Realistic Problem Instances Research Articles

A bi-objective model for planning and managing rail-truck intermodal transportation of hazardous materials

We propose a bi-objective optimization framework for routing rail-truck intermodal shipments with hazardous materials, when shippers and receivers have access to alternate intermodal terminals. A tabu-search based solution methodology is developed, which together with the optimization framework is applied to realistic size problem instances to gain managerial insights. Our analysis indicates that drayage accounts for a significant portion of transport risk and that it can be reduced by scheduling direct and faster trains; and, that the mix of intermodal trains depends on the interest of the decision-makers, where the resulting traffic can facilitate planning emergency response systems.

Freight transportation in railway networks with automated terminals: A mathematical model and MIP heuristic approaches

In this paper we propose a planning procedure for serving freight transportation requests in a railway network with fast transfer equipment at terminals. We consider a transportation system where different customers make their requests (orders) for moving boxes, i.e., either containers or swap bodies, between different origins and destinations, with specific requirements on delivery times. The decisions to be taken concern the route (and the corresponding sequence of trains) that each box follows in the network and the assignment of boxes to train wagons, taking into account that boxes can change more than one train and that train timetables are fixed. The planning procedure includes a pre-analysis step to determine all the possible sequences of trains for serving each order, followed by the solution of a 0–1 linear programming problem to find the optimal assignment of each box to a train sequence and to a specific wagon for each train in the sequence. This latter is a generalized assignment problem which is NP-hard. Hence, in order to find good solutions in acceptable computation times, two MIP heuristic approaches are proposed and tested through an experimental analysis considering realistic problem instances.

Large neighbourhood search algorithms for the founder sequence reconstruction problem

The reconstruction of founder genetic sequences of a population is a relevant issue in evolutionary biology research. The problem consists in finding a biologically plausible set of genetic sequences (founders), which can be recombined to obtain the genetic sequences of the individuals of a given population. The reconstruction of these sequences can be modelled as a combinatorial optimisation problem in which one has to find a set of genetic sequences such that the individuals of the population under study can be obtained by recombining founder sequences minimising the number of recombinations. This problem is called the founder sequence reconstruction problem. Solving this problem can contribute to research in understanding the origins of specific genotypic traits. In this paper, we present large neighbourhood search algorithms to tackle this problem. The proposed algorithms combine a stochastic local search with a branch-and-bound algorithm devoted to neighbourhood exploration. The developed algorithms are thoroughly evaluated on three different benchmark sets and they establish the new state of the art for realistic problem instances.

Haplotype Inference by Pure Parsimony: A Survey

Given a set of genotypes from a population, the process of recovering the haplotypes that explain the genotypes is called haplotype inference. The haplotype inference problem under the assumption of pure parsimony consists in finding the smallest number of haplotypes that explain a given set of genotypes. This problem is NP-hard. The original formulations for solving the Haplotype Inference by Pure Parsimony (HIPP) problem were based on integer linear programming and branch-and-bound techniques. More recently, solutions based on Boolean satisfiability, pseudo-Boolean optimization, and answer set programming have been shown to be remarkably more efficient. HIPP can now be regarded as a feasible approach for haplotype inference, which can be competitive with other different approaches. This article provides an overview of the methods for solving the HIPP problem, including preprocessing, bounding techniques, and heuristic approaches. The article also presents an empirical evaluation of exact HIPP solvers on a comprehensive set of synthetic and real problem instances. Moreover, the bounding techniques to the exact problem are evaluated. The final section compares and discusses the HIPP approach with a well-established statistical method that represents the reference algorithm for this problem.

A lead-time based approach for planning rail–truck intermodal transportation of dangerous goods

The remarkable growth in intermodal transportation over the past two decades has not been matched by a comparable level of academic activity, especially in the context of transporting hazardous materials (hazmats). In this paper, we present a first attempt for the development of an analytical framework for planning rail–truck intermodal transportation of hazmats. A bi-objective optimization model to plan and manage intermodal shipments is developed. To represent the current practice, the routing decisions in the model are driven by the delivery-times specified by the customers. An iterative decomposition based solution methodology which takes advantage of the problem structure is provided. A realistic problem instance based on the intermodal service network in eastern US is solved. This framework is used for developing a number of managerial insights, and for generating elements of the risk-cost frontier.

