Method and apparatus for numerical control

Abstract

Method and apparatus for numerical control to produce contouring by simultaneous coordinated motions along two or more axes, characterized by the creation of desired velocity signals supplied as the primary inputs to velocity servos coupled to drive a member relatively along the respective axes. Although capable of practice by and susceptible of embodiment in other apparatus, a digital computer is physically organized and conditioned by a master program, and associated with an adjustable interrupt clock which measures off successive time periods .DELTA.T, so as to supply a changeable velocity command number signal to each servo proportional to the distance per .DELTA.T through which motion along the associated axis is to occur. This velocity command number is algebraically added to a theoretical position number to signal dynamically during each period .DELTA.T the axis position at which the member should reside; and the latter number is used with a dynamically signaled actual position number to re-compute and signal during each period the position error. The position errors are used as a trimming or adjusting term in the signals applied to the respective axis servos. When the signaled error exceeds a first threshold on any axis, the periods .DELTA.T are lengthened to reduce all axis velocities; when it exceeds a second threshold, the system is put into a holding mode. On the other hand, as the error is reduced below the second threshold, the system returns to the running mode and as the error is reduced below the first threshold the periods .DELTA.T are reduced to increase all axis velocities back to programmed values. The input signal applied to each axis servo includes a further historical adjustment term incrementally changed during any period when the error exceeds a predetermined value --all to the objective of making the axis servos (a) operate with substantially zero following error, (b) correct themselves if the position error becomes at all appreciable, or (c) stop if that error becomes intolerable. Because of the accuracy of velocity control, the system terminates any path segment for one block of command data when the number of elapsed periods .DELTA.T equals a precomputed number of time periods .DELTA.T required for the controlled member to travel the segment length at the commanded velocity expressed in distance per period. No comparison of command end point coordinates with actual end point coordinates is required. Automatic acceleration and deceleration is performed by starting the execution of a block of command data at the commanded velocity, and beginning an incrementing of the velocity command number during each period .DELTA.T after the number of elapsed periods becomes equal to the total of the periods for the entire block less the quantity required for the velocity to change from one block velocity to the next at the chosen incrementing or acceleration value.