Stochastic Model for Machining Processes—Optimal Decision-Making and Control

Abstract

An automatic screw machine process is used as an example to investigate machining operations. Experiments were conducted and a stochastic model for the time series on observed performance of the machine was identified. With the stochastic model, it is possible to forecast the future performance. A simple cost model is assumed for the machining operation and an operating criterion of minimum expected cost per period of operation is used to determine the optimal operating policy. A rolling policy method was used to obtain the optimal operating policy for an indefinite planning horizon. The automatic screw machine process was simulated on a digital computer to determine its performance. The stochastic portion of the deviations from the target was corrected, and the extent of correction was shown to be determined by the cost parameters.