Decision Making in Systems Engineering and Management, 3rd Edition

Chapter 10Stochastic Tradespace Analysis

The only thing that makes life possible is permanent, intolerable uncertainty; not knowing what comes next.

—Ursula K. Le Guin, Author

10.1 INTRODUCTION

In practice, nearly all multiple criteria value modelling (MCVM) tradeoff applications involve three primary decision support efforts: down‐selecting post‐screening alternatives, reducing the choice set to isolate Pareto efficient alternatives, and selecting a single alternative that maximizes value return against a tradeoff dimension. Both deterministic and stochastic tradespace analysis adhere to this pattern, the difference between them being an explicit treatment of uncertainty.

A satisfactory selection decision could be made on deterministic tradespace results alone, and certainly has been in the past. Stochastic tradespace analysis provides a more robust understanding of the complications associated with the decision that uncertainty imposes, taking what appears to be a straightforward decision from a deterministic analysis to one that introduces risk both with respect to system performance and to the decision maker.

As discussed in Chapter 9, the Decision Making phase of the systems decision process (SDP) places an even heavier emphasis on the decision maker who must choose among competing system alternatives. There, the discussion centered on using a deterministic approach in which cost and value data and model parameters were assumed to be known with certainty, even if this ...