CN110888451B - A fault-tolerant control method and system for a multi-rotor UAV - Google Patents

Abstract

The invention relates to a fault-tolerant control method and a fault-tolerant control system for a multi-rotor unmanned aerial vehicle, wherein the flight of the unmanned aerial vehicle is controlled by adopting an improved attitude control algorithm of linear active disturbance rejection control so as to ensure the robustness of the unmanned aerial vehicle in the flight process; when detecting that partial motor of unmanned aerial vehicle is unusual, construct trouble matrix R _i (ii) a Based on the fault matrix R _i Establishing a fault model on line; based on the fault matrix R _i Obtaining control distribution information of all motors on the unmanned aerial vehicle; and controlling the flight of the unmanned aerial vehicle under the fault model by adopting the attitude control algorithm of the improved linear active disturbance rejection control, and controlling the unmanned aerial vehicle according to the control distribution information of the motor so as to achieve the required attitude and height. The control method and the control system improve the fault-tolerant capability of the multi-rotor unmanned aerial vehicle, and ensure that the multi-rotor unmanned aerial vehicle has larger load capacity and higher stability.

Description

A fault-tolerant control method and system for a multi-rotor UAV

technical field

The invention relates to the technical field of unmanned aerial vehicles, in particular to a method and system for fault-tolerant control of a multi-rotor unmanned aerial vehicle.

Background technique

In recent years, with the continuous advancement of science and technology, unmanned aerial vehicles, especially in the field of multi-rotor unmanned aerial vehicles with more than four rotors, have achieved rapid development. A multi-rotor UAV is an aircraft equipped with airborne equipment such as data processing and transmission systems, sensors, automatic control systems, and communication systems. It can perform certain steady-state control and flight, and has certain autonomous flight capabilities. At present, multi-rotor aircraft has been widely used in fields such as agriculture, forestry and plant protection, power inspection, logistics and transportation, which greatly facilitates people's production and life.

When the multi-rotor unmanned aerial vehicle fails, the flight state will change suddenly, resulting in unpredictable consequences. Therefore, it is necessary to design a fault-tolerant control method to improve the fault-tolerant ability of the multi-rotor UAV, so as to ensure that the multi-rotor unmanned The machine has greater load capacity and higher stability.

Contents of the invention

The technical problem to be solved by the present invention is to provide a multi-rotor UAV fault-tolerant control method and system, which improves the fault-tolerant capability of the multi-rotor UAV and ensures that the multi-rotor UAV has a larger load capacity and a higher load capacity. stability.

In order to solve the problems referred to above, the technical scheme that the present invention takes is:

On the one hand, a fault-tolerant control method for a multi-rotor UAV is provided, which includes:

The attitude control algorithm of the improved linear active disturbance rejection control is used to control the flight of the UAV, so as to ensure the robustness of the UAV during flight;

When some motors of the UAV are detected to be abnormal, a fault matrix R _i is constructed, where i is an integer greater than or equal to 0 and less than or equal to the number of all motors in the UAV;

Establishing a fault model online based on the fault matrix R _i ;

Obtain the control allocation information of all motors on the UAV based on the fault matrix R _i ;

Adopt the attitude control algorithm of the improved linear ADRC to control the flight of the unmanned aerial vehicle under the fault model, and control the unmanned aerial vehicle according to the control distribution information of the motor to achieve the required attitude and high.

As a further improvement of the present invention, the attitude control algorithm of the improved linear active disturbance rejection control includes:

Arrange the transition process: use the following formula to convert the input sudden change signal into a slow change signal through the second-order link, and then make the output signal reach the desired input signal:

其中，其中，G(s)代表二阶环节的传递函数，T代表二阶环节的时间常数，s代表代表传递函数中的变量符号；

Wherein, G(s) represents the transfer function of the second-order link, T represents the time constant of the second-order link, and s represents the variable symbol in the transfer function;

Linear expansion state observer: use the following state space equations and formulas to realize real-time tracking of variables in the model:

其中，x₁,x₂,x₃分别代表所系统的状态变量，

b₀代表估计的控制增益，w代表外部扰动，y代表所述模型的输出，u代表所述模型的输入；

Among them, x ₁ , x ₂ , x ₃ respectively represent the state variables of the system,

b ₀ represents the estimated control gain, w represents the external disturbance, y represents the output of the model, and u represents the input of the model;

其中，z₁,z₂,z₃分别代表所述线性扩张状态观测器的系统状态变量，β₁,β₂,β₃分别代表所述线性扩张状态观测器的增益。

Wherein, z ₁ , z ₂ , z ₃ respectively represent the system state variables of the linearly extended state observer, and β ₁ , β ₂ , β ₃ represent the gain of the linearly extended state observer respectively.

As a further improvement of the present invention, when the part of the motor of the drone is detected to be abnormal, the fault matrix R _i is constructed, including:

Real-time detection of all motors of the drone;

When some motors of the UAV are detected to be abnormal, the ratio of the output of the faulty motor to the output without fault is calculated, and a fault matrix R _i is constructed according to the ratio.

