Networks Flashcards
arcs
edges, links,branches
in the course:
arcs are directed
links are undirected
arc can have a capacity : max of flow that can pass through them
net flow
the difference in flow of both direction
it is possible to add a fictional flow in the opposite direction to reduce the net flow
spanning tree
A spanning tree is a subset of Graph G, which has all the vertices covered with minimum possible number of edges. Hence, a spanning tree does not have cycles and it cannot be disconnected.
spanning trees represent the BFS of the network simplex. any set of n-1 arc of a minimum cost problem is a spanning tree as there cant be any undirected cycles (otherwise this leads to averages of other BFS)
types of node
supply/source: flow out of node (will have a + flow)
demand/sink: flow into node (will have a - flow)
transshipment node: no net flow, as much in then out (will have a 0 flow)
minimum cost flow problem
all the flow from the supply nodes need to reach the demand nodes with minimum costs (if there would be mismatch: add dummy supply or dummy demand).
Arcs have capacities that can limit some paths.
directed network
xij flow from i to j cij cost from i to j uij capacity from i to j bi net flow at i objective minimizing total cost sum i sum j cij xij subject to sum j xij - sum j xji = bi for each i ( flow difference = net flow)
if there is a lower bound : xij’ = xij - Lij -> xij = xij’ + Lij. replace everywhere xij with this
Integral Property: if all numbers are integral our solution will also be integral
shortest path problem
finding edges from point to point b with smallest associated cost
maximum flow problem
need to bring as much flow from the supply node to the demand node but arcs have capacities that limit it.
min cut max flow theorem: the size of the max flow is equal to the minimum total capacity of any set of arcs whose removal destroys all directed path from source to sink
transportation to minimum cost
supply node for each source. demand node for each destination. no arc capacity. cost remains the same per unit of flow. arcs between each source and destination.
supply + demain -
There are no transhipment nodes (nodes that don’t supply or demand), the supply and demand are directly connected
assignment to minimum cost
same as transportation:
supply node for each source. demand node for each destination. no arc capacity. cost remains the same per unit of flow. arcs between each source and destination.
BUT number of supply nodes = number of demand nodes AND supply, demands are 1
shortest path to minimum cost
have arcs going in both direction. have supply and demand of 1
Integrality property will ensure this is a path.
max flow to minimum cost
have all cost=0 as no cost is associated with max flow problem. 1 supply , 1 demand node. Choose an upper bound flow to add to supply and demand node. Add arc (dummy edge) from supply to demand node with unlimited capacity and a cost of M
number of basic variables
if there are n nodes then there are n-1 basic variables and these form a spanning tree.