KR20100076348A - Biped robot and method for controlling of walking biped robot - Google Patents

Abstract

PURPOSE: A biped robot and a walking control method for the same are provided to enable the robot to walk on an uneven surface without using information about the surface. CONSTITUTION: A biped robot comprises a hip joint(15), two left and right legs(LL,RL), and a controller. The two left and right legs are connected to the hip joint. The controller controls the ends of the left and right legs to follow an elliptical orbit in a coordinate system including the hip joint. The center of the elliptical orbit is determined based on the difference between the target speed and the measured speed of the hip joint.

Description

Biped robot and method for controlling of walking biped robot}

The present invention relates to a group 2 robot and a method for controlling walking of the group 2 robot.

Robot refers to a mechanical device that can perform a movement similar to the movement of life by using electrical or magnetic action. Such robots are widely used in the industry for automation and unmanned production, and research and development are underway in various fields. One of them is to control the walking of robots.

The robot moves on the ground using wheels largely, the leg (leg) moves like a bipedal animal (such as a human or a monkey), the body itself is a moving part ( For example, the way snakes move. Among the above methods, the walking method using the legs has a wide range of applicability because of its excellent adaptability to move on various types of ground.

However, in walking of a biped robot having two legs, such as a humanoid robot (Humanoid), there is a difficulty in maintaining a stable posture so that the robot does not fall when moving, and some methods for overcoming this are known. For example, a method using a zero moment point (hereinafter referred to as ZMP), a method by learning, and the like.

In the method using ZMP, ZMP means that the total inertia force of the entire group 2 robot becomes zero, and is a method of maintaining stability of the group 2 robot mainly using dynamics. The ZMP method is a very useful method to satisfy the movement and stability of the biped robots, but the accurate dynamic modeling of the biped robots is important, and it is difficult to implement walking on an irregular ground or fast walking or running. many.

In the case of the learning method, a fully trained biped robot can walk well on a flat or irregular ground and can walk fast. However, basically, because the biped robots have multiple degrees of freedom, it takes a lot of time to learn the biped robots, and there is a difficulty in learning about each case for walking or fast walking on various grounds.

An object of the present invention is to provide a biped robot with improved walking ability. In addition, the present invention provides a control method for improving the walking ability of a biped robot.

The biped robot according to the present embodiment for solving the above problems is the hip joint; Two left and right legs connected to the hip joint; And controlling the two left and right leg ends to follow an elliptic trajectory on a coordinate system based on the hip joint, wherein a center point of the elliptic trajectory is a value that is updated by using a difference between a measurement speed of the hip joint and a target speed of the hip joint. It includes a control unit.

According to another exemplary embodiment, a gait control method of a biped robot includes determining coordinates of two left and right leg ends; And controlling the two left and right leg ends according to the coordinates, wherein the coordinates are values in a coordinate system based on the center of gravity of the group 2 robot, and the two left and right leg ends are elliptical in the coordinate system. Control to follow the track, wherein the center point of the elliptic track is determined using the difference between the measuring speed of the center of gravity and the target speed of the center of gravity.

According to the walking control method of the group 2 robot and the group 2 robot according to the present embodiment, the group 2 robot can walk irregularly without using the ground information measuring the situation of the ground.

In addition, the walking control method of the group 2 robot according to the present embodiment can be applied even if a large number of joints are present on the legs of the group 2 robot.

Hereinafter, a walking control method for a group 2 robot and a group 2 robot according to the present embodiment will be described in detail with reference to the accompanying drawings.

1 is a diagram briefly modeling a group 2 robot according to the present embodiment. As shown in FIG. 1, the biped robot has a shape having only a leg part except a part corresponding to a human torso and a foot. In the biped robot, the center of gravity is at the hip joint connecting the leg portions. The center of gravity may be present in the other part of the group 2 robot, but in the present embodiment, to show the functionality of the method for controlling the walking of the group 2 robot, as described above, the leg portion without the body and feet We want to use a robot with a bay and a center of gravity at the hip joint.

Referring back to FIG. 1, the biped robot includes four links (11, 12, 13, 14), three joints (15, 16, 17), three motors (not shown), and a plurality of encoders ( It is composed of a not shown), a sensor (not shown) and a control unit (not shown).

