JP2010269385A - Method of creating s-shaped acceleration/deceleration orbit and articulated robot system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a method of creating an S-shaped acceleration/deceleration orbit shortening the operating cycle time of a robot and reducing excessive load on mechanisms and motors while suppressing vibrations produced during acceleration or deceleration to the minimum, and also to provide an articulated robot system. <P>SOLUTION: A temporary S-shaped acceleration/deceleration orbit for a reference joint axis is created, wherein a time tm1 when the velocity is maximized in a section where the acceleration becomes the maximum, or a time tm2 when the velocity is maximized in a section where the acceleration becomes the minimum is calculated. Then, using a preset dynamics model, an estimated torque τ of each joint at the time tm1 or the time t2m is calculated. Subsequently, while maintaining the time of the S-shaped acceleration reaching time constant, the modified maximum or minimum acceleration of the reference joint axis is found such that the estimated torque τ found using the dynamics model becomes a preset target torque τ<SB>tg</SB>. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a method for generating an S-curve acceleration / deceleration trajectory and an articulated robot system. In particular, while suppressing vibration during acceleration / deceleration as much as possible, the operation cycle time of the robot is shortened and an excessive load on the mechanism and motor is reduced. The present invention relates to an S-shaped acceleration / deceleration trajectory generation method and an articulated robot system.

  In general, when an arm of an articulated robot is rotated about each joint axis, control is performed so that a graph showing the relationship between the time from the start of rotation and the angular velocity draws an S-shape (S-shape acceleration / deceleration control). Things have been done. That is, the angular acceleration is gradually increased from 0 at the start of rotation, and then is set to constant angular acceleration for a certain time, and then the angular acceleration is gradually decreased to 0. In this state, after rotating the arm at a constant angular velocity for a certain time, the angular acceleration is gradually decreased from 0 and rotated at a negative constant angular acceleration for a certain time. Thereafter, the angular acceleration is gradually increased to zero.

  On the other hand, Patent Document 1 discloses a technique for obtaining a limit acceleration / deceleration time such that the generated torque on the i-th axis becomes the allowable maximum torque.

JP-A-7-261822

  However, in the case of the technique described in Patent Document 1, the entire acceleration time is expanded and contracted in order to change the angular acceleration. When this technique is applied to S-curve acceleration / deceleration trajectory control, the time is expanded and contracted including the S-curve acceleration arrival time (the time from the start of rotation until the equiangular acceleration is reached). As a result, depending on the operation pattern, the S-curve acceleration arrival time may be shorter than the “arm natural frequency”. Thereby, vibration may occur. In this case, the positioning accuracy is lowered during braking, and it becomes very difficult to use.

  The present invention has been made in consideration of such points, and by keeping the S-curve acceleration arrival time constant, while suppressing vibration during acceleration / deceleration as much as possible, the operation cycle time of the robot can be shortened, and the mechanism and motor An object of the present invention is to provide an S-shaped acceleration / deceleration trajectory generation method and an articulated robot system that can reduce an excessive load on the robot.

  The present invention relates to an S-curve acceleration / deceleration trajectory generation method for generating an S-curve acceleration / deceleration trajectory for driving each joint axis for a robot having a plurality of joint axes, based on the motion target position and motion pattern of the robot. Determining a movement time of each joint axis to the operation target position, determining a joint axis having the longest movement time as a reference joint axis, and generating a provisional S-shaped acceleration / deceleration trajectory for the reference joint axis; In the provisional S-curve acceleration / deceleration trajectory, a step of calculating a time tm1 at which the speed is maximum in a section where the acceleration is maximum, or a time tm2 at which the speed is maximum in a section where the acceleration is minimum. And calculating the estimated torque of each joint axis at time tm1 or time tm2 using a preset dynamic model, and reaching the maximum acceleration after starting acceleration for each joint axis The estimated torque obtained using the dynamic model is set in advance while keeping the maximum acceleration arrival time until starting or the minimum acceleration arrival time from starting deceleration until reaching the minimum acceleration. So as to obtain the corrected maximum acceleration or the corrected minimum acceleration of the reference joint axis, and the S-shaped acceleration / deceleration of the reference joint axis based on the corrected maximum acceleration or the corrected minimum acceleration of the reference joint axis thus obtained. A method for generating an S-curve acceleration / deceleration trajectory comprising a step of generating a modified S-curve acceleration / deceleration trajectory modified so that the trajectory has the corrected maximum acceleration or the corrected minimum acceleration.

