KR101419844B1 - Skid steering remote controlled unmanned spray capable of compensating odometric sensor noise with extended kalman filter - Google Patents

Abstract

A skid steering-type remote-controllable unmanned sprayer includes: wheels installed at the both lateral sides; a chemical spray device to spray a chemical through a chemical spray nozzle; and a boom to locate the chemical spray nozzle to spray the chemical to a target. The sprayer may include: a manipulation unit which includes switches for a user to manually or automatically control the driving, posture, and chemical spraying of the skid steering-type remote-controllable unmanned sprayer and generates selection signals according to the manipulation of the switches; a sensor unit which includes sensors for sensing the driving status and speed of the skid steering-type remote-controllable unmanned sprayer and neighboring obstacles and generates sensor signals by the sensors; a control unit which calculates pre-compensation position data, which is not yet compensated, from the sensors, compensates the observation errors included in the pre-compensation position data using the expanded Kalman filter to calculate post-compensation position data, determines the next target position according to the current position determined based on the post-compensation position data, the detected moving object guiding lines, and a certain operation schedule, and generates a wheel control signal for operating the wheels to move to the next target position; and a wheel driving unit which drives the wheels based on the wheel control signal.

Description

TECHNICAL FIELD [0001] The present invention relates to a skid steering type remote control unmanned air conditioner that corrects traveling sensor noise using an extended Kalman filter. [0002] The present invention relates to a skid steering type remote control unmanned air conditioner,

The present invention relates to a remote control unmanned controller, and more particularly, to a remote control unmanned controller having a skid-steered method.

The skid-steered steering system is a system in which the number of revolutions of the left and right wheels of the moving body is different or the steering is performed by braking only one wheel and the other wheel by skid in a desired direction without a separate steering device. Compared with a rack and pinion gear steering system that adjusts the direction of travel with a steering wheel as used in a typical automobile, it is widely used in the field of civil engineering and the like because of its simple, lightweight and durable structure.

In recent years, there has been an attempt to utilize unmanned vehicles for certain tasks in the agricultural sector to reduce costs and improve productivity. Particularly, spraying a plant with regularly planted fruit or vegetable crops on a large area of cultivated land is somewhat difficult for people to work continuously because it treats toxic pesticides in some cases. If you can do this with a remotely controlled unmanned controller, you will be very safe and productive.

In this case, the remote control unattended sprayer needs to be lightweight and simple in structure, since it has to operate for a long time with its own power source while moving along a narrow and long travel route between the crops to be controlled at a slow speed. Therefore, the skid steering method is a suitable steering method for the remote control unmanned controller.

However, the position of the skid steering system can be specified with high accuracy at a short distance, but it becomes difficult to control precisely when the position measurement error is accumulated due to a long moving distance.

Therefore, there is a need for a method capable of correcting a traveling position error of a skid-steered remote control sprayer.

[1] Korean Patent Publication No. 1999-0038243 (published Jun. 5, 1999) [2] Korean Patent Publication No. 2005-0018440 (published on February 23, 2005)

SUMMARY OF THE INVENTION It is an object of the present invention to provide a remote control unmanned sprayer of a skid-steered method for correcting traveling sensor noise using an extended Kalman filter.

A remote control unmanned sprayer of the skid-steered type according to one aspect of the present invention includes wheels mounted on both sides, a drug spraying mechanism for spraying a drug through a drug spray nozzle, And a boom for positioning the boom,

An operation unit including switches for manually or automatically controlling the running, posture, and medicine spraying of the skid-steered remote control unmanned sprayer by a user, and generating selection signals according to the operation of the switches;

A sensor unit including sensors for detecting a running state, a speed, and an obstacle in the vicinity of the skid steered remote control unmanned controller, and generating sensor signals by the sensors;

Calculating an uncorrected pre-correction positional information from the sensor signals, calculating an after-correction positional information by correcting an observation error included in the pre-correction positional information by an extended Kalman filter, A control unit for determining a next target position according to the determined current position, the detected moving object lead, and a predetermined progress schedule, and generating a wheel control signal for driving the wheels to move to the next target position; And

And a wheel driving unit for driving the wheels based on the wheel control signal.

According to one embodiment,

An automatic mode selection switch for selecting a driving mode or a medicine dispensing mode to be performed in an automatic mode or a manual mode, a forward / backward selection switch for selecting a forward running, a neutral or reverse running, a leftward, A steering selection switch for selecting a direction, a speed control lever for selecting acceleration, constant speed or deceleration, or a control switch for selecting the height, direction, or spray pressure of the medicine spraying mechanism.

