KR20030067023A - Walking control method for robot - Google Patents

Abstract

PURPOSE: A method for controlling a walking operation of a robot is provided to perform stably the walking operation on a slope way by using a slope correction. CONSTITUTION: A stable walking process performs to determine previously a stable walking track. A slope way correction process is performed to climb a slope of predetermined degrees when a robot walks on the determined stable walking track and faces a slope way. The robot walks on a corrected track which is corrected by the slope way correction process. The stable walking process includes the first setup process for setting a relative angle between a body of the robot and the ground and the second setup process for setting the relationship between the center of gravity and a support polygon.

Description

Walking Control Method of Robots {WALKING CONTROL METHOD FOR ROBOT}

The present invention relates to a walking control method for a robot, and more particularly, to a walking control method for a robot to smoothly perform an inclined plane walking.

The walking control method of general, group 2 and group 4 robots is largely classified into two types as follows.

The first is to implement it in high-end robots that are studied in academia or a group of industries. The stability of walking is determined through ZMP (Zero Moment Point) and ZMP is controlled in real time.

This method has the advantage of adapting to walking in real time even on uneven surfaces such as obstacles and slopes, such as Honda ASIMO.

However, in order to calculate the ZMP, a sensor, such as a pressure sensor or a load cell, capable of measuring the reaction force applied to the sole of the robot by the ground, must be provided.

In addition, US Patent 5404086 proposes a method of adapting gait to the inclination angle or the uneven ground by using the reaction force of the ground and the inclination sensor, but this method requires a considerable amount of calculation and can actively control the ZMP. Must have an algorithm

Second, in the case of commercially available quadruped robots of the Sony AIBO class, there is a method of realizing walking that can stably maintain the ZMP trajectory, and simply regenerating it in a real robot, which does not deal with external environmental changes. In particular, there is a problem in that it is not possible to cope with this effectively in the case of facing the slope while walking on a general flat surface, and thus cannot perform stable walking.

The present invention has been made in order to solve the above problems, by compensating for the inclination caused by the inclination angle with respect to the robot body when facing the inclined surface, the walking trajectory corrected according to the angle degree of the inclined surface is reproduced to perform stable walking The purpose is to provide a walking control method for a robot.

1 is a flow chart for the walking control method of the present invention robot.

FIG. 2 is an operation flowchart showing slope correction for tilt angle adaptation in FIG.

3 is a view showing the relationship between the center of gravity of the robot and the support polygon in FIG.

4 is a view showing a state of calculating the center of gravity movement amount in FIG.

Fig. 5 shows the principle of inclined plane adaptation in Fig. 2;

Figure 6 shows an embodiment of inclined plane adaptation in Figure 2;

The present invention for achieving the above object is a first process for setting a stable walking trajectory; A second process of correcting an inclined surface for adapting an inclination angle when walking on the preset walking trajectory and facing an inclined surface; And performing a third process of walking with the correction trajectory calculated by the slope correction.

Hereinafter, with reference to the accompanying drawings, the operation and effect of the walking control method of the robot according to the present invention will be described in detail.

1 is an operation flowchart of a walking control method of a robot of the present invention, the first process of setting a stable walking trajectory as shown in the drawing; A second process of correcting an inclined surface for adapting an inclination angle when walking on the preset walking trajectory and facing an inclined surface; And a third process of walking with the correction trajectory calculated by the inclined plane correction.

FIG. 2 is an operation flowchart showing the inclination plane correction for the inclination angle adaptation, and the first step of detecting the inclination angle through the tilt sensor as shown in FIG. A second step of calculating a weight-based movement amount for correcting the inclination angle if the inclination angle is different from the preset inclination angle when planning the walking; The third step of calculating the angle value of each joint with respect to the center of gravity corrected using the inverse kinematics, it will be described the operation of the present invention configured as described above.

First, the stable walking trajectory is set offline, that is, the relative angle between the robot body and the ground is set, and the relationship between the center of gravity and the support polygon is set to set the stable walking trajectory, and also the center and support of the robot body. The coordinates of the end of each leg with respect to the whole coordinate system in one cycle of walking are determined by the relationship between the polygons and the separation and support of each leg from the ground.

Here, the support polygon means a polygon generated by a leg that is in contact with the ground and supports the load of the body. In order to implement walking stably, the center of gravity of the body is set to exist in the support polygon.

When making the robot's walking trajectory off-line, the relationship between the center of gravity of the robot and the support polygon is as shown in FIG. 3. In the static / semistatic state, when the orthographic projection of the ground of the center of the robot exists in the support polygon, The robot does not fall and implements stable walking.

In FIG. 3, the first and sixth figures show the start and the end of one cycle of walking, where the black dots represent the state where the soles are in contact with the ground and the white dots are spaced apart from the ground. The quarter point also means the center of gravity.

Therefore, as shown in the first and sixth figures, the support polygon is composed of squares when all four legs are in contact with the ground, and the center of gravity is in the squares, so they are never inverted in the static state.

And, as shown in the second to fifth pictures, when one leg is spaced for walking, the support polygon consists of a dark line and has a triangular shape. During one cycle of walking, the four legs are alternately spaced one by one. Move the center of gravity.

On the other hand, walking may be basically defined as an operation of moving the body in a stable manner without conduction of the body.

Therefore, in order to implement the walking stably, the center of gravity of the body as shown in Figure 3 must be moved inside the support polygon while moving the center of gravity.

