Dependability & Maintainability Theory and Methods 4. State Enumeration Andrea Bobbio Dipartimento di Informatica Università del Piemonte Orientale, “A. Avogadro” 15100 Alessandria (Italy) [email protected] http://www.mfn.unipmn.it/~bobbio/IFOA/ A. Bobbio IFOA, Reggio Emilia, June 2003 17-18, 2003 Reggio Emilia, June 17-18, 1 State space Consider a system with n binary components. We introduce an indicator variable x i : 1 component i up x i= 0 component i down The state of the system can be identified as a vector x = (x 1, x 2, . . . . , x n ) . The state space (of cardinality 2 n ) is the set of all the possible values of x. A. Bobbio Reggio Emilia, June 17-18, 2003 2 2-component system A. Bobbio Reggio Emilia, June 17-18, 2003 3 3-component system A. Bobbio Reggio Emilia, June 17-18, 2003 4 Characterization of system states The system has a binary behavior. We introduce an indicator variable for the system y: 1 system up y = 0 system down For each state s corresponding to a single value of the vector x = (x 1, x 2, . . . . , x n ) . 1 system up y = (x)= 0 system down y = (x) is the structure function Characterization of system states The structure function y = (x) depends on the system configuration The state space can be partitioned in 2 subsets: A1 A2 2-component system A1 A2 A. Bobbio Reggio Emilia, June 17-18, 2003 7 3-component system a) A1 A3 A2 b) A1 A2 A3 A. Bobbio Reggio Emilia, June 17-18, 2003 8 State probability Define: Pr{x i(t) = 1} = R i (t) Pr{x i(t) = 0} = 1 - R i (t) Suppose components are statistically independent; The probability of the system to be in a given state x = (x 1, x 2, . . . . , x n ) at time t is given by the product of the probability of each individual component of being up or down. P {x(t)} = Pr{x 1(t)} · Pr{x 2(t)} · … ·Pr{x n(t)} A. Bobbio Reggio Emilia, June 17-18, 2003 9 A1 2-component system A1 A2 A2 A. Bobbio Reggio Emilia, June 17-18, 2003 10 3-component system A. Bobbio Reggio Emilia, June 17-18, 2003 11 Dependability measures A. Bobbio Reggio Emilia, June 17-18, 2003 12 Dependability measures A. Bobbio Reggio Emilia, June 17-18, 2003 13 2-component series system A1 A. Bobbio Reggio Emilia, June 17-18, 2003 A2 14 2-component parallel system A1 A2 A. Bobbio Reggio Emilia, June 17-18, 2003 15 3-component system a) A1 A3 A2 b) A1 A2 A3 A. Bobbio Reggio Emilia, June 17-18, 2003 16 A1 Voter A2 A3 3-component system 2:3A. majority voting Bobbio Reggio Emilia, June 17-18, 2003 17 5 component systems A. Bobbio Reggio Emilia, June 17-18, 2003 18 Non series-parallel systems with 5 components A1 Independent identically distributed components A. Bobbio Reggio Emilia, June 17-18, 2003 A4 A2 A3 A5 19