According to one embodiment,

An ultrasonic sensor or optical sensors provided at the left and right sides of the front and rear sides of the front and rear to recognize the peripheral state, a tilt sensor for detecting the posture, a speedometer for detecting the speed, or an extension direction of the boom equipped with the medicine sprayer And a boom angle sensor for detecting the boom angle.

According to one embodiment,

A sensor signal processor for calculating pre-correction positional information, moving object inducing line direction, and posture information having an observation error from the sensor signals to an uncorrected state;

An extended Kalman filtering unit for calculating post-correction position information by correcting the pre-correction position information having the observation error by the extended Kalman filtering based on the pre-correction position information having the observation error, the moving object guide line direction and posture information;

A position controller for determining a current position of the skid steerable remote control unmanned controller based on the post-correction position information, and determining a next target position according to the current position, the moving object guide line direction, and a given progress schedule;

A wheel control section for generating a wheel control signal for controlling the number of revolutions of each of the wheels so that the skid-steered remote control unmanned controller moves from a current position to a next target position; And

And a medicine dispensing control unit for controlling a medicine dispensing apparatus driving unit for adjusting the posture and operation of the boom and the medicine dispensing apparatus.

According to one embodiment, the control unit

Further comprising a risk judging unit for judging whether or not a risk element such as a topography or an obstacle is present in front of the sensor based on the sensor signals,

If the risk judging unit judges a risk factor such as rollover or collision, the wheel control unit may operate to generate the wheel control signal for avoidance start based on the risk factor and the posture information.

According to one embodiment, the expanding Kalman filtering unit may be expressed by the following equation

The extended Kalman filtering is performed according to the following equation:

는 상태 변수 x의 사후적 추정 값이고, y는 관측 값이며, Q는 프로세스 오차 w의 공분산이다.Where K is the Kalman gain, P is the covariance of the estimation error w, H is the relational matrix between the observed value y and the state variable x, R is the covariance matrix of the observed error of the observed value y, A is the state transition matrix ,

Is the posterior estimate of the state variable x, y is the observed value, and Q is the covariance of the process error w.

According to the skid-steered remote control unmanned controller of the present invention, the travel error can be corrected with high accuracy, so that it can be used in a narrow running route, and can be run precisely according to the terrain.

1 is a block diagram illustrating an electrical and mechanical structure of a skid-steered remote control unattended sprayer according to an embodiment of the present invention.
FIG. 2 and FIG. 3 are conceptual diagrams illustrating kinematic models of a skid-steered remote control unmanned controller according to an embodiment of the present invention in terms of the relationship between free object motion, wheel and moving object velocity.
FIGS. 4 and 5 are graphs illustrating compensated position measurement results and absolute errors of the skid-steered remote control unmanned controller according to an embodiment of the present invention, respectively.

For the embodiments of the invention disclosed herein, specific structural and functional descriptions are set forth for the purpose of describing an embodiment of the invention only, and it is to be understood that the embodiments of the invention may be practiced in various forms, The present invention should not be construed as limited to the embodiments described in Figs.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. The same reference numerals are used for the same constituent elements in the drawings and redundant explanations for the same constituent elements are omitted.

1 is a block diagram illustrating an electrical and mechanical structure of a skid-steered remote control unattended sprayer according to an embodiment of the present invention.

1, a skid steering type remote control unmanned controller (hereinafter also referred to as an unmanned controller) 10 includes an operating portion 11, a sensor portion 12, a control portion 13, a wheel driving portion 14, A medicine dispensing device drive unit 15, wheels 16, a medicine dispensing mechanism 17, and a boom 18. [

The operation unit 11 may include switches 111 to 115 for controlling the running, posture, and drug application of the unmanned detergent 10 by the user.

For example, the operation unit 11 may include automatic mode selection switches 111 for selecting a mode in which a running operation or a medicine application operation is performed in an automatic mode or a manual mode, a mode selection switch 111 for selecting a forward running, A steering selection switch 113 for selecting a steering direction for left, neutral, or right turning; a lever 114 for controlling the speed to select acceleration, constant speed or deceleration; Or a spray control switch 115 for selecting spray pressure or the like.