As a further improvement of the present invention, the online establishment of a fault model based on the fault matrix R _i includes:

The fault model is established online using the following formula:

分别代表大地坐标系下的位置加速度，

分别代表在大地坐标系下所述无人机的飞行器姿态角的角加速度，

θ,ψ分别代表横滚角、俯仰角和偏航角，I_x,I_y,I_z分别代表所述无人机机身在三个方向的转动惯量，m代表所述无人机的质量，g代表重力加速度，U_R,U_P,U_Y,U_T分别代表所述无人机的电机均无故障时的横滚力矩、俯仰力矩、偏航力矩以及升力，f_p,f_q,f_r,f_z分别表示横滚力矩误差、俯仰力矩误差、偏航力矩误差以及升力误差。in

Represent the position acceleration in the earth coordinate system,

represent the angular acceleration of the aircraft attitude angle of the unmanned aerial vehicle under the earth coordinate system respectively,

θ, ψ represent the roll angle, pitch angle and yaw angle respectively, I _x , I _y , I _z represent the moments of inertia of the UAV fuselage in three directions respectively, and m represents the mass of the UAV , g represents the acceleration of gravity, U _R , U _P , U _Y , U _T respectively represent the rolling moment, pitching moment, yaw moment and lift when the motors of the UAV are not faulty, f _p , f _q , f _r , f _z represent roll moment error, pitch moment error, yaw moment error and lift error respectively.

As a further improvement of the present invention, the control distribution information of all motors on the drone is obtained based on the fault matrix R _i , including:

The following formula is used to obtain the optimized allocation matrix N ^f , and the optimized allocation matrix N ^{f is} used as the control allocation information of all motors on the drone:

Nf = ^{Af -1} ;

N ^f = A ^fT (A ^f A ^fT ) ^-1 ; wherein, A ^f represents the control efficiency matrix after a partial motor failure, and A ^fT represents the transposition of A ^f ;

A ^f = AR _i ; where, A represents the control efficiency matrix before failure;

其中，b为升力系数，l为所述无人机的轴距，d为反扭矩系数。

Wherein, b is the lift coefficient, l is the wheelbase of the UAV, and d is the counter torque coefficient.

On the other hand, a fault-tolerant control system for a multi-rotor UAV is provided, which includes:

The first control module is used to control the flight of the unmanned aerial vehicle by adopting the attitude control algorithm of the improved linear active disturbance rejection control, so as to ensure the robustness of the unmanned aerial vehicle during flight;

A fault matrix construction module, used to construct a fault matrix R _i when detecting abnormality of some motors of the drone, where i is an integer greater than or equal to 0 and less than or equal to the number of all motors in the drone;

A fault model building module, configured to build a fault model online based on the fault matrix R _i ;

An allocation information acquisition module, configured to obtain control allocation information of all motors on the UAV based on the fault matrix R _i ;

The second control module adopts the attitude control algorithm of the improved linear active disturbance rejection control to control the flight of the UAV under the fault model, and controls the UAV according to the control distribution information of the motor, so as to achieve Desired attitude and height.

As a further improvement of the present invention, the fault matrix building block includes:

The transition process unit is arranged to convert the input sudden change signal into a slow change signal through the second-order link by using the following formula, and then make the output signal reach the desired input signal:

其中，其中，G(s)代表二阶环节的传递函数，T代表二阶环节的时间常数，s代表代表传递函数中的变量符号；

Wherein, G(s) represents the transfer function of the second-order link, T represents the time constant of the second-order link, and s represents the variable symbol in the transfer function;

The Linear Extended State Observer unit is used to implement real-time tracking of variables in the model using the following state space equations and formulas:

其中，x₁,x₂,x₃分别代表所系统的状态变量，

b₀代表估计的控制增益，w代表外部扰动，y代表所述模型的输出，u代表所述模型的输入；

Among them, x ₁ , x ₂ , x ₃ respectively represent the state variables of the system,

b ₀ represents the estimated control gain, w represents the external disturbance, y represents the output of the model, and u represents the input of the model;

其中，z₁,z₂,z₃分别代表所述线性扩张状态观测器的系统状态变量，β₁,β₂,β₃分别代表所述线性扩张状态观测器的增益。

Wherein, z ₁ , z ₂ , z ₃ respectively represent the system state variables of the linearly extended state observer, and β ₁ , β ₂ , β ₃ represent the gain of the linearly extended state observer respectively.

As a further improvement of the present invention, the fault model building module includes:

A detection unit is used to detect all motors of the drone in real time;

The fault model building unit is used to calculate the ratio of the output of the faulty motor to the output when there is no fault when some motors of the drone are detected to be abnormal, and construct a fault matrix R _i according to the ratio.

As a further improvement of the present invention, the fault model building module includes:

The fault model establishing unit is used to establish the fault model online by adopting the following formula:

分别代表大地坐标系下的位置加速度，

分别代表在大地坐标系下所述无人机的飞行器姿态角的角加速度，

θ,ψ分别代表横滚角、俯仰角和偏航角，I_x,I_y,I_z分别代表所述无人机机身在三个方向的转动惯量，m代表所述无人机的质量，g代表重力加速度，U_R,U_P,U_Y,U_T分别代表所述无人机的电机均无故障时的横滚力矩、俯仰力矩、偏航力矩以及升力，f_p,f_q,f_r,f_z分别表示横滚力矩误差、俯仰力矩误差、偏航力矩误差以及升力误差。in

Represent the position acceleration in the earth coordinate system,

represent the angular acceleration of the aircraft attitude angle of the unmanned aerial vehicle under the earth coordinate system respectively,

θ, ψ represent the roll angle, pitch angle and yaw angle respectively, I _x , I _y , I _z represent the moments of inertia of the UAV fuselage in three directions respectively, and m represents the mass of the UAV , g represents the acceleration of gravity, U _R , U _P , U _Y , U _T respectively represent the rolling moment, pitching moment, yaw moment and lift when the motors of the UAV are not faulty, f _p , f _q , f _r , f _z represent roll moment error, pitch moment error, yaw moment error and lift error respectively.

