KR20170018218A - Method and system of controlling wearable robot - Google Patents

Abstract

Provided are a method and a system of controlling a wearable robot, wherein the method comprises: a measurement step of measuring joint torque on each joint of a robot, calculating intention torque which a wearer intends in each joint through the same; an adding step of adding the intention torque in each joint to draw an imaginary intention force on a dead end of a robot; and a distribution step of distributing the intention force to control torque to control a drive unit in each joint.

Description

TECHNICAL FIELD [0001] The present invention relates to a method and system for controlling a wearable robot,

The present invention relates to a wearable robot control method and system capable of improving the performance of a robot control method by converting a torque measured by a robot into an acceleration and then determining an accurate operation intention of the wearer through application of an impedance model.

In order to reduce the load acting on the human body and to increase the convenience in handling such a high-load object, a worker wears a robot so that a slight operation force A wearable robot has been developed which can easily handle a heavy object by manipulating the robot with only the robot.

In order to properly control the robot, it is necessary to mount an appropriate sensor on the robot to accurately grasp the intended torque of the user and to control the robot according to the user's intention.

Therefore, a control method that extracts user's intended intention torque easily and quickly through the change of the current of each joint motor of the robot and the measurement value of the acceleration sensor mounted on the robot by various control techniques for grasping the wearer's intention and controlling the robot .

However, the wearer's motion intention torque calculated in the above-described manner may be different from the motion intention of the correct wearer because it is an independent generated torque of each joint with respect to an instantaneous wearer's motion intention. In addition, the conventional method is not efficient in terms of the weight and the cost of the robot since a separate acceleration sensor is required.

Therefore, there is a need for a control method that can precisely grasp the motion intention of the wearer without disposing a separate sensor such as an acceleration sensor on the robot and control the robot appropriately according to the wearer's intention.

It should be understood that the foregoing description of the background art is merely for the purpose of promoting an understanding of the background of the present invention and is not to be construed as an admission that the prior art is known to those skilled in the art .

An object of the present invention is to provide a wearable robot control method and system for allowing a wearer to wear a wearable robot and grasp the correct operation intention of the wearer during the operation thereof and appropriately control the wearer's operation intention.

According to still another aspect of the present invention, there is provided a method of controlling a wearable robot, comprising: measuring a joint torque at each joint of a robot and calculating an intention torque of a wearer at each joint; A summation step of summing the intention torques at each joint to derive a virtual intention force at the robot end; And a distribution step of distributing the intentional force to a control torque for controlling the driving part in each joint.

The joint torque can be measured using the torque cell provided in each joint, and the intent torque is calculated by summing up the joint torque torque and the kinetic model torque value. Here, the kinetic model torque value is obtained by summing the torque value by gravity, the torque value by centrifugal force, and the torque value by Coriolis force.

In the present invention, the imaginary intention force at the end of the robot is calculated by deriving the acceleration at the end of the robot using the summed intention torque and using the following equation.

: 가상의 의도힘,

: 로봇 끝단의 위치, 속도, 가속도,

: 임피던스 모델 파라미터

: Virtual intention force,

: Position, velocity, acceleration,

: Impedance model parameters

The above-mentioned acceleration of the end of the robot is derived using the following equation.

: 로봇 끝단의 가속도, J: 자코비안 행렬, M(q): 질량 행렬, V(q): 원심력과 코리올리힘 행렬, G(q): 중력행렬,

: 의도토크,

: 로봇 관절의 각속도

G (q): Gravitational matrix, V (q), V (q): acceleration of the robot end, J: Jacobian matrix,

: Intention torque,

: Angular velocity of robot joint

The control torque is a sum of a value obtained by applying a Jacobian transform to the intention force and a control compensation torque value compensating for a loss torque that can be generated in a robot operation. Therefore, the control compensation torque value is a value And the joint damper torque value.

