WO1994006602A1 - Robot - Google Patents

Robot Download PDF

Info

Publication number
WO1994006602A1
WO1994006602A1 PCT/AU1993/000481 AU9300481W WO9406602A1 WO 1994006602 A1 WO1994006602 A1 WO 1994006602A1 AU 9300481 W AU9300481 W AU 9300481W WO 9406602 A1 WO9406602 A1 WO 9406602A1
Authority
WO
WIPO (PCT)
Prior art keywords
robot
leg
servo
define
channel
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/AU1993/000481
Other languages
English (en)
Inventor
Lin Guo
Kevin John Rogers
Robin John Kirkham
Maciej Wlodzimierz Rogozinski
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Commonwealth Scientific and Industrial Research Organization CSIRO
Original Assignee
Commonwealth Scientific and Industrial Research Organization CSIRO
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Commonwealth Scientific and Industrial Research Organization CSIRO filed Critical Commonwealth Scientific and Industrial Research Organization CSIRO
Priority to AU48111/93A priority Critical patent/AU4811193A/en
Publication of WO1994006602A1 publication Critical patent/WO1994006602A1/fr
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B62—LAND VEHICLES FOR TRAVELLING OTHERWISE THAN ON RAILS
    • B62D—MOTOR VEHICLES; TRAILERS
    • B62D57/00—Vehicles characterised by having other propulsion or other ground- engaging means than wheels or endless track, alone or in addition to wheels or endless track
    • B62D57/02—Vehicles characterised by having other propulsion or other ground- engaging means than wheels or endless track, alone or in addition to wheels or endless track with ground-engaging propulsion means, e.g. walking members
    • B62D57/024—Vehicles characterised by having other propulsion or other ground- engaging means than wheels or endless track, alone or in addition to wheels or endless track with ground-engaging propulsion means, e.g. walking members specially adapted for moving on inclined or vertical surfaces
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B23—MACHINE TOOLS; METAL-WORKING NOT OTHERWISE PROVIDED FOR
    • B23Q—DETAILS, COMPONENTS, OR ACCESSORIES FOR MACHINE TOOLS, e.g. ARRANGEMENTS FOR COPYING OR CONTROLLING; MACHINE TOOLS IN GENERAL CHARACTERISED BY THE CONSTRUCTION OF PARTICULAR DETAILS OR COMPONENTS; COMBINATIONS OR ASSOCIATIONS OF METAL-WORKING MACHINES, NOT DIRECTED TO A PARTICULAR RESULT
    • B23Q5/00—Driving or feeding mechanisms; Control arrangements therefor
    • B23Q5/22—Feeding members carrying tools or work
    • B23Q5/34—Feeding other members supporting tools or work, e.g. saddles, tool-slides, through mechanical transmission
    • B23Q5/50—Feeding other members supporting tools or work, e.g. saddles, tool-slides, through mechanical transmission feeding step-by-step

