Parallel Numerical Algorithms for Optimisation in the presence of Uncertainty


Uncertainty: a natural componet of real world situations.
e.g. exchange rate fluctuations, sensor inaccuracy.
mean variance approach: taking into account decision maker's preferences towards risk
Mathematical Formulation:
Min {(1-a)P(u) + aR(a) | constraints}
Approaches adopted:
  • P(u) = Performance Index
  • R(u) = Sensitivity of System
  • Constraints = General Linear / Non-linear Stochastic
  • Can handle scenario-based or continuous systems