KR102294829B1 - System and flight control method for unmanned aerial vehicle with variable load - Google Patents

Abstract

In accordance with an embodiment of the present invention, a system for controlling the flight of an uncrewed aerial vehicle having a variable load, comprises: a modeling unit for inducing a motion equation by performing dynamic modeling on an uncrewed aerial vehicle; a controller design unit for designing a sliding mode controller for the uncrewed aerial vehicle using a dynamic model or the motion equation obtained from the modeling unit; and an adaptive sliding mode controller obtained by reflecting an adaptive rule obtained from an adaptive rule inducing unit to the sliding mode controller. The adaptive rule inducing unit induces the adaptive rule by reflecting disturbances including a change in mass or load of the uncrewed aerial vehicle, and the adaptive sliding mode controller can be obtained by applying the adaptive rule to the sliding mode controller. Accordingly, designed performance requirements can be satisfied.

Description

SYSTEM AND FLIGHT CONTROL METHOD FOR UNMANNED AERIAL VEHICLE WITH VARIABLE LOAD

The present invention relates to an unmanned aerial vehicle flight control system and method having a variable load, and more particularly, to an unmanned aerial vehicle flight control system and method having a variable load that can minimize changes in control performance due to the loading or removal of a payload is about

Recently, as the problem of ground traffic congestion becomes more serious, the value of multicopter-type unmanned aerial vehicles as a means of air transportation is attracting attention. In particular, the quadcopter-type unmanned aerial vehicle (drone) has a distributed propulsion structure, so it is evaluated as suitable as a means of urban air transportation due to its low noise, low risk of partial failure, and vertical take-off and landing and stationary flight.

In the case of a quadcopter-type unmanned aerial vehicle for transport, it can carry a significant weight compared to the mass of the fuselage, and the control performance changes when a payload is loaded or removed from the mass of the fuselage, and the z-axis direction It becomes unstable due to the sudden external force acting on it.

However, when the P-PID controller of the existing dual loop structure currently used is designed to have the control performance required for the fixed parameter, it may be suitable for the fixed parameter, but the control performance is dependent on the change of the parameter or the control performance is poor. There is a lower limit. That is, in the case of a P-PID controller designed to have the control performance required for the nominal mass, there is a problem in that the designed control performance is not guaranteed when a parameter is changed.

In addition, in the case of the existing sliding mode controller (SMC), robust performance against parameter uncertainty and disturbance has been proven, but there is a problem in that chattering and steady state errors exist when a large load is changed. In addition, when there is a sudden load change in the vertical downward direction in a stationary flight state, there is a problem that an altitude error occurs due to this.

Accordingly, there is a growing demand for flight control technology that can minimize the change in control performance due to the loading or removal of the payload.

The present applicant has proposed the present invention in order to solve the above problems.

Korean Patent Registration No. 10-0963394 (2010.06.04.)

The present invention has been proposed to solve the above problems, and provides a flight control system and method for an unmanned aerial vehicle having a variable load that can improve the altitude and position tracking performance of an unmanned aerial vehicle to which the variable load is applied.

The present invention provides an unmanned aerial vehicle flight control system and method having a variable load capable of securing basic control performance for a flight situation in which system parameters change significantly.

An unmanned aerial vehicle flight control system having a variable load according to an embodiment of the present invention for achieving the above object includes: a modeling unit for inducing a motion equation by performing dynamic modeling on the unmanned aerial vehicle; a controller design unit for designing a sliding mode controller for the unmanned aerial vehicle using the dynamic model or motion equation obtained from the modeling unit; and an adaptive sliding mode controller obtained by reflecting the adaptive rule obtained by the adaptation rule induction unit to the sliding mode controller, wherein the adaptation rule inducing unit reflects the disturbance including a change in mass or load of the unmanned aerial vehicle to reflect the adaptive rule , and the adaptive sliding mode controller can be obtained by applying the adaptive rule to the sliding mode controller.

The modeling unit derives an equation of motion of the unmanned aerial vehicle in consideration of internal and external disturbances applied to the unmanned aerial vehicle, wherein the internal disturbance includes a change in mass or load acting on the unmanned aerial vehicle, and the external disturbance is The influence of wind acting on the unmanned aerial vehicle may be considered.

The adaptive sliding mode controller maintains robustness against disturbance of the sliding mode controller designed by the controller design unit and applies the adaptive rule obtained from the adaptive rule induction unit to the sliding mode controller to estimate unknown parameters and control variables can be changed.

The adaptation rule inducing unit may induce an adaptation rule in order to estimate a change in mass or load applied to the unmanned aerial vehicle.

The adaptive law derivation unit may define a Liafnov function, which is always positive, in order to obtain a control input that minimizes an error between the nominal mass and the estimated mass of the unmanned aerial vehicle.