Definitions

  • the present invention relates to a robot for industrial use.
  • Most robots used in the manufacturing industry have a fixed base and carry a tool which the robot moves relative to the base in order to perform work on a particular part.
  • Positional parameters for movement of the tool to perform work on the part are specified with respect to the base of the robot and the robot is preprogrammed to move the tool in accordance with the parameters.
  • Parts are normally transported to the robot and accurately positioned relative to the robot base in a specially designed fixture. Robots of this type have proved to be adequate for processing small and medium sized parts in relatively high numbers but become impractical and expensive as the size of the part increases.
  • a robot often needs to operate over most of the surface of a part and as the size increases, inevitably the length of the robot's links increase, which increases the complexity and cost of the robot. Also this makes it more difficult to precisely manipulate the tool and the reduced stiffness of the robot can introduce inaccuracies in positioning of the tool.
  • Robots which have been developed to walk fall within one of three categories.
  • the first includes robots with several legs, each with multiple joints. They are complex and expensive and are used primarily for scientific, experimental or entertainment purposes, and are not practical or economic for industrial use.
  • the second category comprises dynamic legged robots which are able to run and hop on a surface and rely on the dynamic exchange of energy between energy storage devices in the robot, such as springs, and the kinetic or potential energy of the robot to move. Robots of this second category have not yet been applied industrially and their motion is relatively erratic and jerky.
  • the third category includes static legged robots which generally have only two legs, one leg fixed to a frame which carries the tool or a payload, and the second leg moveable relative to the frame. The robots of this type move across a part in a discontinuous or stop-start manner. The attitude of the frame with respect to the surface of the part also changes discontinuously as the robot moves across a curved surface. Discontinuous motion is undesirable for many tasks.
  • a robot comprising a body, and two legs without active joints, said legs being mounted to said body such that the legs can be independently translated and rotated with respect to said body.
  • said legs may each include at least one foot.
  • the feet are adapted to retract and extend relative to the respective legs to engage and disengage a surface.
  • the robot may further include abutting means which extend from and set the attitude of said body, wherein said feet on engaging said surface are adapted to retract to ensure the ends of said abutting means abut said surface.
  • the ends of said abutting means are configured to move over said surface and the motion of the legs is such that, at any given time during movement of said robot over said surface, said at least one foot of at least one of the legs engages said surface and said ends substantially continuously abut said surface. Retraction of the engaging feet is such that contact is maintained between the surface and said ends.
  • said feet each include a surface engaging shoe having means for adhering to said surface.
  • said shoe is pivotal within a body of the foot to facilitate tangential engagement of said shoe with said surface.
  • the robot may include a tool mounted on said body and which is moveable relative to said body.
  • the robot includes means for providing position information to said robot, and means for controlling said moveable legs in response to said information.
  • controlling means further controls movements of said tool relative to said body in response to the position information.
  • said information providing means comprises means for recognising a feature of the surface and determining said position information for said feature relative to said robot, so as to cause said robot to track said feature.
  • controlling means includes a pair of motors for each moveable leg which act on a shaft of the respective moveable leg, wherein uniform control of said motors effects translational movement of said shaft and differential control of said motors effects rotational movement of said shaft relative to said body.
  • the feet further includes ram means for retracting and extending said shoes of said feet.
  • said robot is able to move across said surface in a continuous motion.
  • the robot is also able to maintain a substantially continuous attitude for said body with respect to said surface, even when said surface is curved.
  • Figure 1 is a diagram of the basic components of a preferred embodiment of a robot
  • Figure 2 is a schematic diagram illustrating six phases of movement of the robot in a straight line
  • Figure 3 are graphs of the velocities of parts of the robot during the six phases of movement
  • Figure 4 is a perspective view of mechanical components of the robot
  • Figure 5A is a cross-sectional view of the toe of the robot in a retracted position
  • Figure 5B is a cross-sectional view of the toe of the robot in an extended position
  • Figure 5C is a cross-sectional view of the toe of the robot in an extended position with a shoe of the toe pivoted;
  • Figure 6 is a pneumatic circuit diagram of a pneumatic circuit for control of the toes of the robot
  • Figure 7 is a plan of a frame and toe pattern for the robot
  • Figure 8 are plans of three alternative toe arrangements
  • Figure 9 is a graph of robot mobility envelopes for alternative toe arrangements
  • Figure 10 is a graph of robot mobility envelopes for alternative shaft separations of the robot
  • Figure 11 is a perspective view of a leg arrangement for the robot
  • Figure 12 is a plan of a preferred leg arrangement
  • Figure 13 is a block diagram of the modules of a preferred control system of the robot.
  • Figure 14 is a block diagram of a motion control system of the robot
  • Figure 15 is a diagram of a robot coordinate system for the control system
  • Figure 16 is a diagram of movement of tracking and directional points in the robot coordinate system
  • Figure 17 is a diagram of a tracking point path and a predicted tool point path traversed during tracking of a feature by the robot;
  • Figure 18 is a diagram of joint trajectory and setpoint points for a smooth joint trajectory of the robot; and Figure 19 is a block diagram of the electrical components of the preferred control system of the robot.
  • a robot 2 as shown in Figure 1, includes a rectangular frame 4, and first and second legs 6 and 8.
  • the legs 6 and 8 each include three respective toes 10.
  • the toes 10 are disposed underneath the frame 4 and for each leg 6 and 8 are positioned to define points of a triangle parallel to the plane of the frame 4.
  • the base of the triangle formed by the toes 10 of the first leg extends along one side 12 of the frame 4 and the base of the triangle formed by the toes 10 of the second leg 8 extends along the opposite side 14 of the frame 4.
  • the toe 10 which forms the apex of each triangle is disposed in the middle of the first side 12 for the second leg 8 and in the middle of the other side 14 for the first leg 6.
  • the centroid of all of the toes 10 relative to a plan view plane of the frame 4 is approximately the centre 16 of the frame 4, and will move as the legs 6 and 8 move.
  • the toes 10 are used to engage the surface 18 of a part over which the robot 2 moves, and the frame 4 is positioned over the surface 18 and the toes 10.
  • the toes 10 of each leg 6 and 8 are connected by respective struts 20 which extend between the toes 10 of each leg 6 and 8, as shown in Figure 1.
  • the struts 20 of the first leg 6 are disposed below the struts 20 of the second leg 8.
  • the struts 20 for a leg 6 or 8 meet at a point near the toe 10 which forms the apex of the triangle defined by the toes 10 of the leg.
  • the legs 6 and 8 each include a shaft 22 which extends upwardly from the intersecting point of the struts 20.
  • the shafts 22 extend up through respective longitudinal slots 24 in the frame 4, each slot 24 being disposed on respective sides of and equidistant from the centre 16 of the frame 4. Movement of the frame 4 and legs 6 and 8 relative to one another is controlled by equipment, described hereinafter, mounted on the frame 4 which acts on the shafts 22. Movement of the shafts 22 is constrained by the equipment to rotation about their respective axis and translatio ⁇ al movement in one plane along the respective slot 24.
  • the legs 6 and 8 therefore have two degrees of freedom.
