Control method of carrier robot

Abstract

To provide a control method capable of suppressing deviation of a target track of a hand on an end side in a belt drive type carrier robot.SOLUTION: A control method of a belt drive type carrier robot includes a first arm, a second arm, and a hand which are mutually rotatably supported and by which the hand is moved horizontally linearly. An angular velocity of the first arm comprises an acceleration section S1 where acceleration occurs when operation starts, a constant speed section S2 where constant speed continues in succession to the acceleration section S1, and a deceleration section S3 where deceleration occurs until operation completes, in succession to the constant speed section S2. At least either the acceleration section S1 or the deceleration zone S3 has: a first section S31 which accelerates at first acceleration or decelerates at first deceleration; and a second section S32 which decelerates at a second acceleration different from the first acceleration and decelerates at a second deceleration different from the first deceleration.SELECTED DRAWING: Figure 5

Description

[Technical field]

[0001] The present invention relates to a control method for a transport robot. [Background technology]

[0002] In the field of semiconductor manufacturing, a transport robot for transporting a workpiece is used when transporting a thin plate-like workpiece such as a wafer. As a transport robot for transporting such a thin plate-like workpiece, for example, a belt-driven horizontal articulated robot is known (see, for example, Patent Document 1). This transport robot includes a rear-stage arm having a drive shaft, a front-stage arm rotatably supported by the rear-stage arm, a hand rotatably supported by the front-stage arm, and a drive unit for driving these two arms and the hand. In this transport robot, a belt is used as a power transmission mechanism for rotating the two arms and the hand. The transport robot is configured such that when changing the posture of the two arms between a retracted posture and an extended posture, the hand that grips the workpiece moves along a linear target trajectory. On the other hand, there is a demand for high-speed transport of wafers and the like in the semiconductor manufacturing process, and the rotation drive of the arm is performed, for example, by trapezoidal acceleration / deceleration control.

[0003] In the semiconductor manufacturing process, the path of a workpiece such as a wafer held by a hand contains narrow areas such as gate valves and cassette entrances, so the hand that holds the workpiece must precisely follow the target trajectory. However, it is known that a hand moving at high speed deviates from the target trajectory that it should follow. It is known that the main cause of the deviation of the hand from the target trajectory is the expansion and contraction of the belt, but it is difficult to completely eliminate the deviation from the trajectory because it is difficult to measure and control the amount of expansion and contraction of the belt itself. By operating the arm and hand at a low speed, the load on the belt can be reduced, and the deviation of the hand from the trajectory can be reduced. However, in semiconductor manufacturing sites, it is required to reduce the deviation from the target trajectory without extending the operating time (takt time). [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2003-188231 A Summary of the Invention [Problem to be solved by the invention]

[0005] The present invention was devised under these circumstances, and its main objective is to provide a control method in a belt-driven transport robot that can suppress deviation of the tip hand from the target trajectory. [Means for solving the problem]

[0006] In order to solve the above problems, the present invention employs the following technical means.

[0007] A control method for a transport robot provided by the present invention is a control method for a belt-driven transport robot comprising a first arm rotatable about a first vertical axis on one end side, a second arm supported at one end side to be rotatable about a second vertical axis with respect to the other end side of the first arm, and a hand supported at one end side to be rotatable about a third vertical axis with respect to the other end side of the second arm, wherein when the first arm is rotated about the first vertical axis by linkage of belts arranged inside each of the first arm and the second arm, the hand is moved horizontally and linearly, the angular velocity of the first arm about the vertical axis when the hand is moved horizontally in a straight line between a retracted posture in which the first arm and the second arm are retracted and an extended posture in which the first arm and the second arm are extended includes an acceleration section in which acceleration occurs from the start of movement, a constant velocity section following the acceleration section in which a constant velocity continues, and a deceleration section following the constant velocity section in which deceleration occurs until the end of the movement, and at least one of the acceleration section and the deceleration section has a first section in which acceleration occurs at a first acceleration or deceleration occurs at a first deceleration, and a second section in which acceleration occurs at a second acceleration different from the first acceleration or deceleration occurs at a second deceleration different from the first deceleration.

[0008] In a preferred embodiment, the first section is continuous with the constant velocity section and the second section is continuous with the first section, and when the acceleration section has the first section and the second section, the first acceleration is smaller than the second acceleration, and when the deceleration section has the first section and the second section, the first deceleration is smaller than the second deceleration.