The adaptation law derivation unit may derive the adaptation law for the estimated mass by using a negative semi-definite sign condition of the Liafnov function.

On the other hand, according to another field of the invention, the present invention provides a flight control method for an unmanned aerial vehicle having a variable load using the above-described unmanned aerial vehicle flight control system having a variable load, the method comprising: performing dynamic modeling of the unmanned aerial vehicle; deriving a condition for the dynamic model of the unmanned aerial vehicle to reach the sliding surface; designing a Liafnov function; determining whether a differential value of the Liafnov function is greater than zero; and inducing an adaptation law; may provide an unmanned aerial vehicle flight control method having a variable load comprising.

In the designing of the Liafnov function, a Liafnov function, which is always positive, may be defined in order to obtain a control input that minimizes an error between the nominal mass and the estimated mass of the unmanned aerial vehicle.

In the step of deriving the adaptation law, the adaptation law for the estimated mass may be derived by using a negative semi-definite condition of the Liafnov function.

The unmanned aerial vehicle flight control system and method having a variable load according to the present invention can maintain control performance even when the load applied to the unmanned aerial vehicle changes and satisfy the designed required performance.

The unmanned aerial vehicle flight control system and method having a variable load according to the present invention can minimize the change in control performance due to the loading or removal of the payload applied to the unmanned aerial vehicle.

An unmanned aerial vehicle flight control system and method having a variable load according to the present invention can minimize an altitude tracking control error that occurs when a load applied to an unmanned aerial vehicle changes.

The unmanned aerial vehicle flight control system and method with variable load according to the present invention considers both the influence of external disturbance acting on the unmanned aerial vehicle and the change of mass, which is an internal disturbance, so that robust flight control is possible against disturbance and parameter change.

1 is a diagram illustrating the configuration of an unmanned aerial vehicle flight control system having a variable load according to an embodiment of the present invention.
2 is a diagram illustrating a coordinate system of an unmanned aerial vehicle according to an embodiment of the present invention.
3 is a diagram for explaining the principle of motion of an unmanned aerial vehicle according to an embodiment of the present invention.
4 is a diagram for explaining a dynamic model of an unmanned aerial vehicle according to an embodiment of the present invention.
5 and 6 are flowcharts for explaining a method for controlling the flight of an unmanned aerial vehicle having a variable load according to an embodiment of the present invention.
7 to 9 are diagrams for verifying the control performance of the flight control system according to an embodiment of the present invention.

Hereinafter, embodiments according to the present invention will be described in detail with reference to the accompanying drawings. However, the present invention is not limited or limited by the examples. Like reference numerals in each figure indicate like elements.

1 is a diagram illustrating the configuration of an unmanned aerial vehicle flight control system having a variable load according to an embodiment of the present invention, FIG. 2 is a diagram illustrating a coordinate system of an unmanned aerial vehicle according to an embodiment of the present invention, and FIG. 3 is this view A diagram for explaining the principle of motion of an unmanned aerial vehicle according to an embodiment of the present invention, FIG. 4 is a diagram for explaining a dynamic model of an unmanned aerial vehicle according to an embodiment of the present invention, and FIGS. 5 and 6 are one of the present invention A flowchart for explaining a method for controlling an unmanned aerial vehicle having a variable load according to an embodiment, FIGS. 7 to 9 are diagrams for verifying the control performance of the flight control system according to an embodiment of the present invention.

In the present invention, the unmanned aerial vehicle 100 refers to a rotary wing unmanned aerial vehicle including a drone, a quadcopter, and a multicopter. Hereinafter, for convenience of description, a case in which the unmanned aerial vehicle 100 is a quadcopter will be described.

Referring to FIG. 1 , an unmanned aerial vehicle flight control system 10 having a variable load according to an embodiment of the present invention includes an unmanned aerial vehicle 100 including a quadcopter; a modeling unit 200 for inducing an equation of motion by performing dynamic modeling on the unmanned aerial vehicle 100; a controller design unit 300 for designing a sliding mode controller for the unmanned aerial vehicle 100 using the dynamic model obtained from the modeling unit 200; and an adaptive sliding mode controller 400 obtained by reflecting the adaptive rule obtained by the adaptation rule inducing unit 500 to the sliding mode controller.

Here, the adaptation rule inducing unit 500 may induce the adaptation rule by reflecting disturbances including variations in mass or load applied to the unmanned aerial vehicle 100 .

The unmanned aerial vehicle flight control system 10 having a variable load according to an embodiment of the present invention (hereinafter referred to as a "flight control system") is an unmanned aerial vehicle 100 used for high-function missions such as cargo and passenger transport system parameters. It can ensure basic control performance (response) even in greatly changing flight conditions. In addition, it minimizes the change in control performance due to the loading of a payload on the unmanned aerial vehicle 100 or the removal of the payload, and an altitude tracking control error of the unmanned aerial vehicle that occurs when the load applied to the unmanned aerial vehicle 100 changes. can be minimized.