As a further improvement of the present invention, the allocation information acquisition module includes:

The allocation information acquisition unit is used to adopt the following formula to obtain the optimized allocation matrix ^Nf , and use the optimized allocation matrix ^Nf as the control allocation information of all motors on the drone:

Nf = ^{Af -1} ;

N ^f = A ^fT (A ^f A ^fT ) ^-1 ; wherein, A ^f represents the control efficiency matrix after a partial motor failure, and A ^fT represents the transposition of A ^f ;

A ^f = AR _i ; where, A represents the control efficiency matrix before failure;

其中，b为升力系数，l为所述无人机的轴距，d为反扭矩系数。

Wherein, b is the lift coefficient, l is the wheelbase of the UAV, and d is the counter torque coefficient.

The beneficial effects produced by adopting the above-mentioned technical scheme are:

The present invention provides a multi-rotor unmanned aerial vehicle fault-tolerant control method and system provided by an embodiment of the present invention. The attitude control algorithm of the improved linear active disturbance rejection control is used to control the flight of the unmanned aerial vehicle to ensure that all Describe the robustness of the unmanned aerial vehicle in the flight process; when detecting the abnormal part of the motor of the unmanned aerial vehicle, construct the fault matrix R _i ; establish the fault model online based on the fault matrix R _i ; based on the fault matrix R _i obtain the control distribution information of all motors on the UAV; adopt the attitude control algorithm of the improved linear active disturbance rejection control to control the flight of the UAV under the fault model, and distribute information according to the control distribution of the motors Control the drone to achieve the desired attitude and altitude. When the motor of the UAV does not fail, the basic control law selected is the attitude control algorithm of the improved linear active disturbance rejection control, which has strong robustness to disturbance. Moreover, when some motors of the UAV fail, based on the failure matrix after the failure, a new flight model-fault model and the control distribution information of all motors can be obtained, so as to ensure that the UAV can operate in this failure mode. While flying, it can also reduce the use of faulty motors, so that the UAV can achieve a stable flight state.

Description of drawings

In order to more clearly illustrate the specific implementation of the present invention or the technical solutions in the prior art, the following will briefly introduce the accompanying drawings that need to be used in the specific implementation or description of the prior art. Obviously, the accompanying drawings in the following description The drawings show some implementations of the present invention, and those skilled in the art can obtain other drawings based on these drawings without creative work.

Fig. 1 is a flowchart of a fault-tolerant control method for a multi-rotor UAV provided by an embodiment of the present invention.

Fig. 2 is a response curve diagram of the roll angle for which the system does not perform distribution optimization when the first type of fault occurs according to an embodiment of the present invention.

Fig. 3 is a response curve diagram of the pitch angle at which the system does not perform distribution optimization when the first type of fault occurs according to an embodiment of the present invention.

Fig. 4 is a response curve diagram of the yaw angle without allocation optimization of the system when the first type of fault occurs according to an embodiment of the present invention.

Fig. 5 is a comparison diagram of the roll angle when the system performs allocation optimization when the first type of failure occurs and when no failure occurs according to an embodiment of the present invention.

Fig. 6 is a comparison diagram of pitch angles when the system performs allocation optimization when the first type of failure occurs and when no failure occurs according to an embodiment of the present invention.

Fig. 7 is a comparison diagram of the yaw angle when the system performs allocation optimization when the first type of failure occurs and when no failure occurs according to an embodiment of the present invention.

Fig. 8 is a response curve diagram of the roll angle for which the system does not perform distribution optimization when the second type of fault occurs according to an embodiment of the present invention.

Fig. 9 is a response curve diagram of the pitch angle for which the system does not perform allocation optimization when the second type of fault occurs according to an embodiment of the present invention.

Fig. 10 is a response curve diagram of the yaw angle without allocation optimization of the system when the second type of fault occurs according to an embodiment of the present invention.

Fig. 11 is a comparison diagram of the roll angle when the system performs allocation optimization when the second type of failure occurs and when no failure occurs according to an embodiment of the present invention.

Fig. 12 is a comparison diagram of pitch angles when the system performs allocation optimization when the second type of failure occurs and when no failure occurs according to an embodiment of the present invention.

Fig. 13 is a comparison diagram of the yaw angle when the system performs allocation optimization when the second type of failure occurs and when no failure occurs according to an embodiment of the present invention.

Fig. 14 is a structural diagram of a fault-tolerant control system for a multi-rotor UAV provided by an embodiment of the present invention.

detailed description

In order to make the purpose, technical solution and advantages of the present invention clearer, the invention will be clearly and completely described below in conjunction with the accompanying drawings and specific embodiments.

At present, multi-rotor UAVs are mainly divided into: quad-rotor UAVs and multi-rotor UAVs with more than four rotors. Quad-rotor UAVs lack the redundancy of rotor components. Once the transmission fails, the flight attitude will change suddenly. Some application areas can cause incalculable consequences. For multi-rotor UAVs with four or more rotors, the control algorithm can be optimized to make the multi-rotor UAV have good fault tolerance, greater load capacity and higher stability, so as to ensure the stability of the UAV. It still has good stability and safety when encountering strong external force interference or part of the motor is damaged, so that it can carry more mission equipment and complete more complex missions.