[0009] In a preferred embodiment, in an extension operation in which the first arm and the second arm are moved from the retracted position to the extended position, the deceleration section has the first section and the second section, and in a retraction operation in which the first arm and the second arm are moved from the extended position to the retracted position, the acceleration section has the first section and the second section.

[0010] In a preferred embodiment, the angular velocity of the first arm in the constant velocity section is smaller than the angular velocity in the constant velocity section in trapezoidal acceleration / deceleration control consisting of a constant acceleration section, a constant velocity section, and a constant deceleration section.

[0011] In a preferred embodiment, the operating time for rotating the first arm from the acceleration section to the deceleration section is substantially the same as the operating time for rotating the first arm from the constant acceleration section to the constant deceleration section in the trapezoidal acceleration / deceleration control, and the angular movement amount of the first arm from the acceleration section to the deceleration section is substantially the same as the angular movement amount of the first arm from the constant acceleration section to the constant deceleration section in the trapezoidal acceleration / deceleration control. Effect of the Invention

[0012] In a belt-driven transport robot, when the first arm, the second arm, and the hand are operated at high speed, this causes the tip hand to deviate from its trajectory. According to the control method for a transport robot of the present invention, by designing a target trajectory with attention paid to the expansion and contraction of the belt, which is the cause of the tip hand's trajectory deviation, it is possible to suppress deviation of the tip hand's target trajectory without installing sensors or the like for acquiring additional information.

[0013] Other features and advantages of the present invention will become more apparent from the following detailed description taken in conjunction with the accompanying drawings. [Brief description of the drawings]

[0014] [Figure 1] 1 is a schematic plan view showing an example of a wafer transport robot capable of implementing a transport robot control method according to the present invention, illustrating an example of a retracted posture in which a first arm and a second arm are retracted. [Diagram 2] 2 illustrates an example of an extended posture in which a first arm and a second arm of the transport robot illustrated in FIG. 1 are extended. [Diagram 3] 2 shows the reference postures of the first arm, the second arm, and the hand in the transfer robot shown in FIG. [Figure 4] This is a graph showing the change in values ​​of -2.4m21 -1 and m22 -1 relative to the robot's posture. [Diagram 5] 13 is a graph showing the angular velocity of a target joint angle trajectory in a straightening motion. [Figure 6] 13 is a graph showing angles of a joint angle target trajectory in a stretching motion. [Figure 7] 13 is a graph showing the angular velocity of a target joint angle trajectory in a contraction motion. [Figure 8] 13 is a graph showing the angle of a target joint angle trajectory in a contraction motion. [Figure 9] FIG. 13 is a diagram showing an evaluation of a joint angle target trajectory. [Figure 10] 13 is a graph showing a hand trajectory waveform in a stretching motion. [Figure 11] 13 is a graph showing a hand trajectory waveform during a contraction motion. [Figure 12] FIG. 13 is a diagram showing the maximum hand trajectory deviation of the tip side hand. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0015] Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the drawings.

[0016] Fig. 1 is a schematic plan view showing an example of a wafer transport robot capable of executing a control method for a transport robot according to the present invention. The wafer transport robot A1 shown in Fig. 1 includes a first arm 1, a second arm 2, a hand 3, and a base unit 9. The first arm 1 is configured as three links.

[0017] The first arm 1 is supported by the base 9 in a position in which the longitudinal direction is horizontal. One end of the first arm 1 can rotate around a first vertical axis O1, and a rotational driving force is transmitted from a driving source (e.g., a servo motor) arranged on the base 9. The second arm 2 is supported by the first arm 1 in a position in which the longitudinal direction is horizontal. One end of the second arm 2 can rotate around a second vertical axis O2 relative to the other end of the first arm 1. The hand 3 has a bifurcated fork-like tip and is supported by the second arm 2 in a horizontal position. The hand 3 is for placing and holding a thin plate-like workpiece of a predetermined size, such as a circular silicon wafer. Although detailed illustration is omitted, in the wafer transport robot A1, the base 9 supporting the first arm 1 has, for example, a rotating mechanism or a multi-joint mechanism, and the first arm 1 can be arranged in a predetermined position and direction.