A wearable robot control system according to the present invention includes: a measurement unit for measuring a torque value of each joint of a robot; A driving unit applying a control torque of each joint of the robot; The joint torque at each joint of the robot is measured to calculate the intended intention torque of the wearer at each joint and the imaginary intention force at the end of the robot is calculated by summing the calculated intention torque, And a control unit for distributing the control torque to control the driving unit in the joint.

The intentional torque of the wearable robot control system according to the present invention is a value obtained by summing up the joint torque value and the kinetic model torque value, and the kinetic model torque value is a torque value by gravity, a torque value by centrifugal force, .

The virtual intention force at the robot end of the wearable robot control system can be calculated using the following equation after deriving the acceleration of the robot end using the summed intention torque.

: 가상의 의도힘,

: 로봇 끝단의 위치, 속도, 가속도,

: 임피던스 모델 파라미터

: Virtual intention force,

: Position, velocity, acceleration,

: Impedance model parameters

Also, the acceleration of the robot end of the wearable robot control system can be derived using the following equation.

: 로봇 끝단의 가속도, J: 자코비안 행렬, M(q): 질량 행렬, V(q): 원심력과 코리올리힘 항, G(q): 중력항,

: 의도토크,

: 로봇 관절의 각속도

: Acceleration of robot end, J: Jacobian matrix, M (q): mass matrix, V (q): centrifugal force and Coriolis force term, G (q)

: Intention torque,

: Angular velocity of robot joint

The control torque distributed to the driving unit from the control unit of the wearable robot control system is a sum of a value obtained by applying Jacobian Transpose to the intention force and a control compensation torque value compensating for the loss torque that may occur during the operation of the robot The control compensation torque value can add up the friction compensation torque value and the joint damper torque value.

As described above, by using the robot control method when worn, the following effects can be obtained.

First, the robot senses the movement of the end of the robot leg using the torque measured at each joint, so that it is possible to accurately detect the motion intention of the wearer, thereby improving the consistency of the robot's motion with the motion intention of the wearer.

Second, the acceleration, speed, and position of each leg end can be calculated using only the algorithm of the present invention without adding a separate device by using the torque measured at the joint, so there is no fear of cost increase by the implementation of the present invention.

1 is a flowchart of a wearable robot control method according to an embodiment of the present invention
2 is a configuration diagram of a wearable robot control system according to an embodiment of the present invention

Hereinafter, preferred embodiments of the present invention will be described with reference to the accompanying drawings.

As shown in FIG. 1, the method of controlling a wearable robot according to the present invention includes a measuring step (S100) of measuring joint torques at respective joints of a robot and calculating intentional torques intended by a wearer at each joint; A summing step S200 of summing the intention torques at the respective joints to derive a virtual intention force at the robot end; And a distribution step (S300) of distributing the intentional force to a control torque for controlling the driving part in each joint.

In the measuring step S100, the joint torque values at the respective joints of the robot are derived. There are various methods for deriving the joint torque values. The most accurate and direct method is to measure using the torque cells provided in each joint. Therefore, the present invention proposes a method of measuring the joint torque value using the torque cell.

In addition, a method of calculating the torque value based on the measurement of the angle and angular velocity of the joint using the encoder provided in the joint may be used. However, such a method is disadvantageous in that the controller needs to separately convert the measured angular and angular velocity values into the torque values, and the accuracy of the conversion may deteriorate during the conversion process.

After measuring the joint torque value of each joint, it is used to calculate the intention torque of the wearer. There are various ways of calculating the intention torque of the wearer. The intention torque may be calculated using the force value after changing the joint torque value described above by force or a method of providing a separate sensor for calculating the intention torque of the wearer may be considered.

In the present invention, a method of calculating a wearer's intention torque by summing a kinematic model torque value to a joint torque value measured differently from the above-described method is proposed.

When a person wears a robot, there are two types of torque values that can be measured in joints. One is the torque value that occurs when the joints move in different directions and the joints are twisted. This can be regarded as a relative torque value according to the wearer's motion, and the above-mentioned joint torque measurement value corresponds to this value. The other is the torque value generated by the gravity force, which is always applied to the human being and the robot, unlike the above-mentioned dynamics model torque value. This value should always be applied as long as a person wears a robot on the earth and can be regarded as an absolute torque value when compared with the previous torque value.

