Sensitivity Analysis of Hardware-In-The-Loop (HWIL) Simulation Systems

J. Pate, K.R. Edlin, K. Parker, and K.I. Kawano (USA)

Keywords

sensitivity analysis, V&V, HWIL, BMD, kill chain

Abstract

Verification and Validation (V&V) of Models and Simulations (M&S) is essential to ensuring timely accreditation decisions. Acceptability criteria for accreditation decisions need broad community consensus to determine when "good enough is good enough." This is especially true for system software and Hardware-In-The Loop (HWIL) simulations. Uncertainties in parameter values used in dynamic simulations lead to difficulties in the a prioi choosing of acceptability criteria for use in an accreditation decision This paper focuses on key parameter variables in the Ballistic Missile Defense kill chain and how sensitivity analysis can be used to determine how sensitive a simulation is to changes in the value of the simulation's inputs, and to changes in the simulation's structure. We will demonstrate how parameter sensitivity analysis was performed in a series of tests where variations in parameter values produced changes in the dynamic behavior of the simulation. We will also illustrate how sensitivity analysis can indicate which acceptability criteria are reasonable choices for use in the accreditation decision. Discovering that the simulation's behavior changes significantly due to a change in some parameter's value can allow the Accreditation Authority to identify threshold values for establishing the acceptability criteria, helping to answer the question: What is "good enough"?

Important Links:



Go Back