Demand Planning Hierarchy Change Propagation

The patent segments the demand planning hierarchy into discrete nodes with defined relationships (parent-child, one-to-one, one-to-many). Each node is treated as an independent unit with its own data structure containing product quantity, disaggregation factor, and flow variables. This segmentation allows efficient localized updates while maintaining global consistency through the mathematical model.

The patent transforms the change propagation problem into a parameter optimization problem by introducing augmenting and relieving flows as mathematical parameters. By solving simultaneous equations that minimize the sum of these flows, the system efficiently determines updated product quantities across all nodes. This parameter-based approach replaces inefficient traditional update methods with a mathematically rigorous optimization framework.

The patent replaces traditional mechanical update propagation methods (sequential or iterative adjustments) with a mathematical substitution approach. By formulating the problem as a system of simultaneous equations with augmenting and relieving flows, the solution can be obtained through direct mathematical computation rather than repeated mechanical updates. This substitution dramatically reduces computational complexity while maintaining precision.

The patent performs preliminary setup by establishing the data structure for each node including all necessary relationships (parent-child, one-to-one, one-to-many) and initializing flow variables before changes are propagated. This preliminary structuring allows the subsequent optimization calculation to proceed efficiently without requiring multiple passes or iterative adjustments, reducing overall computational complexity.