Heterogeneous computing scheduling with evolutionary algorithms

This work presents sequential and parallel evolutionary algorithms (EAs) applied to the scheduling problem in heterogeneous computing environments, a NP-hard problem with capital relevance in distributed computing. These methods have been specifically designed to provide accurate and efficient solutions by using simple operators that allow them to be later extended for solving realistic problem instances arising in distributed heterogeneous computing (HC) and grid systems. The EAs were codified over MALLBA, a general-purpose library for combinatorial optimization. Efficient numerical results are reported in the experimental analysis performed on well-known problem instances. The comparative study of scheduling methods shows that the parallel versions of the implemented evolutionary algorithms are able to achieve high problem solving efficacy, outperforming traditional scheduling heuristics and also improving over previous results already reported in the related literature.

Location discovery in Wireless Sensor Networks using metaheuristics

Wireless Sensor Networks (WSN) monitor the physical world using small wireless devices known as sensor nodes, with high precision and in real time, without the intervention of a human operator. Location information plays a critical role in many of the applications where WSN are used. Though a simple and effective solution could be to equip every node with self-locating hardware such as a GPS, the resulting cost renders such a solution unefficient. A widely used self-locating mechanism consists in equipping a small subset of the nodes with some GPS-like hardware, while the rest of the nodes employ reference estimations (received signal strength, time-of-arrival, etc.) in order to determine their locations. The task of determining the node locations using node-to-node distances combined with a set of known node locations is referred to as location discovery (LD). The main difficulty found in LD is the presence of distance estimation errors, which result in node positioning errors. We describe in this work an error model for the estimations, and propose a two-stage search procedure that combines minimization of an error norm function with maximization of a maximum likelihood function to solve the problem. We perform an empirical study of the performance of several variants of the guiding functions, and several metaheuristics used to solve real LD problem instances. Finally, we test our proposed technique against the single phase techniques in order to evaluate its performance.

Optimizing daily agent scheduling in a multiskill call center

We examine and compare simulation-based algorithms for solving the agent scheduling problem in a multiskill call center. This problem consists in minimizing the total costs of agents under constraints on the expected service level per call type, per period, and aggregated. We propose a solution approach that combines simulation with integer or linear programming, with cut generation. In our numerical experiments with realistic problem instances, this approach performs better than all other methods proposed previously for this problem. We also show that the two-step approach, which is the standard method for solving this problem, sometimes yield solutions that are highly suboptimal and inferior to those obtained by our proposed method.

A dynamic model for the failure replacement of aging high-voltage transformers

As the electric transmission system in the U.S. ages, mitigating the risk of high-voltage transformer failures becomes an increasingly important issue for transmission owners and operators. This paper introduces a model that supports these efforts by optimizing the acquisition and the deployment of high-voltage transformers dynamically over time. We formulate the problem as a Markov Decision Process which cannot be solved for realistic problem instances. Instead we solve the problem using approximate dynamic programming using three different value function approximations, which are compared against an optimal solution for a simplified version of the problem. The methods include a separable, piecewise linear value function, a piecewise linear, two-dimensional approximation, and a piecewise linear function based on an aggregated inventory that is shown to produce solutions within a few percent with very fast convergence. The application of the best performing algorithm to a realistic problem instance gives insights into transformer management issues of practical interest.

Constraint-based personalised configuring of product and service bundles

The configuring of product bundles such as tourism packages, financial services or compatible skin care products is a synthesis task that requires the support of a knowledgeable information system. We propose a personalised constraint-based configuration approach that is capable of solving such tasks in e-commerce environments. Here, a knowledge-based configurator is hybridised with collaborative methods from the domain of recommender systems in order to exploit user preferences to guide the problem solving process through large product spaces. The system implementation is based on a service-oriented architecture and the Choco open source constraint solver. It supports a model-driven approach for knowledge base design, acquisition and maintenance. An evaluation of the system suggests that it can solve realistic problem instances within an acceptable period of time.

Fuzzy goal programming models for analyzing demand coverage

The main objective of demand coverage models is to locate servers so that a given demand space is appropriately covered. Most existing models assume that demand is located at specific points within an area and that coverage is evaluated by certain quantifiable criteria. However, in realistic applications, the concept of coverage may also include qualitative aspects. Moreover, the location of the servers may be determined on the basis of more than one objective. In this paper, we present a number of fuzzy goal programming models for demand coverage. We consider three objectives: (a) maximization of total coverage, (b) maximization of minimum coverage, and (c) minimization of distance to servers of uncovered demand points. Through a series of realistic problem instances, we demonstrate that the proposed models provide satisfactory solutions with respect to all three objectives.

