Nonlinear Programming

Optimization of single mixed refrigerant natural gas liquefaction plant with nonlinear programming

by Muhammad Shariq Khan

Asia-Pacific Journal of Chemical Engineering
Supplement: Selected Papers from the 13th Asia-Pacific Confederation of Chemical Engineering (APCChE) Congress, 5–8 October 2010, Taipei, Taiwan
Volume 7, Issue Supplement S1, pages S62–S70, May 2012

Article first published online: 8 NOV 2011
DOI: 10.1002/apj.642

The liquefaction of natural gas (NG) in a mixed refrigerant (MR) system is an energy-demanding process. Much energy is... more The liquefaction of natural gas (NG) in a mixed refrigerant (MR) system is an energy-demanding process. Much energy is wasted because of its irreversibilities and its nonoptimal execution. The most important factors affecting this process's performance are the refrigerant's composition and flow rate, the suction and evaporation pressures, and the extent of refrigerant vaporization. They should be adjusted to optimize the overall operation. The adjustment of one of these variables will affect the other because of their highly nonlinear interactions. This work reports the optimization of a single MR (SMR) process of NG liquefaction. The SMR process was modeled in the UniSim Design commercial process plant simulator, and the model was optimized for compression energy with nonlinear programming (NLP) while satisfying constraints. The base case for optimization was selected by mesh searching, and case study demonstrates that NLP can reduce energy use and improve the process's efficiency. © 2011 Curtin University of Technology and John Wiley & Sons, Ltd.

Total Discount Policy and Two Warehouses Strategy to Store Raw Materials with Economic Order Quantity Model

by Saeid Mokaram

A. A. Taleizadeh, S. Mokaram, N. Shafii and M. Zarei, “Total Discount Policy and Two Warehouses Strategy to Store Raw Materials with Economic Order Quantity Model”, Journal of Applied Sciences, 9: 1267-1275, 2009.

This study introduced an Economic Order Quantity (EOQ) model with payment in advance to purchase high-price raw... more This study introduced an Economic Order Quantity (EOQ) model with payment in advance to purchase high-price raw materials. We relax and change some assumptions that were considered in earlier researches. At first we considered transportation cost as a linear function. Total discount policy is considered instead of incremental discount one. Also we developed model based on two warehouses strategy to store raw material in which holding cost is different for each of warehouses. We show that the model of this problem is shown to be a mixed-integer-nonlinear-programming type and in order to solve it, a simulated annealing approach is used. At the end, a numerical example is given to demonstrate the applicability of the proposed methodology in real world inventory control problems.

Social cognitive optimization for nonlinear programming problems

by Xiao-Feng Xie

A freight transshipment network model for forecasting commodity and cyclic commercial vehicle flows

by Joseph Chow

Chow, J.Y.J., Ritchie, S.G., 2012. A freight transshipment network model for forecasting commodity and commercial vehicle flows. Proceedings of the Transportation Research Board 91st Annual Meeting (forthcoming).

A freight forecast model that assigns commodity flows to cyclic commercial vehicles is proposed in this study. The... more A freight forecast model that assigns commodity flows to cyclic commercial vehicles is proposed in this study. The commercial vehicles are formulated to traverse in cycles and include loading and unloading costs at zone centroids. Empty hauls can be tracked, as can transshipment flows by commodity type and by inbound and outbound modes. A linear programming formulation is proposed as well as nonlinear objectives for link and transshipment congestion. An inverse nonlinear programming approach using Karush-Kuhn-Tucker conditions is formulated to calibrate the congestion parameters of this model such that observed flow variables are optimal. Because the forward problem is convex and composed of only equality or non-negativity constraints, it can be readily solved with classical nonlinear optimization methods instead of treating the inverse problem as a nonlinear complementarity problem. The models are tested on a 6-node network with up to 54 transshipment activities. The model is shown to be sensitive to supply side changes on links and transshipment facilities or to fuel cost changes. The appendix includes an inverse traffic assignment problem using the inverse nonlinear programming method.