Risk analysis of cost and schedule of complex engineering systems

Risk is inherent in engineering systems. A kernel element of risk that requires adequate treatment is uncertainty represented by plural outcomes and associated future likelihoods. This research examines treatment of uncertainty and provides methodologies for risk analysis of cost and schedule of complex engineering systems. The uncertainty intrinsic in a system influences requisite risk analysis methodology. Probabilistic, ambiguous, or aleatory uncertainty entails stochastic methods. Cognitive, vague, or epistemic uncertainty requires fuzzy sets and logic. Simulation overcomes difficulty, or known intractability, associated with mathematical formulation of analytical models of complex engineering systems. Methodology for fuzzy stochastic risk analysis by simulation of cost and schedule networks is provided. Engineering decisions involving risk require decision models with systematic frameworks to consider pertinent facets of decision problems. Methodology covering continuous, discrete, static, and dynamic aspects of problems for risk-based decision analysis is provided. Classical decision analysis assumes probabilities associated with consequences are known numerically. With complex engineering systems, available probabilities can be unreliable, incomplete, or imprecise and be known as fuzzy rather than crisp numbers. Methodology for the use of fuzzy probabilities in decision analysis is given. A systematic framework consisting of influence diagrams and decision trees is used to incorporate distributions associated with discrete or continuous probabilities and consequences. Dynamic decision models are given that change over time by the updating of variables as additional information is acquired. Fuzzy-Bayesian methodology is presented for updating the state-of-knowledge of cost and schedule information as it is piecewise accumulated and for determination of the value of additional information before acquisition. This research provides a generalized methodology for risk analysis of cost and schedule of complex engineering systems. Methodology includes simulation of deterministic, stochastic, fuzzy, and fuzzy stochastic activities in networks with nodes combining AND, IOR, or EOR entrances and deterministic or stochastic exits. Methodology is given for risk-based decision analysis covering continuous, discrete, static, and dynamic aspects of a decision problem, and using fuzzy-Bayesian methods for updating information and determining the value of information. Methodology provided is general, covering all kinds of uncertainty, and is validated by examples and selected case studies.