Based on this, the multi-rotor UAV involved in the present invention refers to a multi-rotor UAV with more than four rotors.

Fig. 1 and Fig. 2 are the flowcharts of the fault-tolerant control method of the multi-rotor UAV provided by the present invention, as shown in Fig. 1 and Fig. 2, the fault-tolerant control method comprises:

S101: Control the flight of the UAV by using the attitude control algorithm of the improved linear ADRC to ensure the robustness of the UAV during the flight.

Among them, the attitude control algorithm of the improved linear ADRC includes:

(1) Arrange the transition process: use the following formula to convert the input mutation signal into a slow-change signal through the second-order link, and then make the output signal reach the desired input signal:

其中，G(s)代表二阶环节的传递函数，T代表二阶环节的时间常数，T可取期望过渡时间的

s代表传递函数中的变量符号，是一个复参数。

Among them, G(s) represents the transfer function of the second-order link, T represents the time constant of the second-order link, and T can be taken as the expected transition time

s represents the variable symbol in the transfer function and is a complex parameter.

(2) Linear expansion state observer: use the following state space equations and formulas to realize real-time tracking of each variable in the model:

其中，x₁,x₂,x₃分别代表系统的状态变量，

b₀代表估计的控制增益，w代表外部扰动，y代表模型的输出，u代表模型的输入；

Among them, x ₁ , x ₂ , x ₃ represent the state variables of the system respectively,

b ₀ represents the estimated control gain, w represents the external disturbance, y represents the output of the model, and u represents the input of the model;

其中，z₁,z₂,z₃分别代表线性扩张状态观测器的系统状态变量，β₁,β₂,β₃分别代表线性扩张状态观测器的增益。

Among them, z ₁ , z ₂ , z ₃ represent the system state variables of the linearly extended state observer respectively, and β ₁ , β ₂ , β ₃ represent the gain of the linearly extended state observer respectively.

It should be noted that the multi-rotor UAV mainly changes the flying attitude of the UAV by changing the speed of the motor corresponding to each rotor. The speed of the motor is changed by changing the duty cycle of the PWM signal. After the change, the lifting force and torque generated by the motor change, and at the same time, according to the position distribution of the multi-rotor UAV's own motor, the torque in the roll direction, the pitch direction and the counter torque in the yaw direction are determined. The change of roll direction and pitch direction produces the linear velocity of the aircraft in the X and Y axis directions, the change of yaw angle is the change of the aircraft heading, and the change of lift force produces the change of the aircraft Z axis, that is, the height.

Among them, the body coordinate system is defined as: the origin is taken at the center of mass of the drone, and the coordinate system is fixedly connected to the body; the X-axis is parallel to the longitudinal axis of the fuselage design, and is in the plane of symmetry of the drone, pointing forward; the Y-axis Pointing to the right perpendicular to the plane of symmetry of the drone; the Z-axis is within the plane of symmetry of the drone and pointing downwards perpendicular to the X-axis. The entire coordinate system conforms to the right-hand rule of the Euler coordinate system. The ground coordinate system, that is, the inertial coordinate system, is defined as follows: using the northeast ground coordinate system, the XE axis points to the north, the YE axis points to the east, and the ZE axis points to the center of the earth. The ground coordinate system is the coordinate system in the environment of the simulation experiment.

The virtual control command generated by the UAV according to the user's operation is assigned as the actual control command of each motor through the control distribution information. When the four motors are normal, the control assignment information is a fixed value and does not change. However, when some motors of the UAV fail, the motors cannot correctly respond to the assigned actual control commands, that is, the motors cannot be adjusted to the corresponding speed for the same control commands, and cannot meet the control requirements. Therefore, it is necessary to optimize the control distribution information redistribute.

S102: When some motors of the UAV are detected to be abnormal, construct a fault matrix Ri, where i is an integer greater than or equal to 0 and less than or equal to the number of all motors in the UAV.

Among them, this step includes:

S1021: Detect all motors of the drone in real time.

S1022: When some motors of the UAV are detected to be abnormal, calculate the ratio of the output of the faulty motor to the output when there is no fault, and construct a fault matrix according to the ratio.

Regarding the method of judging the abnormality of some motors, in one possible implementation, during the flight of the drone, the user can push the joystick of the remote control to send control instructions to the flight control system of the drone, and the drone will fly After the control system receives the control command, it outputs a certain amount of control to each ESC through the internal attitude calculation link and attitude control link, and then each ESC outputs a PWM value corresponding to the control amount to control the ESC. The speed of the connected motor, that is, the PWM value output by each ESC corresponds to the control command. Therefore, when the PWM value output by a certain ESC does not correspond to the control command, it is determined that the motor corresponding to the ESC is faulty.

When some motors of the drone are detected to be abnormal, the flight control system of the drone acquires the output values of all the motors, that is, the efficiency of the motors. Wherein, in the embodiment of the present invention, the efficiency of a motor without a fault is set as 1; and the efficiency of a motor with a fault is a quantized value of the output of the faulty motor relative to the normal output. The efficiency of each motor is represented by k ₁ , k ₂ ...k _i respectively, then the constructed fault matrix R _i is:

R _i =diag[k ₁ , k ₂ . . . k _i ].