The four links 11, 12, 13 and 14 are left thigh links 11, right thigh links 13, left tibia links 12 and right shin links. (Right Tibia Link, 14), the three joints (15, 16, 17) is composed of the left knee joint 16, the right knee joint 17 and the hip joint (coxa, 15).

One side of the left thigh link 11 and one side of the right thigh link 13 are connected to the hip joint 15 in a state spaced apart from each other on a central axis of the hip 15. The left knee joint 16 is connected to the other side of the left thigh link 11, and the right knee joint 17 is connected to the other side of the right thigh link 13. One end of the left shin link 12 is connected to the left knee joint 16, and the other end of the left shin link 12 provides an end of the left leg LL. One end of the right shin link 14 is connected to the right knee joint 17, and the other end of the right shin link 14 provides an end of the right leg RL. The left / right leg end means a part in contact with the ground when the biped robot walks.

Here, the left leg LL means a portion including the left thigh link 11, the left knee joint 16, the left shin link 12, the right leg RL is the right thigh link 13, right It means the part including the knee joint 17, the right shin link (14).

The left knee joint 16 is provided with a left knee joint motor LM, the right knee joint 17 has a right knee joint motor RM, and the hip joint 15 is provided with a hip joint motor CM. The left knee joint motor LM may rotate the left shin link 12 at a predetermined angle θ2 with the left thigh link 11 as a reference axis according to a control signal input from the controller. The right knee joint motor RM may rotate the right shin link 14 at a predetermined angle θ3 with the right thigh link 13 as a reference axis according to a control signal input from the controller. In addition, the hip joint motor CM may rotate the left thigh link 11 and the right thigh link 13 to be spaced apart by a predetermined angle θ 1 according to a control signal input from the controller.

An encoder is attached to the motors CM, LM, and RM, respectively, to measure the rotation angles θ1, θ2, θ3 and rotational speeds of the joints 15, 16, and 17, and provide them to the controller. Can be.

The sensor is attached to the group 2 robot to measure the moving direction speed of the group 2 robot to provide to the controller. The sensor may be attached to, for example, the hip joint 15 of the biped robot, and measures a moving speed in a direction horizontal to the ground of the hip joint 15 (hereinafter referred to as a measurement speed of the hip joint), and controls the controller. Can be provided as

) 및 위치(

)를 계산하고, 각 모터(CM, LM, RM)의 제어신호(

)를 생성하여 각각의 모터(CM, LM, RM)로 제공하는 부분이다. 또한, 상기 센서로부터 상기 고관절(즉, 무게 중심)의 측정속도를 제공받고, 후술할 고관절의 목적속도와 상기 고관절의 측정속도 차이를 이용하여 타원궤도의 중심점을 계산 및 갱신한다. 상기 타원궤도에 대해서는 후술한다. 이러한 제어부는 PD(Proportional-plus-Derivative)경로추정 제어기를 포함할 수 있다.The controller receives rotation angles (θ1, θ2, θ3) and rotational speeds of the joints 15, 16, and 17 from each encoder, and determines the speeds of the left and right leg ends (

) And location (

) And control signals (CM, LM, RM) of each motor

) Is generated and provided to each motor (CM, LM, RM). In addition, the sensor receives the measurement speed of the hip joint (ie, the center of gravity) from the sensor, and calculates and updates the center point of the elliptical track using the difference between the target speed of the hip joint and the measurement speed of the hip joint, which will be described later. The elliptical trajectory will be described later. Such a controller may include a proportional-plus-derivative path estimation controller.

Hereinafter, a method of controlling the walking of the biped robot by the controller will be described. First, when the principle of the walking control method of the biped robot according to the present embodiment will be described schematically, the controller controls the left and right leg ends of the biped robot to follow the elliptical trajectory in the coordinate system based on the center of gravity. At this time, the center point of the elliptical orbit is updated every moment according to the measurement speed of the center of gravity. In the present embodiment, since the center of gravity is located at the hip joint, the speed of the hip joint is measured.

First, it will be described that the biped robot can walk if the left and right leg ends of the biped robot are controlled to move along the elliptical trajectory in the coordinate system based on the hip joint 15. To this end, the gait pattern of the biped robot is analyzed below.