  • the shafts are rotated and translated independently so as to move the legs 6 and 8 and the robot 2 across the surface 18, as desired.
  • the shafts 22 are not raised and lowered to engage or disengage the surface 18 but instead the toes 10 of a leg 6 or 8 are retracted or extended.
  • the robot 2 further includes four legs 28 fixed to each corner of the frame and which each extend down from the frame 4.
  • the lower ends 30 of the legs 28 include a wheel or a roller 32 which abuts and moves across the surface 18.
  • the legs 28 define the attitude of the frame 4 with respect to the surface 18.
  • the toes 10 of a leg 6 or 8 are retracted so as to maintain contact between the surface 18 and the rollers 32 of the legs 28.
  • the first and second legs 6 and 8 both engage the surface 18, and the motors for each leg rotate at W 1 to drive the frame forward at velocity V r with respect to the surface 18.
  • the frame 4 can only have straight line motion when both legs 6 and 8 engage the surface 18 as angular motion is not possible because the centres of both shafts 22 are not coincident.
  • the frame 4 continues to be driven forward at V r by the motors of the second leg 8.
  • the first leg 6 is driven forward at a high velocity V reset > V r until it reaches a reset position in front of said second leg and its motors are stopped.
  • the frame continues to be driven forward at V r by the motors of the leg 8 and the first leg 6 moves at V r with the frame 4.
  • the motors of the first leg 6 are driven so the first leg 6 is made stationary with respect to the surface 18.
  • the motors of the second leg 8 continue to drive the frame forward at the velocity V r .
  • the toes 10 of the first leg are extended to engage the surface 18 in front of the toes 10 of the second leg 8.
  • One step of the walking cycle is completed and the six phases are repeated for the next step with the roles of the first leg 6 and the second leg 8 being reversed. Any number of successive steps can be executed to move the robot over a part.
  • the shafts 22 of the robot 2, as shown in Figure 4, are each held within two aligned cylindrical bearings 60 which extend from the side of a first nut 62.
  • the nuts 62 are mounted on tracks 64 of the frame 4 which extend longitudinally with respect to the frame.
  • One track 64 for each nut 62 may be used as illustrated, or alternatively a pair of tracks 64 can be used for each nut 62 to enhance the mechanical stiffness of the structure.
  • the tracks 64 effectively define the two slots 24.
  • the nuts 62 are each driven along the respective track 64 by a first ball screw 66 which lies parallel with the track 64 and is driven by a first servo motor 68.
  • a second servo motor 70 is mounted on the outside of the first servo motor 68 for each of the shafts 22, and the second servo motor 70 drives a second ball screw 72, which is parallel to the first ball screw 66.
  • the rotation of the second ball screw 72 drives a second nut 74 along the screw 72.
  • An arm 76 extends over the first nut 62 and connects the second nut 74 to the respective shaft 22.
  • One end of the arm 76 is fixed to the head 78 of the shaft 22 and the other end is pivotally mounted on the second nut 74.
  • Translational movement of a shaft 22 is achieved by driving the respective pair of motors 68 and 70 in unison, i.e. in the same direction, and the translational movement is confined by the track 64.
  • Differential movement of the first and second nuts 62 and 64 causes the arm 76 to pivot about its mounting on the second nut 74 and forces the shaft 22 to rotate within the bearings 60. Rotational movement of a shaft 22 is therefore controlled by differential rotation of the pair of motors 68 and 70.
  • Pneumatic cylinders 80 are mounted on each of the toes 10 to retract and extend the toes 10 independently.
  • the ends 30 of the fixed legs 28 each include rollers 32.
  • a toe 10 is bolted to a leg 6 or 8 underneath a respective pneumatic cylinder 80.
  • the toe 10 includes a toe body 82 having an open end which receives an open ended cup 84.
  • the open end of the cup 84 receives a shoe 86 of the toe 10.
  • the closed end 88 of the cup 84 is pushed downwards by a washer 94 attached to the piston 90 by a bolt 92, when the piston 90 is extended.
  • the shoe 86 has a partly spherical convex outer surface 100 which abuts the rim of the cup 84, and the shoe 86 is held in the cup 84 by a tensile link 96.
  • the link 96 has spherical ends that are received by the bolt 92 and a connecting part 98 on the top of the shoe, respectively. Contact is maintained between the surface 100 and the cup 84 by three springs 102 connected between the top of the shoe 86 and the closed end 88 of the cup 84.
  • the link 96 allows the shoe 86 to be retracted by the cylinder 90, which also causes the cup 84 to be retracted within the toe body 82.
  • a clearance 105 exists between the washer 94 and the top of the closed end 88 of the cup 84, and the link 96 is in tension.
  • the clearance 105 and the spherical ends of the link 96 allow the tension in the link 96 to be removed and the shoe 86 to pivot within the cup 84 when the washer 94 is pressed against the closed end 88 during extension of the piston 90.
  • the clearance 105 decreases as the shoe 86 pivots.
  • the curved outer surface 100 of the shoe 86 also facilitates the pivotal movement of the shoe 86.
  • Three pins 104 extend from the top 108 of the toe body 82 through the closed end 88 of the cup 84 to abut the top of the shoe 86 when it is retracted into the body 82.
  • the pins 104 reset the orientation of the shoe 86 so that its surface engaging face 106 is parallel with respect to the closed ends 88 and 108 of the cup 84 and the body 82.
  • the end of the pin 104 also includes a flange 110 which engages the end 88 of the cup 84 when fully extended by the piston 90 of the cylinder 80.
  • the ability of the shoe 86 to pivot within the cup 84, as shown in Figure 5, is particularly advantageous.
  • the shoe 86 rotates about a centre of rotation which is below the surface 18 which the shoe 86 engages. This configuration ensures that when contact between the shoe 86 and the surface 18 occurs, the shoe is able to rotate rather than jamb with contact only on one edge. The shoe 86 is therefore always able to achieve a tangential contact with the surface 18.
  • Rotation of the shoe 86 within the cup 84 als o allows the toes 10 to maintain the initial tangential contact with the surface 18 as the frame is moved with respect to the surface when the toes 10 are magnetically clamped to the surface 18 during a walk cycle.
  • the shoe 86 includes an annular hollow 112 in which a coil is wound about a central core pole 114.
  • the coil 37 can be energised to magnetise the shoe 86 as an electromagnet.
  • the robot 2 uses electromagnetic toes 10 for operation on ferromagnetic surfaces which may be hot, for example, from welding or cutting.
  • the coil is shaped to maximise turns while minimising local magnetic saturation in the shoe and this is achieved because the cross-sectional area of the magnetic path from the central pole 114 to the outer pole 116 is approximately constant.
  • the surface engaging face 106 is formed primarily by a pole piece 118 which is inserted into the central pole 114.
  • the pole piece 118 increases the circumference of the central pole 114 and allows the robot 2 to operate on surfaces of parts which are relatively thin without magnetic saturation.
  • the pole piece 118 also retains the coil in the shoe 86.
  • an electronic circuit may be used to monitor the change in inductance of the coil in the shoe 86 when the shoe contacts a ferromagnetic surface.
  • the circuit superimposes a relatively small low frequency AC signal on the DC signal supplied to energise the coil.
  • the phase shift between the superimposed AC signal and the monitored AC voltage of the coil is digitally timed and compared with a threshold value.
  • a leg 6 or 8 is only considered to be secured to the surface 18 when the phase shift for each of the coils of the leg's three toes 10 exceeds the threshold value.
  • the circuit may also be used to monitor the inductance of the coil 37 so as to determine when the respective shoe 86 abuts the surface 18 and the coil can be energised to secure engagement.
  • the toes 10 can, however, also simply be energised after the respective cylinder 80 has been extended the full length, which may take 150 ms.
  • the pressure in the upper chamber 122 of the cylinders 80 is maintained constant, typically 1 bar.
  • the pressure in the bottom chamber 124 is atmospheric pressure when the cylinder piston 90 is extended.
  • the pressure in the lower chamber 124 is set to a higher value depending on the weight and attitude of the robot 2, and other external forces.
