Classifications

• • B—PERFORMING OPERATIONS; TRANSPORTING
  • B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
  • B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
  • B25J9/00—Programme-controlled manipulators
  • B25J9/16—Programme controls
  • B25J9/1602—Programme controls characterised by the control system, structure, architecture
  • B25J9/1607—Calculation of inertia, jacobian matrixes and inverses

Definitions

• the present disclosure relates to multi-axis machines and, more particularly, to systems and methods of dealing with kinematic singularities during machine operation.
• the statements in this section merely provide background information related to the present disclosure and may not constitute prior art.
• a multi-axis machine may be software-driven to move one or more machine tools relative to a work piece.
• a multi-axis kinematic linkage of the machine is capable of moving a tool relative to a plurality of predetermined coordinate axes.
• a plurality of translational and/or rotational joints may be operable singly and/or cooperatively to move one or more links to position a tool at a desired location.
• a multi-axis linkage may be moved to a position (i.e. "a singularity", or "a singular point” ) in which one or more degrees of freedom are lost.
• a singular point one or more joints may be incapable of moving a tool as instructed by the software.
• the present disclosure in one implementation, is directed to a method of operating a multi-axis machine.
• a kinematic linkage of the machine is monitored.
• the linkage includes a plurality of joints and links.
• the linkage is monitored to detect an approach by one or more joints of the linkage toward singularity.
• the monitoring is performed using a manipulator matrix.
• a degree ⁇ of the approached singularity is determined.
• the one or more joints approaching singularity are identified.
• the method includes obtaining n virtual joints, replacing the one or more identified joints with the virtual joints in the manipulator matrix to modify the manipulator matrix, and using the modified manipulator matrix to determine position changes for links of the linkage.
• Figure 1 is a diagram of a numerically controlled (NC) processing system in accordance with one implementation of the present disclosure
• Figure 2 is a frontal perspective view of a portion of a multi-axis kinematic linkage of a machine controlled in accordance with one implementation of the present disclosure
• Figure 3 is a diagram of a singular value decomposition of a manipulator
• Figures 4A and 4B are a flow diagram of a method of operating a multi-axis machine in accordance with one implementation of the disclosure.
• the present disclosure in one implementation, is directed to a method of automatically determining in real time (1 ) whether a machine linkage is approaching a singularity and how close the linkage is to the singularity, (2) which joint(s) of the linkage are approaching the singularity, and (3) which degrees of freedom would be lost at the singularity.
• a numerically controlled (NC) processing system in accordance with one implementation of the present disclosure is indicated generally in Figure 1 by reference number 20.
• the system 20 includes one or more multi-axis machines 24, one of which is shown in Figure 1.
• a machine 24 may be configured, for example, to perform multi- axis machining operations on a work piece (not shown). Additionally or alternatively, the system 20 may include other types of multi-axis machines capable of component placement, for example, a numerically controlled assembly robot.
• the system 20 includes one or more computers 28, one of which is shown in Figure 1.
• the computer 28 includes a suitably programmed computer processor 32 and memory 36.
• the computer 28 may be programmed in accordance with one implementation of the disclosure to monitor and control operation of the multi-axis machine 24.
• the computer 28 may be used to execute software for adjustably compensating the positioning of the multi-axis machine 24, as further described below.
• Software compensation may be used, e.g., to adjust for certain "as-built” conditions of the machine 24 in order to bring operation of the machine 24 within tolerances specified relative to an ideal or "perfect” machine.
• a user interface 40 such as a computer terminal including a keyboard, pointing device, and monitor, is provided by which a user may communicate with the system 20. It should be noted that the present disclosure could be implemented relative to many and various kinds of installations, including but not limited to materials processing installation configurations described in U.S. Patent Publication No. 20060271241. It shall be understood by those skilled in the art that many different types and configurations of multi-axis machines, computers, processors, input-output devices, communication systems, etc., could be included in NC processing systems in accordance with various implementations of the disclosure.
• a portion of a multi-axis kinematic linkage of the machine 24 is indicated generally in Figure 2 by reference number 100.
• the linkage 100 includes six movable links 108a-108f interconnected by six joints 112a-112f. Joints 112a-112e are revolute (i.e. rotary) and joint 112f is prismatic (i.e. sliding).
• the linkage 100 is operable by the computer 28 to position and control a tip portion 116 along a predefined path during a work process.
• the tip portion 116 may have, e.g., a cutting tool or other type of tool (not shown) mounted thereon.
• the linkage 100 is exemplary only.
• Machines 24 may have linkages more or less complex than the linkage 100 and may include more, fewer and/or different types of links and/or joints and/or may provide more, fewer and/or different degrees of freedom and/or types of degrees of freedom.
• the linkage 100 is driven by the computer 28 via a drive apparatus (not shown), e.g., one or more electric motors.
• a link 108 As a link 108 is driven, the positioning of other links 108 may be affected.
• J ⁇ Q ⁇ X so that a change in the Cartesian position X of the tip portion 116 is related to a change in the link positions Q.
• the Jacobian matrix J(Q) becomes progressively more ill- conditioned.
• the Jacobian matrix J(Q) becomes rank- deficient.
• singular value decomposition is used to determine (1) whether the linkage 100 is approaching a singularity and how close the linkage 100 is to the singularity, (2) which joint(s) 112 of the linkage are approaching the singularity, and (3) which degrees of freedom would be lost at the singularity.
• the diagonal elements of S may be referred to as singular values of A.
• the columns of U may be referred to as left singular vectors.
• the columns of V i.e., the rows of V ⁇
