Abstract
Sensitivity can be defined as the degree of variability of the output parameters of a system with respect to the variability of the input parameters. This work establishes the link between the sensitivity functions and the Volterra kernel functions, which are an expansion of nonlinear impulse response functions. The Volterra kernel values may be extracted directly from the sample estimates of the statistical moments obtained from the time series data. The three most commonly used definitions of sensitivity are given, and sensitivity analysis is introduced as an important tool in the validation process of the simulation model. Stochastic sensitivity is presented as a complementary technique, and the relationships with the other forms of sensitivity analysis are discussed. As an example, the results from a stochastic sensitivity analysis are compared with the results from a differential sensitivity analysis, and the appropriate range of usage of the stochastic technique is discussed.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call