  • the higher pressure setting ensures the toe 10 pulls the respective leg 6 or 8 towards the surface 18 with a predetermined force independent of the length of extension of the toe 10.
  • respective solenoid valves 128 are de-energised, as shown in Figure 6, to vent the lower chambers 124.
  • the coils of the shoes 86 are energised and the lower chambers 124 are pressurised to the desired higher value.
  • the frame 4 is pulled toward the surface 18 and contact between the wheels 32 and the surface 18 is maintained.
  • the pressure in the lower chambers 24 causes the cylinders 80 to retract the shoes 86.
  • the valves 128 are inexpensive compact three-way valves which the circuit 122 uses to achieve two-way force control of the cylinders 80.
  • a regulator 130 sets the pressure in the upper chambers 122, and therefore the extension force of the cylinders 80.
  • An electronically set regulator 132 may be used to facilitate control of the cylinder pulling force in response to operational factors including the attitude of the robot 2 with respect to horizontal.
  • the circuit 122 also includes for each leg 6 and 8 a differential pressure switch 134, a non-return valve 136 and a two-way valve 138 to vent the upper chambers 122 of each leg when their pressure exceeds the setting of the regulator 130 by a predetermined value, for example 0.2 bar. This overcomes any inability of the regulator 130 to vent flow arising from retraction of the cylinders 80.
  • the force between a shoe 86 and the surface 18 is the difference between the magnetic force and the pulling force exerted on the shoe 86 by the respective cylinder 80. Traction of the robot 2 is maximised by maximising the shoe to surface force, which requires that the cylinder pulling force be a minimum that is consistent with maintaining contact between the wheels 32 and the surface 18. This also reduces the contact force between the wheels 32 and the surface 18 and improves the ability of the robot 2 to travel over irregularities, such as bumps, of the surface 18.
  • the mobility of the robot 2 is constrained by the limitations on the strokes of the nuts 62 and 74 on the ball screws 66 and 72 and by interference between components of the robot 2 during movement.
  • Computer simulations have been performed to select a design which minimises interference.
  • the robot 2 has dimensions for the frame 4 and legs 6 and 8, as shown in Figure 7, then three possible toe arrangements A, B and C could be used, as shown in Figure 8, where the centres of the shafts 22 of legs 6 and 8 are O 1 and O 2 , respectively, and are chosen to be 100 mm apart.
  • a plan view pattern for the wheels 32 and toes 10 is selected to provide the longest first step, initially considering only toe/toe and toe/wheel interference.
  • a graph 130 of the maximum step length 1, achievable for different orientations, ⁇ , of the frame 4 relative to the clamped or engaging leg 6 or 8 is shown in Figure 9 for each of the alternative toe arrangements A, B and C. From the graph 130 it is apparent that arrangement A offers the best mobility unless rotations of the robot 2 are extremely important for a given task.
  • the separation of the two shafts 22 is determined after considering multiple steps of the robot 2 and the graph 132 is shown in Figure 10 for the results obtained for four different separations S. Positive separations occur for the arrangement A as shown in Figure 8 and negative values for S occur when the orientation of the centres of the shafts 22 are reversed.
  • Figure 12 shows the plan of a frame and leg assembly for the robot 2 which maximises mobility.
  • Possible collision spots include first spots 140 between the front and back toes 10 of the first leg 6 and the structure of the second leg 8, and second spots 142 between the shaft 22 of the first leg 6 and the structure of the second leg 8.
  • the robot 2 includes an electrical control system 250, as shown in Figure 19, controlled by either first control software shown in Figure 13 or second control software listed on pages 32 to 94.
  • the robot 2 has a tool (obscured) mounted underneath the frame 4. The tool is used to perform work on the part over which the robot 2 moves.
  • the first control software 200 is able to monitor a feature of the part so the robot 2 can track the feature and position the tool relative to the feature, as desired.
  • the second control software positions the tool relative to the feature after receiving information concerning the position of the feature relative to the frame 4.
  • the control system 250 includes DC servo motors 68, 70, such as the Electrocraft S19-1A and S586-1A, which drive ball screws 66 and 72 of the robot 2, respectively.
  • the motors 68, 70 are each provided with a tachometer 254 and an optical encoder 256, which is used to monitor the movement and position of a respective nut 62, 74.
  • the motors 68, 70 are driven by respective servo drivers 258, such as the Galil SSA-8/80 which is provided with analogue velocity and current loops.
  • the solenoid valves 128 and the magnetic coils 37 are powered by an I/O solenoid valve and magnet drive unit 260.
  • the servo drivers 258 and the magnet and solenoid drive 260 are both controlled by a controller 252 via a 30 metre long umbilical cable 262 which transmits electrical and pneumatic power to the robot 2.
  • the umbilical cable 262 also returns the feedback signals from the encoders 256 to the controller 252, together with signals from microswitches which are used to limit travel of the drive shafts of the servo motors 68, 70.
  • Digital proportional integral derivative (PID) position control of the nuts 62 and 74 is performed by two Industry Pack servo control modules 274 produced by a GreenSpring Computers Inc., using the information provided by the encoders 256.
  • the servo control modules 274 include two channels and are based on the National Semiconductor LM628 chip.
  • Two Industry Pack digital I/O modules 270 remotely control the servo drivers 258 and the magnetic and solenoid drive 260.
  • Control of the modules 270, 274 is provided by an MV147 central processing unit 266 via a VME (Versa Module Euro card) bus 268.
  • the CPU 266 is used to run control software on the VxWorks real time operating system. Alternatively, the control software can be run off-line on Unix workstations connected to the CPU 266 for development and testing.
  • the controller 252 can be configured, if desired, for mounting on the robot 2.
  • the first control software 200 takes into account that the robot 2 may provide an additional axis of the tool movement transverse to frame 4. This allows the tool to have a more complex path and rapid short stroke local motions as may be required to weave a welding torch during a welding operation.
  • the stroke of the transverse tool axis is determined by the minimum radius of curvature on the surface 18 of the feature to be tracked and any additional motion related to the process to be performed, such as welding torch weave. For example, sharper feature curves increase the stroke of the tool axis as the robot moves inside a feature corner while it is moved along a tracked feature.
  • Process data 212 from the tool may also be used to provide data on the location of the feature relative to the tool. For example, in welding this is used as the basis for through-arc sensing for seam tracking systems of conventional robots.
  • the first control software 200 uses the optical sensor 202 which includes a light source and video camera mounted on the frame 4.
  • the optical sensor 202 captures an image of a light stripe, projected transverse ahead of the robot 2 and across the feature.
  • the shape of the light stripe image indicates the location and cross-section of the feature.
  • the optical sensor 202 performs standard image processing techniques on the image to provide data on the location of the feature with respect to the robot and the feature dimensions to a motion control system (MCS) 204.
  • MCS motion control system
  • Other feature sensors may be employed, such as mechanical probes and flux sensors.
  • the MCS 204 also receives data on the state of the robot, motion parameters and a desired traversal velocity which may be inputted by an external operator or process controller 206. On the basis of all the entered data, the MCS 204 determines the desired positions and rotations of the legs 6 and 8 and generates joint position commands which are fed to a joint servo system (JSS) 208.
  • JSS joint servo system
  • joint refers to the interface between two parts which constrains relative movement between the parts to one type of motion (i.e., translation or rotation).
  • the permitted motion is normally actively controlled by a servomechanism.