  According to the present invention, the step of obtaining the corrected maximum acceleration or the corrected minimum acceleration of the reference joint axis uses a dynamic model, so that the maximum or minimum acceleration of the reference joint axis in the provisional S-shaped acceleration / deceleration trajectory and the time tm1 Alternatively, an acceleration change rate that is a ratio of the corrected maximum acceleration or the corrected minimum acceleration of each joint axis when the estimated torque of each joint axis at time tm2 matches a preset target torque of each joint axis is set to all joint axes. Calculating the minimum acceleration change rate for each joint axis calculated, and calculating the minimum acceleration change rate for the joint joint axis calculated in the provisional S-curve acceleration / deceleration trajectory And calculating a corrected maximum acceleration or a corrected minimum acceleration of the reference joint axis by multiplying the maximum acceleration or the minimum acceleration. It is a trajectory generation method.

  The present invention, in an articulated robot system having a plurality of joint axes, includes a robot main body and a control device that controls the robot main body, and the control device operates based on an operation target position and an operation pattern of the robot. A movement time of each joint axis to the target position is obtained, a reference joint axis determination processing unit that determines a joint axis having the longest movement time as a reference joint axis, and a provisional S-curve acceleration / deceleration trajectory for the reference joint axis are generated. In the first S-shaped acceleration / deceleration trajectory generation unit and the provisional S-shaped acceleration / deceleration trajectory, the time tm1 at which the speed is maximum is calculated among the sections where the acceleration is maximum, or the speed is determined among the sections where the acceleration is minimum. A time calculation unit that calculates the maximum time tm2, and a calculation processing unit that calculates an estimated torque of each joint axis at the time tm1 or the time tm2, using a preset dynamic model; Dynamics while maintaining the maximum acceleration arrival time from the start of acceleration for each joint axis until reaching the maximum acceleration, or the minimum acceleration arrival time from the start of deceleration to the arrival of the minimum acceleration. The angular acceleration change processing unit for obtaining the corrected maximum acceleration or the corrected minimum acceleration of the reference joint axis so that the estimated torque obtained using the model becomes a preset target torque, and the reference joint axis obtained in this way Generation of a second S-shaped acceleration / deceleration trajectory that generates a modified S-shaped acceleration / deceleration trajectory modified so that the S-shaped acceleration / deceleration trajectory of the reference joint axis has the corrected maximum acceleration or the corrected minimum acceleration based on the corrected maximum acceleration or the corrected minimum acceleration This is an articulated robot system characterized by having a part.

  According to the present invention, the estimated torque obtained using the dynamic model is set in advance while the maximum acceleration arrival time or the minimum acceleration arrival time (S-shaped acceleration arrival time) of each joint axis is kept constant. The corrected maximum acceleration or the corrected minimum acceleration of the reference joint axis is obtained so that As a result, it is possible to optimize the acceleration according to the operation pattern and load conditions of the robot while suppressing the occurrence of vibration during acceleration / deceleration. Further, in the S acceleration / deceleration trajectory operation, the joint axis of the robot can be smoothly accelerated and decelerated without lowering the positioning accuracy.

1 is a schematic configuration diagram showing an articulated robot system according to an embodiment of the present invention. The flowchart which shows the S-shaped acceleration / deceleration orbit generation method by one embodiment of this invention. The graph which shows the relationship between the time and the angular velocity of a reference joint axis, and the relationship between the time and the angular acceleration of a reference joint axis in a provisional S-shaped acceleration / deceleration trajectory. The graph which shows the relationship between the time and the angular velocity of a reference joint axis at the time of acceleration, and the relationship between the time and the angular acceleration of a reference joint axis.