[0018] The first arm 1, the second arm 2, and the hand 3 are configured as three links and are driven by a belt. Although not shown, a pair of pulleys having a first vertical axis O1 and a second vertical axis O2 as rotation axes and an endless belt wound around the pair of pulleys are arranged inside the first arm 1. Also, a pair of pulleys having a second vertical axis O2 and a third vertical axis O3 as rotation axes and an endless belt wound around the pair of pulleys are arranged inside the second arm 2. Then, for example, a servo motor (drive source) outputs torque according to a command torque given from a controller, and rotates the pulley of the first arm 1 having the first vertical axis O1 as its rotation axis. The rotational motion of the pulley is transmitted to the belt, and the first arm 1 rotates around the first vertical axis O1, the second arm 2 rotates around the second vertical axis O2, and the hand 3 rotates around the third vertical axis O3. In this manner, the first arm 1, the second arm 2, and the hand 3 operate. Here, the rotation angle of the first arm 1 about the first vertical axis O1, the rotation angle of the second arm 2 about the second vertical axis O2, and the rotation angle of the hand 3 about the third vertical axis O3 are designed to be 1:2:1, as described below. With the first arm 1, the second arm 2, and the hand 3 configured in this manner, only the horizontal transport operation is performed, and the first arm 1, the second arm 2, and the hand 3 are linked to each other, so that the hand 3 is moved horizontally in a straight line. The servo motor can only measure the rotation angle and rotation angular velocity of the first arm 1, and sensors and the like for measuring other information cannot be arranged in the wafer transport robot A1 due to cost maintenance, mechanical constraints, the operating environment of the robot, and other constraints.

[0019] When the first arm 1, the second arm 2, and the hand 3 are operated at a sufficiently low speed, it can be considered that the hand 3 on the tip side extends or retracts on a horizontal linear trajectory (on the y-axis in Figs. 1 and 2 described later). Deviation from the horizontal linear trajectory, that is, when the x-coordinate (the coordinate of the x-axis, which is the horizontal axis perpendicular to the y-axis) of the hand 3 on the tip side takes a value other than 0, is called trajectory deviation. In this case, when the first arm 1, the second arm 2, and the hand 3 are operated at high speed, it causes the hand 3 on the tip side to deviate from the trajectory. The inventors of the present invention focused on "stretching and contraction of the belt," which is one of the causes of the trajectory deviation, and completed the present invention. From the kinematics of the robot, it can be considered that "the belt is stretched ⇒ a joint angle error occurs," and that "a joint angle error occurs ⇒ deviation from the trajectory of the hand 3 on the tip side occurs." In the present invention, attention was paid to the joint angle error caused by the stretching and contraction of the belt. By formulating, modeling and analyzing the robot's behavior as an equation of motion and by conducting preliminary experiments with an actual robot, the following two points became clear. (1) The degree to which the belt stretches (the amount of joint angle error caused primarily by belt stretch) varies depending on the robot's posture. (2) The degree to which joint angle errors affect trajectory deviation also differs depending on the posture. Taking the above two points into consideration, we proposed an evaluation index formula that combines "the likelihood of joint angle errors occurring" with "the difference in the degree of influence." In designing the target trajectory, we found a method to design a target trajectory such that when the posture has a small evaluation index value, it is assumed that the posture is unlikely to deviate from the trajectory and is operated with a large angular acceleration, and when the posture has a large evaluation index value, it is assumed that the posture is likely to deviate from the trajectory and is operated with the smallest possible angular acceleration. The method of the present invention can reduce the amount of trajectory deviation while maintaining the takt time.

[0020] A method for controlling the transport robot of this embodiment will be described below.

[0021] <Robot coordinate system> The robot coordinate system will be described with reference to FIG. 1 and FIG. 2. In FIG. 1 and FIG. 2, the direction as seen from the back of the robot is the y-axis direction, and the horizontal direction perpendicular to the y-axis direction is the x-axis direction. The driving operation of the wafer transport robot A1, for example, repeats contraction and extension of the first arm 1 and the second arm 2 (hereinafter, both arms 1 and 2 are collectively referred to as "arms" as appropriate) between an arbitrary contracted position (hereinafter, referred to as "contracted position" as appropriate) and an arbitrary extended position (hereinafter, referred to as "extended position" as appropriate). FIG. 1 shows an example of the contracted position in which the first arm 1 and the second arm 2 are contracted. FIG. 2 shows an example of the extended position in which the first arm 1 and the second arm 2 are extended. The horizontal transport mechanism consisting of the first arm 1, the second arm 2, and the hand 3 constitutes three links, and here, for convenience of explanation, the first arm 1 will be described as link 1, the second arm 2 as link 2, and the hand 3 as link 3 as appropriate. The reference posture is the posture in which links 1 and 2 are overlapped on the x-axis, and link 3 is overlapped toward the positive direction of the y-axis. Figure 3 shows the reference postures of links 1 to 3 in the wafer transport robot A1. The angular displacement of link i (i is any one of 1 to 3) from the reference posture is expressed as the joint angle θ of each link. i [rad]. θ in the reference position i is 0 [rad], and the counterclockwise direction is the positive direction. The "retracted posture" described above with reference to FIG. 1 is a posture that is θ i The “stretched posture” described above with reference to FIG. 2 is a posture in which θ i is large (i.e., positive). i is designed to satisfy the following geometric constraints when the belt does not stretch.

