Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

Accelerated Discrete Adjustment of the Parameters of a Process

IP.com Disclosure Number: IPCOM000078012D
Original Publication Date: 1972-Oct-01
Included in the Prior Art Database: 2005-Feb-25
Document File: 2 page(s) / 42K

Publishing Venue

IBM

Related People

Etchison, KL: AUTHOR

Abstract

This mechanism provides a nonlinear adjustment factor. The size of the adjustment factor increases geometrically for a linear increase in the amount of adjustment required, then decreases geometrically for a linear increase in the stability of the process. Mechanisms used previously were: Fixed adjustment factor: some fixed number that is a compromise between quick reaction and fine approximation. Fixed percentage: reaction and fineness of approximation depended on the value of the input. Stops: when change is required, an adjustment is made to the next predetermined level. This mechanism exhibits improvements over previous mechanisms in: Speed with which the optimum value is approached, and Fineness of resolution toward the optimum.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 74% of the total text.

Page 1 of 2

Accelerated Discrete Adjustment of the Parameters of a Process

This mechanism provides a nonlinear adjustment factor. The size of the adjustment factor increases geometrically for a linear increase in the amount of adjustment required, then decreases geometrically for a linear increase in the stability of the process. Mechanisms used previously were: Fixed adjustment factor: some fixed number that is a compromise between quick reaction and fine approximation. Fixed percentage: reaction and fineness of approximation depended on the value of the input. Stops: when change is required, an adjustment is made to the next predetermined level. This mechanism exhibits improvements over previous mechanisms in: Speed with which the optimum value is approached, and Fineness of resolution toward the optimum.

This mechanism works by multiplying or dividing the adjustment factor, based on a history of the number of times adjustment is made in the same direction. Each time the direction of adjustment changes, the adjustment value is divided by a predetermined value (n). The n+1 time, and subsequent times, adjustment is made in the same direction, the adjustment factor is multiplied by a second fixed value (m). The values specified for m and n are chosen to control the speed of the reaction; m to control how quickly the proper vicinity is reached, and n to control how quickly the approximation becomes fine.

This mechanism can be used in an algorithm that controls the output of a ...