FIG. 2 is a diagram showing the walking of the biped robot seen in the global coordinates (denoted by {0}) with respect to the ground. Here, the global reference coordinate system {0} refers to a coordinate system having an origin at a predetermined place on the ground, and comprising an axis parallel to the ground (x axis) and an axis perpendicular to the ground (y axis). Referring to FIG. 2, the end of the left leg LL of the group 2 robot moves from B ¹ to B ² while the end of the right leg RL of the group 2 robot is stopped in contact with the point A of the ground. . In this process, the hip joint 15 moves in a curve from C ¹ to C ² .

As described above, when the walking pattern of the biped robot is examined in the global reference coordinate system {0}, both the ends and the hips of the other leg are walking while performing the curved motion while the ends of one leg are in contact with and fixed to the ground.

By the way, the gait of the biped robot described above is not the global reference coordinate system {0} but the center of gravity, that is, the local coordinate system (denoted by {B}) based on the hip joint, and the hip joint 15 is in a fixed position. The left and right ends of the leg may move in an elliptical orbit in some cases. Here, the local coordinate system {B} refers to a coordinate system composed of an axis parallel to the ground and an axis perpendicular to the ground, with the hip joint 15 of the biped robot as the origin.

For the sake of understanding, consider the case where the biped robot walks on a treadmill. FIG. 3 is a diagram showing the movement of the group 2 robot based on the global reference coordinate system {0} when the group 2 robot walks on the treadmill 31. Referring to FIG. 3, while the left leg LL end is curved from point B to point A, the right leg RL end is linearly moving from point A to point B, in this case the hip joint 15. Is exercising only up and down. That is, in this case, the movement path of the end of the left leg LL and the movement path of the end of the right leg RL are continuously connected, and the displacement amount in the vertical direction of the left / right leg end with respect to the hip joint 15 is symmetrically. It happens. In this case, looking at the biped robot in the local coordinate system {B} based on the hip joint 15, the left and right leg ends are moving in an elliptical orbit.

점을 중심점으로 타원궤도를 그리게 된다.4 is a diagram showing the gait of the biped robot according to the present embodiment on a coordinate system based on the hip joint. As shown in FIG.

An elliptical orbit is drawn from the point as the center point.

As described above, when the control unit controls the left and right leg ends of the biped robot to follow the elliptical trajectory on the local coordinate system {B} based on the hip joint 15, it can be seen that the biped robot can walk. . That is, while the left and right leg ends exercise the relative path with respect to the hip joint 15 while drawing an elliptical trajectory, one of the left and right leg ends contacts the ground, and at this time, the ground and the leg ends When the leg end is fixed to the ground by a restraint force (friction force), the biped robot proceeds forward by the driving force to follow the relative path, so that the walking is possible.

As described above, the biped robot tracks the velocity of the biped robot on the {0} coordinate system when the left and right leg ends follow an elliptical trajectory in the local coordinate system {B} and walk as a result. 5 (a) is a view showing a case in which the left and right leg ends of the biped robot moves an ellipse orbit having a length of 2a in the {B} coordinate system, and FIG. 5 (b) shows the same walking process in the {0} coordinate system. Drawing.

As shown in FIG. 5 (a), when the left and right leg ends of the biped robot move on the {B} coordinate system with an elliptical orbit having a long axis length of 2a as the period T (T is a constant), FIG. 5 (b) As shown, the group 2 robot moves in the x-axis direction at a speed of 4a / T. Therefore, the traveling speed with respect to the ground of the group 2 robot moving the relative path as shown in Figure 5 (a) can be seen as 4a / T. If T is given as a function of time T (t), the velocity of the biped robot will be given as a function of time.

However, when the biped robot actually walks, the biped robot is not further controlled by the biped robot due to uneven ground or any external force acting on the biped robot. It is generally not possible to maintain this 4a / T traveling speed.

The failure of the biped robot to maintain the speed of 4a / T means that the left and right leg ends of the biped robot cannot maintain the elliptical trajectory on the local coordinate system based on the center of gravity. In other words, it can be seen that the biped robot tries to fall without maintaining its posture. Therefore, the method of maintaining the speed which the said group 2 robot should follow becomes the method of making the said group 2 robot keep walking without falling.