The coordinate system used in the flight control system 10 according to an embodiment of the present invention is an inertial coordinate system and a body (aircraft) coordinate system, as shown in FIG. 2 . The inertial coordinate system uses the ENU coordinate system, and in the fuselage coordinate system, the forward direction of the unmanned aerial vehicle 100 is the X-axis and the vertical upward direction is the Z-axis, and the Y-axis direction is defined by the right-hand rule.

, 동체 좌표계 기준 무인비행체(100)의 선속도를

으로 정의하고, 오일러 각속도

와 동체 좌표계의 각속도를

라고 정의하면 관성 좌표계 기준 속도와 동체 좌표계 기준 속도 사이의 관계식은 [수학식 1]과 같다.The position vector of the unmanned aerial vehicle (100, quadcopter) based on the inertial coordinate system

, the linear velocity of the unmanned aerial vehicle 100 based on the fuselage coordinate system

, and Euler angular velocity

and the angular velocity of the body coordinate system

If defined, the relational expression between the reference speed of the inertial coordinate system and the reference speed of the body coordinate system is [Equation 1].

The coordinate transformation matrices R and J between the inertial coordinate system and the moving body coordinate system are shown in [Equation 2] and [Equation 3].

When a quadcopter is used as the unmanned aerial vehicle 100 , the quadcopter may be composed of two pairs of propellers and four motors facing each other in a distributed propulsion structure. One pair of motors facing each other rotates in the same direction, and the other pair of motors facing each other rotate in the same direction, but rotate in opposite directions to the other pair of motors to cancel the torque.

3 shows the principles of hovering, rolling, pitching, and yawing rotational motions of the cruciform quadcopter. The thickness of the arrow indicating the rotation direction of the propeller in FIG. 3 means the rotation speed. That is, when the thickness of the arrow is large, it means that the rotation speed is faster.

The quadcopter rises and falls by the sum of the thrust generated by the rotation of each rotor (propeller + motor), and the difference in motor rotation speed creates a moment that generates rolling and pitching motion, and as the quadcopter is tilted, the thrust component It is decomposed into a lateral force and a Z-axis force.

The flight control system 10 according to an embodiment of the present invention models the unmanned aerial vehicle 100 having the above-described coordinate system, that is, a quadcopter, and controls the flight of the unmanned aerial vehicle using the result.

The unmanned aerial vehicle 100 according to an embodiment of the present invention can be used to transport cargo or passengers. It can affect altitude retention.

The cargo transport method includes a method of mounting a gripper on the unmanned aerial vehicle to position the payload near the center of gravity, and a method of increasing the degree of freedom by connecting the unmanned aerial vehicle and the payload with a cable. In the case of the unmanned aerial vehicle 100 according to the present invention, a method of mounting a gripper to position the payload near the center of gravity is used.

On the other hand, when the unmanned aerial vehicle 100 is used for cargo transportation, etc., a change in mass (load) due to the payload acts as an internal disturbance to the unmanned aerial vehicle 100. It acts as an external disturbance to the aircraft 100 .

In order to control the flight of the unmanned aerial vehicle 100 when a variable load is applied to the unmanned aerial vehicle 100, which is a quadcopter, first, the flight control system 10 according to an embodiment of the present invention controls the flight of the unmanned aerial vehicle 100. modeling will be performed. In this case, the flight control system 10 may perform dynamic modeling of the unmanned aerial vehicle 100 in consideration of the variable load. To this end, the flight control system 10 according to an embodiment of the present invention may include a modeling unit 200 for inducing an equation of motion by performing dynamic modeling on the unmanned aerial vehicle 100 .

The modeling unit 200 derives an equation of motion of the unmanned aerial vehicle 100 in consideration of internal and external disturbances applied to the unmanned aerial vehicle 100 , and the internal disturbance is a change in mass acting on the unmanned aerial vehicle 100 . Including, the external disturbance may consider the influence of the wind acting on the unmanned aerial vehicle (100).

Referring to FIG. 4 , the modeling unit 200 may include a control assignment unit 210 , a motor dynamics unit 220 , an aerodynamic unit 230 , and a rigid body dynamics unit 240 .

_{The thrust f i} and anti-torque τ _i generated aerodynamically by the rotation of the i-th rotor are proportional to the square of the rotational speed ω _i and can be expressed as in [Equation 4], where k _T is the thrust coefficient and k _D is the value experimentally obtained through the motor test as the drag coefficient.

The control input for changing the altitude and attitude of the unmanned aerial vehicle 100, which is a quadcopter, is expressed as [Equation 5] by the thrust and anti-torque generated by the rotation of each rotor, and the distance (l) from the center of gravity to the rotor. can

On the other hand, when the unmanned aerial vehicle 100 is a quadcopter, the equation of motion of the unmanned aerial vehicle 100 may be derived under the following assumption.