  Hereinafter, an embodiment of the present invention will be described with reference to FIGS. Here, FIG. 1 to FIG. 4 are diagrams showing an embodiment of the present invention.

(Configuration of articulated robot system)
First, the overall configuration of the articulated robot system will be described with reference to FIG. Here, a system including an articulated robot having three joint axes will be described as an example of the articulated robot system.

  As shown in FIG. 1, the articulated robot system 100 includes a robot body 10 and a control device 20 that is connected to the robot body 10 and controls the robot body 10.

Among these, the robot body 10 is provided in order from the base 11 fixed to the floor surface and the base end side (base 11) toward the front end side, and through _three mutually different joint axes J _{1 to} J _3. There are three arm bodies (first arm body 12, second arm body 13, and third arm body 14) connected to each other.

Among the first arm member 12 is rotatable relative to the first joint axis J ₁ base 11 around the. A second joint axis J ₂ is interposed between the first arm body 12 and the second arm body 13, and the second arm body 13 is rotatable about the second joint axis J ₂ . Further, a third joint axis J ₃ is interposed between the second arm body 13 and the third arm body 14, and the third arm body 14 is rotatable about the third joint axis J ₃ .

  On the other hand, the control device 20 includes a RAM 30 and a CPU 40 connected to the RAM 30.

  Of these, the RAM 30 stores teaching information 31, S-shaped acceleration / deceleration parameter information 32, model parameter information 33, and target torque information 34. Hereinafter, an outline of such information will be briefly described.

  The teaching information 31 includes information such as an operation target position (for example, a teaching point) and an operation pattern (for example, an interpolation method) of the robot, which are determined based on an operation program set by the user.

The S-curve deceleration parameter information 32, temporary used in generating the S-curve deceleration trajectory, maximum angular velocity of each joint axis J ₁ through J _3, the maximum angular acceleration of each joint axis J ₁ through J ₃ and minimum angular acceleration, and the maximum angular acceleration arrival time and the minimum angular acceleration time of arrival of each joint axis J ₁ through J ₃ (or less, S-shape also called acceleration arrival time) are included, such as information. Note that the various information included in the S-shaped acceleration / deceleration parameter information 32 is set to a predetermined value for each of the joint axes J _{1 to} J _{3 in} advance, regardless of the operation program described above.

The model parameter information 33 includes model parameters (arm mass (Kg), link offset amount (mm), center of gravity offset (mm), moment of inertia around the center of gravity (Kg · m ² ), etc.), and each joint axis J _{1 to} J Information such as friction coefficient of ₃ is included.

The target torque information 34 includes information on the target torque of each joint axis J _{1 to} J ₃ set in advance for each joint axis J _{1 to} J ₃ .

  On the other hand, the CPU 40 obtains the movement time of each joint axis to the target position given by the program, determines the joint axis having the longest movement time as the reference joint axis, and the reference joint axis determination unit 48. A first S-shaped acceleration / deceleration trajectory generation unit 41 that generates a provisional S-shaped acceleration / deceleration trajectory, a time calculation unit 42 that calculates time tm1 and time tm2, which will be described later, and an arithmetic process that performs arithmetic processing using a dynamic model Part 43.

  The CPU 40 also calculates an S-shaped reference joint axis based on the angular acceleration change processing unit 44 for obtaining the corrected maximum angular acceleration and the corrected minimum angular acceleration of the reference joint axis, and the corrected maximum angular acceleration and the corrected minimum angular acceleration of the reference joint axis. A second S-shaped acceleration / deceleration trajectory generation unit 45 that generates a modified S-shaped acceleration / deceleration trajectory modified so that the acceleration / deceleration trajectory has the corrected maximum angular acceleration and the corrected minimum angular acceleration is provided.