The following describes how the biped robot can be modeled when walking. First, a control method for stably walking without maintaining the speed of the biped robot according to the modeling will be described conceptually, and a detailed walk control method will be described later.

FIG. 6 is a model for the case where the biped robot walks according to the present embodiment. FIG. Referring to FIG. 6, the walking biped robot may be modeled as an inverted pendulum having a rotating joint. The biped robot is modeled as an inverted pendulum 62 connected to a rotary joint 63 on a cart 61, where the cart 61 generates a driving force generated by the biped robot by contact with the ground. The inverted pendulum 62 represents the entire group 2 robot.

In FIG. 6, Vx is a traveling speed in the x-axis direction measured by a sensor attached to the center of gravity of the group 2 robot, which may be a measurement speed of the hip joint, and measures a measured value of the actual speed of the group 2 robot. it means. Hereinafter, Vx will be referred to as the measurement speed of the center of gravity.

Vxd means the traveling speed of the group 2 robot calculated when the left and right leg ends of the group 2 robot move the elliptical orbit described above with reference to FIG. 5. The traveling speed can also be calculated for the hip joint 15, which is the center of gravity, hereinafter, Vxd will be referred to as the target speed of the center of gravity.

Then, the attitude of the biped robot can be estimated by comparing Vx and Vxd. 7 and 8 conceptually show the attitude and control method according to Vx and Vxd of the group 2 robot according to the present embodiment.

Fig. 7 (a) shows the case where the measured speed Vx of the hip, which represents the actual speed of the group 2 robot, is greater than the target speed Vxd of the hip joint obtained by calculation. In this case, as shown in FIG. 7 (a), it can be seen that the group 2 robot is inclined forward, that is, the group 2 robot is falling forward.

In this case, in order to continue walking without the biped robot falling down, it is necessary to control the end of the bridge to go on the ground more than the existing path in order to continue walking. This means that the control unit of the biped robot should control the left and right leg ends to draw an elliptical trajectory with respect to the center point updated in the x-axis direction.

When the control unit is controlled as described above, as shown in FIG. 7 (b), rotational momentum occurs in the direction of pushing the biped robot backward, and the inverted pendulum 62 is laid down on the modeling. . The rotation momentum decreases the Vx and the Vxd increases. Therefore, the biped robot can maintain a stable posture without falling. For example, when the group 2 robot descends the inclined plane, as shown in FIG. 7 (a), the group 2 robot will be larger than Vxd. Therefore, in this case, the control method should be performed in such a manner as to push the biped robot back as shown in FIG.

On the contrary, when the measurement speed Vx of the hip joint in the group 2 robot is smaller than the target speed Vxd of the hip joint, it can be seen that the group 2 robot falls down as shown in FIG. In this case, as shown in FIG. 8 (b), the end of the left and right leg ends of the biped robot should be moved backward so that the Vxd is reduced and the rotational momentum is increased to increase the Vx. do. This means that the control unit of the biped robot should control the left and right leg ends to draw an elliptical trajectory with respect to the center point updated in the direction opposite to the x-axis. This will keep the biped robots falling backwards forward to maintain a stable posture. For example, when the group 2 robot climbs the inclined surface, as shown in FIG. 8 (a), the group 2 robot will be smaller than Vxd. Therefore, in this case, the control method should be performed by pushing the biped robot forward as shown in FIG.

6 to 8, the controller of the biped robot controls two left and right leg ends connected to the hip joint 15 to follow the elliptical trajectory on the {B} coordinate system. The center point is updated every second according to the difference between the measurement speed Vx of the hip joint and the target speed Vxd of the hip joint, and controls to follow the new elliptical trajectory again according to the updated center point.

When the center point of the elliptical orbit is represented by a formula, it is expressed by Equation 1.

는 타원 궤도 중심점의 갱신전 값을 의미하고,

는 상기 타원 궤도 중심점의 갱신후 값을 의미한다.

는 상기 갱신전 시점에 대해 측정되는 고관절의 측정속도이고,

는 상기 갱신전 시점에 대한 고관절의 목적속도이다. 그리고, k 및 m은 미리 정해진 소정의 값이다. 예컨대, 상기 k는 0.2, m은 -0.475(meter)로 주어질 수 있다.In Equation 1

Denotes the value before the update of the elliptical orbit center,

Denotes a value after updating of the elliptic orbit center point.