For example: the multi-rotor UAV is a six-rotor UAV, and the six motors on the six-rotor UAV are respectively marked as 1-6. When k ₁ and k ₂ , the fault matrix of the hexacopter UAV is:

S103: Establish a fault model online based on the fault matrix R _i .

The fault model is established online using the following formula:

分别代表大地坐标系下的位置加速度，

分别代表在大地坐标系下无人机的飞行器姿态角的角加速度，

θ,ψ分别代表横滚角、俯仰角和偏航角，I_x,I_y,I_z分别代表无人机机身在三个方向的转动惯量，m代表无人机的质量，g代表重力加速度，U_R,U_P,U_Y,U_T分别代表无人机的电机均无故障时的横滚力矩、俯仰力矩、偏航力矩以及升力，f_p,f_q,f_r,f_z分别表示横滚力矩误差、俯仰力矩误差、偏航力矩误差以及升力误差。in,

Represent the position acceleration in the earth coordinate system,

Represent the angular acceleration of the UAV's aircraft attitude angle in the earth coordinate system,

θ, ψ represent the roll angle, pitch angle and yaw angle respectively, I _x , I _y , I _z represent the moments of inertia of the UAV fuselage in three directions, m represents the mass of the UAV, and g represents the gravity Acceleration, U _R , U _P , U _Y , U _T respectively represent the rolling moment, pitching moment, yaw moment and lift when the motors of the UAV are not faulty, f _p , f _q , f _r , f _z respectively Indicates roll moment error, pitch moment error, yaw moment error and lift error.

可通过传感器得到，传感器包括测量加速度的三轴加速度计和测量角速度的三轴陀螺仪。the above

Available through sensors that include a three-axis accelerometer that measures acceleration and a three-axis gyroscope that measures angular velocity.

f _p , f _q , f _r , f _z can be obtained from the fault matrix R _i . It should be noted that f _p , f _q , f _r , f _z are related to the number of rotors of the UAV and the layout type of the rotors. In general Next, once the number of rotors and the type of rotors are determined, f _p , f _q , f _r , f _z can be obtained.

In a possible implementation, when the multi-rotor UAV is an "X" type six-rotor UAV, the six motors on the six-rotor UAV are respectively marked as 1-6, and each axis is clamped The angle is 60°, and the following methods can be used to obtain the rolling moment U _R , pitching moment U _P , yaw moment U _Y and lift U _T when the motors of the UAV are not faulty:

可采用以下方式得到无人机的部分电机出现故障时的的横滚力矩

俯仰力矩

偏航力矩

以及升力

Then when the No. 1 motor fails, its motor efficiency is k ₁ , and other motors do not fail, the fault matrix R ₆ is:

The rolling moment when some motors of the UAV fail can be obtained in the following way

pitching moment

Yaw moment

and lift

分别代表1-6号电机未出故障时每个电机所对应的转速。Among them, the above

Respectively represent the corresponding speed of each motor when the No. 1-6 motors are not faulty.

Afterwards, use the following formulas to obtain roll moment error f _p , pitch moment error f _q , yaw moment error f _r and lift error f _z :

S104: Obtain control distribution information of all motors on the UAV based on the fault matrix R _i .

Using the following formula, the optimized distribution matrix N ^f is obtained, and the optimized distribution matrix N ^f is used as the control distribution information of all motors on the UAV:

Nf = ^{Af -1} ;

N ^f = A ^fT (A ^f A ^fT ) ^-1 ; wherein, A ^f represents the control efficiency matrix after a partial motor failure, and A ^fT represents the transposition of A ^f ;

A ^f = AR _i ; where, A represents the control efficiency matrix before failure;

其中，b为升力系数，l为无人机的轴距，d为反扭矩系数。

Among them, b is the lift coefficient, l is the wheelbase of the UAV, and d is the counter torque coefficient.

S105: Use the attitude control algorithm of the improved linear ADRC to control the flight of the UAV under the fault model, and control the UAV according to the control distribution information of the motor to achieve the required attitude and height.

Adjust the speed of each motor according to the actual control instructions, and then make each motor drive the UAV to achieve the required attitude and height. Attitude includes pitch, roll and yaw.

For example, the following formula can be used to adjust the speed of each motor:

τ ^f ＝[U _R U _P U _Y U _T ] ^T ; where, τ ^f represents the moment matrix;

其中，

表示当无人机部分电机异常时，每个电机所对应的转速。

in,

Indicates the corresponding speed of each motor when some motors of the UAV are abnormal.

In the embodiment of the present invention, the UAV flight control system distributes virtual control instructions to the actuator according to the user operation, and the speed of the motor in the actuator changes, the UAV flight control system adopts the improved linear active disturbance rejection control. During the flight process of the UAV controlled by the attitude control algorithm, the control distribution information of each motor is optimized based on the new control distribution information; and the rolling moment error f _p , The pitch moment error f _q , the yaw moment error f _r and the lift error f _z make the height and attitude of the UAV change, reaching the desired height and attitude of the UAV.

In addition, the embodiment of the present invention uses the six-rotor UAV as the object, and simulates and verifies the fault-tolerant control method, as shown in Figures 2-13, and Figures 2-13 are all given the desired roll angle and pitch The resulting plots are given a desired altitude of 1m for a 15° angle and a yaw angle of 15°.