(1) A quadcopter is a rigid body and is symmetrical about the xz and yz planes of the body coordinate system.

(2) The center of the body coordinate system coincides with the center of gravity of the quadcopter.

(3) The amount of change in mass moment of inertia due to mass change is very small.

The action of thrust and moment force generated by motor rotation and the angular velocity measured by the IMU are based on the body coordinate system. .

The translational motion equation is derived as [Equation 7] from Newton's second law in order to represent it based on the inertial coordinate system. In [Equation 7], F _t and F _g represent a force generated by thrust and a force caused by gravity, respectively.

으로 간단해지지만, 본 발명의 경우에는 무인비행체(100) 즉, 쿼드콥터에 가변 하중이 작용하기 때문에 [수학식 8]과 같이 고려된다.If there is no change in the mass acting on the unmanned aerial vehicle 100,

In the case of the present invention, it is considered as in [Equation 8] because a variable load acts on the unmanned aerial vehicle 100, that is, the quadcopter.

When considering the external disturbance with boundary conditions and the internal disturbance due to a change in mass, a translational motion equation as in [Equation 9] is finally obtained.

On the other hand, the controller design unit 300 of the flight control system 10 according to an embodiment of the present invention uses the dynamic model of the unmanned aerial vehicle 100 obtained from the modeling unit 200 in a sliding mode for the unmanned aerial vehicle 100 . Design the controller. That is, the controller design unit 300 may be provided to design a sliding mode controller using the equation of motion of the unmanned aerial vehicle 100 obtained by the modeling unit 200 .

It is difficult to know the model of the quadcopter that is the unmanned aerial vehicle 100 perfectly, and uncertainty about the model always exists. The sliding mode controller is robust to model uncertainty and external disturbance with boundary conditions, and can be applied to a system with uncertainty.

Meanwhile, the sliding mode controller (SMC) defines a stable sliding surface having desired characteristics in phase portrait and designs a control input so that state vectors of the system are located on the sliding surface. It consists of an arrival mode and a sliding mode. When the sliding mode is entered, the system has robust performance against model uncertainty and disturbance. Therefore, designing a stable sliding surface and the condition to reach the sliding surface are important, and the arrival condition can be obtained by the Lyapunov theory.

The Lyapunov function candidate for obtaining the arrival condition is defined as [Equation 10].

을 만족하는 제어 입력을 가함으로써 s(x)=0인 슬라이딩 표면에 도달할 수 있다. 이를 유지하도록 제어 입력을 설계하면 상태 벡터는 불확실성의 영향을 적게 받으며, 슬라이딩 표면을 타고 phase portrait의 원점으로 미끄러져 안정한 상태가 된다.When the state variable x in [Equation 11] does not lie on the sliding surface,

A sliding surface with s(x)=0 can be reached by applying a control input that satisfies If the control input is designed to maintain this, the state vector is less affected by uncertainty, and it slides on the sliding surface to the origin of the phase portrait and becomes a stable state.

The position dynamics model of the quadcopter that is the unmanned aerial vehicle 100 in the modeling unit 200 [Equation 9] may be expressed as [Equation 12].

In [Equation 12], f(x) and g(x) are defined as [Equation 13] and [Equation 14], respectively.

In addition, the state vector and the control input vector are as shown in [Equation 15] and [Equation 16], respectively.

The path tracking control of the unmanned aerial vehicle 100 is a problem of minimizing the position error, and the position error vector e is defined as [Equation 17].

With K ₁ as a positive gain matrix, the sliding surface is defined as in [Equation 18] so that the error dynamics have the response characteristics of the first-order system.

A Lyapunov function, which is always positive for the sliding surface, is defined as in [Equation 19].

When the differential value of the Liafnov function shown in [Equation 20] is always negative definite, the stability of the sliding surface can be guaranteed. A control input to which a negative positive sign condition is applied to ensure stability is designed as in [Equation 21].

In [Equation 21], k ₂ is a gain matrix used as a control input to create the desired error dynamics shown in [Equation 22].

In [Equation 21], k ₃ is the size of the sign function used for the control input to draw into the sliding surface when there is an influence of disturbance, and is always a positive matrix.

The Signum function is a discontinuous function and has the advantage of being robust against disturbance by implementing immediate compensation, but it may cause chattering when the modeling error is large. For this, the hyperbolic tangent function is used instead of the signum function. The finally designed control input U is shown in [Equation 23].

The shape of the hyperbolic tangent function is determined by λ, and the larger the λ value, the closer to the signum function, the more likely the chattering phenomenon will occur.

