Showing papers in "Journal of Mathematical Modelling and Algorithms in 2004"

An Ant Colony Optimization Algorithm for Shop Scheduling Problems

Abstract: We deal with the application of ant colony optimization to group shop scheduling, which is a general shop scheduling problem that includes, among others, the open shop scheduling problem and the job shop scheduling problem as special cases. The contributions of this paper are twofold. First, we propose a neighborhood structure for this problem by extending the well-known neighborhood structure derived by Nowicki and Smutnicki for the job shop scheduling problem. Then, we develop an ant colony optimization approach, which uses a strong non-delay guidance for constructing solutions and which employs black-box local search procedures to improve the constructed solutions. We compare this algorithm to an adaptation of the tabu search by Nowicki and Smutnicki to group shop scheduling. Despite its general nature, our algorithm works particularly well when applied to open shop scheduling instances, where it improves the best known solutions for 15 of the 28 tested instances. Moreover, our algorithm is the first competitive ant colony optimization approach for job shop scheduling instances.

The analysis of evolutionary algorithms on sorting and shortest paths problems

Abstract: The analysis of evolutionary algorithms is up to now limited to special classes of functions and fitness landscapes. E.g., it is not possible to characterize the set of TSP instances (or another NP-hard combinatorial optimization problem) which are solved by a generic evolutionary algorithm (EA) in an expected time bounded by some given polynomial. As a first step from artificial functions to typical problems from combinatorial optimization, we analyze simple EAs on well-known problems, namely sorting and shortest paths. Although it cannot be expected that EAs outperform the well-known problem specific algorithms on these simple problems, it is interesting to analyze how EAs work on these problems. The following results are obtained: - Sorting is the maximization of “sortedness” which is measured by one of several well-known measures of presortedness. The different measures of presortedness lead to fitness functions of quite different difficulty for EAs. - Shortest paths problems are hard for all types of EA, if they are considered as single-objective optimization problems, whereas they are easy as multi-objective optimization problems.

Helper-Objectives: Using Multi-Objective Evolutionary Algorithms for Single-Objective Optimisation

Abstract: This paper investigates the use of multi-objective methods to guide the search when solving single-objective optimisation problems with genetic algorithms. Using the job shop scheduling and travelling salesman problems as examples, experiments demonstrate that the use of helper-objectives (additional objectives guiding the search) significantly improves the average performance of a standard GA. The helper-objectives guide the search towards solutions containing good building blocks and help the algorithm escape local optima. The experiments reveal that the approach works if the number of simultaneously used helper-objectives is low. However, a high number of helper-objectives can be used in the same run by changing the helper-objectives dynamically. The experiments reveal that for the majority of problem instances studied, the proposed approach significantly outperforms a traditional GA.

A No-Free-Lunch Theorem for Non-Uniform Distributions of Target Functions

Abstract: The sharpened No-Free-Lunch-theorem (NFL-theorem) states that, regardless of the per- formance measure, the performance of all optimization algorithms averaged uniformly over any finite set F of functions is equal if and only if F is closed under permutation (c.u.p.). In this paper, we first summarize some consequences of this theorem, which have been proven recently: The number of subsets c.u.p. can be neglected compared to the total number of possible subsets. In particular, problem classes relevant in practice are not likely to be c.u.p. The average number of evaluations needed to find a desirable (e.g., optimal) solution can be calculated independent of the optimization algorithm in certain scenarios. Second, as the main result, the NFL-theorem is extended. Neces- sary and sufficient conditions for NFL-results to hold are given for arbitrary distributions of target functions. This yields the most general NFL-theorem for optimization presented so far. Mathematics Subject Classifications (2000): 90C27, 68T20.

The linear ordering problem: Instances, search space analysis and algorithms

Abstract: The linear ordering problem is an NP-hard problem that arises in a variety of applications. Due to its interest in practice, it has received considerable attention and a variety of algorithmic approaches to its solution have been proposed. In this paper we give a detailed search space analysis of available benchmark instance classes that have been used in various researches. The large fitness-distance correlations observed for many of these instances suggest that adaptive restart algorithms like iterated local search or memetic algorithms, which iteratively generate new starting solutions for a local search based on previous search experience, are promising candidates for obtaining high performing algorithms. We therefore experimentally compared two such algorithms and the final experimental results suggest that, in particular, the memetic algorithm is a new state-of-the-art approach to the linear ordering problem.