Further, in the CPU 40, based on the target position calculation unit 46 for calculating the target position of each joint axis J _{1 to} J ₃ for each processing cycle based on the corrected S-curve acceleration / deceleration trajectory of the reference joint axis, based on the target position. A servo unit 47 that operates the robot body 10 is provided.

(Reference joint axis determination method and S-curve acceleration / deceleration trajectory generation method)
Next, a reference joint axis determination method and an S-shaped acceleration / deceleration trajectory generation method according to the present embodiment will be described with reference to FIGS.

First, the reference joint axis determination processing unit 48 of the CPU 40 includes a robot motion target position and motion pattern included in the teaching information 31 and information included in the S-shaped acceleration / deceleration parameter information 32, specifically, each joint axis J based on the maximum angular velocity of ₁ through J _3, determine the travel time of each joint axis J ₁ through J ₃ to the target position given by the program, determined on the basis joint axis long joint axis of the most moving time ( Step S0 in FIG.

Next, the first S-shaped acceleration / deceleration trajectory generating unit 41 of the CPU 40 includes the robot operation target position and operation pattern included in the teaching information 31, and information included in the S-shaped acceleration / deceleration parameter information 32, specifically, each joint. maximum velocity of the axis J ₁ through J _3, the maximum angular acceleration and minimum angular acceleration, as well as information on the maximum angular acceleration arrival time and the minimum angular acceleration time of arrival of each joint axis J ₁ through J ₃ of each joint axis J ₁ through J ₃ Based on the above, a provisional S-shaped acceleration / deceleration trajectory with respect to the reference joint axis is generated (step S1 in FIG. 2). The method for obtaining the provisional S-shaped acceleration / deceleration trajectory can be performed in the same manner as the conventional S-shaped acceleration / deceleration trajectory calculation method.

A graph based on this provisional S-shaped acceleration / deceleration trajectory is shown in FIG. In FIG. 3, the upper graph shows the relationship between the time from the movement start time (T ₀ ) and the angular velocity of the reference joint axis. The lower graph shows the relationship between the time from the movement start time (T ₀ ) and the angular acceleration of the reference joint axis.

Subsequently, the time calculation unit 42 of the CPU 40 in the provisional S-curve acceleration / deceleration trajectory has a maximum angular acceleration during acceleration (the interval between time T ₁ and time T _{2 in} FIG. 3). Time tm1 (time T _{2 in} FIG. 3) at which the angular velocity is maximum is calculated. Similarly, the time calculating unit 42 of the CPU40, among the sections angular acceleration during deceleration is minimized (section between time T ₅ and time T ₆ in FIG. 3), the time tm2 (Figure 3 the angular velocity is maximum At time T ₅₎ is calculated (step S2 in FIG. 2).

Next, the arithmetic processing unit 43 of the CPU 40 calculates estimated torques of the joint axes J _{1 to} J ₃ at time tm1 and time tm2 using the model parameter information 33 and a preset dynamic model ( Step S3 in FIG. As described above, the model parameter information 33 includes model parameters (arm mass (Kg), link offset amount (mm), center of gravity offset (mm), moment of inertia around the center of gravity (Kg · m ² ), etc.), and It includes information such as the coefficient of friction of the joint axis J ₁ through J _3.

Here, the dynamic model is expressed by the following equation.
τ = M (Θ) Θ " + B (Θ) [Θ a 'Θ b'] + C (Θ) [Θ '2] + G (Θ) + D (Θ') + τ Fc sgn (Θ ') ··· (1 )

