocity vector relative to the fixed coordinate system of the drone, is the roll angular velocity vector relative to the fixed coordinate system of the UAV, is the pitch angular velocity vector relative to the fixed coordinate system of the UAV, is the yaw angular velocity vector relative to the fixed coordinate system of the UAV, is the angular acceleration vector relative to the fixed coordinate system of the drone, is the diagonal inertia matrix, where is the moment of inertia in the roll angle direction, is the moment of inertia in the pitch angle direction, is the moment of inertia in the yaw angle direction. In the embodiment of the present invention, represent The derivative of , for example, yes The derivative of , assuming that the axis of the fuselage fixed frame coincides with the inertial axis of the quadrotor drone, It is a static transformation matrix from the quadcopter frame to the earth fixed frame, which is defined as follows:

[0131] ,

[0132] Get the dynamic transformation matrix of the relationship between Euler angular rate and angular velocity:

[0133] in,

[0134] .

[0135] Usually, in order to facilitate the design of the controller, the changes in the roll angle and pitch angle are very small, and the changes in the roll angle and pitch angle are set to be less than the set change threshold. Therefore, is close to the unit matrix. The angular velocity can be replaced by the Euler angular rate. Based on this, the dynamic transformation matrix of the relationship between the Euler angular rate and the angular velocity is substituted into the quadrotor UAV dynamics model, and the following results are obtained:

[0136] ,

[0137] in, For quadrotor drones Axial acceleration, The acceleration in the axial direction and The acceleration in the axial direction, are the roll acceleration, pitch acceleration, and yaw acceleration of the quadrotor drone.

[0138] Accordingly, the method of constructing a motion model of a quadrotor drone based on the impact of actuator failure and the total disturbance may include: establishing a multiplication or addition uncertainty model in the actuator output based on the impact of common actuator failures; determining the constraint conditions of the control quantity based on the extreme conditions of the failure; and constructing a motion model of the quadrotor drone based on the multiplication or addition uncertainty model and the total disturbance.

[0139] First, the unknown external disturbance to the quadrotor UAV system can be expressed as: , where D1, D2...D6 correspond to the external disturbances of the six system states of the quadrotor drone.

[0140] Combining the quadrotor's own resistance, the external unknown disturbance D, and the unknown function in the system model, the total disturbance to the quadrotor is expressed as: ,in:

[0141] .

[0142] For the total disturbance of the quadrotor, is continuously differentiable, and is bounded, that is ,in is a positive constant, It is a quad-rotor drone in The total disturbance in the axial direction is It is a quad-rotor drone in The total disturbance in the axial direction is It is a quad-rotor drone in The total disturbance in the axial direction is is the total disturbance of the quadrotor drone in the roll angle direction, is the total disturbance of the quadrotor drone in the pitch angle direction, It is the total disturbance received by the quadrotor UAV in the yaw angle direction.

[0143] The rotors, acting as the actuators of the quadrotor control system, are subject to the possibility of failures such as low power supply voltage or blade damage. The impact of common actuator failures can be modeled as multiplicative or additive uncertainty in the actuator output, as shown below:

[0144] ,

[0145] in In order to eliminate the influence of external uncertainty on the position loop control quantity and It is the attitude control quantity of the quadrotor drone without the influence of external uncertainty, where is the roll angle direction control quantity, is the pitch angle control quantity, is the yaw angle direction control quantity, Indicates the multiplicative loss of position loop thrust caused by the fault, It represents the multiplicative loss of torque in the roll angle direction caused by the fault, It represents the multiplicative loss of the pitch angle torque caused by the fault, represents the multiplicative loss of torque in the yaw angle direction, The additive parameter representing the position loop deviation caused by the fault, It is an additive parameter used to represent the pitch angle deviation caused by the fault. It is an additive parameter used to represent the pitch angle deviation caused by the fault. It is an additive parameter used to represent the yaw angle deviation caused by the fault. Since the flight control system is not equipped with a fault detection or identification device, the fault parameter and The real-time value of is unknown.