[0022]

number

[0023] In the drive control of the wafer transport robot A1, a torque of θ1 can be freely applied. Positioning control that controls the angle of θ1, including PID control rules, is incorporated as a control means for applying this torque of θ1. The time series data of the target value applied to this positioning control is called the target trajectory of the joint angle θ1 (hereinafter, appropriately, "target joint angle trajectory").

[0024] However, in reality, an angle error occurs between the above geometric constraints and the actual joint angle due to the expansion and contraction of the belt. The joint angle error between links 1 and 2 is θ e1 [rad], the joint angle error between links 2 and 3 is θ e2 If we define the joint angle as [rad], the actual joint angle can be expressed as follows using the above geometric constraints:

[0025]

number

[0026] The state variable is q=[θ1θ e1 θ e ] T Then, the coordinates (x, y) of the robot's end effector (hand 3) are expressed by the following forward kinematics definition: (f1(q), f2(q)) = (x, y).

[0027]

number

[0028] Since the product is designed so that l1 = l2 is satisfied, θ e1 =θ e2 If θ = 0, the position of the robot's hand (hand 3) moves along the y-axis as shown in Figures 1 and 2. e1 ,θ e2When ≠0, deviation of the hand position trajectory in the x-axis direction (hereinafter referred to as "hand trajectory deviation") may occur. Note that l1, l2, and l3 are the lengths of link 1, link 2, and link 3, respectively, and refer to the length from the first vertical axis O1 to the second vertical axis O2, the length from the second vertical axis O2 to the third vertical axis O3, and the length from the third vertical axis O3 to the hand in Fig. 1.

[0029] <Factor 1 of deviation of hand trajectory (stretching of belt)> Joint angle error θ caused by belt stretching e1 ,θ e2 When θ is not 0, the geometric constraint is broken and the hand 3 at the tip side deviates from the trajectory (hand tip trajectory deviation). e1 ,θ e2 are 10 -5 ,10 -10 Therefore, the main cause of deviation of the hand trajectory is θ e1 and θ e2 is θ e1 It can be said that it is sufficiently small compared to. From the equation of motion, the joint angle error θ e1 The equation for angular acceleration is approximated as follows:

[0030]

number

[0031] However, m ij -1 is the inverse inertia matrix M obtained from the equation of motion -1 There are (i,j) elements in (q), and the subscripts 1,2 on the right hand side of number 4 mean taking the first and second elements of the vector.

[0032] Figure 4 shows -2.4m 21 -1 and m 22 -1and the robot joint angle θ1. It can be seen that similar values ​​are taken at many points. There is a difference between the two values ​​when θ1 is around -1[rad] and 1[rad], but from the perspective of the likelihood of the hand trajectory deviation occurring, it can be evaluated as shown in the following equation 5.

[0033]

number

[0034] As a result, "m 22 -1 The larger (q), the greater the joint angle error θ e1 The joint angle error θ e1 It can be seen that the likelihood of occurrence can be estimated.

[0035] <Factor 2 of deviation from hand trajectory (degree of influence of factor 1, belt stretch, on deviation from hand trajectory)> The forward kinematics showing the relationship between the joint angles and the coordinates of the hand (hand 3) is represented as f(q), the x-coordinate component of the forward kinematics is represented as f1, and the y-coordinate component is represented as f2. In this case, the position of the hand (hand 3) at a given time t is represented as f(q(t)), and its x-component is also represented as f1(q(t)). By differentiating both sides of x(t) = f1(q(t)) twice with respect to time t, the following equation 6 is obtained.