  • the robot 2 has four actively controlled or active joints which position the legs 6 and 8 relative to the frame 4.
  • the legs 6 and 8 each have two degrees of freedom relative to the frame 4 and each degree of freedom is controlled by an active joint.
  • Each of toes 10 may be considered to include one passive joint which allows rotation of shoe 86 and one semi-passive joint relating to extension or retraction of the toe.
  • the JSS 208 includes standard hybrid or digital servo control mechanisms 274,
  • the JSS also supplies joint positions signals which provide information on the position of the legs 6 and 8.
  • the MCS 204 further controls changeover of the legs 6 and 8 and determines the desired position of the tool along the tool axis and generates a predicted tool position command for a tool control system (TCS) 210.
  • TCS 210 also receives data on the feature from the process data 212 discussed previously.
  • the predicted tool position command corresponds to placing the tool over the feature position predicted from the joint position commands or joint position signals and the feature data obtained by the optical sensor 202.
  • the TCS 210 performs fine positioning of the tool over the feature, and with the feature data obtained from the process data 212, the TCS 210 is able to operate in a closed-loop mode to reduce errors in tool position.
  • the TCS 210 generates a tool position command for a tool servo system (TSS) 212 which includes a standard hybrid or digital servo mechanism for controlling the motor which drives the tool transverse to the frame 4.
  • TSS tool servo system
  • the motion parameters used by the MCS 204 are maximum velocities and accelerations for the motors of the robot 2, a joint clock interval T, (the sampling period of the JSS 208), a measurement period T f of the optical sensor 202, a joint trajectory smoothing interval T a discussed hereinafter, and the time required for engaging and disengaging the toes 10 from the surface 18. Except the latter, all of the time intervals are multiples of the joint clock interval T j .
  • the MCS 204 and the TCS 210 are processing systems implemented in software on the CPU 266 of the robot 2, which preferably runs under a real time operating system such as VxWorks.
  • a path planning module can replace the optical sensor 202, if the robot 2 is required to execute a predetermined motion with respect to the surface rather than track a feature, and the module would provide data of a similar form to the MCS 204.
  • the MCS 204 comprises four software modules, as shown in Figure 14, a supervisor module 220, a feature trajectory generator (FTG) 222, a joint setpoint generator (JSG) 226, and a joint trajectory generator (JTG) 228, which are described hereinafter.
  • a supervisor module 220 a feature trajectory generator (FTG) 222
  • JSG joint setpoint generator
  • JTG joint trajectory generator
  • a tracking point which is the point the MCS 204 attempts to move towards the measured feature point in such a way that the feature remains within the range of the optical sensor 202.
  • the tracking point shifts along the line which is perpendicular to and bisects an optical sensor detection line, which corresponds to the light stripe shone across the feature by the optical sensor 202.
  • the tracking point shifts towards the sensor detection line to extract movement from the robot 2 to keep the feature in view of the sensor. If the feature is near the centre of the sensor detection line, the tracking point shifts away from the sensor detection line, towards the tool axis.
  • the centre of the tool axis is positioned close to the feature, and tracking is more closely aligned with the tool.
  • the axis of the tool is preferably located below the motors 68 and 70 at the front of the robot 2.
  • the tracking strategy relies on the tracking point being in front of the centre of the shaft 22 of the clamped leg 6 or 8 at all times.
  • a predicted tool point which is the intersection of the tool axis and the predicted path of the feature. It is desirable to keep this point near the centre of the tool axis to minimise the axis stroke and interference between the tool and other robot components. This is achieved when the tracking point is shifted towards the tool axis.
  • the position of the tracking point is determined by the supervisor 220 which seeks a compromise between the usually conflicting objectives of keeping the feature within the sensor detection line and keeping the tool close to the centre of the tool axis.
  • the robot coordinate system is defined as having its origin at the centre of the shaft of the clamped leg (O 1 or O 2 in Figure 8) and its x axis parallel to the longitudinal axis of the frame 4.
  • the tracking point has coordinates x t , y t and a feature point has coordinates x f , y f , as shown in Figure 15.
  • the y f coordinates are determined by the optical sensor 202 and adjusted for whichever leg is currently clamped.
  • the x f coordinates result from movement of the shaft of the clamped leg along the frame 4.
  • Feature measurements are performed as the robot moves and coordinates in the RCS of a sequence of feature points are created by the FTG 222. These are made available for the joint setpoint generator 226, as shown in Figure 14.
  • T f Selection of the feature sensing interval T f depends on the tool positioning requirements and the feature radius. Decreasing T f improves the accuracy and allows paths of smaller radius to be tracked.
  • the method of generating the joint trajectories requires at least two points of the feature, which implies that d st ⁇ 2V r T f , where d st is the x-direction distance between the sensor detection line and the tracking point and V r is the transversal velocity.
  • the joint setpoint generator (JSG) 226 computes angles and distance setpoints such that the rotation and translation of frame with respect to the clamped leg moves the tracking point towards the next feature point while maintaining constant tool traversal velocity.
  • the angle between the x axis and the tracking point (x t , y t ) is denoted ⁇ (positive anticlockwise), as shown in Figure 16.
  • the directional point (x d , y d ) results from the following procedure:
  • the translational and rotational setpoints for the joints of the clamped leg required to move the tracking point from its present position (x t , y t ) towards the directional point (x d , y d ) are calculated as follows. Firstly, the rotation ⁇ is calculated: Next, the translation distance D is calculated. If the tracking point coincides with the predicted tool point (x p , y p ), then
  • D is the distance between the predicted tool point, after rotation by ⁇ , and a point identified as follows.
  • the line parameters are
  • the joint trajectory generator (JTG) 228 produces joint trajectories (translational and angular movement commands) for feature tracking, leg velocity matching (for leg changeover), leg halting, and leg resetting. Furthermore, the JTG 228 computes the predicted tool position command to position the tool along the tool axis over the expected feature location.
  • the dotted line represents the feature to be tracked.
  • the pointsZ fe ⁇ ture , B feature , C feature and D feature are the points at which the feature is measured at intervals of T f by the optical sensor while the robot moves. In the RCS, these points constitute the feature trajectory produced by the FTG 222.
  • the JSG 226 computes the joint setpoints ⁇ and D based on the feature trajectory produced by the FTG 222.
  • the joint setpoints are generated within the period
  • the JTG 228 computes joint trajectories for transition between subsequent joint setpoints.
  • the following preview joint trajectory generation method is used and is referred to as the preview joint trajectory generation method because the coordinates of the point C feature are needed before the motion segment aiming at point B feature is completed.
  • j s (T) denotes the joint setpoint ( ⁇ or D) required to move the tracking point from point T to point S, the coordinates of which are expressed in the RCS associated with point T; j s denotes the joint trajectory value when the tracking point is at point S.
  • the time interval t a is referred to as the joint trajectory smoothing interval.
  • the joint setpoints, j c ' (B') , required to move the tracking point from point B' towards point C feature are computed using the coordinates of C feature in the RCS associated with point B'. These setpoints move the tracking point to the final point C'.
  • the algorithm is initialised after robot start-up and after each leg changeover.
  • the trajectory from time -T f to -t a has a trapezoidal velocity profile (constant acceleration segment followed by constant velocity segment) as indicated by the thick dotted line in Figure 18.
