Network Flow Problems
Directed graph with source node and sink node. The edges have different weights.
Send max amount of flow.
constraints [
follow edges direction,
cannot send more than an edges capacity,
flow concentration - total amount of flow egres must be == ingres on each node - this is also Kirchoff's law
]
Formulate as LP
Decision Variables: how much flow each edge is going to carry
Objective Function: Max flow leaving the source or the flow entering the sink (they will be the same)
Constraints [
x1...xn >= 0,
# capacity constraint for every edge...
x1 <= capacity constraint,
# Kirchoff's law
x1 + x2 + x6 = x4 + x5 # etc - depends on the graph
]
Flow problems are typically always feasible and bounded.