Is the measuring speed of the hip joint measured with respect to the time before the update,

Is the target speed of the hip joint with respect to the time before the update. And k and m are predetermined values. For example, k may be given as 0.2 and m may be given as −0.475 (meter).

)을 구할 수 있다. 이러한 과정을 먼저 개략적으로 설명하면 관절변수를 이용하여 구동관절에서의 토크값(

)를 구하고, 상기 구동관절에서의 토크값(

)을 θ에 대한 토그값(

)으로 변환하는 과정을 거친다. Hereinafter, the above-described walking control method of the group 2 robot will be described in more detail. To do this, we first derive a dynamics model on the {B} coordinate system based on the hip joint (15). The torque value to be provided to each motor (CM, LM, RM) by the controller through the dynamic model (

) Can be obtained. First of all, this process will be described in brief.

) And the torque value (

) Is the toggle value for θ (

Is converted to).

FIG. 9 is a diagram illustrating a coordinate system and variables for dynamic modeling of a group 2 robot applying the walking control method according to the present embodiment. Referring to FIG. 9, global coordinates based on the ground are represented by {0}, and local coordinate systems based on the hip joint 15, which is the center of gravity of the biped robot, are represented by {B}. have. The local coordinate system {B} may be represented by a 2 x 1 position vector q _b with respect to the global reference coordinate system {0}.

The kinematic equation for the biped robot can be expressed by Equation 2 using joint variables qi (1 ≦ i ≦ 4, i is a natural number). Here, the joint variable q1 is an angle formed by the left thigh link 11 in the counterclockwise direction with respect to the x axis of the {B} coordinate system, and q2 is the counterclockwise direction in the longitudinal axis of the left thigh link 11. The left shin link 12 is at an angle. The joint variable q3 is an angle formed by the right thigh link 13 in the counterclockwise direction with respect to the x axis of the {B} coordinate system, and q4 is the right shank link 14 with the reference axis in the longitudinal direction of the right thigh link 13. The angle is counterclockwise.

In Equation 2, the torque value τ for the joint variable is represented by the joint variable qi as a variable, Mt is the inertia matrix, q = [q1, q2, q3, q4] ^T , and bt is the Coriolis force and This term takes into account the effects of centrifugal force and gravity.

The inertia matrix Mt and the bt can be expressed as Equation 3.

In Equation 3, the subscripts of M _bb , M _bq , M _qb , and M _qq are expressed according to the elements of the inertia matrix and the calculated vector. This notation is a commonly used notation for robots whose base is not fixed to the ground, such as mobile robots, underwater robots, and space robots.

Then, τ can be expressed as Equation 4.

In Equation 4, M = M _qq -M _qb M _bb ^-1 M _bq , b = b _q -M _qb M _bb ^-1 b _b .

However, since the biped robot actually has three joints, a process of converting a model for four joint variables q into a model for three θ is required.

Equation 5 is an expression representing a conversion relationship between θ and q .

Since a relationship such as Equation 5 is established between θ and q , Khatib proposed 'Dynamically consistent generalized inverse' (O.Khatib, "A unified Approach for Motion and Force Control of Robot Manipulators: The Operational Space Formulation," IEEE J.). (Robot.Autom., Vol. RA-3, no. 1, pp. 43-53, 1987), the dynamics model for θ is obtained by equation (6).

는 2족 로봇의 제어부가 각 관절(15, 16, 17)에 부착된 모터(LM, RM, CM)에 제공하는 토크값을 θ를 변수로 하여 나타낸 것을 의미한다. In Equation 6

Denotes a torque value provided by the controller of the group 2 robot to the motors LM, RM, and CM attached to the joints 15, 16, and 17 as a variable θ .

In the walking control method of the group 2 robot according to the present embodiment, since the position of the left and right leg ends with respect to the local coordinate system {B} based on the center of gravity of the group 2 robot is important, Equation 6 The control method using the dynamic model at the left and right leg ends is applied using the result of.