Substituting [Equation 23] into [Equation 20], for the differential expression of the Liafnov function for the sliding surface,

일 때 음의 정 부호 조건을 만족하여 Asymptotically stable함을 증명할 수 있다.Because of

When , it can be proved that it is asymptotically stable by satisfying the negative positive condition.

The controller design unit 300 of the flight control system 10 according to an embodiment of the present invention may design the control rule of the sliding mode controller as in [Equation 23] through the above-described process.

However, in the case of the present invention, the quadcopter, which is the unmanned aerial vehicle 100, can be used for transportation, and in the case of transportation, a considerably heavier weight can be loaded compared to the mass of the unmanned aerial vehicle, and this weight is caused by a sudden external force in the z-axis direction ( internal disturbance).

The sliding mode controller obtained from the controller design unit 300 shows strong control performance against disturbances with a certain amount of parameter fluctuations and boundary conditions, but chattering occurs when a load change above the boundary value is applied to the unmanned aerial vehicle 100. There is a limit that the control performance is not sufficient because it may occur or a steady-state error may occur. Accordingly, the flight control system 10 according to an embodiment of the present invention applies the adaptive control to appropriately change the controller variable as the unknown parameter of the control target changes to satisfy the control performance. (400) can be designed.

The adaptive sliding mode controller 400 can obtain the adaptive rule obtained by the adaptive rule derivation unit 500 by reflecting the adaptive rule obtained in the sliding mode controller.

The adaptive sliding mode controller 400 estimates and controls unknown parameters by using the adaptive law obtained from the adaptive law induction unit 500 while maintaining the robustness against disturbance of the sliding mode controller designed by the controller design unit 300 . Variables can be changed. As described above, the flight control system 10 according to an embodiment of the present invention applies the adaptive control rule to the sliding mode controller obtained by the controller design unit 300, thereby securing control performance for unknown parameter changes. The flight of the unmanned aerial vehicle 100 having a variable load is controlled by using the mode controller 400 .

In order to design the adaptive sliding mode controller 400 , the adaptive rule inducing unit 500 may derive an adaptive rule for estimating a change in mass applied to the unmanned aerial vehicle 100 .

라고 정의하여, [수학식 23]에서 유도된 제어 입력에 g(x)의 추정치

를 사용하면 [수학식 25]와 같은 제어 입력을 설계할 수 있다.The adaptation law obtained by the adaptation law induction unit 500 is a rule for estimating the change in mass applied to the unmanned aerial vehicle 100 . Let the total mass (ie, nominal mass) including the change in the mass of the payload applied to the unmanned aerial vehicle 100 be m _t , and an estimate of the mass

, the estimated value of g(x) in the control input derived from [Equation 23]

can be used to design a control input as in [Equation 25].

이다.In [Equation 25]

am.

를 최소화하는 제어 입력을 얻기 위해 항상 양수인 리아프노브(Lyapunov) 함수를 [수학식 26]과 같이 정의한다.A sliding surface is reached, and the estimation error

In order to obtain a control input that minimizes , a Lyapunov function, which is always a positive number, is defined as in [Equation 26].

Differentiating the Liafnov function of [Equation 26] gives [Equation 27].

라고 가정하고, 슬라이딩 표면의 미분식을 대입하여 [수학식 27]을 전개하면 [수학식 28]과 같다.The instantaneous mass change is negligible and

, and [Equation 27] is developed by substituting the differential equation of the sliding surface to become [Equation 28].

When [Equation 28] is substituted with [Equation 29] to simplify [Equation 28], the differential expression of the Ryafnov function is arranged as [Equation 30].

If the negative semidefinite condition of the Liafnov function is used to ensure stability, [Equation 31], which means the law of adaptation to the estimate of mass, is obtained.

When the estimated value of [Equation 31] is substituted into the differential expression of the Liafnov function of [Equation 30], [Equation 32] becomes [Equation 32], which satisfies only the semi-signal condition.

Therefore, only the Liafnov stability is guaranteed, but according to LaSalle's invariance principle, _{if m t} is within the boundary condition, the Liafnov function V ₂ is a decreasing function, so it converges to a constant, thereby proving that it is asymptotically stable.

) 사이의 오차(

)를 최소화하는 제어 입력을 얻기 위해 항상 양수인 리아프노브 함수([수학식 26] 참조)를 정의할 수 있다.As described above, the flight control system 10 according to an embodiment of the present invention uses the adaptive law induction unit 500 to determine the nominal mass (m _t ) and the estimated mass (

) between (

), it is possible to define a Liafnov function (refer to [Equation 26]), which is always positive, in order to obtain a control input that minimizes it.

The flight control system 10 according to an embodiment of the present invention includes the unmanned aerial vehicle 100 together with the first to third terms (switching function) on the right side of [Equation 26], which means the arrival condition to the sliding surface. By considering the fourth term on the right-hand side, which means the variable load (external disturbance) applied to , at once, and defining the always positive Liafnov function, the law of adaptation for the estimate of mass can be obtained.