  • the smoothed joint trajectory is evaluated at steps 3 and 6 of the above algorithm.
  • the smoothed trajectory for the translational joint consists of absolute linear displacement j trans lin which increases with forward robot motion.
  • the smoothed trajectory for the rotational joint consists of absolute angular displacements j rot ang which increase for anticlockwise movement of frame about the shaft of the clamped leg.
  • the drive arrangement employed to achieve rotation requires conversion of the rotational trajectory j rot ang to a translational trajectory j rot lin as follows:
  • J rot lin (t) j trans lin ( t) ⁇ r tan ( ⁇ (t) )
  • ⁇ (t) j rot ang (t) - j rot ang (t - T j )
  • T j is the joint trajectory update period
  • r is the distance between the centres of the ball screws for translation and rotation. The positive sign is used for the leg whose shaft is at the right side of frame and the negative sign is used for the other leg.
  • leg changeover involves matching the leg velocities, leg change-over, halting the lifted leg and resetting it to its foremost position and initial orientation.
  • the stationary lifted leg which is at its foremost position, has to be accelerated to the traversal velocity with respect to the frame.
  • the rotational velocity of the frame about the clamped leg is decelerated to zero and the translational velocity of the frame with respect to the clamped leg is changed to the traversal velocity, as discussed previously.
  • the joint position commands generated by the JTG 228 for leg changeover have trapezoidal velocity profiles. Each motion segment is completed at a multiple of the joint clock interval T j .
  • the supervisor module 220 supervises and controls tasks performed by other modules of the MCS 204. In particular, the supervisor 220 determines the tracking point, controls leg changeover, and adjusts the traversal velocity.
  • the tracking point ( x t , y t ) is located on the line bisecting the tool axis and the optical sensor detection line. Whenever the joint trajectories are generated, x t is determined by:
  • x w (t) is the x coordinate of the tool axis
  • d sw is the x distance between the sensor detection line and the tool axis
  • e mx is half the length of the sensor detection line
  • k p and k d are adjustable parameters
  • y f (t) and y a (t) are the y coordinates of the feature at the sensor detection line and of the middle of the sensor detection line, respectively
  • [x] denotes the integer part of x.
  • This algorithm shifts the tracking point towards the sensor detection line when the feature moves out from the middle of the sensor, and towards the tool axis when the feature moves towards the middle of the detection line.
  • the algorithm is a proportional-derivative type with positive k,, and k d weighting the "proportional" and "derivative” terms, respectively.
  • the supervisor 220 also controls leg changeover. At each joint clock interval T j , the supervisor evaluates the time t leg required to complete leg velocity matching, leg changeover, and to halt the lifted leg. A safety margin (minimum one joint clock interval) is added to this to determine t leg . If the motion commanded by the joint trajectories during t leg cannot be fully executed due to ball-screw limits or mobility constraints, the supervisor 220 issues the leg change command.
  • the desired traversal velocity may be set by the operator or by process considerations.
  • the supervisor determines the traversal velocity, V r , used to generate joint setpoints. If the tracking point is near the optical sensor and the feature moves out of the sensor detection zone, the traversal velocity can be reduced to decrease the distance between the measured feature points. This reduces the radius of the feature which the robot can track.
  • the second control software is a simplified implementation of the first control software 200, in that feature sensoring is omitted, and, as there is no feature measurement, a feature trajectory generator 222 is not required. Movement of the tool is also restricted to a simple engage and disengage operation and therefore an elaborate tool control system 210 is not required.
  • the second control software runs continuously to move a point on the robot, such as the position of the tool, along a path to a target point, the vehicle frame coordinates of which are either provided by a command or a remote joystick 280 to the controller 252.
  • the second control software comprises a number of C procedures, discussed hereinafter, listed on pages 32 to 85. Typical header files which define system parameters, variables and functions for the procedures are listed on pages 86 to 94.
  • the servo motors 68, 70 are calibrated using the microswitches.
  • a "null" move to the target point is performed by making the target point coincident with the tool's initial position. The robot only begins to move when it is given a different target point.
  • positions for the servo motors 68, 70 are determined using a trajectory generator which plans motion of the legs 6 and 8 within mobility constraints, and controls the legs 6 and 8 during the changeover phases.
  • the trajectory generator is called every servo interval, which is approximately 20 ms, and is able to accept a new target point at any time except during leg changeover.
  • the existing target point is used as a basis for moving the robot.
  • the target point includes, in addition to the coordinates for a path point, the desired speed of travel of the tool or tool point.
  • the trajectory generator calls an inverse kinematic module to determine the desired position and orientation of the shafts 22 required to move the robot 2 towards the target point.
  • a range checking module is then called to determine whether the selected shaft positions and orientations are within the mobility constraints of the robot 2. If the constraints are not met, a closer target point is chosen midway between the unachievable target point and the current robot position. The process is repeated until the desired shaft positions and orientations are within the mobility constraints.
  • the number of servo intervals required to reach the target point is calculated from the desired speed and the servo motor positions required. At each servo interval the number of counts for each encoder 256, which need to be received to indicate that the desired nut positions have been reached, is determined and movement of the robot 2 effected.
  • the number of servo intervals remaining before the target point is reached first equals the number of intervals comfortably required for leg changeover, changeover is initiated. Changeover is left until the end of a planned movement to reduce unnecessary changeovers which are time consuming. For example, target points may change before changeover, requiring a new movement. Movement of the legs 6 and 8 during changeover follows the trapezoidal velocity profiles discussed previously with reference to Figure 3.
  • the robot 2 is initialised by sav_init in sav.c. This:
  • Encoder counters for the controller 252 are zeroed and then all servo motors 68, 70 are slowly turned to drive the nuts 62, 74 towards the motors 68, 70 until the limit switches operate. These switches disable further drive of the motors 68, 70 in this direction, and the motors 68, 70 stop, but the position targets continue their slow change.
  • Each encoder counter is read when its limit switch actuates and when all have actuated, position commands to move each servo motor 68, 70 back to its switch position are issued.
  • Each servo is then commanded to move another 1.25 revolutions in this direction.
  • its counter contents are saved, a predetermined offset added and a new move started to this position.
  • the servo motors 68, 70 are all in their home position.
  • the encoder counters are then re-zeroed and the servo position commands reset to zero. This completes servo calibration and all subsequent servo positions are referenced to these home positions. Control returns to sav_init.
  • the trajectory generator software is then initialised by traj_init in traj.c. This involves creation of mutex semaphores and initialisation of the homogeneous transform from the datum of the robot to the tool.
  • Target points set using the control software in the specification include the desired velocity and the required position of the robot tool relative to its current position. A constant target point therefore causes continuous motion, provided the commanded velocity is non-zero.
  • the function traj_settar in traj.c sets target points. It can be called from a VxWorks shell by typing traj_settar V, x, y, where V is the desired velocity in m/s, x and y are the transverse and axial components of the position of the target point in millimetres.
