This preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: EE236A (Fall 200708) Lecture 8 Network optimization • network flows • minimum cost network flow problem • extreme flows • shortest path problem • bipartite matching 8–1 Networks network (directed graph): m nodes connected by n directed arcs • arcs are ordered pairs ( i, j ) • we assume there is at most one arc from node i to node j • we assume there are no selfloops (arcs ( i, i ) ) arcnode incidence matrix A ∈ R m × n : A ij = 1 arc j starts at node i 1 arc j ends at node i otherwise column sums of A are zero: 1 T A = 0 reduced arcnode incidence matrix ˜ A ∈ R ( m 1) × n : the matrix formed by the first m 1 rows of A Network optimization 8–2 example ( m = 6 , n = 8 ) 1 2 3 4 5 6 1 2 3 4 5 6 7 8 A = 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Network optimization 8–3 Network flow flow vector x ∈ R n • x j : flow (of material, traffic, charge, information, . . . ) through arc j • positive if in direction of arc; negative otherwise total flow leaving node i : n summationdisplay j =1 A ij x j = ( Ax ) i i x j A ij = 1 x k A ik = 1 Network optimization 8–4 External supply supply vector b ∈ R m • b i : external supply at node i • negative b i represents external demand from the network • must satisfy 1 T b = 0 (total supply = total demand) i x j A ij = 1 x k A ik = 1 b i balance equations: Ax = b reduced balance equations: ˜ Ax = ( b 1 , . . . , b m 1 ) Network optimization 8–5 Minimum cost network flow problem...
View
Full Document
 Spring '07
 Dr.Vandenber
 Graph Theory, Shortest path problem, acyclic networks

Click to edit the document details