[0146] Based on the above, the quadcopter model can be rewritten as:

[0147] ,

[0148] Using the above steps, the motion model of the quadrotor drone can be obtained.

[0149] Step 120: Obtain the expected value of the motion target of the quadrotor drone, convert the expected value of the motion target into a smoothing process, and calculate the expected value of the system.

[0150] Exemplarily, the method may include converting the expected value of the moving target into a smoothing process and calculating the expected value of the system, including:

[0151] According to the desired roll angle , determine new x-axis, y-axis and z-axis control inputs; determine the state space expression of the quadrotor drone based on the new x-axis, y-axis and z-axis control inputs, and determine the desired orbit; define the desired vector using the desired roll angle, the desired pitch angle and the desired orbit; input the desired vector into the nonlinear tracking differentiator to obtain the smoothed system expected value.

[0152] Receive the expected value of the motion target of the quadrotor drone, that is, Target expected value in the axis direction , Target expected value in the axis direction , Target expected value in the axis direction and the target expected value of the yaw angle First, the model axis, Axis and The axis control quantities are defined as , as shown below:

[0153] ,

[0154] Since the system's actuator has a fault, considering the control quantity with unknown parameters, a new axis, Axis and The axis control input is and , respectively satisfying:

[0155] ,

[0156] According to the above control variables, the state space expression of the quadrotor drone can be rewritten as:

[0157] ;

[0158] in:

[0159] , ;

[0160] , , , ;

[0161] , ;

[0162] ;

[0163] ;

[0164] , The intermediate amount calculated by the above formula is, is the system state vector composed of the six-dimensional state of the UAV system, is the derivative of the UAV system state vector, expressed as the velocity and angular velocity of the system, is the vector of the drone's gravity conversion, is the vector transformed from the multiplicative coefficient of the actuator fault, It is the matrix of actuator fault additive coefficient transformation.

[0165] According to the new axis, Axis and Axis control input and , we can get the expected position of the quadrotor drone and the expected angle at each position, and get the motion trajectory, that is, the expected orbit.

[0166] ,

[0167] According to the desired roll angle and the desired pitch angle , control input It can be defined as:

[0168] .

[0169] The desired trajectory and the desired attitude angle are defined as a new vector: ,

[0170] The nonlinear tracking differentiator is as follows:

[0171] ,

[0172] Among them, the nonlinear function The definition is as follows:

[0173] ,

[0174] in, is the required deflection angle The tracking signal, yes The differential signal of and are the speed factor and filter factor of the tracking differentiator, , is a nonlinear function The intermediate parameters are used to calculate the nonlinear function. Since the control space is 6-dimensional, all subscripts in the above formula are All ,For example .

[0175] The above-mentioned nonlinear tracking differentiator uses the state value of the quadrotor drone as input to obtain the differential signal of the state value. The above-mentioned nonlinear function can convert the expected value into a smooth process to avoid overshoot during tracking.

[0176] Step 130: Design a super-twisted expansion state observer to observe the total disturbance in the system.

[0177] Although various disturbances inside and outside the UAV system are given in the previous steps, there are still some unknown disturbances, so it is necessary to add the position disturbances existing in the system.

[0178] In this embodiment, the method may include: expanding the state space expression of the quadrotor drone system; using the designed super-twisted expanded state observer, adjusting the adjustable gain of the expanded state observer, obtaining the estimated error system of the expanded state observer, and calculating the unknown disturbance existing in the observation system.

[0179] First, the state space expression of the quadrotor UAV system is expanded:

[0180] ,

[0181] in, ,, .

[0182] The super-twisted expansion state observer is set as follows:

[0183] ,

[0184] in, is the output state of the extended state observer, is the velocity observation value of the extended state observer, is the perturbation observation value of the extended state observer, is the adjustable gain of the extended state observer, assuming that the estimated error state is and ,

[0185] The following is the detailed form:

[0186] ,

[0187] ,

[0188] Using the state space expression of the expanded quadrotor UAV system and the super-distorted expanded state observer, the estimated error system of the expanded state observer can be obtained:

[0189] ,

[0190] in, , , The above method can be used to obtain the existing unknown disturbance.