The selection signals generated by the operation of the switches 111 to 115 of the operation unit 11 are transmitted to the control unit 13. [

The sensor unit 12 senses the moving object guiding line to recognize the running state of the unmanned controller 10, detects the speed of the unmanned controller 10, detects an obstacle around the unmanned controller 10, And sensors 121 to 125 for sensing the operating state of the spraying mechanism 17 and the boom 18.

The sensor unit 12 includes ultrasonic sensors 121 and optical sensors 122 installed on the right and left sides of the front of the body and on the left and right sides of the body for recognizing the surroundings and an inclination angle sensor 123 A speedometer 124 for determining the speed of the moving object or a boom angle sensor 125 for detecting the direction of the boom 18 equipped with the medicine spraying mechanism 17, (121 to 125) and transmits the sensor signal to the control unit (13).

For example, the ultrasonic sensor 121 may measure the distance of a moving object guide line or an adjacent object using a difference in arrival time (TDoA) until a plurality of ultrasonic signals are reflected by an object, have.

The light sensor 122 can, for example, recognize a three-dimensional arrangement of moving object lead or adjacent objects by emitting a structured light pattern and analyzing a distorted pattern of light reflected on the object.

The inclination angle sensor 123 can identify the three-dimensional slope of the moving object using, for example, an acceleration sensor or a gravity sensor.

The speed meter 124 can estimate the speed at which the center of gravity of the unmanned controller 10 moves based on the number of revolutions per hour of the wheels.

The boom angle sensor 125 can be implemented, for example, as a potentiometer. When the knob of the potentiometer is configured to rotate in correspondence with the extending direction of the boom 18 mounted with the medicine dispensing mechanism 17, The angle of the boom 18 can be determined in comparison with the reference voltage.

The control unit 13 detects the presence and direction of the moving object guide line based on the sensor signals, acquires the position information or the attitude information of the unmanned controller 10, and based on the obtained position information or attitude information and the selection signals, Drives the wheel drive unit 14 and the medicine sprinkler drive unit 15 of the vehicle 10.

Here, the positional information can be obtained, for example, by comparing the travel distance obtained from the speed meter 124 of the unmanned controller 10 moving in accordance with a given travel route with respect to the travel route.

The attitude information is based on the tilted angle of the moving object with respect to the ground surface and the direction of the moving object guiding line through the ultrasonic sensor 121, the optical sensor 122, the inclination angle sensor 123, or the boom angle sensor 125 The moving direction of one moving body, and the like.

The control unit 13 may include a sensor signal processing unit 131, an extended Kalman filtering unit 132, a position control unit 133, a wheel control unit 134, and a medicine dispensing control unit 135.

Specifically, the control unit 13 includes a sensor signal processing unit 131 for calculating the pre-correction position information, the moving object guide line direction, and the posture information of the non-corrected state from the received sensor signal, An extended Kalman filtering unit 132 for calculating the position information after correction by the Kalman filtering to calculate the position information after correction, a current position of the unmanned controller 10 on the basis of the corrected position information, A position controller 133 that determines the next target positions of the unmanned controller 10 according to a schedule, a controller 13 that decelerates, stops, accelerates, rotates left, or rotates the unmanned controller 10 so that the unmanned controller 10 moves from a current position to a next target position. A wheel control unit 134 for generating a wheel control signal for starting a steering operation, a post-stop and a post-stop, And a medicine dispensing control unit 135 for controlling the medicine dispensing device driving unit 16 for adjusting the position, direction, pressure, spray amount, and the like of the medicine dispensing mechanism 17.

The extended Kalman filtering unit 132 corrects the pre-correction position information having the observation error by the extended Kalman filtering based on the pre-correction position information, the moving body guide line direction and the posture information having an observation error, Location information can be generated.

The kinematic modeling and the extended Kalman filtering algorithm of the unattended sprayer 10 necessary for the extended Kalman filtering for the pre-correction position information including the observation error is described in more detail below.

Once the post-correction position information is obtained, the wheel controller 134 calculates the post-correction post-correction position information using the post-correction position information and the posture information, Can be moved. The wheel driving section 14 of the unmanned controller 10 can be controlled according to the wheel control signal and the unmanned controller 10 can move or change its direction as the wheel driving section 14 moves the wheels 16 .

According to the embodiment, in addition, the control unit 13 may further include a risk judging unit 136, respectively.

The danger judgment section 136 can judge, based on the sensor signals, whether there is a dangerous element such as a terrain, an obstacle or the like at the front.