[0036]

number

[0037] It is known from numerical experiments that the first term on the right hand side of the acceleration equation (1) in Equation 6 is sufficiently small. e1 ,θ e2 is 10 -5 It is known that the following values ​​are taken, and since l1 = l2 is satisfied structurally, the following approximation formula (7) holds true.

[0038]

number

[0039] As a result, the acceleration equation (1) of Equation 6 that the expansion and contraction of the belt exerts on the x-axis, which is the direction in which the hand trajectory deviates, is approximated as Equation 8 below.

[0040]

number

[0041] Therefore, in the present invention, the value of the element of the following equation 9 obtained from the robot's equation of motion is used as an evaluation index in designing a joint angle target trajectory, and it has been discovered that a joint angle target trajectory can be designed with two or more stages of trapezoidal acceleration / deceleration such that large acceleration is achieved at a joint angle θ1 where the evaluation index value is small.

[0042]

number

[0043] Next, a procedure for designing a joint angle target trajectory in the control method for the transport robot in this embodiment will be described. From preliminary experiments and trends in evaluation indices, it has been found that the influence of changes in design parameters on deviation of the hand trajectory is greatest in the following order: (1) Maximum angular velocity during constant speed motion (2) Angular acceleration with the arm extended (3) Angular acceleration when the arm is retracted Therefore, the joint angle target trajectory according to the present embodiment is designed by modifying the joint angle target trajectory based on the conventional trapezoidal acceleration / deceleration control in consideration of the above factors (1) to (3) while referring to the formula (9) of the evaluation index having the graph shape shown in FIG. Specifically, the following procedure was used to design the target joint angle trajectory. Step (1) By lowering the maximum angular velocity during the constant-velocity motion, the maximum deviation of the hand trajectory is reduced. However, this operation increases the takt time. Step (2) Design a trapezoidal acceleration / deceleration trajectory with two or more stages by lowering the angular acceleration when the arm is in an extended position, and make adjustments to reduce the maximum deviation of the hand trajectory. Step (3): Increase the angular acceleration with the arm retracted, and shorten the takt time to an acceptable level.

[0044] The robot's hand (hand 3) trajectory is measured for each of the joint angle target trajectory designed according to this embodiment and the joint angle target trajectory based on trapezoidal acceleration / deceleration of the conventional method, and the maximum hand trajectory deviation is compared. The comparative experiment is performed using a product (UTM-R3700F manufactured by Daihen Corporation) that is actually used at the manufacturing site. The experimental operation is performed by setting the joint angle θ1 to the minimum joint angle θ of the robot. 1s =-1.037[rad] to maximum joint angle θ 1f = 1.334[rad] and the maximum joint angle θ 1f From the minimum joint angle θ 1s The trajectory tracking control law uses the PI control law built into the robot as is, and the gains are also common. The conventional method, which is the subject of comparison, uses angular acceleration α = 15.31 [rad / s 2 ], and the maximum angular velocity was set to 95% of the mechanical limit (see number 10 below).

[0045]

number

[0046] The joint angle target trajectory designed by this embodiment is designed by modifying the joint angle target trajectory designed by the conventional method. When designing the joint angle target trajectory by this embodiment, it is designed to be within ±0.01 [sec] of the tact time of 0.845 [sec] of the joint angle target trajectory by the conventional trapezoidal acceleration / deceleration, and it is evaluated whether the deviation amount of the hand trajectory can be reduced while maintaining the tact time.

[0047] <Design of joint angle target trajectory according to this embodiment> (1) Changing the maximum angular velocity during constant speed motion The maximum angular velocity for the extension and retraction movements is determined by considering both the increase in takt time due to the reduction in the maximum angular velocity and the reduction in deviation from the hand trajectory. In the conventional target trajectory using trapezoidal acceleration / deceleration, the maximum angular velocity is set to 95% of the machine limit, whereas in this embodiment, the maximum angular velocity is set to 85% of the machine limit for the two movements (extension and retraction). (2) Change in angular acceleration near the extended position In the stretching operation, the maximum angular velocity is 2α / 3 [rad / s] from θ1 = 0.659 [rad] to θ1 = 1.309 [rad]. 2 ] and then decelerates slowly to the maximum joint angle θ 1f Until it reaches 1.65α [rad / s 2 ] was set to decelerate significantly. Also, in the contraction motion, the maximum joint angle θ 1f From θ1=1.294[rad] to 1.65α[rad / s 2 ] and then accelerates at α / 2[rad / s] until it reaches the maximum angular velocity θ1=0.470[rad]. 2 ] is set to accelerate slowly. (3) Changing the angular acceleration near the contracted position In order to shorten the tact time that has increased due to the above factors (1) and (2), the actuator is operated with a large angular acceleration near the contraction position. 2 ], 1.65α[rad / s 2 ] was set to accelerate with a large angular acceleration.