• a diagram of singular value decomposition of a manipulator Jacobian matrix J in accordance with one implementation of the disclosure is indicated generally in Figure 3 by reference number 200. Referring to Figures 2 and 3, differential motions of the joints 112 are related to their respective effects on positioning of the tip portion 116 by the manipulator matrix J.
• Each column 204 of J is associated with a corresponding joint 112.
• a condition number of the manipulator matrix J may be calculated by dividing a largest singular value 236 by a smallest singular value 236.
• the condition number may range from 1 for a perfectly conditioned matrix to infinity for a singular matrix.
• a reciprocal of the condition number is used, which ranges from 1 for a perfectly conditioned matrix to 0 for a singular matrix.
• a flow diagram of a method of operating the machine 24 in accordance with one implementation of the disclosure is indicated generally in Figures 4A and 4B by reference number 300.
• operation 308 kinematic joint positions for moving the tip portion 116 along a prescribed path are calculated.
• operation 312 a manipulator Jacobian J corresponding to the calculated joint 112 positions is calculated and then decomposed using SVD.
• operation 316 a reciprocal of a condition number is determined.
• the condition number reciprocal is compared to a predefined threshold value, e.g., 0.001. If the condition number reciprocal is greater than the threshold value, then it is assumed that the linkage 100 is not approaching any singularities, and control passes to operation 340.
• a predefined threshold value e.g., 0.001. If the condition number reciprocal is greater than the threshold value, then it is assumed that the linkage 100 is not approaching any singularities, and control passes to operation 340.
• a degree n of the singularity is determined in the following manner.
• n "virtual joint(s)" are obtained by taking the last n left singular vector(s) 208 and multiplying them by the largest singular value 236.
• the column(s) 204 of J associated with the joint(s) 112 determined in operation 328 are replaced with the "virtual joint(s)" obtained in operation 332.
• a nominal inverse kinematics calculation is used to obtain an initial link position vector Q).
• an as-built forward kinematics calculation and the link position vector Q 1 are used to obtain an actual tip position vector X.
• a ⁇ X vector is calculated, indicating a difference between a desired tip position vector and an actual tip position vector.
• a ⁇ Q vector is calculated using the Jacobian matrix obtained in operation 312 (and as the Jacobian matrix may have been modified in operation 336).
• the ⁇ Q vector calculated in operation 352 and the link position vector Qi are used to obtain a new link position vector Q i+ i.
• convergence is tested by comparing an absolute difference between vectors Q i+ i and Qi to a predetermined convergence criterion ⁇ . If convergence has not yet been reached, the vector Qj + i is used to update the vector Qi and control returns to operation 344.
• convergence is reached in operation 360, it then is determined in operation 364 whether the condition number reciprocal was less than the threshold value, that is, whether the linkage is approaching singularity.
• value(s) representing the "virtual joint(s)" are removed from the vector Q M , which then is used in operation 372 to compensate the non-singular joints 112. Because SVD is performed only once for each point and is not part of the iterative solution process, the foregoing methods and systems are less computationally expensive than linear quadratic regulation (LQR) or damped least squares.
• LQR linear quadratic regulation
• left singular vectors provide ideal "virtual joint(s)", in that they move the linkage in exactly the singular direction(s), regardless of the position or complexity of the machine.
• Another advantage is that by scaling the "virtual joint(s)" by the largest singular value, the resulting condition of the new manipulator Jacobian is optimized, thereby assuring a stable, rapidly converging solution, e.g., in operations 340 through 360 of the foregoing method 300.
• implementations of the foregoing method require no a priori knowledge of the locations of singularities for the machine, so the same algorithm can be used on a variety of machines without change.
• Implementations of the present disclosure provide a means of improving the positioning accuracy of automation equipment that works near its singularity. Where accuracy can be improved, costs of acquisition of new capital equipment, and maintenance costs for existing equipment, can be lowered. Furthermore, various implementations of the present disclosure can provide a single method that works on several different machine geometries, thereby reducing a need for software development and maintenance to deal with new machine geometries.
• Calibrated machine tools can be provided which are capable of producing more accurate work. Thus higher precision parts can be made with less waste material. When implemented in connection with aircraft production, the production of higher- precision parts can result in reduced weight and improved performance of aircraft. Implementations of the disclosure can result in reduced set-up time and more productive use of machine tools, thereby reducing the cost of machining.
• Various implementations of the foregoing methods and systems make it possible to use software to compensate machines that have complex singularity configurations (e.g., a six-axis revolute robot having rotating joints). Compensation can be provided "on the fly” for a wide range of machine configurations, even without precise a priori knowledge of the location or nature of all singularities for a given machine. The same core compensation software can be used on a wide range of machine configurations without a need to rewrite the software based on machine geometry.

Landscapes

• Engineering & Computer Science (AREA)
• Physics & Mathematics (AREA)
• Mathematical Physics (AREA)
• Automation & Control Theory (AREA)
• Robotics (AREA)
• Mechanical Engineering (AREA)
• Numerical Control (AREA)
• Manipulator (AREA)

Priority Applications (3)

Applications Claiming Priority (2)

Publications (2)

Family

ID=39136516

Family Applications (1)

Country Status (5)

Families Citing this family (13)

Family Cites Families (18)

• 2006
  • 2006-08-31 US US11/514,072 patent/US8010235B2/en not_active Expired - Fee Related
• 2007
  • 2007-08-27 CN CN2007800306231A patent/CN101505925B/zh not_active Expired - Fee Related
  • 2007-08-27 EP EP07837375.0A patent/EP2073959B1/en not_active Not-in-force
  • 2007-08-27 JP JP2009526668A patent/JP5393461B2/ja not_active Expired - Fee Related
  • 2007-08-27 WO PCT/US2007/018820 patent/WO2008027334A2/en not_active Ceased

Also Published As

Similar Documents

Legal Events