For example, if a risk factor such as rollover or collision is determined in the risk judgment section 136, the wheel control section 134 of the control section 13 determines whether the risk element identified and the posture information calculated in the sensor signal processing section 131 A wheel control signal is generated so as to avoid danger such as rollover or collision.

Based on the wheel control signal of the wheel control unit 134, the number of revolutions of each of the wheels 16 is increased or decreased so that the vehicle can be steered quickly, rapidly or suddenly, steered to the left or right, The wheel driving portion 14 of the unmanned controller 10 can be driven so that proper avoidance maneuvering can be performed by performing the maneuvering such as turning.

On the other hand, the wheel drive section 14 can be implemented, for example, with motors directly connected to the wheel 16. [

Accordingly, the wheel control signal of the wheel controller 134 can be generated in the form of a plurality of pulse width modulation (PWM) signals, for example, so as to individually and precisely control the number of revolutions of the wheels.

FIG. 2 and FIG. 3 are conceptual diagrams illustrating kinematic models of a skid-steered remote control unmanned controller according to an embodiment of the present invention in terms of the relationship between free object motion, wheel and moving object velocity.

Referring to FIG. 2, a kinetic model of the unmanned detergent 10 directed in the x _c direction in the inertial coordinate system (X _g , Y _g ) is illustrated to represent the macroscopic position of the unmanned detergent 10.

In order to express the instantaneous movement of the unmanned detergent dispenser 10, the local coordinate system xc, yc, which is based on the traveling direction of the unmanned detergent dispenser 10 and its orthogonal direction, .

로 정의된다고 가정하면, X, Y는 무게 중심 COM의 관성 좌표계 좌표이고, θ는 관성 좌표계를 기준으로 국부 좌표계의 방위 차이를 의미한다. 또한 본 명세서에서 행렬 또는 벡터의 위 첨자에 쓰인 T는 행렬 또는 벡터의 전치(transpose)를 의미함을 유의한다.If the generalized positioning coordinates q of the unmanned controller 10 are

, X and Y are the coordinates of the inertial coordinate system of the center of gravity COM, and θ is the azimuth difference of the local coordinate system with respect to the inertial coordinate system. Also note that T used in superscripts of a matrix or vector herein refers to the transpose of a matrix or vector.

3, in the inertial coordinate system (X _g , Y _g ), the relationship between the center of gravity COM and the velocity vectors moving around the center of the respective centers around the instantaneous center of rotation (ICR) is illustrated.

로 정의되는데, 여기서 v_x(t), v_y(t)는 각각 국부 좌표계의 x_c 축과 y_c 축에 따른 무게 중심 COM의 속도 성분들을 의미한다. 또한 무인 방제기(10)의 진행 방위각의 각속도 w_c는

와 같이 정의된다.The linear velocity v (t) of the center of gravity COM

Where v _x (t) and v _y (t) are velocity components of the center of gravity COM along the x _c and y _c axes of the local coordinate system, respectively. The angular velocity w _c of the azimuth angle of the unmanned controller 10

Respectively.

2 and 3, the velocity of the unmanned controller 10 based on the inertial coordinate system can be expressed as Equation 1 based on the instantaneous velocity components of the unmanned controller 10.

과 같이 정의되고, 순간 회전 중심(ICR)에서 무인 방제기(10)의 무게 중심 COM 까지 벡터 d_C는

로 정의된다. 여기서 d_Cx, d_Cy는 국부 좌표계 좌표 값들이고, 도 3과 같이 x_ICR, y_ICR로 표현하면

와 같다.3, vectors d _i up to the center of each of the four wheels (P _i , i = 1, 2, 3, 4) in the instantaneous center of rotation (ICR)

And the vector d _C from the instantaneous rotation center (ICR) to the center of gravity COM of the unmanned controller 10 is defined as

. Here, d _Cx and d _Cy are local coordinate system coordinate values, and expressed as x _ICR , y _ICR as shown in FIG. 3

.

The relationship between these vectors d _i , d _C , distance and velocities, and angular velocity is given by Equation 2 below.

From Equation (2), the velocity constraint differential equation can be expressed as Equation (3).

로 간소화하여 다시 쓸 수 있다.Equation (3) is a non-holonomic operational constraint, which is expressed by the following equation (4) using Equation (1)

Can be simplified and rewritten.

는 일반화 좌표이고

는 일반화 속도(generalized velocity)이다.

Is the generalized coordinate

Is the generalized velocity.