In addition, in the equation (1) of the dynamic model, the meaning of each term is as follows.
τ: Estimated torque matrix for each joint axis (matrix with n rows and 1 column)
Θ, Θ ', Θ ": Joint axis angle matrix, joint axis angular velocity matrix, joint axis angular acceleration matrix (n-by-1 matrix)
Θ ' ² : Angular velocity product matrix of the same joint axis (matrix with n rows and 1 column)
Θ _a ′ · Θ _b ′: angular velocity product matrix with different axes (matrix with {n · (n−1) / 2} rows and one column)
M (Θ): Inertia (mass) matrix (matrix with n rows and n columns)
B (Θ): Coriolis force term matrix (matrix with n rows {n · (n−1) / 2} columns)
C (Θ): Centrifugal force matrix (matrix with n rows and n columns)
G (Θ): Gravity matrix (matrix with n rows and n columns)
D (Θ '): viscous frictional force matrix (matrix with n rows and 1 column)
τ _Fc sgn (Θ ′): Coulomb frictional force matrix (matrix with n rows and 1 column)
In the above, n represents the number of joint axes J _{1 to} J ₃ (in this embodiment, n = 3).

By the way, among the equations (1) of the dynamic model, the Coriolis force term matrix, the centrifugal force matrix, the gravity matrix, the viscous friction force matrix, and the Coulomb friction force matrix are not functions of the angular acceleration Θ ″. The matrix is summed to be τ _other (n × 1 matrix) In this case, the dynamic model (1) described above can be expressed as follows.

τ = M (Θ) Θ ”+ τ _other (Θ, Θ ′) (2)
(However, τ _other (Θ, Θ ′) = B (Θ) [Θ _a 'Θ _b '] + C (Θ) [Θ ′ ² ] + G (Θ) + D (Θ ′) + τ _Fc sgn (Θ ′))

Subsequently, the angular acceleration changing processing unit 44 of the CPU 40 starts the acceleration for each joint axis J _{1 to} J ₃ and reaches the maximum angular acceleration arrival time (in FIG. 3, time T ₀ ). time) between time T ₁ while maintaining a constant, so that the target torque estimated torque is set in advance determined by using a dynamic model, obtains the corrected maximum angular acceleration of the reference joint axis.

Similarly, the angular acceleration change processing unit 44 of the CPU 40 starts the deceleration for each joint axis J _{1 to} J ₃ and then reaches the minimum angular acceleration arrival time (in FIG. 3, as time T ₄ ). time) between time T ₅ while holding constant, so that the target torque estimated torque is set in advance determined by using a dynamic model, obtains the corrected minimum angular acceleration of the reference joint axis.

  Specifically, the corrected maximum angular acceleration and the corrected minimum angular acceleration of the reference joint axis are obtained as follows.

First, at time tm1 and time tm2, it is considered that target torque (τ _tg ) (matrix of n rows and 1 column) is generated by changing only the angular acceleration while keeping the S-curve acceleration arrival time constant.

Here, the angle matrix, the angular velocity matrix, and the angular acceleration matrix of the joint axes J _{1 to} J ₃ at time tm1 (time T _{2 in} FIG. 3) are respectively represented by Θ _Acc , Θ _Acc ′, Θ _Acc ″ (each n rows). 1 column matrix).

Further, the angle matrix, angular velocity matrix, and angular acceleration matrix of the joint axes J _{1 to} J ₃ at time tm2 (time T _{5 in} FIG. 3) described above are respectively Θ _Dec , Θ _Dec ′, Θ _Dec ″ (n rows 1 Column matrix).

Further, a matrix including the angular acceleration change rate for _obtaining the corrected maximum angular acceleration and the corrected minimum angular acceleration is α _Acc (acceleration) and α _Dec (deceleration) (each matrix of n rows and 1 column).

Note that the angular acceleration change rate during acceleration (each component of α _Acc ) is the joint axis J in the provisional S-shaped acceleration / deceleration trajectory (see the dotted lines in FIGS. 3, 4A, and 4B). _{1 to} J ₃ maximum angular accelerations (each component of Θ _Acc ) and the estimated torque (τ) of each joint axis J _{1 to} J ₃ at time tm1 are preset targets for each joint axis J _{1 to} J ₃ It is a ratio to the corrected maximum angular acceleration (each component of Θ _{Acc_N} ″ described later) of each joint axis J _{1 to} J ₃ when it corresponds to the torque (τ _tg ) (that is, τ = τ _tg ) (FIG. 4 ( b)). That is, the angular acceleration change rate during acceleration is {(each component of Θ _{Acc — N} ″) / (each component of Θ _Acc ″)}.