A New Approach for Air Cargo Network Planning

A central element of the air cargo planning process is the generation of optimal flight schedules. A flight schedule simultaneously defines the market potential of an airline and allocates its resources. The schedule design process is a difficult and time-consuming task that involves and affects virtually all business units. Because of its complexity, the process is traditionally decomposed into several steps that are executed in a sequential manner. In this paper, we present novel model formulations and solution procedures which have been developed in the course of a feasibility study for a decision support system (DSS) for a pragmatic approach to “freighter network planning” at one of the top international cargo carriers. We formulate two integrated models that combine the three planning steps: flight selection, aircraft rotation planning, and cargo routing. The aim of the schedule optimization is to maximize the network-wide profit by determining the best combination from a list of mandatory and optional flights, assigning the selected flights to aircrafts and identifying optimal cargo flows. Both model formulations are embedded in a solution procedure that builds on the column generation technique with shortest path algorithms for solving the subproblems. The applicability of the models in a DSS is demonstrated on realistic problem instances that match the requirements specified in the feasibility study.

PSO-based algorithm for home care worker scheduling in the UK

This paper presents the novel application of a collaborative population-based meta-heuristic technique called Particle Swarm Optimization (PSO) to the scheduling of home care workers. The technique is applied to a genuine situation arising in the UK, where the provision of community care service is a responsibility of the local authorities. Within this provision, optimization routes for each care worker are determined in order to minimize the distance traveled providing that the capacity and service time window constraints are not violated. The objectives of this paper are twofold; first to exploit a systematic approach to improve the existing schedule of home care workers, second to develop the methodology to enable the continuous PSO algorithm to be efficiently applied to this type of problem and all classes of similar problems. For this problem, a particle is defined as a multi-dimensional point in space which represents the corresponding care activities and assignment priority. The Heuristic Assignment scheme is specially designed to transform the continuous PSO algorithm to the discrete job schedule. The Earliest Start Time Priority with Minimum Distance Assignment (ESTPMDA) technique is developed for generating an initial solution which guides the search direction of the particle. Local improvement procedures (LIP), i.e. insertion and swap, are embedded in the PSO algorithm in order to further improve the solution quality. The proposed methodology is implemented, tested, and compared with existing solutions on a variety of real problem instances.

Understanding Algorithm Performance on an Oversubscribed Scheduling Application

The best performing algorithms for a particular oversubscribed scheduling application, Air Force Satellite Control Network (AFSCN) scheduling, appear to have little in common. Yet, through careful experimentation and modeling of performance in real problem instances, we can relate characteristics of the best algorithms to characteristics of the application. In particular, we find that plateaus dominate the search spaces (thus favoring algorithms that make larger changes to solutions) and that some randomization in exploration is critical to good performance (due to the lack of gradient information on the plateaus). Based on our explanations of algorithm performance, we develop a new algorithm that combines characteristics of the best performers; the new algorithm's performance is better than the previous best. We show how hypothesis driven experimentation and search modeling can both explain algorithm performance and motivate the design of a new algorithm.

Combined Network Design and Multiperiod Pricing: Modeling, Solution Techniques, and Computation

In this paper we describe an efficient algorithm for solving novel optimization models arising in the context of multiperiod capacity expansion of optical networks. We assume that the network operator must make investment decisions over a multiperiod planning horizon while facing rapid changes in transmission technology, as evidenced by a steadily decreasing per-unit cost of capacity. We deviate from traditional and monopolistic models in which demands are given as input parameters, and the objective is to minimize capacity deployment costs. Instead, we assume that the carrier sets end-to-end prices of bandwidth at each period of the planning horizon. These prices determine the demands that are to be met, using a plausible and explicit price-demand relationship; the resulting demands must then be routed, requiring an investment in capacity. The objective of the optimization is now to simultaneously select end-to-end prices of bandwidth and network capacities at each period of the planning horizon, so as to maximize the overall net present value of expanding and operating the network. In the case of typical large-scale optical networks with protection requirements, the resulting optimization problems pose significant challenges to standard optimization techniques. The complexity of the model, its nonlinear nature, and the large size of realistic problem instances motivates the development of efficient and scalable solution techniques. We show that while general-purpose nonlinear solvers are typically not adequate for the task, a specialized decomposition scheme is able to handle large-scale instances of this problem in reasonable time, producing solutions whose net present value is within a small tolerance of the optimum.