는 항상 행렬

의 영 공간(null space) 내에 존재하므로, 수학식 4로부터 상태 공간 표현식을 설계하기 위해 의사 속도(pseudo-velocity) 행렬

와 회전 행렬

를 도입하면, 수학식 5와 같이 다시 정리할 수 있다.Generalization speed

Is always a matrix

(X, y) in the null space of the pseudo-velocity matrix

And a rotation matrix

The following equation (5) can be rearranged.

는 극좌표계를 직교좌표계로 변환하기 위한 회전 행렬,

은 의사 속도 행렬이다.

A rotation matrix for transforming a polar coordinate system into a rectangular coordinate system,

Is the pseudo velocity matrix.

The kinetic characteristics of the unmanned controller 10 thus modeled are linearized to be suitable for extended Kalman filtering.

Linearization can be implemented by removing nonlinear components from the relationship between the input and output of the system. Linearization makes it possible to design the control unit and contribute to the stability of the system.

를 가지고 다음과 같이 정의될 수 있다.The state space model related to the motion characteristics of the unmanned controller (10)

Can be defined as follows.

는 상태 공간에 대한 제어 입력 값이다.here

Is the control input value for the state space.

The motion characteristics of the unmanned controller 10 can be summarized as shown in Equation (7) by Equations (1), (5) and (6).

,

는 각각 직교 관성 좌표계에서 의사 속도 성분 값들이고,

는 순간 회전 중심과 무게 중심까지의 거리이며,

,

는 제어 입력 값들이다.

,

Are the pseudo velocity component values in the orthogonal inertial coordinate system, respectively,

Is the distance from the moment of rotation to the center of gravity,

,

Are control input values.

,

는 직교 관성 좌표계에서 무게 중심의 좌표이고,

,

는 위치 오차가 가지는 백색 가우시안 잡음이며,

는 이러한 성분들로써 선형화된 상태 공간 모델의 위치 변수이다.

,

Is the coordinates of the center of gravity in the orthogonal inertial coordinate system,

,

Is the white Gaussian noise of the position error,

Is the positional variable of the linearized state space model with these components.

The extended Kalman filtering algorithm obeys the nonlinear system represented by the state change function f (x) of the state variable x to remove the error v from the observed state y and obtain the true x And can be expressed as Equation (8).

In this case, w denotes a process error and v denotes a measurement error. h (x) is a function indicating the relationship between the state variable x and the observed state y.

를 상태 변수 x의 사후적 추정값(posteriori estimate value)라고 할 경우에, 확장형 칼만 필터를 구현하기 위한 상태 변화 행렬 A와 관계 행렬 H는 다음과 같이 자코비안 행렬로 얻을 수 있다.

Is a posteriori estimate value of the state variable x, the state transition matrix A and the relational matrix H for implementing the extended Kalman filter can be obtained by a Jacobian matrix as follows.

과 추정 오차 공분산(covariance)

은 다음 수학식 10과 같다.For a given state initial value x (0), the first posterior state estimate

And the estimated error covariance

Is expressed by the following equation (10).

Where E [x] is the expected value function of x.

는 다음 수학식 11과 같은 반복 연산에 의해 매번 입력되는 관측 값 y의 오차를 제거한 사후적 추정 상태

를 지속적으로 연산하는 방식으로 보정된다.Estimated state through extended Kalman filtering

Is a posterior estimation state in which an error of an observation value y input every time by an iterative calculation as shown in the following Equation (11)

Is continuously corrected.

는 상태 변수 x의 사후적 추정 값이고, y는 관측 값이며, Q는 프로세스 오차 w의 공분산이다.Where K is the Kalman gain, P is the covariance of the estimation error w, H is the relational matrix between the observed value y and the state variable x, R is the covariance matrix of the observed error v, A is the state transition matrix,

Is the posterior estimate of the state variable x, y is the observed value, and Q is the covariance of the process error w.

FIGS. 4 and 5 are graphs illustrating compensated position measurement results and absolute errors of the skid-steered remote control unmanned controller according to an embodiment of the present invention, respectively.

Referring to FIG. 4, as a result of simulating a case where the unmanned detergent dispenser 10 moves linearly along a guide line at a constant speed, it is assumed that the reference line indicated by a thin straight line, the uncompensated position information indicated by a thin solid line, And the post-correction position information (EKF) corrected by the extended Kalman filter indicated by a bold solid line.

It can be seen that the post-correction position information as measured by the sensor has relatively severe noise, but the post-correction position information corrected by the extended Kalman filter is corrected to be close to the reference line of the position change.