Therefore, in the present invention, a method of summing the two torque values to accurately reflect the wearer's intention in calculating the intention torque of the wearer is proposed.

The kinetic model torque value may exist in various forms, but generally, the torque value due to gravity, the torque value due to centrifugal force, and the torque value due to the Coriolis force are summed up.

: 동역학모델토크값, M(q): 질량행렬, q'': 관절 각가속도, V(q,q'): 원심력과 코리올리힘 행렬, G(q): 중력행렬

(Q, q '): centrifugal force and Coriolis force matrix, G (q): gravity matrix,

In the above equation, the mass matrix is the sum of the masses of the robot and the human, and the mass matrix value is multiplied by the angular velocity of the joint to obtain the torque value according to the rotational motion of the joint. G (q) is a term used to reflect the gravity acting on the earth as a gravity matrix. V (q, q ') is the centrifugal force and Coriolis force matrix value. The reasons for reflecting this value in the kinetic model torque value are as follows.

The centrifugal force is intended to reflect the centrifugal force generated by the joint rotating motion. The Coriolis force is often referred to as the turning force, which is the apparent force that appears when observing a moving object in a rotating kinetic system. As the Earth rotates, the rotating force of rotation in both the northern and southern hemispheres will act. Of course, in the northern hemisphere where the Republic of Korea is located, there is a directing force in the direction of movement of the object to the right, although the directional force will hardly act in the equatorial region. Therefore, the present invention also reflected the Coriolis force for accurate torque value calculation.

If the measurement step S100 is completed according to the above-described method, the process proceeds to the summing step S200 as shown in FIG. Control methods for obtaining the intention torque value of the existing wearer are mostly the method of obtaining the wearer's intention torque using the torque cell measurement torque mounted on each joint and controlling each wearer's intention torque by each joint.

However, the wear intention torque of the wearer may be different from the wear intention of the wearer because it is an independent occurrence torque of each joint with respect to the intentional motion intention of the wearer. Accordingly, the present invention proposes a method of collecting the operation intention of the wearer by summing the operation intention torque measured at each joint, so that in this summation step (S200) Adds the intention torque of the joint.

After summing, the virtual intention force at the end of the robot should be calculated using the summed intent torque value. In the present invention, a method of obtaining a virtual intention force through a virtual impedance model is presented, and detailed methods are as follows.

: 가상의 의도힘,

: 로봇 끝단의 위치, 속도, 가속도,

: 임피던스 모델 파라미터

: Virtual intention force,

: Position, velocity, acceleration,

: Impedance model parameters

In the above equation, the impedance model parameters may be the same as the transformation coefficients set in the robot controller for calculating the virtual intention force, and may be various values depending on the type and state of the robot. However, the position, velocity, and acceleration values of the end of the robot are determined according to the motion of the robot. Therefore, in order to obtain the accurate virtual intention force in this equation, it is important to accurately obtain the position, velocity, and acceleration of the end of the robot. Therefore, the prior art has a separate sensor capable of measuring acceleration values to accurately calculate position, velocity, and acceleration values.

However, in the present invention, a method using the following formula is proposed as a method of obtaining the acceleration of the end of the robot without a separate acceleration sensor.

: 로봇 끝단의 가속도, J: 자코비안 행렬, M(q): 질량 행렬, V(q): 원심력과 코리올리힘 행렬, G(q): 중력행렬,

: 의도토크,

: 로봇 관절의 각속도

G (q): Gravitational matrix, V (q), V (q): acceleration of the robot end, J: Jacobian matrix,

: Intention torque,

: Angular velocity of robot joint

The above equation is based on the above-described kinetic model torque value, but it is basically an equation that starts at F = ma. However, since the previously calculated torque value must be converted to the force region, the Jacobian matrix value is applied to this.