Similarly, the angular acceleration change rate during deceleration (each component of α _Dec ) is calculated for each joint axis in the provisional S-shaped acceleration / deceleration trajectory (see the dotted lines in FIGS. 3, 4A, and 4B). J ₁ and the minimum angular acceleration through J ₃ (each component of theta _Dec "), for each joint axis J ₁ through J ₃ at time tm2 estimated torque (tau) is the joint axis J ₁ through J ₃ which is set in advance It is a ratio with the corrected minimum angular acceleration (each component of Θ _{Dec_N} ″ described later) of each joint axis J _{1 to} J ₃ when it matches the target torque (τ _tg ) (that is, τ = τ _tg ). In other words, the angular acceleration change rate during deceleration is {(each component of Θ _{Dec_N} ″) / (each component of Θ _Dec ″)}.

4A shows the relationship between the time from the movement start time (T ₀ ) and the angular velocities of the joint axes J _{1 to} J ₃ during acceleration, and FIG. 4B shows the acceleration. The relationship between the time from the movement start time (T ₀ ) and the angular acceleration of each joint axis J _{1 to} J ₃ is shown. In FIGS. 4A and 4B, the provisional S-curve acceleration / deceleration trajectory graph described above is indicated by a dotted line, and a corrected S-curve acceleration / deceleration trajectory graph described later is indicated by a solid line.

  In this case, the dynamic model (2) described above can be rewritten as follows.

τ _tg = M (Θ _Acc ) α _Acc Θ _Acc ″ + τ _other (Θ _Acc , Θ _Acc ′) (3)
τ _tg = M (Θ _Dec ) α _Dec Θ _Dec ”+ τ _other (Θ _Dec , Θ _Dec ') (4)

By transforming Equations (3) and (4), the angular acceleration change rate matrix (α _Acc , α _Dec ) can be obtained as follows.

In this manner, the (5) by using the formula, each joint axis J ₁ through J ₃ during acceleration of the angular acceleration change rate matrix sought (alpha _Acc), all of the joint axis J ₁ through J ₃ The angular acceleration change rate during acceleration (each component of α _Acc ) is calculated (step S4 in FIG. 2).

Similarly, by using the above equation (6), the angular acceleration change rate matrix (α _Dec ) at the time of deceleration of each joint axis J _{1 to} J ₃ is obtained, and at the time of deceleration of all the joint axes J _{1 to} J ₃ . An angular acceleration change rate (each component of α _Dec ) is calculated (step S4 in FIG. 2).

Next, the angular acceleration change processing unit 44 of the CPU 40 selects the joint axis having the minimum value from the angular acceleration change rates (α _Acc ) at the time of acceleration of all the joint axes J _{1 to} J _3. The angular acceleration change rate is set to the minimum angular acceleration change rate (α _{Acc_MIN} ) (step S5 in FIG. 2).

Similarly, the angular acceleration change processing unit 44 of the CPU 40 selects the joint axis having the smallest value from the angular acceleration change rates (α _Dec ) at the time of deceleration of all the joint axes J _{1 to} J _3. The angular acceleration change rate is set to the minimum angular acceleration change rate (α _{Dec_MIN} ) (step S5 in FIG. 2).

Subsequently, the maximum correction of the reference joint axis is obtained by multiplying the minimum angular acceleration change rate (α _{Acc_MIN} , α _{Dec_MIN} ) by the maximum angular acceleration and minimum angular acceleration of the reference joint axis in the provisional S- _curve acceleration / deceleration trajectory, respectively. The angular acceleration and the corrected minimum angular acceleration are obtained (step S6 in FIG. 2).

