Sampling Techniques

Abstract

Sampling is a statistical procedure used for uncertainty analysis and stochastic modeling in many applications related to process design, operation, control and optimization. The increased efforts for pollution prevention and sustainability necessitate a more contemporary approach for process design, in which objectives such as reducing environmental and health impacts, increasing reliability, safety and controllability, reducing risk and achieving better profitability are considered simultaneously. In this multifaceted approach to process design and operation, uncertainties need to be considered and included in process models and simulations. These uncertainties can be static or dynamic. A review of sampling techniques such as Monte Carlo sampling, importance sampling, Latin Hypercube Sampling (LHS) and Hammersley Sequence Sampling (HSS) are provided here. These sampling techniques have a wide range of use in chemical and process synthesis, process scheduling, supply chain management, process control and reliability as well as risk management. Apart from uncertainty analysis, sampling and sampling accuracy plays an important role in discrete, stochastic, and multi‐objective optimization algorithms. This review also describes the recent developments in this area. In addition, future trends in process design and sampling techniques are also presented.