[0191] Step 140 , utilizing the design of the adaptive rate to obtain an estimated value of the fault in the system model, and calculating the control parameters of the system based on the estimated value.

[0192] Exemplarily, it may include: obtaining unknown faults and corresponding parameters in the system by the adaptive law, and according to the position faults and corresponding parameters in the system: using the designed global fast terminal sliding surface to design control quantities to express the control quantities of the quadrotor drone position loop and the quadrotor drone attitude loop control quantities.

[0193] For example, the following control variables are designed for the drone:

[0194] ,

[0195] in, is the control quantity of the quadrotor UAV position loop, where For quadrotor drones The control amount in the axis direction, for The control amount in the axis direction and for The control amount in the axis direction, is the attitude loop control variable of the quadrotor drone, where is the control quantity of the roll angle direction of the quadrotor drone, is the control quantity in the pitch angle direction, is the control quantity in the yaw angle direction, and is the designed global fast terminal sliding surface, where For quadrotor drones Sliding surface in the axial direction, for Sliding surface in the axial direction, for Axial synovial surface, is the sliding surface of the quadrotor UAV in the rolling angle direction, is the sliding surface in the pitch angle direction and is the sliding membrane surface in the yaw angle direction, and are the position and velocity tracking errors. and is the tracking error of angular velocity and angular acceleration. and are greater than 0, and is a positive odd number, and , and yes , is the control gain of the controller, which is a strictly positive constant, , , , . Figure 2 The invention provides a quadrotor drone with a multi-source uncertainty sliding mode adaptive control method. Axis direction position loop control quantity diagram; Figure 3 The invention provides a quadrotor drone with a multi-source uncertainty sliding mode adaptive control method. Axis direction position loop control quantity diagram; Figure 4 The invention provides a quadrotor drone with a multi-source uncertainty sliding mode adaptive control method. Axis direction position loop control quantity diagram; Figure 5 It is a graph of attitude loop control quantities in the roll angle direction of a quadrotor drone in the multi-source uncertainty sliding mode adaptive control method for a quadrotor drone provided by an embodiment of the present invention; Figure 6 It is a graph of attitude loop control quantities in the pitch angle direction of a quadrotor drone in the multi-source uncertainty sliding mode adaptive control method for a quadrotor drone provided by an embodiment of the present invention; Figure 7 This is the attitude loop control quantity diagram of the yaw angle direction of the quadrotor drone in the multi-source uncertainty downward mode adaptive control method of the quadrotor drone provided by the embodiment of the present invention. Figure 2-Figure 7 It can be seen that the control variables of the position loop and attitude loop of the quadrotor drone change over time. In the tracking control of the quadrotor drone, the system can be stabilized by designing a suitable sliding surface.

[0196] ,

[0197] in:

[0198] , ;

[0199] , ;

[0200] , , , .

[0201] in, 、 、 and is the unknown parameter of the fault in the system 、 、 and The online estimation of is designed by the following adaptive law:

[0202] ,

[0203] in, is a positive constant, is the transpose of the position loop sliding mode surface quantity, It is the transpose of the attitude loop sliding mode surface quantity.

[0204] Using the above method, the control quantity of the system can be calculated and , using the calculated control quantity, smooth adaptive stable control can be achieved when the quadrotor UAV is under multi-source uncertainty factors.