Referring to FIG. 1 , the flight control system 10 according to an embodiment of the present invention includes a disturbance reflecting unit 600 interlocking with the adaptation rule induction unit 500 so that the adaptation rule induction unit 500 considers a variable load. can do.

The adaptation law derivation unit 500 may derive an adaptation law for the estimated mass (refer to [Equation 31]) by using the negative semi-definite sign condition of the Liafnov function. The adaptive law induction unit 500 obtains the adaptive sliding mode controller 400 considering the variable load by applying the adaptive law as in [Equation 31] to the sliding mode controller.

5 and 6 are flowcharts illustrating a method for controlling the flight of an unmanned aerial vehicle having a variable load (hereinafter, referred to as a "flight control method") according to an embodiment of the present invention. A flight control method according to an embodiment of the present invention refers to a flight control method by the above-described flight control system.

5, the flight control method according to an embodiment of the present invention is a flight control method by the above-described flight control system 10, the unmanned aerial vehicle modeling step (S100); Sliding mode controller design step (S200); and an adaptation rule induction step (S300).

The unmanned aerial vehicle modeling step ( S100 ) is performed by the modeling unit 200 , and dynamic modeling may be performed on the quadcopter type unmanned aerial vehicle 100 to derive a motion equation.

The sliding mode controller design step (S200) is performed by the controller design unit 300, and the sliding mode controller can be designed using the dynamic model or motion equation of the unmanned aerial vehicle 100 obtained in the unmanned aerial vehicle modeling step (S100). have.

The adaptation rule induction step (S300) is performed by the adaptation rule induction unit 500, and by inducing the adaptation rule and using it in the sliding mode controller, when a load change above the boundary value is applied to the unmanned aerial vehicle 100, chattering is It is possible to prevent a problem in which the control performance of the sliding mode controller is insufficient, such as occurrence or steady state error. In this way, the adaptive sliding mode controller can be designed and used for flight control by the adaptive rule induction step (S300).

6 is a flowchart showing in more detail a flight control method according to an embodiment of the present invention. 6, in the flight control method according to an embodiment of the present invention, in the unmanned aerial vehicle flight control method having a variable load using the above-described unmanned aerial vehicle flight control system 10 having a variable load, the unmanned aerial vehicle ( 100) performing dynamic modeling for (S1100); deriving a condition for the dynamic model of the unmanned aerial vehicle 100 to reach the sliding surface (S1210); designing a Liafnov function (S1230); determining whether the differential value of the Liafnov function is greater than 0 (S1250); and deriving an adaptation law (S1300).

The step (S1100) of performing dynamic modeling on the unmanned aerial vehicle 100 is performed by the modeling unit 200, and dynamic modeling is performed on the quadcopter type unmanned aerial vehicle 100 to perform the equation of motion. can be derived.

The step of deriving the condition for reaching the sliding surface (S1210) is performed by the controller design unit 300, and the sliding mode is performed using the dynamic model or motion equation of the unmanned aerial vehicle 100 obtained in the unmanned aerial vehicle modeling step (S100). The controller can be designed.

In the designing of the Liafnov function ( S1230 ), a Liafnov function, which is always positive, may be defined in order to obtain a control input that minimizes an error between the nominal mass and the estimated mass of the unmanned aerial vehicle 100 . To this end, the step of designing the Liafnov function (S1230) is performed by the adaptive law derivation unit 500, and the Liafnov function, which is always positive, is designed, but the Liafnov function expressed by [Equation 26] is defined. can do.

The step (S1250) of determining whether the differential value of the Liafnov function is greater than 0 is a step of determining the stability of the adaptive sliding mode controller 400, and the differentiation of the Liafnov function expressed by [Equation 27] If the value is greater than 0, it returns to the step of designing the Liafnov function (S1230), and if it is not greater than 0, the step of deriving the adaptation law (S1300) is performed.

The step of deriving the adaptation law ( S1300 ) is performed by the adaptation law inducing unit 500 , and an adaptation law for the estimated mass may be derived using the negative semi-definite condition of the Liafnov function.

In the step of deriving the adaptation law ( S1300 ), an adaptation law for an estimate of the mass expressed by Equation 31 is obtained.

The adaptive rule is used in the sliding mode controller to obtain the adaptive sliding mode controller 400, and the adaptive sliding mode controller 400 is used to control the flight of the unmanned aerial vehicle 100 (S1400).

On the other hand, the present applicant controls the unmanned aerial vehicle 100 using an adaptive sliding mode controller in order to consider the variable load applied to the flight control system 10 , that is, the unmanned aerial vehicle 100 according to an embodiment of the present invention. I tested the system. 7 to 9 show such verification procedures and results.

7 is a view for explaining a stationary flight scenario for verifying the flight control system 10 according to an embodiment of the present invention.