Improved Hidden Markov Model training for multiple sequence alignment by a particle swarm optimization—evolutionary algorithm hybrid

Multiple sequence alignment (MSA) is one of the basic problems in computational biology. Realistic problem instances of MSA are computationally intractable for exact algorithms. One way to tackle MSA is to use Hidden Markov Models (HMMs), which are known to be very powerful in the related problem domain of speech recognition. However, the training of HMMs is computationally hard and there is no known exact method that can guarantee optimal training within reasonable computing time. Perhaps the most powerful training method is the Baum–Welch algorithm, which is fast, but bears the problem of stagnation at local optima. In the study reported in this paper, we used a hybrid algorithm combining particle swarm optimization with evolutionary algorithms to train HMMs for the alignment of protein sequences. Our experiments show that our approach yields better alignments for a set of benchmark protein sequences than the most commonly applied HMM training methods, such as Baum–Welch and Simulated Annealing.

Plant co-ordination in pharmaceutics supply networks

The production of active ingredients in the chemical-pharmaceutical industry involves numerous production stages with cumulative lead times of up to two years. Mainly because of rigorous purity requirements and the need of extensive cleaning of the equipment units, production is carried out in campaigns, i.e. multiple batches of the same product type are produced successively before changing to another product type. Each campaign requires a specific configuration of equipment units according to the recipes of the particular chemical process. In the chemical-pharmaceutical industry, production stages are often assigned to different locations, even different countries. Hence the co-ordination of plant operations within the resulting multi-national supply network is of major importance. A key issue is the co-ordination of campaign schedules at different production stages in the various plants. In practice, it is almost impossible to determine exact optimal solutions to the corresponding complex supply network problem with respect to overall logistics costs. In order to reduce the required computational effort, we introduce several aggregation schemes and a novel MILP model formulation which is based on a continuous representation of time. Moreover, we propose an iterative near-optimal solution procedure which can be successfully applied to even exceptionally large real life problem instances. The applicability of the approach suggested is shown using a case study from industry.

Planning effective cellular mobile radio networks

This paper deals with the automatic selection and configuration of base station sites for mobile cellular networks. An optimization framework based on simulated annealing is used for site selection and for base-station configuration. Realistic path-loss estimates incorporating terrain data are used. The configuration of each base station involves selecting antenna type, power control, azimuth, and tilt. Results are presented for several design scenarios with between 250 and 750 candidate sites and show that the optimization framework can generate network designs with desired characteristics such as high area coverage and high traffic capacity. The work shows that cellular network design problems are tractable for realistic problem instances.

Scheduling in dual gripper robotic cells for productivity gains

We consider single part-type problems. Since all parts produced are identical, it is sufficient to determine the sequence of moves performed by the robot. The processing constraints define the cell to be a flowshop. The objective is the minimization of the steady-state cycle time to produce a part, or equivalently the maximization of the throughput rate. We study the problem of scheduling robot moves in dual gripper robot cells functioning in a bufferless environment. We develop an analytical framework for studying dual gripper robotic cells and examine the cycle time advantage of using a dual gripper rather than a single gripper robot. It is shown that an m-machine dual gripper robot cell can have at most double the productivity of its single gripper counterpart. We also propose a practical heuristic algorithm to compare productivity for given cell data. Computational testing of the algorithm on realistic problem instances is also described.

Tabu Search for Frequency Assignment in Mobile Radio Networks

The main goal of the Frequency Assignment Problem in mobile radio networks consists of assigning a limited number of frequencies to each radio cell in a cellular network while minimizing electromagnetic interference due to the reuse of frequencies. This problem, known to be NP-hard, is of great importance in practice since better solutions will allow a telecommunications operator to manage larger cellular networks. This paper presents a new Tabu Search algorithm for this application. The algorithm is tested on realistic and large problem instances and compared with other methods based on simulated annealing, constraint programming and graph coloring algorithms. Empirical evidence shows that the Tabu algorithm is very competitive by giving the best solutions to the tested instances.