[0205] The embodiment of the present invention provides a multi-source uncertain downward mode adaptive control method for a quadrotor unmanned aerial vehicle. The method combines the self-resistance of the quadrotor unmanned aerial vehicle with external unknown disturbances and unknown functions in the system model to calculate the total disturbance experienced by the quadrotor unmanned aerial vehicle; constructs a motion model of the quadrotor unmanned aerial vehicle according to the influence of actuator failure and the total disturbance; obtains the expected value of the motion target of the quadrotor unmanned aerial vehicle; converts the expected value of the motion target into a smoothing process to calculate the expected value of the system; establishes a state space expression of the quadrotor unmanned aerial vehicle system; designs a super-distorted expansion state observer for observing unknown disturbances in the system; utilizes the design of an adaptive rate to obtain an estimated value of the fault in the system model, and calculates the control parameters of the system based on the estimated value. By combining the quadrotor's inherent resistance with unknown external disturbances and unknown functions in the system model, and taking into account the effects of actuator failures, the quadrotor's motion model is modified. The control variables are optimized based on the various actuator failure parameters. The desired attitude angle is determined based on the optimized control variables, and the desired trajectory is determined based on the desired attitude angle. The trajectory vector is defined using the desired trajectory and smoothed. An optimized hyper-distorted extended state observer is used to obtain the extended state observer's estimated error system, and the adaptive rate is used to design the system's internal fault parameters, ultimately obtaining the quadrotor's control variables. The extended state observer and sliding mode adaptive controller are designed to address the characteristics of the unknown disturbances, addressing the issue of insufficient disturbance rejection.

[0206] Figure 8 This is a position state curve diagram of a quadrotor drone using the quadrotor drone multi-source uncertainty downward mode adaptive control method provided by an embodiment of the present invention; Figure 9 This figure shows the angular state curves of a quadrotor drone using the multi-source uncertainty downward mode adaptive control method for a quadrotor drone provided by an embodiment of the present invention. It can be seen that when using the multi-source uncertainty downward mode adaptive control method for a quadrotor drone provided by an embodiment of the present invention, the drone's state can converge smoothly within a finite time.

[0207] Note that the above are only preferred embodiments of the present invention and the technical principles employed. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and that various obvious changes, readjustments, and substitutions can be made by those skilled in the art without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in detail through the above embodiments, the present invention is not limited to the above embodiments and may include many other equivalent embodiments without departing from the concept of the present invention. The scope of the present invention is determined by the scope of the appended claims.