Among them, Fig. 2-Fig. 4 respectively show the response curves of roll angle, pitch angle and yaw angle when the efficiency of No. 1 motor is 1/5, and the system is not optimized for distribution. It can be seen from Figures 2 to 4 that when the No. 1 motor has a power effect of 1/5 and the UAV fault-tolerant control method provided by the embodiment of the present invention is not used, the three attitude angles cannot follow the input expected attitude angles to achieve stability.

Figures 5-7 are the comparison diagrams of the roll angle, pitch angle and yaw angle when the efficiency of the No. 1 motor is 1/5, the distribution optimization of the system and the failure-free time, where (a) line represents the system The three attitude angles for allocation optimization, the line (b) represents the three attitude angles without failure. It can be seen from Fig. 5-Fig. 7 that when the UAV fault-tolerant control method provided by the embodiment of the present invention is adopted, compared with the flight condition when there is no fault, although the unsteady time is longer, the expected value can still be tracked very well. and reach final stability.

Figures 8-10 are the response curves of roll angle, pitch angle, and yaw angle when the efficiency of the No. 1 motor is 1/2, and the system is not optimized for distribution. From Figures 8 to 10, it can be seen that when the power effect of the No. 1 motor is 1/2 and the UAV fault-tolerant control method provided by the embodiment of the present invention is not used, the three attitude angles cannot follow the input expected attitude angles to reach stability and Vibration is obvious.

Figures 11-13 are the comparison diagrams of the roll angle, pitch angle and yaw angle when the efficiency of the No. 1 motor is 1/2, the distribution optimization of the system and the failure-free time, where the (a) line represents the system The three attitude angles for allocation optimization, the line (b) represents the three attitude angles without failure. From Figures 11 to 13, it can be seen that when the UAV fault-tolerant control method provided by the embodiment of the present invention is adopted, compared with the flight condition without failure, although the unsteady time is longer, the expected value can still be tracked very well and reach final stability.

A fault-tolerant control method for a multi-rotor UAV provided by an embodiment of the present invention uses an improved linear active disturbance rejection control attitude control algorithm to control the flight of the UAV to ensure the robustness of the UAV during flight; When some motors of the UAV are detected to be abnormal, the fault matrix R _i is constructed; the fault model is established online based on the fault matrix R _i ; the control distribution information of all motors on the UAV is obtained based on the fault matrix R _i ; the improved linear active disturbance rejection The attitude control algorithm of the control controls the flight of the UAV under the fault model, and controls the UAV according to the control distribution information of the motor to achieve the required attitude and height. When the motor of the UAV does not fail, the basic control law selected is the attitude control algorithm of the improved linear active disturbance rejection control, which has strong robustness to disturbance. Moreover, when some motors of the UAV fail, based on the failure matrix after the failure, a new flight model-fault model and the control distribution information of all motors can be obtained, so as to ensure that the UAV can operate in this failure mode. While flying, it can also reduce the use of faulty motors, so that the UAV can achieve a stable flight state.

Figure 14 is a structural diagram of a multi-rotor UAV fault-tolerant control system provided by an embodiment of the present invention, as shown in Figure 14, which includes:

The first control module 1401 is used to control the flight of the UAV by adopting the attitude control algorithm of the improved linear ADRC to ensure the robustness of the UAV during the flight;

The fault matrix construction module 1402 is used to construct a fault matrix R _i when some motors of the drone are abnormal, and i is an integer greater than or equal to 0 and less than or equal to the number of all motors in the drone;

A fault model building module 1403, configured to build a fault model online based on the fault matrix R _i ;

Distribution information acquisition module 1404, for obtaining the control distribution information of all motors on the UAV based on the fault matrix R _i ;

The second control module 1405 uses the attitude control algorithm of the improved linear ADRC to control the flight of the UAV under the fault model, and controls the UAV according to the control distribution information of the motor to achieve the required attitude and altitude.

Wherein, the first control module 1401 and the second control module 1405 may be the same control module.

In a possible implementation, the fault matrix construction module 1402 includes:

The transition process unit is arranged to convert the input sudden change signal into a slow change signal through the second-order link by using the following formula, and then make the output signal reach the desired input signal:

其中，其中，G(s)代表二阶环节的传递函数，T代表二阶环节的时间常数，s代表传递函数中的变量符号；

Wherein, G(s) represents the transfer function of the second-order link, T represents the time constant of the second-order link, and s represents the variable symbol in the transfer function;

The Linear Extended State Observer unit is used to implement real-time tracking of variables in the model using the following state space equations and formulas:

其中，x₁,x₂,x₃分别代表所系统的状态变量，

b₀代表估计的控制增益，w代表外部扰动，y代表模型的输出，u代表模型的输入；

Among them, x ₁ , x ₂ , x ₃ respectively represent the state variables of the system,

b ₀ represents the estimated control gain, w represents the external disturbance, y represents the output of the model, and u represents the input of the model;

其中，z₁,z₂,z₃分别代表线性扩张状态观测器的系统状态变量，β₁,β₂,β₃分别代表线性扩张状态观测器的增益。

Among them, z ₁ , z ₂ , z ₃ represent the system state variables of the linearly extended state observer respectively, and β ₁ , β ₂ , β ₃ represent the gain of the linearly extended state observer respectively.

In a possible implementation, the fault model building module 1403 includes:

A detection unit for real-time detection of all motors of the drone;

The fault model building unit is used to calculate the ratio of the output of the faulty motor to the output when there is no fault when some motors of the UAV are detected to be abnormal, and construct a fault matrix R _i according to the ratio.