Referring to FIG. 7 , Phase 1 is a flight control system using a dual loop PID controller (P-PID) and an adaptive sliding mode controller according to an embodiment of the present invention for an unmanned aerial vehicle having a nominal mass of 2.6 kg. (10) is the section in which the gain value is adjusted so that the transient response and settling time have similar performance. Phase 2 is a section in which the unmanned aerial vehicle ascends with a total weight of 3.6 kg by loading a weight (1 kg) that is about 40% of the unmanned aerial vehicle's mass. Phase 3 is a section in which the payload (that is, a weight that is less than about 40% of the mass of the aircraft) is dropped while the unmanned aerial vehicle with a total weight of 3.6 kg is hovering.

In Phase 1, the unmanned aerial vehicle ascends or lands (Ascending/Landing), in Phase 2 it ascends (Ascending), and in Phase 3, it performs a hovering operation.

In FIG. 7, "Added payload 1kg" means to load a weight of about 1kg to the unmanned aerial vehicle, and "dropped payload 1kg" means to separate and drop the weight of about 1kg from the unmanned aerial vehicle. In FIG. 7 , the horizontal axis means time and the unit is seconds (s).

8 is a simulation result of a flight test scenario of an unmanned aerial vehicle, in which (a) of FIG. 8 is a comparison result of altitude control performance, and (b) is a result of mass tracking performance of the present invention.

The flight scenario consists of three stages divided according to the mass change in order to clearly compare the controller performance under the same conditions (see Fig. 7). In Phase 1, the nominal mass of the aircraft (unmanned aerial vehicle) of 2.6 kg is used, and after reaching the target altitude of 1 m, it descends and lands. At the start of Phase 2, a weight of 1 kg, which is approximately 40% of the mass of the aircraft, is loaded at 20 seconds, then ascends to the target altitude of 1 m, and at 30 seconds at the beginning of Phase 3, the additional load is removed to equal the nominal mass. , the altitude command at this time is maintained at 1m.

Referring to (a) of FIG. 8 , the transient response and settling time of the present invention (ASMC) using a PID controller (P-PID) and an adaptive sliding mode controller for a unit step input of a target altitude of 1 m as a simulation result of Phase 1 It can be seen that the gain value is adjusted to have similar performance. As shown in Fig. 8(a), in Phase 1, both controllers reach the settling state within 5 seconds, but when the unmanned aerial vehicle rises to 3.6 kg after loading a weight of 1 kg in 20 seconds, PID (P- PID) increases the time it takes to reach the target altitude (see Reference), whereas in the present invention (see ASMC) using the adaptive sliding mode controller, the time taken to reach the target altitude is shorter than that of PID. It can be seen that the performance is maintained. When the weight is separated in Phase 3, the altitude rises instantaneously. It can be seen that in the present invention, the time taken to reach the target altitude after weight separation is shorter than the PID, as well as the altitude fluctuation is less.

Referring to (b) of FIG. 8 , the present invention using the adaptive sliding mode controller provides a mass value within 1 to 2 seconds even though a load is added at 20 seconds at the start of Phase 2 and additional gravity acts in the vertical downward direction. It can be seen that it follows and instantaneously generates a larger control input. The mass tracking performance is adjusted by the _{tracking gain k m value.}

Referring to (a) of FIG. 8 , at the moment when the load of 1 kg at which Phase 3 starts is removed, in the case of the PID controller (see P-PID), a path (altitude) error of about 0.4 m occurs and converges. Although it takes about 10 seconds, in the case of the present invention (see ASMC) using the adaptive sliding mode controller, a slight path error occurs within 0.1 m, but it can be seen that the convergence is quickly achieved within 5 seconds.

Referring to FIG. 8 , when a variable load is applied, the convergence time of the P-PID controller increases to about 10 seconds, whereas in the case of the present invention using the adaptive sliding mode controller, the performance is improved by generating a large control input by rapidly following the mass. can be found to be maintained. In addition, it can be seen that when a sudden external force (variable load) is applied, the present invention using the adaptive sliding mode controller generates a smaller altitude error compared to the P-PID controller and converges to the target altitude quickly.

In addition, the applicant performed a flight experiment to verify the performance of the flight control system 10 according to an embodiment of the present invention, and the result is shown in FIG. 9 . Figure 9 (a) is a flight test altitude response comparison result, Figure 8 (b) shows the flight test mass estimate.

The static flight experiment was performed with a gain value adjusted to have the same performance when the target altitude was 1 m and a step command was applied as in the flight simulation, and the nominal mass value was 2.6 kg.