5, in the same simulation, the absolute values (Uncompensated) of the error between the reference position and the pre-correction position information indicated by a thin solid line appear relatively large, whereas the absolute values (Uncompensated) It can be seen that the absolute values (EKF) of the error between the position information after the correction and the reference line of the position change are greatly reduced.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed exemplary embodiments, but, on the contrary, Modification is possible. Accordingly, the spirit of the present invention should be understood only in accordance with the following claims, and all of the equivalent or equivalent variations will fall within the scope of the present invention.

Skid Steering Remote Control Unmanned Detectors (10)
11 Operation part 12 Sensor part
13 control unit 131 sensor signal processing unit
132 extended Kalman filtering unit 133 position control unit
134 Wheel control unit 135 Medicament dispensing control unit
136 Risk Assessment
14 wheel drive part 15 Drug spraying device drive part
16 wheels 17 Chemical Sprayer
18 Boom

Claims (6)

A skid steer-type remote control unmanned detergent including wheels mounted on both sides, a medicine spraying mechanism for spraying a medicine through a spray nozzle, and a boom for positioning the spray nozzle for spraying a desired target ,
An operation unit including switches for manually or automatically controlling the running, posture, and medicine spraying of the skid-steered remote control unmanned sprayer by a user, and generating selection signals according to the operation of the switches;
A sensor unit including sensors for detecting a running state, a speed, and an obstacle in the vicinity of the skid steered remote control unmanned controller, and generating sensor signals by the sensors;
Calculating an uncorrected pre-correction positional information from the sensor signals, calculating an after-correction positional information by correcting an observation error included in the pre-correction positional information by an extended Kalman filter, A control unit for determining a next target position according to the determined current position, the detected moving object lead, and a predetermined progress schedule, and generating a wheel control signal for driving the wheels to move to the next target position; And
And a wheel driving unit for driving the wheels based on the wheel control signal. The apparatus according to claim 1,
An automatic mode selection switch for selecting a driving mode or a medicine dispensing mode to be performed in an automatic mode or a manual mode, a forward / backward selection switch for selecting a forward running, a neutral or reverse running, a leftward, And a control switch for selecting a height, a direction, or a spraying pressure of a lever for controlling the speed for selecting acceleration, constant speed or deceleration, or a medicine spraying mechanism. Skid Steering Remote Control Unmanned Sprayer. [2] The apparatus according to claim 1,
An ultrasonic sensor or optical sensors provided at the left and right sides of the front and rear sides of the front and rear to recognize the peripheral state, a tilt sensor for detecting the posture, a speedometer for detecting the speed, or an extension direction of the boom equipped with the medicine sprayer And a boom angle sensor for detecting the boom angle of the skid steerer. The apparatus of claim 1,
A sensor signal processor for calculating pre-correction positional information, moving object inducing line direction, and posture information having an observation error from the sensor signals to an uncorrected state;
An extended Kalman filtering unit for calculating post-correction position information by correcting the pre-correction position information having the observation error by the extended Kalman filtering based on the pre-correction position information having the observation error, the moving object guide line direction and posture information;
A position controller for determining a current position of the skid steerable remote control unmanned controller based on the post-correction position information, and determining a next target position according to the current position, the moving object guide line direction, and a given progress schedule;
A wheel control section for generating a wheel control signal for controlling the number of revolutions of each of the wheels so that the skid-steered remote control unmanned controller moves from a current position to a next target position; And
And a medicine dispensing control unit for controlling a medicine dispensing apparatus driving unit for adjusting the posture and operation of the boom and the medicine dispensing apparatus. 5. The apparatus of claim 4,
Further comprising a risk judging unit for judging whether or not a risk element such as a topography or an obstacle is present in front of the sensor based on the sensor signals,
Wherein if the risk judging unit judges a risk factor such as rollover or collision, the wheel control unit operates to generate the wheel control signal for avoiding starting based on the risk factor and the attitude information. Method remote control unmanned detergent. The apparatus of claim 1, wherein the expanding Kalman filtering unit comprises:

The extended Kalman filtering is performed according to the following equation:
Where K is the Kalman gain, P is the covariance of the estimation error w, H is the relational matrix between the observed value y and the state variable x, R is the covariance matrix of the observed error of the observed value y, A is the state transition matrix ,

Is a posterior estimation value of the state variable x, y is an observation value, and Q is a covariance of the process error w.

Images