Tuning Parameters Using VisTHAA Applied to a Metaheuristic Algorithm That Solves the Order Picking Problem

Tuning Parameters Using VisTHAA Applied to a Metaheuristic Algorithm That Solves the Order Picking Problem

Luis Rodolfo Garcia Nieto (National Institute of Technology of Mexico, Mexico & Technological Institute of Ciudad Madero, Mexico), Claudia Gómez-Santillán (National Institute of Technology of Mexico, Mexico & Technological Institute of Ciudad Madero, Mexico), Laura Cruz-Reyes (National Institute of Technology of Mexico, Mexico & Technological Institute of Ciudad Madero, Mexico), Nelson Rangel-Valdez (National Institute of Technology of Mexico, Mexico & Technological Institute of Ciudad Madero, Mexico) and Héctor J. Fraire-Huacuja (National Institute of Technology of Mexico, Mexico & Technological Institute of Ciudad Madero, Mexico)
DOI: 10.4018/978-1-5225-8131-4.ch005

Abstract

The problem of product transportation has taken great importance due to the demand of products in the market that has to be delivered, and solving the transportation problem is a complex task because it involves the solution of three NP-hard problems. A warehouse involves many operations, such as receiving, locating items, sending an order, among others. For the problem of location of elements several techniques of an arrangement of elements to be able to increase the efficiency of the delivery of orders have been used. In the literature, some studies have studied this problem through the variable cost and size bin packing problem (VCSBPP), which consists of the location of elements in containers of varying sizes and costs. In this chapter, the authors propose an algorithm based on a metaheuristic algorithm, which is called variable neighborhood search for optimizing the VCSBPP solution. This chapter includes an analysis of the factors involved in the solution of the VCSBPP through techniques of parameter setting, in a tool of analysis of heuristic algorithms called VisTHAA.
Chapter Preview
Top

Introduction

Today the virtual network has become a fundamental piece for the product transaction because they allow the demand for products of customers. In a transaction is involved a warehouse since it is fundamental in the relationship between the customers and the suppliers, allowing the guarantee of the storage and the delivery of the items included in the transaction. Manufacturers and distributors today are judged not only by the quality of the product but also by how quickly and efficiently they deliver to customers.

The managers focus on the resources and attention in optimizing the performance of all aspects of the supply chain, the traditional warehouses and distribution centers are recognized as areas where significant performance improvements can be achieved. Most of the traditional models focus on objective functions as the minimization of costs or maximization of throughput, without explicitly considering the customer's point of view.

In the work of Lambert, Stock and Ellram (1998) and other from de Koster, Le-Duc and Roodbergen (2007), they explain that the operations in a warehouse are divided into four main functions, which are: the reception of elements, the transfer of items, the delivery of orders and the shipment of products. Currently, the warehouses have in a way immersed operational problems, like what is the arrangement of lots and the selection of orders.

The Order picking Problem (OPP) is based on the collection of items in a storage warehouse to meet customer demand; this problem is related to different processes that aim at the satisfaction of the client. This collection process tends to be prone to several problems, for example: not being able to locate in a warehouse the requested elements of an order, the use of obsolete strategies for the location of objects in a warehouse, not prioritize the selection of elements of an order and its delivery, misuse of storage space, among others. For the collection of elements have been identified factors affecting the localization of elements in a warehouse, for example, the volume and weight factor of the objects, store capacity or poor metaheuristic implementations for the location of objects.

The OPP is subdivided into three sub-problems for its study; among them are location or storage of items, selection, and shipment of orders. Considering that one of the important factors for the collection of items in a warehouse is to develop a good object-localization strategy, this can be done through various solution strategies such as be deterministic, heuristic, metaheuristics, and hybrid algorithms.

According to Bartz-Beielstein, Chiarandini, Paquete and Preuss (2010), applications have been identified with the various algorithmic strategies that have given a solution to the subproblem of the location of items in stores, which will be studied through the Variable Cost and Size Bin Packing Problem (VCSBPP) which focuses on finding the best solution to the location of items in variable-size containers.

Taking into account that all algorithms need configuration parameters, it is important to identify the best values that will be assigned to the parameters. This area of study is known as the parameter adjustment problem, which studies the selection of the best values of the parameters that regulate the behavior of the algorithms. It is well known that the lack of proper adjustment in the parameters consumes a lot of human and computational resources. Due to its relevance in the scientific and technological fields, the problem has received a lot of attention (Hwang, Baek & Lee, 2007). Over time researchers have created some tools for parameter configuration, within which we can find in the work of Adenso-Díaz and Laguna (2006), Uğur and Aydin (2009), Uğur (2008), Cruz-Reyes et al. (2013), among others.

In this work will focus on the tuning of parameters of algorithms that give a solution to the subproblem of locating or storing items in the warehouses, this problem in the literature has been studied through the VCSBPP. To provide a solution, there will be an analysis of the characteristics of: the instances of the problem, the VCSBPP, and the VNS-VCSBPP solver algorithm; we will be finding the important factors and the levels to work with. All the information is loaded in VisTHAA, which is the parameter adjustment tool to find the best initial configurations in the different instances. The strategy that will be applied to find the configurations is the competences of Hoeffding (1963).

Complete Chapter List

Search this Book:
Reset