[0048] Fig. 5 to Fig. 8 show the joint angle target trajectories set as described above. Fig. 5 shows the angular velocity of the joint angle target trajectory in an extension motion. Fig. 6 shows the joint angle of the joint angle target trajectory in an extension motion. Fig. 7 shows the angular velocity of the joint angle target trajectory in a contraction motion. Fig. 8 shows the joint angle of the joint angle target trajectory in a contraction motion.

[0049] <Evaluation of joint angle target trajectory> We will check to what extent the joint angle target trajectory created in this embodiment takes the evaluation index into consideration. In the evaluation of the joint angle target trajectory, the number of steps in the acceleration / deceleration region of the joint angle target trajectory is n, the time of the i-th step is ti, and the following evaluation formula 11 is used.

[0050]

number

[0051] Figure 9 shows the evaluation values ​​calculated using the evaluation formula for the joint angle target trajectory based on conventional trapezoidal acceleration / deceleration control and the joint angle target trajectory created in this embodiment. It can be seen that the evaluation values ​​are smaller in this embodiment than in the joint angle target trajectory based on conventional trapezoidal acceleration / deceleration control, in both the arm extended posture and the retracted posture. This confirms that a design index that accelerates and decelerates greatly can be applied in postures where deviation from the hand trajectory is unlikely to occur, with a small evaluation index.

[0052] <Evaluation based on maximum deviation of hand trajectory of tip end hand 3> Figures 10 and 11 are graphs of the hand trajectory waveform of the tip hand 3 when the joint angle target trajectory according to the conventional method and the joint angle target trajectory according to this embodiment are used, measured using a transfer robot (UTM-R3700F manufactured by Daihen Corporation). Figure 10 shows the hand trajectory waveform in an extension motion. Figure 11 shows the hand trajectory waveform in a retraction motion. Figure 12 also shows the maximum hand trajectory deviation of the tip hand 3 in an extension motion and a retraction motion. The measured values ​​shown in Figure 12 show that a reduction in hand trajectory deviation of approximately 18% was achieved in an extension motion and approximately 29% in a retraction motion.

[0053] Next, the operation of this embodiment will be described with reference to FIGS.

[0054] Fig. 5 shows the angular velocity of the joint angle target trajectory in the control method of this embodiment and the trapezoidal acceleration / deceleration control of the conventional method in a stretching motion. In Fig. 5, the conventional trapezoidal acceleration / deceleration control consists of a constant acceleration section S7, a constant velocity section S8, and a constant deceleration section S9. The constant acceleration section S7 is a section in which acceleration occurs at a constant acceleration from the start of the motion. The constant velocity section S8 is a section in which a constant velocity continues. The constant deceleration section S9 is a section in which deceleration occurs at a constant deceleration until the end of the motion.

[0055] The control method of this embodiment includes an acceleration section S1, a constant speed section S2, and a deceleration section S3. The acceleration section S1 is a section in which acceleration occurs from the start of the movement. The constant speed section S2 follows the acceleration section S1 and is a section in which a constant speed continues. The deceleration section S3 follows the constant speed section S2 and is a section in which deceleration occurs until the end of the movement.

[0056] The angular velocity in the constant velocity section S2 (angular velocity of the first arm 1) is smaller than the angular velocity in the constant velocity section S8. In the design of the joint angle target trajectory of this embodiment, the angular velocity in the constant velocity section S8 is 95% of the mechanical limit, whereas the angular velocity in the constant velocity section S2 is 85% of the mechanical limit. Therefore, the angular velocity in the constant velocity section S2 is approximately 89% of the angular velocity in the constant velocity section S8, which is a decrease of approximately 11%.