Claims

1. A multi-source uncertainty downward mode adaptive control method for a quadrotor drone, characterized in that: include: The total disturbance to the quadrotor is calculated by combining the quadrotor's own resistance, external unknown disturbances, and unknown functions in the system model. The motion model of the quadrotor is constructed based on the influence of actuator failure and the total disturbance. Obtaining an expected value of a motion target of the quadrotor drone; Convert the expected value of the moving target into a smoothing process and calculate the expected value of the system; Establish the state space expression of the quadrotor UAV system; Design a super-distorted expansion state observer to observe unknown disturbances in the system; By using the design of adaptive rate, the estimated value of the fault in the system model is obtained, and the control parameters of the system are calculated based on the estimated value; The total disturbance to which the quadrotor drone is subjected is calculated by combining the quadrotor drone's own resistance, external unknown disturbances, and unknown functions in the system model, including: Establish a dynamic model of a quadrotor drone; Based on the quadrotor UAV dynamics model, a quadrotor UAV dynamics model with multi-source uncertainty Lagrangian form is established; The method of constructing a motion model of a quadrotor drone based on the influence of actuator failure and total disturbance includes: Modeling multiplicative or additive uncertainty in actuator output based on the effects of common actuator faults; Construct a motion model of a quadrotor drone based on a multiplicative or additive uncertainty model and a total disturbance; The quadrotor UAV dynamics model includes: ， ， ， in, is the position of the quadrotor drone, are the roll angle, pitch angle and yaw angle of the quadrotor drone, are the speeds of the quadrotor drone on the x-axis, y-axis, and z-axis respectively. are the roll angular velocity, pitch angular velocity and yaw angular velocity of the quadrotor drone, is the total weight of the quadrotor drone, is the acceleration due to gravity, is the control quantity of the quadrotor drone, where is the position loop control quantity of the quadrotor drone, is the attitude loop control value of the quadrotor drone, is the propeller's moment of inertia, For the The speed of the propeller, is the drag coefficient, is the resulting force vector relative to the Earth-fixed frame, is the velocity vector relative to the Earth-fixed frame, is the acceleration vector relative to the Earth-fixed frame, is the resultant moment vector, is the angular velocity vector relative to the fixed coordinate system of the drone, is the angular acceleration vector relative to the fixed coordinate system of the drone, is the diagonal inertia matrix; When the axis of the fuselage fixed frame coincides with the main inertial axis of the quadrotor drone, is a static transformation matrix from the quadcopter frame to the earth-fixed frame: ， The dynamic transformation matrix is ​​established through the relationship between Euler angular rate and angular velocity: ,in: ， in, For the conversion matrix, set the change of roll angle and pitch angle to be less than the set change threshold so that Close to the identity matrix, using Euler angular rate instead of angular velocity, we get: ， in, , For quadrotor drones Axial acceleration, The acceleration in the axial direction and The acceleration in the axial direction, are the roll acceleration, pitch acceleration, and yaw acceleration of the quadrotor drone; The quadrotor UAV dynamics model based on the quadrotor UAV dynamics model is established in a multi-source uncertainty Lagrangian form, including: The unknown external disturbance to the quadrotor UAV system is expressed as: , where D1, D2…D6 correspond to the external disturbances of the six system states of the quadrotor drone; Combining the quadrotor's own resistance, the external unknown disturbance D, and the unknown function in the system model, the total disturbance to the quadrotor is expressed as: ,in: ， Assumption: For the total disturbance of the quadrotor, is continuously differentiable, and is bounded, that is ,in is a positive constant; Accordingly, the multiplication or addition uncertainty model in the actuator output is established according to the influence of common actuator faults, including: Create the following model: ， in In order to eliminate the influence of external uncertainty on the position loop control of the quadrotor drone, It is the attitude control quantity of the quadrotor drone without the influence of external uncertainty, where is the roll angle direction control quantity, is the pitch angle control quantity, is the yaw angle direction control quantity, Indicates the multiplicative loss of position loop thrust caused by the fault, It represents the multiplicative loss of torque in the roll angle direction caused by the fault, It represents the multiplicative loss of the pitch angle torque caused by the fault, represents the multiplicative loss of torque in the yaw angle direction, The additive parameter representing the position loop deviation caused by the fault, It is an additive parameter used to represent the pitch angle deviation caused by the fault. It is an additive parameter used to represent the pitch angle deviation caused by the fault. It is an additive parameter used to represent the yaw angle deviation caused by the fault. The fault parameter and The real-time value of is unknown, and τ is the attitude loop control variable; Accordingly, the motion model of the constructed quadrotor drone is as follows: 。 2. The method according to claim 1, characterized in that The method of converting the expected value of the moving target into a smoothing process and calculating the expected value of the system includes: According to the desired yaw angle , determine the new x-axis, y-axis and z-axis control inputs; Determine the state space of the quadrotor based on the new x-axis, y-axis, and z-axis control inputs to determine the desired trajectory; A desired vector is defined using a desired roll angle, a desired pitch angle, and a desired trajectory; The expected vector is input into a nonlinear tracking differentiator to obtain a smoothed system expected value.

3. The method according to claim 2, wherein the designing of a super-distorted expansion state observer for observing unknown disturbances in the system comprises: Expanding the state space expression of the quadrotor UAV system; By using the designed super-distorted extended state observer, adjusting the adjustable gain of the extended state observer, the estimated error system of the extended state observer is obtained, and the unknown disturbance in the observation system is calculated.

4. The method according to claim 3, characterized in that The design of the adaptive rate is used to obtain an estimated value of the fault in the system model, and the control parameters of the system are calculated based on the estimated value, including: The unknown faults and corresponding parameters in the system are obtained by the adaptive law, and the control quantities of the position faults and corresponding parameters in the system are expressed by the designed global fast terminal sliding surface to express the control quantities of the position loop and the attitude loop of the quadrotor drone.