In a possible implementation, the fault model building module 1403 includes:

The fault model establishing unit is used to establish the fault model online by adopting the following formula:

分别代表大地坐标系下的位置加速度，

分别代表在大地坐标系下无人机的飞行器姿态角的角加速度，

θ,ψ分别代表横滚角、俯仰角和偏航角，I_x,I_y,I_z分别代表无人机机身在三个方向的转动惯量，m代表无人机的质量，g代表重力加速度，U_R,U_P,U_Y,U_T分别代表无人机的电机均无故障时的横滚力矩、俯仰力矩、偏航力矩以及升力，f_p,f_q,f_r,f_z分别表示横滚力矩误差、俯仰力矩误差、偏航力矩误差以及升力误差。in

Represent the position acceleration in the earth coordinate system,

Represent the angular acceleration of the UAV's aircraft attitude angle in the earth coordinate system,

θ, ψ represent the roll angle, pitch angle and yaw angle respectively, I _x , I _y , I _z represent the moments of inertia of the UAV fuselage in three directions, m represents the mass of the UAV, and g represents the gravity Acceleration, U _R , U _P , U _Y , U _T respectively represent the rolling moment, pitching moment, yaw moment and lift when the motors of the UAV are not faulty, f _p , f _q , f _r , f _z respectively Indicates roll moment error, pitch moment error, yaw moment error and lift error.

In a possible implementation manner, the allocation information obtaining module 1404 includes:

The allocation information acquisition unit is used to obtain the optimized allocation matrix ^Nf by using the following formula, and use the optimized allocation matrix ^Nf as the control allocation information of all motors on the drone:

Nf = ^{Af -1} ;

N ^f = A ^fT (A ^f A ^fT ) ^-1 ; wherein, A ^f represents the control efficiency matrix after a partial motor failure, and A ^fT represents the transposition of A ^f ;

A ^f = AR _i ; where, A represents the control efficiency matrix before failure;

其中，b为升力系数，l为无人机的轴距，d为反扭矩系数。

Among them, b is the lift coefficient, l is the wheelbase of the UAV, and d is the counter torque coefficient.

A fault-tolerant control system for a multi-rotor UAV provided by an embodiment of the present invention uses an improved linear active disturbance rejection control attitude control algorithm to control the flight of the UAV to ensure the robustness of the UAV during flight; When some motors of the UAV are detected to be abnormal, the fault matrix R _i is constructed; the fault model is established online based on the fault matrix R _i ; the control distribution information of all motors on the UAV is obtained based on the fault matrix R _i ; the improved linear active disturbance rejection The attitude control algorithm of the control controls the flight of the UAV under the fault model, and controls the UAV according to the control distribution information of the motor to achieve the required attitude and altitude. When the motor of the UAV does not fail, the basic control law selected is the attitude control algorithm of the improved linear active disturbance rejection control, which has strong robustness to disturbance. Moreover, when some motors of the UAV fail, based on the failure matrix after the failure, a new flight model-fault model and the control distribution information of all motors can be obtained, so as to ensure that the UAV can operate in this failure mode. While flying, it can also reduce the use of faulty motors, so that the UAV can achieve a stable flight state.

Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: it can still be Modifications are made to the technical solutions described in the foregoing embodiments, or equivalent replacements are made to some of the technical features; these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims (8)

1. A multi-rotor unmanned aerial vehicle fault-tolerant control method, is characterized in that, it comprises: The attitude control algorithm of the improved linear active disturbance rejection control is used to control the flight of the UAV, so as to ensure the robustness of the UAV during flight; When some motors of the UAV are detected to be abnormal, a fault matrix R _i is constructed, where i is an integer greater than or equal to 0 and less than or equal to the number of all motors in the UAV; Establishing a fault model online based on the fault matrix R _i ; Obtain the control allocation information of all motors on the UAV based on the fault matrix R _i ; Adopt the attitude control algorithm of the improved linear ADRC to control the flight of the unmanned aerial vehicle under the fault model, and control the unmanned aerial vehicle according to the control distribution information of the motor to achieve the required attitude and high; The online establishment of a fault model based on the fault matrix R _i includes: The fault model is established online using the following formula:

in

Represent the position acceleration in the earth coordinate system,

represent the angular acceleration of the aircraft attitude angle of the unmanned aerial vehicle under the earth coordinate system respectively,

θ, ψ represent the roll angle, pitch angle and yaw angle respectively, I _x , I _y , I _z represent the moments of inertia of the UAV fuselage in three directions respectively, and m represents the mass of the UAV , g represents the acceleration of gravity, U _R , U _P , U _Y , U _T respectively represent the rolling moment, pitching moment, yaw moment and lift when the motors of the UAV are not faulty, f _p , f _q , f _r , f _z represent roll moment error, pitch moment error, yaw moment error and lift error respectively. 2. a kind of multi-rotor UAV fault-tolerant control method according to claim 1, is characterized in that, the attitude control algorithm of described improved linear active disturbance rejection control comprises: Arrange the transition process: use the following formula to convert the input sudden change signal into a slow change signal through the second-order link, and then make the output signal reach the desired input signal:

Wherein, G(s) represents the transfer function of the second-order link, T represents the time constant of the second-order link, and s represents the variable symbol in the transfer function; Linear expansion state observer: use the following state space equations and formulas to realize real-time tracking of variables in the model:

Among them, x ₁ , x ₂ , x ₃ respectively represent the state variables of the system,

b ₀ represents the estimated control gain, w represents the external disturbance, y represents the output of the model, and u represents the input of the model;

Wherein, z ₁ , z ₂ , z ₃ respectively represent the system state variables of the linearly extended state observer, and β ₁ , β ₂ , β ₃ represent the gain of the linearly extended state observer respectively. 3. A kind of multi-rotor unmanned aerial vehicle fault-tolerant control method according to claim 1, is characterized in that, described when detecting the part motor abnormality of described unmanned aerial vehicle, constructing fault matrix R _i , comprising: Real-time detection of all motors of the drone; When some motors of the UAV are detected to be abnormal, the ratio of the output of the faulty motor to the output without fault is calculated, and a fault matrix R _i is constructed according to the ratio. 4. a kind of multi-rotor unmanned aerial vehicle fault-tolerant control method according to claim 1, is characterized in that, described based on described failure matrix R _i obtains the control distribution information of all motors on the described unmanned aerial vehicle, comprising: The following formula is used to obtain the optimized allocation matrix N ^f , and the optimized allocation matrix N ^{f is} used as the control allocation information of all motors on the drone: Nf = ^{Af -1} ; N ^f = A ^fT (A ^f A ^fT ) ^-1 ; wherein, A ^f represents the control efficiency matrix after a partial motor failure, and A ^fT represents the transposition of A ^f ; A ^f = AR _i ; where, A represents the control efficiency matrix before failure;

Wherein, b is the lift coefficient, l is the wheelbase of the UAV, and d is the counter torque coefficient. 5. A multi-rotor unmanned aerial vehicle fault-tolerant control system, is characterized in that, it comprises: The first control module is used to control the flight of the unmanned aerial vehicle by adopting the attitude control algorithm of the improved linear active disturbance rejection control, so as to ensure the robustness of the unmanned aerial vehicle during flight; A fault matrix construction module, used to construct a fault matrix R _i when detecting abnormality of some motors of the drone, where i is an integer greater than or equal to 0 and less than or equal to the number of all motors in the drone; A fault model building module, configured to build a fault model online based on the fault matrix R _i ; An allocation information acquisition module, configured to obtain control allocation information of all motors on the UAV based on the fault matrix R _i ; The second control module adopts the attitude control algorithm of the improved linear active disturbance rejection control to control the flight of the UAV under the fault model, and controls the UAV according to the control distribution information of the motor, so as to achieve desired attitude and altitude; The fault model establishment module includes: a fault model establishment unit, which is used to establish a fault model online by using the following formula:

in

Represent the position acceleration in the earth coordinate system,

represent the angular acceleration of the aircraft attitude angle of the unmanned aerial vehicle under the earth coordinate system respectively,

θ, ψ represent the roll angle, pitch angle and yaw angle respectively, I _x , I _y , I _z represent the moments of inertia of the UAV fuselage in three directions respectively, and m represents the mass of the UAV , g represents the acceleration of gravity, U _R , U _P , U _Y , U _T respectively represent the rolling moment, pitching moment, yaw moment and lift when the motors of the UAV are not faulty, f _p , f _q , f _r , f _z represent roll moment error, pitch moment error, yaw moment error and lift error respectively. 6. a kind of multi-rotor unmanned aerial vehicle fault-tolerant control system according to claim 5, is characterized in that, described failure matrix building block comprises: The transition process is arranged to convert the input sudden change signal into a slow change signal through the second-order link by using the following formula, and then make the output signal reach the desired input signal:

Wherein, G(s) represents the transfer function of the second-order link, T represents the time constant of the second-order link, and s represents the variable symbol in the transfer function; The Linear Extended State Observer unit is used to implement real-time tracking of variables in the model using the following state space equations and formulas:

Among them, x ₁ , x ₂ , x ₃ respectively represent the state variables of the system,

b ₀ represents the estimated control gain, w represents the external disturbance, y represents the output of the model, and u represents the input of the model;

Wherein, z ₁ , z ₂ , z ₃ respectively represent the system state variables of the linearly extended state observer, and β ₁ , β ₂ , β ₃ represent the gain of the linearly extended state observer respectively. 7. a kind of multi-rotor unmanned aerial vehicle fault-tolerant control system according to claim 5, is characterized in that, described failure model building module comprises: A detection unit is used to detect all motors of the drone in real time; The fault model building unit is used to calculate the ratio of the output of the faulty motor to the output when there is no fault when some motors of the drone are detected to be abnormal, and construct a fault matrix R _i according to the ratio. 8. A kind of multi-rotor UAV fault-tolerant control system according to claim 5, is characterized in that, described distribution information acquisition module comprises: The allocation information acquisition unit is used to obtain the optimized allocation matrix N ^f by using the following formula, and use the optimized allocation matrix N ^f as the control allocation information of all motors on the drone: Nf = ^{Af -1} ; N ^f = A ^fT (A ^f A ^fT ) ^-1 ; wherein, A ^f represents the control efficiency matrix after a partial motor failure, and A ^fT represents the transposition of A ^f ; A ^f = AR _i ; where, A represents the control efficiency matrix before failure;

Wherein, b is the lift coefficient, l is the wheelbase of the UAV, and d is the counter torque coefficient.

Images