[0057] In the extension motion shown in FIG. 5, the deceleration section S3 has a first section S31 and a second section S32. In the first section S31, deceleration occurs at a first deceleration, and in the second section S32, deceleration occurs at a second deceleration different from that of the first section S31. The first section S31 is continuous with the constant velocity section S2, and the second section S32 is continuous with the first section S31. The first deceleration in the first section S31 is smaller than the second deceleration in the second section S32. In designing the joint angle target trajectory of this embodiment, the second deceleration in the second section S32 is 1.65α [rad / s 2 ], and the first deceleration in the first section S31 is smaller than the second acceleration, 2α / 3 [rad / s 2 ].

[0058] The tact time (operation time from the start of the acceleration section to the end of the deceleration section) in this embodiment (extension operation) is set, as described above, to be within ±0.01 [sec] of the 0.845 [sec] of conventional trapezoidal acceleration / deceleration control, and is set substantially the same as the conventional method.

[0059] In addition, in both the present embodiment (stretching motion) and the trapezoidal acceleration / deceleration control, the minimum joint angle θ 1s and the maximum joint angle θ 1f Therefore, the change amount of the joint angle θ1 (angle movement amount of the first arm 1) from the acceleration section S1 to the deceleration section S3 in this embodiment (extending operation) is the same as the change amount of the joint angle θ1 (angle movement amount of the first arm 1) from the constant acceleration section S7 to the constant deceleration section S9 in the trapezoidal acceleration / deceleration control of the conventional method.

[0060] Fig. 7 shows the angular velocity of the joint angle target trajectory in the control method of this embodiment and the trapezoidal acceleration / deceleration control of the conventional method in a contraction motion. In Fig. 7, the conventional trapezoidal acceleration / deceleration control consists of a constant acceleration section S7, a constant velocity section S8, and a constant deceleration section S9. The constant acceleration section S7 is a section in which acceleration occurs at a constant acceleration from the start of the motion. The constant velocity section S8 is a section in which a constant velocity continues. The constant deceleration section S9 is a section in which deceleration occurs at a constant deceleration until the end of the motion.

[0061] The control method of this embodiment includes an acceleration section S1, a constant speed section S2, and a deceleration section S3. The acceleration section S1 is a section in which acceleration occurs from the start of the movement. The constant speed section S2 follows the acceleration section S1 and is a section in which a constant speed continues. The deceleration section S3 follows the constant speed section S2 and is a section in which deceleration occurs until the end of the movement.

[0062] The angular velocity in the constant velocity section S2 (angular velocity of the first arm 1) is smaller than the angular velocity in the constant velocity section S8. In the design of the joint angle target trajectory of this embodiment, the angular velocity in the constant velocity section S8 is 95% of the mechanical limit, whereas the angular velocity in the constant velocity section S2 is 85% of the mechanical limit. Therefore, the angular velocity in the constant velocity section S2 is approximately 89% of the angular velocity in the constant velocity section S8, which is a decrease of approximately 11%.

[0063] In the contraction motion shown in FIG. 6, the acceleration section S1 has a first section S11 and a second section S12. In the first section S11, acceleration occurs at a first acceleration, and in the second section S12, acceleration occurs at a second acceleration different from that of the first section S11. The first section S11 is continuous with the constant velocity section S2, and the second section S12 is continuous with the first section S11. The first acceleration in the first section S11 is smaller than the second acceleration in the second section S12. In designing the joint angle target trajectory of this embodiment, the second acceleration in the second section S12 is 1.65α [rad / s 2 ], and the first acceleration in the first section S11 is smaller than the second acceleration, α / 2 [rad / s 2 ].

[0064] The tact time (operation time from the start of the acceleration section to the end of the deceleration section) in this embodiment (shrinking operation) is set, as described above, to be within ±0.01 [sec] of the 0.845 [sec] of the conventional trapezoidal acceleration / deceleration control, and is set substantially the same as the conventional method.

[0065] In addition, in both the present embodiment (contraction operation) and the trapezoidal acceleration / deceleration control, the minimum joint angle θ 1s and the maximum joint angle θ 1f Therefore, the amount of change in the joint angle θ1 (the amount of angular movement of the first arm 1) from the acceleration section S1 to the deceleration section S3 in this embodiment (contraction operation) is the same as the amount of change in the joint angle θ1 (the amount of angular movement of the first arm 1) from the constant acceleration section S7 to the constant deceleration section S9 in the trapezoidal acceleration / deceleration control of the conventional method.