Referring to (a) of FIG. 9 , the payload is dropped in 10 seconds after ascending with an additional weight of 1 kg mounted. In the case of PID (see P-PID), it took about 10 seconds to converge to the target altitude (Reference) because gravity was generated in the vertical downward direction by the additional weight, but the present invention using the adaptive sliding mode controller (see ASMC) ), as can be seen from (b) of FIG. 9 , the total mass value increased within 1.2 seconds was estimated, and a large control input was generated according to the increased mass value to compensate for the generated gravity and rapidly ascended to the target altitude within 5 seconds. It can be seen that reaching

Referring to Phase 3 of FIG. 9 (a), in the case of PID in 10 seconds of dropping the payload, an altitude error of 0.4 m occurs, and it takes more than 10 seconds to reach the target altitude again without rapidly reducing the error, whereas In the case of the present invention using the adaptive sliding mode controller, an altitude error of about 0.1 m occurs, but it can be seen that it converges to the target altitude within 5 seconds. This is because, in the case of PID, the control input cannot be rapidly reduced, whereas in the case of the present invention using the adaptive sliding mode controller, the estimated mass value is from 3.6 kg to about 2.5 kg as shown in FIG. converges within 1 or 2 seconds, resulting in a rapid decrease in control input.

As described above, in one embodiment of the present invention, specific matters such as specific components, etc., and limited embodiments and drawings have been described, but these are only provided to help a more general understanding of the present invention, and the present invention is not limited to the above embodiments. It is not limited, and various modifications and variations are possible from these descriptions by those of ordinary skill in the art to which the present invention pertains. Accordingly, the spirit of the present invention should not be limited to the described embodiments, and not only the claims described below, but also all those with equivalent or equivalent modifications to the claims will fall within the scope of the spirit of the present invention.

10: Unmanned aerial vehicle flight control system with variable load
100: unmanned aerial vehicle
200: modeling unit
300: controller design unit
400: adaptive sliding mode controller
500: adaptive law induction unit
600: disturbance reflection unit

Claims (9)

a modeling unit for inducing an equation of motion by performing dynamic modeling on an unmanned aerial vehicle;
a controller design unit for designing a sliding mode controller for the unmanned aerial vehicle using the dynamic model or motion equation obtained from the modeling unit; and
and an adaptive sliding mode controller obtained by reflecting the adaptive rule obtained from the adaptive rule derivation unit to the sliding mode controller,
The adaptation law induction unit induces the adaptation law by reflecting disturbances including changes in the mass or load of the unmanned aerial vehicle,
The adaptive sliding mode controller is an unmanned aerial vehicle flight control system having a variable load, characterized in that it is obtained by applying the adaptive rule to the sliding mode controller.
According to claim 1,
The modeling unit derives an equation of motion of the unmanned aerial vehicle in consideration of internal and external disturbances applied to the unmanned aerial vehicle;
The internal disturbance includes a change in mass or load acting on the unmanned aerial vehicle, and the external disturbance considers the influence of wind acting on the unmanned aerial vehicle flight control system having a variable load.
3. The method of claim 2,
The adaptive sliding mode controller,
Maintaining the robustness against disturbance of the sliding mode controller designed by the controller design unit and applying the adaptation rule obtained from the adaptation rule induction unit to the sliding mode controller to estimate unknown parameters and change control variables Unmanned aerial vehicle flight control system with variable load.
4. The method of claim 3,
The adaptive law induction unit,
An unmanned aerial vehicle flight control system having a variable load, characterized in that inducing the adaptation law to estimate a change in mass or load applied to the unmanned aerial vehicle.
5. The method of claim 4,
The adaptive law induction unit,
An unmanned aerial vehicle flight control system having a variable load, characterized in that always positive Liafnov function is defined to obtain a control input that minimizes an error between the nominal and estimated mass of the unmanned aerial vehicle.
6. The method of claim 5,
The adaptive law induction unit,
An unmanned aerial vehicle flight control system having a variable load, characterized in that the adaptive law for the estimated mass is derived using the negative semi-definite sign condition of the Liafnov function.
In the flight control method of an unmanned aerial vehicle having a variable load using the system according to any one of claims 1 to 6,
performing dynamic modeling on the unmanned aerial vehicle;
deriving a condition for the dynamic model of the unmanned aerial vehicle to reach the sliding surface;
designing a Liafnov function;
determining whether a differential value of the Liafnov function is greater than zero; and
deriving an adaptation law;
An unmanned aerial vehicle flight control method having a variable load, comprising:
8. The method of claim 7,
In the step of designing the Liafnov function,
An unmanned aerial vehicle flight control method with a variable load, characterized in that always positive Liafnov functions are defined to obtain a control input that minimizes an error between the nominal mass and the estimated mass of the unmanned aerial vehicle.
9. The method of claim 8,
In the step of deriving the adaptation law,
An unmanned aerial vehicle flight control method with a variable load, characterized in that the adaptive law for the estimated mass is derived by using the negative semi-definite sign condition of the Liafnov function.

Images