On the other hand, the operation in the robot is achieved by feedback control of the potentiometer signal to the motor mounted on each joint. Accordingly, the trajectory of the walking cycle of each leg end should be changed to the angle trajectory information of each rotating joint constituting the leg. To do this, we need to solve inverse kinematics.

The inverse kinematics problem can be performed off-line and the result (angle trajectory) can be stored in the robot's memory, and the robot's walking is realized by calculating in real time only with the relationship between the predefined center of gravity and the support polygon.

Subsequently, the robot moves on a flat walking trajectory, and when the robot meets the inclined plane, the robot performs the inclination angle correction for adapting to the inclined plane. The basic of the inclination adaptation algorithm for the inclination angle correction is shown in FIG.

As shown on the left, the robot walking on the flat surface realizes perfect walking because it has already walked on the flat surface.However, if the flat walking plan is applied to the inclined surface as shown in the middle figure, the center of gravity and support polygon and There is an error in the relationship between the robot and the robot, which can not be implemented in the planned walk, can be flipped or flipped. To solve this problem, implement a method of tilting the body as shown in the figure on the right.

The basis of the tilt is to make the legs parallel to the direction of gravity, so that the parallel relationship between the initially planned ground and the body is maintained even when the tilt is implemented as shown in the left figure.

That is, after detecting the inclination angle through the tilt sensor attached to the robot, if the inclination angle is different from the preset inclination angle when planning the walk, calculate the center of gravity movement to correct the inclination angle.

Here, as shown in Figure 4, the amount of movement of the center of gravity along the inclined surface is implemented by the following equation.

[Equation]

Here, x is the moving direction of the robot, y is the direction perpendicular to the traveling direction, z represents the vertical direction, Is the pitching of the robot, Represents rolling.

Then, the inverse kinematics is used to calculate the angle value of each joint with respect to the corrected center of gravity, which is used to convert the trajectory of each gait cycle into the angle trajectory information of each rotational joint constituting the leg. will be.

Then, by the inclined plane correction, the walking is performed with the calculated correction trajectory, that is, through the inclination angle adaptation algorithm, the robot implements the planned walking on the flat surface through the inclination of the body even on the inclined plane.

The inclined walking algorithm may be performed not only in real time during the walking period but also intermittently as follows. As shown in FIG. 6, the three legs are separated from the ground while the legs land on the ground. Since it is stably supported, without performing the inclination angle correction, if the leg is landed on the inclined surface, the center of gravity movement of the robot body is sensed by sensing the change in the orientation of the robot body generated from the tilt sensor.

Therefore, when the next leg is spaced apart, the walking trajectory corrected according to the angle degree of the inclined surface is reproduced.

That is, the present invention, when facing the inclined surface, by performing the center of gravity movement of the body using the orientation information of the robot body obtained through the tilt sensor, to perform a stable walking on the inclined surface.

The specific embodiments or examples made in the detailed description of the present invention are intended to clarify the technical contents of the present invention to the extent that they should not be construed as limited to these specific embodiments and should not be construed in consultation. Various changes can be made within the scope of.

As described in detail above, the present invention, when facing an inclined surface during walking, has the effect of implementing stable walking on the inclined surface by performing the walking by moving the center of gravity of the robot body by the calculated correction trajectory by the inclined surface correction.

Claims (9)

A first step of presetting a stable walking trajectory; A second step of performing an inclined plane correction to adapt to an inclined angle when the person walks with the preset walking trajectory and faces an inclined plane; A walking control method of a robot, characterized in that performed in a third process of walking with a correction trajectory calculated by the inclined plane correction. The method of claim 1, wherein the first process comprises: A first step of setting a relative angle between the robot body and the ground; And a second step of establishing a relationship between the center of gravity and the support polygon. The method of claim 2, wherein the support polygon, A walking control method of a robot, characterized by a polygon generated by a leg that touches the ground and supports a load of a body. 3. The method of claim 2, further comprising the step of setting the coordinates of each leg end relative to the entire coordinate system in one cycle of walking through the relationship between the center of gravity of the robot body and the support polygon and the separation and support relationship from the ground of each leg. Walking control method of the robot, characterized in that. The method of claim 2, wherein the second step is In order to stably implement walking, the robot's walking control method, characterized in that the center of gravity of the body is set to exist in the support polygon. The method of claim 1, wherein the second process comprises: Detecting a tilt angle through a tilt sensor; A second step of calculating a weight-based movement amount for correcting the inclination angle if the inclination angle is different from the preset inclination angle when planning the walking; A walk control method for a robot, comprising a third step of calculating an angle value of each joint with respect to a corrected center of gravity using inverse kinematics. The method of claim 6, wherein the inverse kinematics A walking control method of a robot, which is used for converting a trajectory of each walking cycle into angle trajectory information of each rotational joint forming a bridge. 7. The walking control method for a robot according to claim 6, wherein the center of gravity movement is calculated by the following equation. [Equation] Here, x is the moving direction of the robot, y is the direction perpendicular to the traveling direction, z represents the vertical direction, Is the pitching of the robot, Represents rolling. The method of claim 1, wherein the third process comprises: When the leg currently spaced from the ground lands on the inclined surface, the walking control method of the robot, characterized in that spaced apart the next leg with a walking trajectory corrected according to the angle information of the inclined surface.