Catálogo de publicaciones - libros

Production planning problems play a vital role in the supply chain management. The methodology of production planning problem can provide the quantity of production and the workforce level at each production plant to fulfil market demand. This paper develops a stochastic programming model with additional constraints. A set of data from a multinational lingerie company in Hong Kong is used to demonstrate the robustness and effectiveness of the proposed model.

Subgradient optimization methods provide a valuable tool for obtaining a lower bound of specially structured linear programming or linear programming relaxation of discrete optimization problems. However, there is no practical rule for obtaining primal optimal solutions from subgradient-based approach other than the lower bounds. This paper presents a class of procedures to recover primal solutions directly from the information generated in the process of using subgradient optimization methods to solve such Lagrangian dual formulations. We also present a hybrid primal dual algorithm based on these methods and some computational results.

We consider the problem of static load balancing with the objective of minimizing the job response times. The jobs that arrive at a central scheduler are allocated to various processors in the system with certain probabilities. This optimization problem is solved using real-coded genetic algorithms. A comparison of this approach with the standard optimization methods are presented.

We reformulate the linear second-order cone complementarity problem into a system of nonlinear equations. Our reformulation is different from others. Our algorithm for the reformulation can start from an arbitrary point. We prove that our algorithm approximates an optimum of the linear second-order cone complementarity problem in finite steps under certain conditions. Finally, we show that the system of nonlinear equations of our reformulation is nonsingular at optimum under certain conditions.

This paper considers a PCB group setup strategy in a single placement machine producing multiple products. The grouping objective is to minimize makespan that is composed of two attributes, the feeder change time and the placement time. Current group setup strategy considers only component similarity in grouping boards. In this paper, we propose an improved group setup strategy which considers both component similarity and geometric similarity using minimum metamorphic distance measures. The proposed group setup strategy is tested on a variety of production environments with existing group setup strategies and board’s similarity coefficients. Test results show that the improved group setup strategy outperforms existing group setup strategies.

The conceptual block layouts generated by layout software packages very often yield detailed layouts with non-smooth material flow paths that have an excessive number of turns. Such non-smooth flow paths are not acceptable or feasible for actual construction. This paper presents a mixed integer programming model to modify a given block layout into a layout that allows the construction of smooth flow paths, while preserving the relative positions of the departments in the original block layout. Results of the computational experiments show that much smoother flow paths are generated in the modified block layouts than those in the original block layouts without significantly increasing the travel-based quality of final layouts.

A company needs to implement several make-to-stock policies apart from a regular make-to-order production, so that the capacity of expensive resources can be fully utilized. The constraints to be considered in such capacity planning problem include finite budget for investing resources, lump demands of customers, decline of products price with time, different product mix for simultaneous manufacturing, time value of capital asset, technology levels of resources, and limited capacity of resources. We focus on the issues of resources acquisition and allocation decision in each production period. The goal is to maximize the long-term profit. This study formulates the problem as a non-linear mixed integer mathematical programming model. A constraint programming based genetic algorithm is developed to solve the problem efficiently.

In a Vendor-Managed Inventory (VMI) system, the supplier or the distributor is authorized to coordinate and consolidate the inventories at the retailers. The advantage of VMI is that the bullwhip effect can be minimized and the stock-out situations can also be reduced. Moreover, it provides a framework for synchronizing transportation decisions and hence reduce the transportation cost significantly. In this paper, we present an analytic model for quantity-time-based dispatching policy. The model discussed here takes into the account of the inventory cost, the transportation cost, the dispatching cost and the re-order cost. Since a new inventory cycle begins whenever there is a dispatching of products, the long-run average costs of the model can be obtained by using the renewal theory. We also derive a closed form solution of the optimal dispatching policy.

This paper presents a method for determining multiple attribute weights when pairwise comparison judgments on alternatives are specified and attribute consequences are captured in imprecise ways. A decision-maker or expert can express holistic pairwise comparisons on alternatives from his/her domain knowledge and decision alternatives are characterized by some tangible or possibly some intangible multiple attributes of which consequences can be represented by imprecise information. In this paper, attribute weights are to be estimated in the direction of minimizing the amount of violations and thus to be as consistent as possible with a decision-maker’s ordered pairs. Multiple attribute weights that were determined with pairwise judgments on a subset of alternatives can be used to prioritize the other remaining alternatives.

This paper deals with the simultaneous determination of initial sales price, secondary sales price and resource capacity needed to be rented in a service rental system. It is assumed that demands are price and waiting time dependent. A deterministic model is developed for establishing the above-mentioned decisions. Analytical results show that optimal solutions can be found as a closed form formulation when the times to change sales prices are exogenous variables. In addition, we found sufficient conditions for optimization in the case where the times to change sales price are decision variables.