를 만족하는 자코비안(Jacobian) J에 의하여

인 동역학 관계식이 존재한다. 따라서, 상기 J의 역행렬이 존재한다면, 상기 f는 수학식 7과 같이 구해진다. The position vector of the left and right leg ends with respect to the {B} coordinate system is represented by u = [u _l ^T , u _r ^T ]. Here, u ₁ and u _r are position vectors of the left and right leg ends of the group 2 robot displayed with respect to the {B} coordinate system, respectively, as shown in FIG. 9. Since the position vector u is a function of the joint variable qi,

By Jacobian J

There is an kinetic equation. Therefore, if the inverse of J exists, f is obtained as shown in Equation 7.

이다.here,

to be.

The torque values input to the motors CM, LM, and RM of the group 2 robot by the controller by Equations 6 and 7 are determined as shown in Equation 8.

The dynamics model derived as in Equation 8 is referred to as a dynamics model in the working space below. The following describes an embodiment of controlling the walking of the biped robot using the dynamic model on the workspace.

First, the coordinates of the left and right leg ends of the biped robot are determined. 10 is an illustration of an elliptical trajectory followed by left and right leg ends of a group 2 robot according to the present embodiment. Referring to FIG. 10, the coordinate may be given by Equation 9 with respect to the {B} coordinate system.

는 4차 다항식으로 주어질 수 있다. 상기

는 반시계 방향으로 회전하며,

를 조절하여 상기 고관절의 목적속도 Vxd를 조절할 수 있다. In Equation 9, the biped robot starts walking in a stationary state and reaches a desired final walking speed (for example, 0.7 m / s).

Can be given as a fourth order polynomial. remind

Rotates counterclockwise,

By adjusting the target speed Vxd of the hip joint can be adjusted.

는 t가 0에서 T로 변하는 동안 0에서 -2π만큼 변해야 한다. 따라서,

의 평 균변화율

는

이다. 이 경우, 2족 로봇의 평균진행속도

는

이므로(도 5 (a), (b) 참조),

로 나타낼 수 있다. 따라서, 2족 로봇의 고관절에 대한 목적속도 Vxd를 정하면, 상기

를 적분하여

도 구할 수 있다. If the u _l , u _r rotates the ellipse orbit one time during the time T,

Must vary from 0 to -2π while t changes from 0 to T. therefore,

Average change rate of

Is

to be. In this case, the average running speed of the biped robot

Is

(See Figures 5 (a) and (b)),

It can be represented as. Therefore, if the target speed Vxd for the hip joint of the biped robot is determined,

By integrating

Also available.

에 따라 상기 T를 변화시킬 수 있고, 그 결과 2족 로봇의 고관절에 대한 목적속도 Vxd도 조절할 수 있다. In other words, the above

According to the present invention, the T may be changed, and as a result, the target speed Vxd of the hip joint of the biped robot may also be adjusted.

를 가지고, 수학식 1에 의해 갱신된 후의 값

로 계속하여 갱신되는 것이다.In addition, x _c in Equation (9) is updated every moment by Equation (1) so that the biped robot does not fall while walking and maintains a stable posture. That is, the x _c value is a value before the update,

With the value after being updated by Equation 1

Will continue to be updated.

로 하여 주어지는 타원궤도를 추종하기 위해 제어부는 상술한 작업공간에서의 동역학 모델을 이용한다. And then update the center point

In order to follow the elliptic orbit given by the controller, the controller uses the dynamic model in the above-described workspace.

In this case, f of Equation 8 can be obtained by Equation 10 using a PD (Proportional-plus-Derivative) path estimation controller.

이다. here,

to be.

,

,

,

는 각각 도 10에 나타난 좌우측 다리 끝단의 가속도, 속도, 위치를 의미하고,

과 u는 실제 측정된 관절각(θ1,θ2,θ3)으로부터 구한 좌우측 다리 끝단의 속도와 위치이다. In Equation 10, kp = 4000, kv =

,

,

,

Denotes acceleration, velocity, and position of the left and right leg ends shown in FIG. 10, respectively.

And u are the speed and position of the left and right leg ends obtained from the actual measured joint angles (θ1, θ2, θ3).