If the acceleration of the end of the robot is obtained by using the above equation, the velocity and the position can be easily obtained. This is because acceleration, velocity and position can be obtained directly by using calculus in the time domain.

Therefore, according to the present invention, the torque value of each of the existing joints is independently measured, and the acceleration, speed, and position information of the end of the robot are obtained by summing the torque values of the respective joints, The robot control precision can be further improved as compared with the prior art.

If the summation step S200 is completed in the above-described manner, the distribution step S300 is performed as shown in FIG. In the distribution step S300, it is necessary to first convert the virtual intention force calculated in the summation step S200 back into the torque region. To do this, Jacobian Transpose should be applied to the intentional force.

Therefore, it is possible to calculate the control torque value of each joint through Jacobian transposing as described above. It is not preferable to apply this value to each joint as it is. Because the values described above are based on the ideal equation, there is a gap between the actual robot motion values. Therefore, in the present invention, the control torque values for each joint are calculated by summing the values obtained by applying the Jacobian transform to the intention force and the control compensation torque values for compensating the loss torque that may occur in the robot operation.

The control compensation torque value may have various values depending on the movement of the wearer or the type of the robot. However, the friction compensation torque value and the joint damper torque value will always affect the control torque value of each joint regardless of the type of robot or the movement of the wearer. Therefore, in the present invention, a method of summing the friction compensation torque value and the joint damper torque value with the control compensation torque value is proposed.

The friction compensation torque value is a torque value for compensating for the frictional force that can be generated according to the motion of the robot. Frictional force can occur in various places. The most representative example is friction force generated between the foot and the ground in a wearable walking robot. In this case, the frictional force value can be obtained by using the friction coefficient of the ground surface and the mass value of the robot and the wearer. In this way, the frictional compensation torque value can be calculated. In addition to this, friction between the robot and the wearer may occur, and frictional forces may be generated in the joints.

The joint damper torque value refers to the torque value by the damper which alleviates the impact in the joint. There is a damper in the robot joint to mitigate the impact due to the motion. The damper is also affected by the movement of the joint of the robot. The joint damper torque value is needed to compensate this value. The calculation of the joint damper torque value may be performed in various ways, but in the case of the present invention, it is most preferable to calculate the joint damper torque value using the torque cell located at each joint.

The wearable robot control system according to the present invention includes: a measuring unit 500 for measuring a torque value of each joint of a robot; A driving unit 600 for applying a control torque of each joint of the robot; The joint torque at each joint of the robot is measured to calculate the intended intention torque of the wearer at each joint and the imaginary intention force at the end of the robot is calculated by summing the calculated intention torque, And a control unit 400 for distributing the control torque to control the driving unit in the joint, and this configuration diagram is shown in FIG.

The measuring unit 500 is generally located at each joint and transmits the signal to the control unit 400. The measuring unit 500 measures a torque cell capable of directly measuring a joint torque value or a joint angle and angular velocity, A possible encoder would be this. The driving unit 600 applies the control torque calculated by the control unit 400 to the joints, and generally exists in the form of a rotary actuator in each joint.

As shown in FIG. 1, the control unit receives the joint torque value from the measuring unit 500, converts the joint torque value into the control torque, and transmits the control torque to the driving unit 600 to apply the control torque to the joint.

The robot control unit 400 can accurately calculate the motion intention of the wearer by using the independent torque value generated at each joint through the system described above, Can be further improved.

Also, the intentional torque of the wearable robot control system according to the present invention adds the kinetic model torque value to the joint torque value as described above, and the kinetic model torque value is determined by the gravitational torque value, the centrifugal torque value, and the Coriolis force Can be set to a value obtained by summing up the torque values by the torque command value?

The virtual intention force at the robot end in the present system can be calculated using the following equation after deriving the acceleration of the robot end using the intention torque summed as described above.

: 가상의 의도힘,

: 로봇 끝단의 위치, 속도, 가속도,

: 임피던스 모델 파라미터

: Virtual intention force,

: Position, velocity, acceleration,

: Impedance model parameters

Also, the acceleration of the robot end can be derived using the following equation.

