PROCESS-RELATED SIMULATION APPLIED TO MANUFACTURING OPTIMIZATION

As processes in production, logistics, chemical engineering, and so on , become more complex, human experience alone is no longer sufficient to optimize the processes purposefully and without any doubts. This is because the varied dependencies, with their typically non-linear rela tionships, cannot be assimilated. Mathematically based simulations are necessary to move from use of trial and error methods to obtain optim um processes. These are integrated into the process to allow direct an d continuous control during changing conditions to achieve an optimum in such things as yield, costs, and time. The simulations can also be used to obtain information about the effects of trying different metho ds before actually testing them. For processes which are subject to ra ndom effects, the results of the mathematical simulations describe the statistical limits of values, which themselves depend on parameters w hich have statistical limits. Changing one parameter shows immediately whether the result would be within the old process limit or not. This allows the effect of changes in parameters to be evaluated before exp ensive and time-consuming work is done. Several different simulations are discussed: one to determine the optimal amount of stock with chang eable market conditions, one to opimize an air dryer used in the food industry for different product requirements, one to run a bio-fermente r plant, and one to obtain information about the effects of changing p arameters in a modern mass-production process which runs under random effect conditions.