  • Control by a joystick is initiated by spawning the joystick task as listed in joystick.c. This reads the joystick and generates the arguments V, x, y and then calls traj_settar, as above. Control by the joystick simulates derivation of data from a feature sensor, possibly a mechanical probe or a light stripe system discussed for the first control system 200.
  • the main loop within servo_task in servo.c continues execution every servo interval awaiting a target point. It is the master which initiates execution of other functions and controls the robot 2. It interfaces with the trajectory generator traj.c via the task tragenjask in tragen.c. Communication between servo.c and tragen.c is via message queues which deliver two structures, state and command. These contain all the data to define the actual and desired state of the robot 2.
  • the first operation is to compute the commanded velocity for each servo motor 68, 70 from the current and previous position commands and compare it with the maximum velocity. If it is within the limits, position and velocity commands are determined and the servo modules 274 instructed accordingly by servojoad and initiated by servo_move.
  • the output signals for magnetic coils 37 and solenoids 128 of each leg are set as required from the command structure. This is done using io_set. As the command structure is invalid during the first pass, the variable serial which is false only on the first pass, inhibits erroneous setting of these outputs.
  • the robot state is read to build the state structure.
  • the servo positions are read into the state structure before servo_load above using lm628_rdrp.
  • the state of each magnet and solenoid is read using io_readback and the adhesion element of the leg's state structure set true for each toe 10 if both its magnetic coil 37 and solenoid 128 are on.
  • the state structure is then sent to tragen.c using msgQSend and servo.c is suspended awaiting the next command message from tragen.c. Its suspension allows tragen_task (which has a lower priority) to run.
  • tragen_task receives the message using msgQReceive, it calls traj.c, passing pointers to the state and an empty command structure.
  • traj.c returns, the command structure is returned to servo.c using another message queue send/receive pair.
  • Both servo.c and tragen.c have timeouts a little longer than the servo interval which halt execution via sav_panic if the messages are not passed in time.
  • the purpose of using separate servo_task and tragen_task tasks with message queues is to isolate the code in traj.c from the mechanism of robot control. This improves the portability between VxWorks and Unix and also potentially allows recovery from failure of the code in traj.c.
  • execution After execution returns to servo.c, the status of the robot power and each servo motor 68, 70 is checked and if an anomaly is found, execution is halted via sav_panic and an appropriate message printed. If not, execution returns to the top of the main loop ready for the next servo interval.
  • traj in traj.c receives the robot state from tragen_task and calls it svacl internally. It computes the command structure, which it calls svcom internally, and returns it to tragen_task. On the first pass (serial is false), traj initialises internal variables for the encoder values, status flags for each leg 6, 8, and homogeneous transforms for each leg 6, 8 relative to the robot datum. Functions for transform manipulation are defined in math.c.
  • the flag sts is set to the value NEWTAR by traj_settar when a new target point has been accepted. If in addition, leg changeover is not occurring, it allows acquisition of a new target point by the function get_newtar.
  • the function traj_tar is then called to compute servo position commands which traj places in the structure svcom.
  • the function traj next executes the logic controlling the operations in leg changeover, providing the velocity demand vd is non-zero.
  • the changeover operations are controlled by the value of elements in the leg structure, leg->no, leg->nm, leg->ns, Ieg->nu, and leg->nn.
  • the operation is executed whenever seg_no reaches an element value.
  • the elements are computed by the function traj_seg from the number of servo interval time segments to be completed before the tool point reaches the current target point.
  • Each of the controlling elements in the leg structure is computed from this value and the number of servo intervals required to execute the individual processes associated with leg changeover.
  • leg->no is the number of trajectory segments remaining to be completed before the solenoid valves 128 of the leg 6 or 8 which is engaging the surface 18 must be actuated to drive the toes 10 down.
  • seg_no reaches leg->no
  • the solenoids 128 of both legs 6 and 8 are set to the value SOLDOWN which drives the legs down.
  • the solenoids 128 of at least one leg 6 or 8 will be in this state, and their setting will have no effect. This practice is used in many places for both solenoids 128 and magnet coils 37 to improve the symmetry of the code.
  • leg changeover is complete, the status of each leg 6, 8 is exchanged and the pointers indicating the clamped leg, leg and lifted leg, leg_up are set. Control is then returned to tragen_task.
  • the function traj_tar computes an achievable target point and calls other functions to compute servo position commands which are passed to traj.c.
  • traj_tar saves the datum transform of the target point to the clamped leg. The transform of each target point is computed with respect to the clamped leg 6 or 8. This is used by traj_stck to check if the target is within the mobility region of the robot.
  • a new target point in the vehicle coordinate system midway between the tool point and the original target point is computed and retested. Closer target points continue to be computed until one within the mobility region is obtained. It is then returned to traj_tar.
  • the function traj_seg is then called and the number of servo interval segments required to move the tool point to the target is computed (the function inv_kin is defined in kine.c). If the result is non-positive, leg changeover is initiated. In either case, traj_step is called.
  • the function traj_step computes servo positions to reach the target point. It calls traj_gtst to compute the position and angle of the shaft 22 of the clamped leg 6 or 8. It then calls traj_axis for each leg. This calculates the servo positions of the clamped leg to achieve this position and angle and the servo positions of the other leg as appropriate. Traj_step also updates the number of servo interval segments done within this step. If the result indicates that the target has been reached, the flag stsl is set to the value NEWSTEP and the transform of the target point converted to be referenced to the newly clamped leg.
  • the function traj_axis computes servo positions for each leg 6, 8. It has arguments which signify the mode of a leg. If the mode is LEGDOWN the leg is clamped to the surface and the function traj_down is called. This calls inv_kin in kine.c to compute the position and angle of shaft 22, then leg_axis to convert to ball screw positions, and axis_enc to compute servo positions in encoder counts.
  • traj_reset is called. This implements a trapezoidal velocity profile to return the leg to its home position. If the mode is LEGUP, and seg_no indicates that the leg does have to accelerate to catch up to the clamped leg, traj_catch is called. This ramps the leg velocity up to that of the clamped leg. If the leg is in the overlapping phase (both legs clamped), traj_overlap is called as the velocities of each leg are matched.
  • col_cllsn is called to check for collisions before the servo positions are passed back to tragen.c for execution.
  • the code in collision.c represents the rollers 32, legs 6, 8, shafts 22, and toes 10 of the robot 2 as simple lines and circles. Each is defined relative to the positions and angles of the shafts 22 of each leg 6, 8.
  • A sqrt((c->px[0]-1->px[1]) * (c->px[0]-1->px[l]) + (c->py[0]-1->py[l]) * (c->py[0]-1->py[l]));
  • a11 atan2(11_y2 - 11_y1, ARMLEN);
  • a12 atan2(12_y1 - 12_y2, ARMLEN);
  • leg1 circles ⁇ leg2 line */
  • leg1 line */ if(col_line_circle(11[0], c2[1])) ⁇ sav_panic("110 collides with c21 ⁇ n"); co_flag 1;
  • V_MIN 0.01 /* speed at which the sav is turning max.
  • y -x * (Y_MAX - Y_MIN)/X_MAX + Y_MAX;
  • y x * (Y_MAX - Y_MIN)/X_MAX + Y_MAX;
  • djx XINC_MAX * sgn(djx);
  • djy YINC_MAX * sgn( djy ) ;
  • a and/or v may be NULL, indicating carry-over
  • mode is the Startup mode; period is the tragen rate in us
  • timeout WAIT_FOREVER
  • servo_task servo %d: serial %d: velocity limit %u ⁇ n
  • timeout (period * sysClkRateGet())/1000000 + 1;