That is, the modified maximum angular acceleration of the reference joint axis during acceleration, theta _{Acc_NB} by placing a "theta _{Acc_B} the maximum angular acceleration of the reference joint axis in provisional S-curve deceleration trajectory", the maximum angular acceleration theta this modification _{Acc_NB} ”
Θ _{Acc_NB} “= α _{Acc_MIN} · Θ _{Acc_B} ” (7)
Is required.

Similarly, if the corrected minimum angular acceleration of the reference joint axis during deceleration is Θ _{Dec_NB} ", and the minimum angular acceleration of the reference joint axis in the provisional S-shaped acceleration / deceleration trajectory is Θ _{Dec_B} ", this corrected minimum angular acceleration Θ _{Dec_NB} "
Θ _{Dec_NB} “= α _{Dec_MIN} · Θ _{Dec_B} ” (8)
Is required.

Subsequently, the second S-shaped acceleration / deceleration trajectory generating unit 45 of the CPU 40 determines the reference joint axis based on the corrected maximum angular acceleration (Θ _{Acc_NB} ″) and corrected minimum angular acceleration (Θ _{Dec_NB} ″) of the changed reference joint axis. A modified S- _curve acceleration / deceleration trajectory modified so that the S- _curve acceleration / deceleration trajectory has a corrected maximum angular acceleration (Θ _{Acc_NB} ″) or a corrected minimum angular acceleration (Θ _{Dec_NB} ″) is generated (step S7 in FIG. 2).

Thereafter, the target position calculation unit 46 of the CPU 40 calculates the target positions of the joint axes J _{1 to} J ₃ for each processing cycle based on the corrected S-shaped acceleration / deceleration trajectory of the reference joint axis. Next, the servo unit 47 operates the robot body 10 based on the target position.

As described above, according to the present embodiment, the dynamic model is used while keeping the S-curve acceleration arrival time (maximum angular acceleration arrival time and minimum angular acceleration arrival time) of each joint axis J _{1 to} J ₃ constant. The corrected maximum angular acceleration and the corrected minimum angular acceleration are obtained so that the estimated torque obtained in this way becomes a preset target torque. As a result, the angular acceleration can be optimized in accordance with the robot operation pattern and load conditions while suppressing the occurrence of vibration during acceleration / deceleration. Further, in the S acceleration / deceleration trajectory operation, the acceleration / deceleration can be smoothly performed without reducing the positioning accuracy.

  In the present embodiment, the corrected maximum angular acceleration and the corrected minimum angular acceleration are obtained in both cases of acceleration and deceleration. However, the present invention is not limited to this, but only in one case of acceleration or deceleration. The corrected maximum angular acceleration or the corrected minimum angular acceleration may be obtained by the method described above, and the corrected S-shaped acceleration / deceleration trajectory may be generated based on either the corrected maximum angular acceleration or the corrected minimum angular acceleration.

Also in this embodiment, each joint axis J ₁ through J ₃ have been described for the case of rotation, respectively, it can be applied the same considerations apply for the joint to linear motion. In this case, the words “angle”, “angular velocity”, and “angular acceleration” in the present embodiment are replaced with “position”, (narrow sense) “velocity”, and (narrow sense) “acceleration”, respectively. be able to.

DESCRIPTION OF SYMBOLS 10 Robot main body 11 Base 12 1st arm body 13 2nd arm body 14 3rd arm body 20 Control apparatus 30 RAM
31 Teaching information 32 S-curve acceleration / deceleration parameter information 33 Model parameter information 34 Target torque information 40 CPU
41 First S-shaped acceleration / deceleration trajectory generation unit 42 Time calculation unit 43 Calculation processing unit 44 Angular acceleration change processing unit 45 Second S-shaped acceleration / deceleration trajectory generation unit 46 Target position calculation unit 47 Servo unit 48 Reference joint axis determination processing unit 100 Articulated type robot system J ₁ through J ₃ joint axis

Claims (3)