On Solving Quickest Time Problems in Time-Dependent, Dynamic Networks

Abstract: In this paper, a pseudopolynomial time algorithm is presented for solving the integral time-dependent quickest flow problem (TDQFP) and its multiple source and sink counterparts: the time-dependent evacuation and quickest transshipment problems. A more widely known, though less general version, is the quickest flow problem (QFP). The QFP has historically been defined on a dynamic network, where time is divided into discrete units, flow moves through the network over time, travel times determine how long each unit of flow spends traversing an arc, and capacities restrict the rate of flow on an arc. The goal of the QFP is to determine the paths along which to send a given supply from a single source to a single sink such that the last unit of flow arrives at the sink in the minimum time. The main contribution of this paper is the time-dependent quickest flow (TDQFP) algorithm which solves the TDQFP, i.e. it solves the integral QFP, as defined above, on a time-dependent dynamic network, where the arc travel times, arc and node capacities, and supply at the source vary with time. Furthermore, this algorithm solves the time-dependent minimum time dynamic flow problem, whose objective is to determine the paths that lead to the minimum total time spent completing all shipments from source to sink. An optimal solution to the latter problem is guaranteed to be optimal for the TDQFP. By adding a small number of nodes and arcs to the existing network, we show how the algorithm can be used to solve both the time-dependent evacuation and the time-dependent quickest transshipment problems.

A General Approach for Constraint Solving by Local Search

Abstract: In this paper, we present a general approach for solving constraint problems by local search. The proposed approach is based on a set of high-level constraint primitives motivated by constraint programming systems. These constraints constitute the basic bricks to formulate a given combinatorial problem. A tabu search engine ensures the resolution of the problem so formulated. Experimental results are shown to validate the proposed approach.

Solving the Probabilistic TSP with Ant Colony Optimization

Abstract: In this paper, we describe new ways to apply Ant Colony Optimization (ACO) to the Probabilistic Traveling Salesperson Problem (PTSP). PTSP is a stochastic extension of the well known Traveling Salesperson Problem (TSP), where each customer will require a visit only with a certain probability. The goal is to find an a priori tour visiting all customers with minimum expected length, customers not requiring a visit simply being skipped in the tour.We show that ACO works well even when only an approximative evaluation function is used, which speeds up the algorithm, leaving more time for the actual construction. As we demonstrate, this idea can also be applied successfully to other state-of-the-art heuristics. Furthermore, we present new heuristic guidance schemes for ACO, better adapted to the PTSP than what has been used previously. We show that these modifications lead to significant improvements over the standard ACO algorithm, and that the resulting ACO is at least competitive to other state-of-the-art heuristics.

Tabu Search Heuristics for the Arc Routing Problem with Intermediate Facilities under Capacity and Length Restrictions

Abstract: This paper deals with the Arc Routing Problem with Intermediate Facilities under Capacity and Length Restrictions(CLARPIF), a variant of the classical Capacitated Arc Routing Problem(CARP), in which vehicles may unload or replenish at intermediate facilities and the length of any route may not exceed a specified upper bound. Three heuristics are developed for the CLARPIF: the first is a constructive procedure based on a partitioning approach while the second and the third are tailored Tabu Search procedures. Computational results on a set of benchmark instances with up to 50 vertices and 92 required edges are presented and analyzed.

The Attribute Based Hill Climber

Abstract: In this paper we introduce the Attribute Based Hill Climber, a parameter-free algorithm that provides a concrete, stand-alone implementation of a little used technique from the Tabu Search literature known as “regional aspiration”. Results of applying the algorithm to two classical optimisation problems, the Travelling Salesman Problem and the Quadratic Assignment Problem, show it to be competitive with existing general purpose heuristics in these areas.

A Multiobjective Hybrid Metaheuristic Approach for GIS-based Spatial Zoning Model

Abstract: This paper presents a multiobjective hybrid metaheuristic approach for an intelligent spatial zoning model in order to draw territory line for geographical or spatial zone for the purpose of space control. The model employs a Geographic Information System (GIS) and uses multiobjective combinatorial optimization techniques as its components. The proposed hybrid metaheuristic consists of the symbiosis between tabu search and scatter search method and it is used heuristically to generate non-dominated alternatives. The approach works with a set of current solution, which through manipulation of weights are optimized towards the non-dominated frontier while at the same time, seek to disperse over the frontier by a strategic oscillation concept. The general procedure and its algorithms are given as well as its implementation in the GIS environment. The computation has resulted in tremendous improvements in spatial zoning.

A Scatter Search Method for the Bi-Criteria Multi-dimensional 0,1-Knapsack Problem using Surrogate Relaxation

Abstract: This paper presents a scatter search (SS) based method for the bi-criteria multi-dimensional knapsack problem. The method is organized according to the usual structure of SS: (1) diversification, (2) improvement, (3) reference set update, (4) subset generation, and (5) solution combination. Surrogate relaxation is used to convert the multi-constraint problem into a single constraint one, which is used in the diversification method and to evaluate the quality of the solutions. The definition of the appropriate surrogate multiplier vector is also discussed. Tests on several sets of large size instances show that the results are of high quality and an accurate description of the entire set of the non-dominated solutions can be obtained within reasonable computational time. Comparisons with other meta-heuristics are also presented. In the tested instances the obtained set of potentially non-dominated solutions dominates the set found with those meta-heuristics.

