Smoothing Production by Inventory—An Application of the Wiener Filtering Theory

In many cases it will be possible to separate a production smoothing problem into two parts. One part describes a feedforward production smoothing which may be considered as a “gross regulation.” The other part then smoothes the remaining regulation errors of the first part. This “fine regulation” is performed by an application of the Wiener Newton Theory which allows us to calculate an optimal feedback production policy. It can be shown that if the output of the feedforward part (regulation error) is Gaussian, then the Wiener Newton Theory may be applied even in the case of non-quadratic cost criteria. It is not required that demand in successive time periods be independent, as is usually assumed in the classical approaches working with dynamic programming.

• 저널Management Science · 17(7) · 472–483
• 토픽Flexible and Reconfigurable Manufacturing Systems · Industrial and Manufacturing Engineering
• DOI10.1287/mnsc.17.7.472
• 저자Christoph Schneeweiß