5. The method according to claim 4, characterized in that The desired yaw angle , determine the new x-axis, y-axis and z-axis control inputs, including: According to the target expected value of the received yaw angle , respectively define the quadrotor drone model axis, Axis and Axis control amount for: ， According to the characteristics of the control quantity of the actuator error with unknown parameters, new axis, Axis and The axis control inputs are and , and meet the following conditions: ， The state space expression of the quadrotor drone is determined based on the new x-axis, y-axis and z-axis control inputs: ， in, , , ， ， ， ， ， ， ， ， , The intermediate amount calculated by the above formula; is the system state vector composed of the six-dimensional state of the UAV system, is the derivative of the system state vector composed of the six-dimensional state of the UAV system, is the derivative of the UAV system state vector, It is the second-order derivative of the system state vector composed of the six-dimensional state of the UAV system, expressed as the speed and angular velocity of the system, is the vector of the drone's gravity conversion, is the vector transformed from the multiplicative coefficient of the actuator fault, is the matrix of actuator fault additive coefficient transformation; Determining the state space of the quadrotor drone based on the new x-axis, y-axis, and z-axis control inputs and determining the desired trajectory includes: Determine the desired roll angle based on the new x-, y-, and z-axis control inputs. and pitch angle , according to the desired roll angle and pitch angle Determine the desired trajectory: ， Defining a desired vector using a desired roll angle, a desired pitch angle, and a desired trajectory includes: Generate the desired vector using the desired trajectory and the desired attitude angle: ， The nonlinear tracking differentiator comprises: ， Among them, the nonlinear function The definition is as follows: ， in, is the required deflection angle The tracking signal, yes The differential signal of and are the speed factor and filter factor of the tracking differentiator, , It is a nonlinear function The intermediate parameters of , used to calculate nonlinear functions.

6. The method according to claim 5, characterized in that The state space expression of the expanded quadrotor UAV system is: ， in, , , The super-twisted expansion state observer is: ， in, is the output state of the extended state observer, is the velocity observation value of the extended state observer, is the perturbation observation value of the extended state observer, is the adjustable gain of the extended state observer, assuming that the estimated error state is and , The detailed form of the superwarp algorithm is: ， ， The designed super-distorted extended state observer is used to adjust the adjustable gain of the extended state observer to obtain the estimated error system of the extended state observer, and the unknown disturbance existing in the observation system is calculated, which is achieved by the following method: ， Accordingly, the unknown faults and corresponding parameters in the system are obtained by the adaptive law, and according to the position faults in the system and the corresponding parameters, the following are included: set up 、 、 and is the unknown parameter of the fault in the system 、 、 and The online estimation of is designed by the following adaptive law: ， in, is a positive constant, is the transpose of the position loop sliding mode surface quantity, It is the transpose of the attitude loop sliding mode face quantity; The control quantities designed by using the designed global fast terminal sliding surface to express the control quantities of the position loop and the attitude loop of the quadrotor drone include: The drone is designed with the following control quantities: ， in, is the control quantity of the quadrotor UAV position loop, where For quadrotor drones Control amount in the axial direction, for The control amount in the axis direction and for Control quantity in the axis direction; is the attitude loop control variable of the quadrotor drone, where is the control quantity of the roll angle direction of the quadrotor drone, is the control quantity in the pitch angle direction and is the control quantity in the yaw angle direction, and is the designed global fast terminal sliding surface, where For quadrotor drones Sliding surface in the axial direction, for Sliding surface in the axial direction, for Axial synovial surface, is the sliding surface of the quadrotor UAV in the rolling angle direction, is the sliding surface in the pitch angle direction, is the sliding membrane surface in the yaw angle direction, and are the position and velocity tracking errors, and is the tracking error of angular velocity and angular acceleration, where and are greater than 0, and yes , is a strictly positive constant of the control gain of the controller, , , , , ， in: ， ， ， ， ， ， ， 。

Images