Resource Constrained Shortest Path Problems in Path Planning for Fleet Management

Abstract: In the management and control of a vehicle fleet on a road network, the problem arises of finding the best route in relation to the mission of the fleet and to the typology of freight or users. Sometimes the route has to be adapted not only to current traffic conditions, but also to the physical, geometric and functional attributes of the roads, related to their urban location and environmental characteristics. This problem is approached here through an extension of the classic Shortest Path problem, named Resource Constrained Shortest Path problem (RCSP), where the resources are related to the road link attributes, assumed as relevant to the specific planning problem. A classification scheme of these attributes is first proposed and RCSP is described and reviewed. Next, a General Resource Constrained Shortest Path problem (GRCSP) is defined, which can be found in all cases where it is necessary to plan, statically or dynamically, the path of a vehicle and to respect a set of constraints expressed in terms of parameters and indices associated with the roads on the network. For this general problem a model has been formulated and a Branch and Cut solution approach is proposed. Computational results obtained on test and real networks during the experimentation of a fleet with low emission vehicles are also given.

New Models of the Generalized Minimum Spanning Tree Problem

Abstract: We consider a generalization of the Minimum Spanning Tree Problem, called the Generalized Minimum Spanning Tree Problem, denoted by GMST. It is known that the GMST problem is NP-hard. We present a stronger result regarding its complexity, namely, the GMST problem is NP-hard even on trees as well an exact exponential time algorithm for the problem based on dynamic programming. We describe new mixed integer programming models of the GMST problem, mainly containing a polynomial number of constraints. We establish relationships between the polytopes corresponding to their linear relaxations. Based on a new model of the GMST we present a solution procedure that solves the problem to optimality for graphs with nodes up to 240. We discuss the advantages of our method in comparison with earlier methods.

A Tabu Search Heuristic for a Full-Load, Multi-Terminal, Vehicle Scheduling Problem with Backhauling and Time Windows

Abstract: The problem considered is the full-load pickup and delivery problem with time windows (PDPTW), and heterogeneous products and vehicles, where the assignment of pickup points to requests is not predetermined. Elements associated with tabu search, such as diversification by reversion to ‘junctions’ and the use of soft aspiration criteria, are embedded into our tabu search implementation. This metaheuristic is evaluated using random instances and selected data from a construction company in the U.K. The obtained results are compared against lower bounds from LP relaxation and also solutions from an existing multi-level heuristic.

A Meta-heuristic with Orthogonal Experiment for the Set Covering Problem

Abstract: This paper reports an evolutionary meta-heuristic incorporating fuzzy evaluation for some large-scale set covering problems originating from the public transport industry. First, five factors characterized by fuzzy membership functions are aggregated to evaluate the structure and generally the goodness of a column. This evaluation function is incorporated into a refined greedy algorithm to make column selection in the process of constructing a solution. Secondly, a self-evolving algorithm is designed to guide the constructing heuristic to build an initial solution and then improve it. In each generation an unfit portion of the working solution is removed. Broken solutions are repaired by the constructing heuristic until stopping conditions are reached. Orthogonal experimental design is used to set the system parameters efficiently, by making a small number of trials. Computational results are presented and compared with a mathematical programming method and a GA-based heuristic.

Planar Maximum Box Problem

Abstract: Given two finite sets of points X + and X − in ℝ d , the maximum box problem asks to find an axis-parallel box B such that B∩X −=∅ and the total number of points from X + covered is maximized. In this paper we consider the version of the problem for d = 2 (and find the smallest solution box). We present an O(n 3 log4 n) runtime algorithm, thus improving previously best known solution by almost quadratic factor.

Finding Robust Solutions Using Local Search

Abstract: This paper investigates how a local search metaheuristic for continuous optimisation can be adapted so that it finds broad peaks, corresponding to robust solutions. This is relevant in problems in which uncertain or noisy data is present. When using a genetic or evolutionary algorithm, it is standard practice to perturb solutions once before evaluating them, using noise from a given distribution. This approach however, is not valid when using population-less techniques like local search and other heuristics that use local search. For those algorithms to find robust solutions, each solution needs to be perturbed and evaluated several times, and these evaluations need to be combined into a measure of robustness. In this paper, we examine how many of these evaluations are needed to reliably find a robust solution. We also examine the effect of the parameters of the noise distribution. Using a simple tabu search procedure, the proposed approach is tested on several functions found in the literature.

Two-Machine Flow Shop Scheduling with Nonregular Criteria

Abstract: We consider a two-machine flow shop problem with a common due date where the objective is to minimize the sum of functions which penalize early as well as tardy completion of jobs. Since the problem is NP-hard in the strong sense, we investigate some general properties of optimal schedules for the problem, we develop lower and upper bounds, derive dominance criteria, and propose an enumerative algorithm for finding an optimal schedule. The performance of the proposed algorithm together with the influence of the individual components is thoroughly discussed.

Unification of some Least Squares Clustering Methods

Abstract: We give a unifying description of several least squares clustering methods. We consider the K-means/Lloyd's algorithm, various local search based methods and clustering techniques based on neural networks. The relations between the various algorithms are explained. Also some computational results of these algorithms are given.

Numerical Modeling of Solitons in a Low Birefringent Optical Fiber

Abstract: A numerical investigation, based on the split-step Fourier transform algorithm of all optical switching of solitons in a low birefringent optical fiber is presented. The numerical algorithm is described in detail.