: 로봇 끝단의 가속도, J: 자코비안 행렬, M(q): 질량 행렬, V(q): 원심력과 코리올리힘 항, G(q): 중력항,

: 의도토크,

: 로봇 관절의 각속도

: Acceleration of robot end, J: Jacobian matrix, M (q): mass matrix, V (q): centrifugal force and Coriolis force term, G (q)

: Intention torque,

: Angular velocity of robot joint

The above-mentioned two formulas are described in detail above, so they are omitted.

The control torque of the wearable control robot system according to the present invention is calculated by summing a value obtained by applying a Jacobian transform to the intention force and a control compensation torque value compensating for a loss torque that can be generated in the robot operation, The value can be calculated by summing up the friction compensation torque value and the joint damper torque value.

Although the invention has been shown and described with respect to specific embodiments thereof, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention as defined by the appended claims. It will be obvious to those who have knowledge of.

S100: Measurement step S200: Summing step
S300: Distribution step 400:
500: measuring unit 600:

Claims (15)

A measuring step of measuring a joint torque at each joint of the robot and calculating the intention torque of the wearer at each joint through the measurement;
A summation step of summing the intention torques at each joint to derive a virtual intention force at the robot end; And
And a distributing step of distributing the intentional force to a control torque for controlling the driving part in each joint. The method according to claim 1,
Wherein the joint torque is measured using a torque cell provided in each joint. The method according to claim 1,
Wherein the intent torque is a value obtained by summing a joint torque value and a kinetic model torque value. The method of claim 3,
Wherein the kinetic model torque value is a sum of a torque value due to gravity, a torque value due to centrifugal force, and a torque value due to a force of Coriolis force. The method according to claim 1,
Wherein the imaginary intention force at the end of the robot is calculated using the following equation after deriving the acceleration of the end of the robot using the summed intention torque.

: Virtual intention force,

: Position, velocity, acceleration,

: Impedance model parameters The method of claim 5,
Wherein the acceleration of the end of the robot is derived using the following equation.

: Acceleration of robot end, J: Jacobian matrix, M (q): mass matrix, V (q): centrifugal force and Coriolis force term, G (q)

: Intention torque,

: Angular velocity of robot joint The method according to claim 1,
Wherein the control torque is a value obtained by summing a value obtained by applying a Jacobian transform to the intention force and a control compensation torque value compensating a loss torque that may be generated in the robot operation. The method of claim 7,
Wherein the control compensation torque value is a sum of a friction compensation torque value and a joint damper torque value. A measuring unit for measuring a torque value of each joint of the robot;
A driving unit applying a control torque of each joint of the robot;
The joint torque at each joint of the robot is measured to calculate the intended intention torque of the wearer at each joint and the imaginary intention force at the end of the robot is calculated by summing the calculated intention torque, And a controller for distributing the control torque to control the driving part in the joint. The method of claim 9,
Wherein the intentional torque is a value obtained by summing up the joint torque value and the kinetic model torque value. The method of claim 10,
Wherein the kinetic model torque value is a sum of a torque value due to gravity, a torque value due to centrifugal force, and a torque value due to a force of Coriolis force. The method of claim 9,
Wherein the imaginary intention force at the end of the robot is calculated using the following equation after deriving the acceleration of the end of the robot using the summed intention torque.

: Virtual intention force,

: Position, velocity, acceleration,

: Impedance model parameters The method of claim 12,
Wherein the acceleration of the end of the robot is derived using the following equation.

: Acceleration of robot end, J: Jacobian matrix, M (q): mass matrix, V (q): centrifugal force and Coriolis force term, G (q)

: Intention torque,

: Angular velocity of robot joint The method of claim 9,
Wherein the control torque is a sum of a value obtained by applying a Jacobian transform to the intention force and a value obtained by adding a control compensation torque value compensating for a loss torque that can be generated during operation of the robot. 15. The method of claim 14,
Wherein the control compensation torque value is a sum of a friction compensation torque value and a joint damper torque value.

Images