11 is a control block diagram for implementing a walking control method of the group 2 robot according to the present embodiment. In FIG. 11, Kin denotes forward kinematics for obtaining the velocity and position of the left and right leg ends from the measured joint angles θ1, θ2, and θ3.

Hereinafter, the experimental results of applying the control method according to the present embodiment of the biped robot will be described.

12 are photographs taken at intervals of 0.2 seconds by a biped robot according to the present embodiment for walking on flat ground according to a walking control method. FIG. 12 shows that the biped robot can walk only by drawing an elliptical orbit at the ends of the left and right legs with respect to the center of gravity on a coordinate system having the center of gravity as the origin.

13 are photographs taken at intervals of 0.2 seconds by a biped robot according to the present embodiment for walking irregular terrain according to the control method. 13 is a stable gait even on irregular terrain by controlling the left and right leg ends to follow the elliptical trajectory with respect to the center point is updated according to the difference between the actual speed of the center of gravity and the target speed when the biped robot walking. Is showing. In FIG. 13, two wooden boards having a height of 0.03m, a length of 0.3m, and a width of 0.2m are fixed to the ground to form irregular terrain. As can be seen in FIG. 13, the biped robot is walking without receiving any information on the position and size of the wooden board and measuring it. The biped robot is only controlled so that the left and right leg ends follow the elliptical trajectory set in the {B} coordinate system, and corrects only the center point of the elliptical trajectory using Equation 1 according to the measurement speed of the hip joint measured by the attached sensor. Do the work.

That is, the biped robot successfully walks on irregular terrain having a height difference of 0.03 m without using information about the ground. As shown in photographs 9 and 10 of FIG. 13, the leg ends of the biped robots are lowered above the wooden boards during the walking process. This indicates that the biped robot can walk even if the leg end of the biped robot in contact with the ground is not necessarily fixed to the ground in contact with the ground. This represents the superiority of the present invention in contrast to conventional techniques, ie that one leg tip of the robot in contact with the ground must be fixed while the other leg tip moves away from the ground.

Although the preferred embodiments of the present invention have been described in detail above, those of ordinary skill in the art to which the present invention pertains can make various modifications of the present invention without departing from the scope of the present invention as defined in the appended claims. It will be appreciated that the modification can be carried out as follows.

1 is a diagram briefly modeling a group 2 robot according to the present embodiment.

FIG. 2 is a diagram showing the walking of the biped robot seen in the global coordinates (denoted by {0}) with respect to the ground.

FIG. 3 is a diagram showing the movement of the group 2 robot based on the global reference coordinate system {0} when the group 2 robot walks on the treadmill 31.

4 is a diagram showing the walking of the biped robot according to the present embodiment on a coordinate system based on the hip joint.

5 (a) is a view showing a case in which the left and right leg ends of the biped robot moves an ellipse orbit having a length of 2a in the {B} coordinate system, and FIG. 5 (b) shows the same walking process in the {0} coordinate system. Drawing.

FIG. 6 is a model for the case where the biped robot walks according to the present embodiment. FIG.

7 is a view conceptually illustrating a posture and a control method according to Vx and Vxd of the group 2 robot according to the present embodiment.

8 is a view conceptually illustrating a posture and a control method according to Vx and Vxd of the group 2 robot according to the present embodiment.

 FIG. 9 is a diagram illustrating a coordinate system and variables for dynamic modeling of a group 2 robot applying the walking control method according to the present embodiment.

10 is an illustration of an elliptical trajectory followed by left and right leg ends of a group 2 robot according to the present embodiment.

11 is a control block diagram for implementing a walking control method of the group 2 robot according to the present embodiment.

12 are photographs taken in a step of 0.2 seconds in which the biped robot walks on a flat surface according to the control method.

FIG. 13 is a photograph of a biped robot walking in irregular terrain according to the control method in units of 0.2 seconds.