In a S-shaped acceleration / deceleration trajectory generation method for generating an S-shaped acceleration / deceleration trajectory for driving each joint axis for a robot having a plurality of joint axes,
Determining the movement time of each joint axis to the movement target position based on the movement target position and movement pattern of the robot, and determining the joint axis having the longest movement time as a reference joint axis;
Generating a provisional S-shaped acceleration / deceleration trajectory for the reference joint axis;
In a provisional S-shaped acceleration / deceleration trajectory, calculating a time tm1 at which the speed is maximum in a section where the acceleration is maximum, or calculating a time tm2 at which the speed is maximum in a section where the acceleration is minimum. ,
Calculating an estimated torque of each joint axis at time tm1 or time tm2 using a preset dynamic model;
Dynamics while maintaining the maximum acceleration arrival time from the start of acceleration for each joint axis until reaching the maximum acceleration, or the minimum acceleration arrival time from the start of deceleration to the arrival of the minimum acceleration. Obtaining a corrected maximum acceleration or a corrected minimum acceleration of the reference joint axis so that the estimated torque obtained using the model becomes a preset target torque;
Based on the corrected maximum acceleration or the corrected minimum acceleration of the reference joint axis thus obtained, the corrected S-shaped acceleration / deceleration trajectory corrected so that the S-shaped acceleration / deceleration trajectory of the reference joint axis has the corrected maximum acceleration or the corrected minimum acceleration. And a method of generating an S-shaped acceleration / deceleration trajectory. The process of obtaining the corrected maximum acceleration or corrected minimum acceleration of the reference joint axis is as follows:
By using the dynamic model, the maximum or minimum acceleration of the reference joint axis in the provisional S-shaped acceleration / deceleration trajectory and the target torque of each joint axis in which the estimated torque of each joint axis at time tm1 or time tm2 is set in advance. A step of calculating an acceleration change rate that is a ratio of the corrected maximum acceleration or the corrected minimum acceleration of each joint axis when matching with the torque for all joint axes;
Of the calculated acceleration change rates of each joint axis, obtaining a minimum minimum acceleration change rate;
Multiplying the minimum acceleration change rate by the maximum acceleration or the minimum acceleration of the reference joint axis in the provisional S-curve acceleration / deceleration trajectory to obtain the corrected maximum acceleration or the corrected minimum acceleration of the reference joint axis. The method for generating an S-shaped acceleration / deceleration trajectory according to claim 1. In an articulated robot system having a plurality of joint axes,
The robot body,
A control device for controlling the robot body,
The control device
A reference joint axis determination processing unit that determines the movement time of each joint axis to the movement target position based on the movement target position and movement pattern of the robot, and determines the joint axis having the longest movement time as the reference joint axis;
A first S-shaped acceleration / deceleration trajectory generating unit that generates a provisional S-shaped acceleration / deceleration trajectory for the reference joint axis;
In a provisional S-curve acceleration / deceleration trajectory, a time tm1 at which the speed is maximum in a section where the acceleration is maximum is calculated, or a time tm2 at which the speed is maximum is calculated in a section where the acceleration is minimum. And
An arithmetic processing unit that calculates an estimated torque of each joint axis at time tm1 or time tm2 using a preset dynamic model;
Dynamics while maintaining the maximum acceleration arrival time from the start of acceleration for each joint axis until reaching the maximum acceleration, or the minimum acceleration arrival time from the start of deceleration to the arrival of the minimum acceleration. An angular acceleration change processing unit for obtaining a corrected maximum acceleration or a corrected minimum acceleration of the reference joint axis so that the estimated torque obtained using the model becomes a preset target torque;
Based on the corrected maximum acceleration or the corrected minimum acceleration of the reference joint axis thus obtained, the corrected S-shaped acceleration / deceleration trajectory corrected so that the S-shaped acceleration / deceleration trajectory of the reference joint axis has the corrected maximum acceleration or the corrected minimum acceleration. And a second S-shaped acceleration / deceleration trajectory generation unit for generating an articulated robot system.

Images