Landscapes

  • Engineering & Computer Science (AREA)
  • Mechanical Engineering (AREA)
  • Chemical & Material Sciences (AREA)
  • Combustion & Propulsion (AREA)
  • Transportation (AREA)
  • Manipulator (AREA)

Abstract

Un robot comporte un corps (4) et deux pieds (6, 8) sans articulations actives. Lesdits pieds sont montés sur ledit corps de manière qu'ils puissent être translatés et tournés indépendamment l'un de l'autre par rapport audit corps.
PCT/AU1993/000481 1992-09-18 1993-09-20 Robot Ceased WO1994006602A1 (fr)

Priority Applications (1)

Application Number Priority Date Filing Date Title
AU48111/93A AU4811193A (en) 1992-09-18 1993-09-20 A robot

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
AUPL482692 1992-09-18
AUPL4826 1992-09-18

Publications (1)

Publication Number Publication Date
WO1994006602A1 true WO1994006602A1 (fr) 1994-03-31

Family

ID=3776428

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/AU1993/000481 Ceased WO1994006602A1 (fr) 1992-09-18 1993-09-20 Robot

Country Status (1)

Country Link
WO (1) WO1994006602A1 (fr)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116061173A (zh) * 2022-11-04 2023-05-05 云南电网有限责任公司保山供电局 一种带电作业的机械臂六自由度冗余任务轨迹规划方法

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4248103A (en) * 1977-09-06 1981-02-03 Imperial Chemical Industries Limited Straight line mechanism
SU932017A1 (ru) * 1980-07-02 1982-05-30 За витель Шарнирно-рычажный дифференциал Э.М.Кудинцева
SU1142271A1 (ru) * 1983-04-29 1985-02-28 Новосибирский электротехнический институт Устройство дл перемещени исполнительного органа
US4565487A (en) * 1981-09-04 1986-01-21 International Robotic Engineering, Inc. System of robots with legs or arms
SU1463467A1 (ru) * 1987-01-13 1989-03-07 Казанский Авиационный Институт Им.А.Н.Туполева Манипул тор
SU1472252A1 (ru) * 1987-08-20 1989-04-15 Азербайджанский политехнический институт им.Ч.Ильдрыма Манипул тор
US4832563A (en) * 1986-08-06 1989-05-23 Fibro Gmbh Portal system
SU1523337A1 (ru) * 1988-04-12 1989-11-23 Новосибирский электротехнический институт Обрабатывающее устройство
US5094311A (en) * 1991-02-22 1992-03-10 Gmfanuc Robotics Corporation Limited mobility transporter

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4248103A (en) * 1977-09-06 1981-02-03 Imperial Chemical Industries Limited Straight line mechanism
SU932017A1 (ru) * 1980-07-02 1982-05-30 За витель Шарнирно-рычажный дифференциал Э.М.Кудинцева
US4565487A (en) * 1981-09-04 1986-01-21 International Robotic Engineering, Inc. System of robots with legs or arms
SU1142271A1 (ru) * 1983-04-29 1985-02-28 Новосибирский электротехнический институт Устройство дл перемещени исполнительного органа
US4832563A (en) * 1986-08-06 1989-05-23 Fibro Gmbh Portal system
SU1463467A1 (ru) * 1987-01-13 1989-03-07 Казанский Авиационный Институт Им.А.Н.Туполева Манипул тор
SU1472252A1 (ru) * 1987-08-20 1989-04-15 Азербайджанский политехнический институт им.Ч.Ильдрыма Манипул тор
SU1523337A1 (ru) * 1988-04-12 1989-11-23 Новосибирский электротехнический институт Обрабатывающее устройство
US5094311A (en) * 1991-02-22 1992-03-10 Gmfanuc Robotics Corporation Limited mobility transporter

Non-Patent Citations (5)

* Cited by examiner, † Cited by third party
Title
DERWENT ABSTRACT, Accession No. 86-040903/06, Class P62; & SU,A,1 142 271, (NOVOS ELECTROTECH INST), 28 February 1985. *
DERWENT ABSTRACT, Accession No. 89-376481/51, Class P62; & SU,A,1 463 467, (KAZAN AVIATION INST), 7 March 1989. *
DERWENT ABSTRACT, Accession No. 90-020983/03, Class P62; & SU,A,1 472 252, (AZERB POLY), 15 April 1989. *
DERWENT ABSTRACT, Accession No. 90-170264/22, Class P62; & SU,A,1 523 337, (NOVOS POLY), 23 November 1989. *
DERWENT ABSTRACT, Accession No. E-7056K/14, Class, P62; & SU,A,932 017, (KUDINTSEV), 30 May 1982. *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116061173A (zh) * 2022-11-04 2023-05-05 云南电网有限责任公司保山供电局 一种带电作业的机械臂六自由度冗余任务轨迹规划方法

Similar Documents

Publication Publication Date Title
US11850738B2 (en) Robotic leg
US10780578B2 (en) Reaching mobile robots
JP3473834B2 (ja) ロボットの制御装置
CN106625665B (zh) 一种可移动式自动寻址的钻铣机器人系统
US10802508B2 (en) Mobile robot
US12172321B2 (en) Work mode and travel mode for mobile robots
CN108582082A (zh) 具有精确回零位功能的机器人、回零位方法、系统和介质
CN111716355A (zh) 机器人绝对定位精度补偿系统及方法
Guo et al. A climbing robot with continuous motion
Kelly et al. Fixed-camera visual servo control for planar robots
CN114643599B (zh) 一种基于点激光和面阵相机的三维机器视觉系统及方法
CN110456729B (zh) 一种轨迹跟踪控制方法及轨迹跟踪系统
WO1994006602A1 (fr) Robot
CN109773781A (zh) 一种无传感器的工业机器人直线软浮动控制方法
Baeten et al. Tool/camera configurations for eye-in-hand hybrid vision/force control
JPS60127987A (ja) ならい制御方法および装置
Alkkiomaki et al. Smooth transition from motion to force control in robotic manipulation using vision
Park et al. Balance control of a biped robot using camera image of reference object
JP2001022418A (ja) 作業ロボットのティーチングデータ補正方法
Liu et al. High-accuracy visual/PSD hybrid servoing of robotic manipulator
Yoshimi et al. Closed-loop visual grasping and manipulation
JPH012883A (ja) ロボット制御装置
Gao et al. Adaptive velocity planning for 6-DOF Robots with fixed tracks
CN120461443B (zh) 机器人控制方法、系统、机器人、设备及介质
US12330317B2 (en) Calibration method and robot system

Legal Events

Date Code Title Description
AK Designated states

Kind code of ref document: A1

Designated state(s): AU CA JP US

AL Designated countries for regional patents

Kind code of ref document: A1

Designated state(s): AT BE CH DE DK ES FR GB GR IE IT LU MC NL PT SE

DFPE Request for preliminary examination filed prior to expiration of 19th month from priority date (pct application filed before 20040101)
121 Ep: the epo has been informed by wipo that ep was designated in this application
122 Ep: pct application non-entry in european phase
NENP Non-entry into the national phase

Ref country code: CA