<Description of Signs of Major Parts of Drawings>

11: left thigh link 12: left shin link

13: right thigh link 14: right shin link

15: hip joint 16: left knee joint

17: right knee joint

Claims (13)

Hip joint; Two left and right legs connected to the hip joint; And The two left and right leg ends are controlled to follow an elliptic orbit on a coordinate system based on the hip joint, wherein a center point of the elliptic orbit is a value determined by using a difference between a measurement speed of the hip joint and a target speed of the hip joint. Two-group robot, comprising a. The method of claim 1, The two left and right legs each have a thigh link connected to the hip joint; A knee joint connected to the thigh link; One side is connected to the knee joint and the other side of the shin link providing the leg end; And a knee joint motor capable of rotating the shin link at a predetermined angle according to a control signal provided by the controller. And the hip joint includes a hip motor capable of rotating each of the two thigh links at a predetermined angle according to a control signal provided by the controller. The method of claim 2, The biped robot further comprises a sensor unit for measuring the measurement speed of the hip joint. The method of claim 3, wherein The center point of the elliptic orbit when the direction horizontal to the ground on the coordinate system is the x axis, and the upper direction perpendicular to the x axis is the y axis.

A biped robot characterized by the fact that it is determined as in Equation (1). Formula (1)

here,

Denotes a value before updating of the elliptic orbit center point,

Denotes a value after updating of the elliptic orbit center point.

Is a measurement speed of the hip joint at the time before the update measured by the sensor unit,

Is the target velocity of the hip joint before the update. And k and m are predetermined values. The elliptic trajectory of claim 4, wherein the elliptic trajectory is given by Equation (2), and the control unit provides a control signal to the hip joint motor and the two knee joint motors to control the left and right leg ends to follow the elliptic trajectory. Two-group robot, characterized in that. Formula (2)

here,

Is the position of the left leg end,

Is the position of the tip of the right leg. a is 1/2 of the long axis length of the elliptic track, b is 1/2 of the short axis length of the elliptic track, and a and b are predetermined values.

Is a value previously given by a fourth order polynomial representing the angle between the straight line connecting the left and right leg ends and the x-axis. And,

Is a value which is updated every second by the said Formula (1). The method of claim 5, The target speed

Above

Group 2 robot, characterized in that determined by. The method of claim 6, The control signal provided by the control unit may include the speed and position of the left and right leg ends derived from Equation (2). Provides a torque value determined using the difference in the speed and position of the left and right leg ends measured from the rotation angles of the hip motor and the left and right knee joint motors so that the two left and right leg ends follow the elliptic track. Two-group robot, characterized in that the control. 8. The bipedal robot according to claim 1, wherein the center of gravity is at the hip joint. In the walking control method of a biped robot including two legs, left and right, Determining coordinates of the ends of the two left and right legs in a coordinate system using the center of gravity of the group 2 robot as an origin; And And controlling the coordinates of the two left and right leg ends to follow an elliptical trajectory in the coordinate system. In the walking control method of a biped robot including two legs, left and right, Determining coordinates of the two left and right leg ends; And And controlling the two left and right leg ends according to the coordinates, wherein the coordinates are values in a coordinate system based on the center of gravity of the biped robot, wherein the two left and right leg ends are elliptical orbits in the coordinate system. While controlling to follow, the center point of the elliptic orbit is determined using the difference between the measuring speed of the center of gravity and the target speed of the center of gravity, the walk control method of a biped robot. The method of claim 10, A center point of the elliptic orbit when the direction horizontal to the ground is the x axis and the upper direction perpendicular to the x axis is the y axis on the coordinate system based on the center of gravity.

The gait control method of a biped robot, characterized in that is determined as in equation (3). Formula (3)

here,

Denotes a value before updating of the elliptic orbit center point,

Denotes a value after updating of the elliptic orbit center point.

Is the measurement speed in the x-axis direction of the center of gravity measured for the time point before the update,

Is the target velocity in the x-axis direction of the center of gravity with respect to the point before the update. And k and m are predetermined values. 12. The walking control method for a biped robot according to claim 11, wherein the left and right leg ends of the biped robot are controlled to follow an elliptic orbit as shown in Equation (4). Formula (4)

here,

Is the position of the left leg end,

Is the position of the tip of the right leg. a is 1/2 of the long axis length of the elliptic track, b is 1/2 of the short axis length of the elliptic track, and a and b are predetermined values.

Is a value previously given by a fourth order polynomial representing the angle between the straight line connecting the left and right leg ends and the x-axis. And,

Is a value which is updated every second by the said Formula (1). 13. The method of claim 12, The target speed

Above

Walking control method of a biped robot, characterized in that determined by.

Images