[0066] In the wafer transport robot A1, when the first arm 1, the second arm 2, and the hand 3 are operated at high speed, deviation of the hand trajectory of the tip side hand 3 is caused. According to the control method for the transport robot of this embodiment, deviation of the hand target trajectory of the tip side hand 3 can be suppressed without installing sensors or the like for acquiring additional information by designing the joint angle target trajectory focusing on the expansion and contraction of the belt, which is the cause of deviation of the hand trajectory of the tip side hand. Furthermore, according to the above control method of this embodiment, it is possible to improve the tracking accuracy of the hand target trajectory without extending the operation time (takt time).

[0067] Although an embodiment of the present invention has been described above, the scope of the present invention is not limited to the above-described embodiment, and all modifications within the scope of the matters described in each claim are all included in the scope of the present invention.

[0068] In the control method described in the above embodiment, the deceleration section S3 in the arm extension operation is configured in two stages (having a first section S31 and a second section), and the acceleration section S1 in the arm retraction operation is configured in two stages (having a first section S11 and a second section), but this is not limited to this. In both the extension operation and the retraction operation, both the acceleration section and the deceleration section may be configured to accelerate and decelerate in two stages (having a first section and a second section). Also, the acceleration section or the deceleration section may be configured to accelerate and decelerate in three or more stages.

[0069] The transport robot on which the control method of the present invention is performed is not limited to the robot exemplified in the above embodiment, and the control method of the present invention can be applied to various belt-driven horizontal articulated robots. [Explanation of symbols]

[0070] A1: wafer transport robot, 1: first arm, 2: second arm, 3: hand, O1: first vertical axis, O2: second vertical axis, O3: third vertical axis, S1: acceleration section, S11: first section, S12: second section, S2: constant speed section, S3: deceleration section, S31: first section, S32: second section, S7: constant acceleration section, S8: constant speed section, S9: constant deceleration section

Claims

1. a first arm rotatable about a first vertical axis on one end side, a second arm supported at one end side to be rotatable about a second vertical axis with respect to the other end side of the first arm, and a hand supported at one end side to be rotatable about a third vertical axis with respect to the other end side of the second arm, A method for controlling a belt-driven transfer robot in which the hand is moved horizontally and linearly when the first arm is rotated about the first vertical axis by linkage of belts arranged inside the first arm and the second arm, the method comprising: an angular velocity of the first arm about the vertical axis when the hand is moved horizontally and linearly between a retracted posture in which the first arm and the second arm are retracted and an extended posture in which the first arm and the second arm are extended includes an acceleration section in which acceleration occurs from a start of an operation, a constant velocity section following the acceleration section and in which a constant velocity continues, and a deceleration section following the constant velocity section and in which deceleration occurs until the end of the operation, A control method for a transport robot, wherein at least one of the acceleration section and the deceleration section has a first section in which the transport robot accelerates at a first acceleration or decelerates at a first deceleration, and a second section in which the transport robot accelerates at a second acceleration different from the first acceleration or decelerates at a second deceleration different from the first deceleration.

2. the first section is continuous with the constant velocity section, and the second section is continuous with the first section, When the acceleration section has the first section and the second section, the first acceleration is smaller than the second acceleration, The method for controlling a transfer robot according to claim 1 , wherein when the deceleration section has the first section and the second section, the first deceleration is smaller than the second deceleration.

3. In an extension operation in which the first arm and the second arm are moved from the contracted position to the extended position, the deceleration section has the first section and the second section, 3. The method of controlling a transfer robot according to claim 2, wherein in a retracting operation for moving the first arm and the second arm from the extended posture to the retracted posture, the acceleration section has the first section and the second section.

4. 4. The control method for a transport robot according to claim 1, wherein the angular velocity of the first arm in the constant velocity section is smaller than the angular velocity in the constant velocity section in a trapezoidal acceleration / deceleration control consisting of a constant acceleration section, a constant velocity section, and a constant deceleration section.

5. an operation time for rotating the first arm from the acceleration section to the deceleration section is substantially the same as an operation time for rotating the first arm from the constant acceleration section to the constant deceleration section in the trapezoidal acceleration / deceleration control, 5. The control method for a transport robot according to claim 4, wherein an angular movement amount of the first arm from the acceleration section to the deceleration section is substantially the same as an angular movement amount of the first arm from the constant acceleration section to the constant deceleration section in the trapezoidal acceleration / deceleration control.

Images