Solving the response time variability problem by means of a genetic algorithm

The response time variability problem (RTVP) is a hard scheduling problem that has recently been defined in the literature and has a wide range of real-world applications in mixed-model assembly lines, multithreaded computer systems, network environments and others. The RTVP arises whenever products, clients or jobs need to be sequenced in such a way that the variability in the time between the points at which they receive the necessary resources is minimized. Since the RTVP is a complex problem, heuristic and metaheuristic techniques are needed to solve it. The best results in the literature for the RTVP have been obtained with a psychoclonal algorithm. We propose a genetic algorithm (GA) that is adapted to solve the RTVP. A computational experiment is carried out and it is shown that, on average, the GA produces better results than the psychoclonal algorithm.     

Mathematical programming modeling of the Response Time Variability Problem

The Response Time Variability Problem (RTVP) is a scheduling problem that has recently been defined in the literature. The RTVP has a broad range of real-life applications from manufacturing to services and information technology. A previous study developed a position exchange heuristic to apply to initial sequences for the RTVP, and a MILP (Mixed Integer Linear Programming) to obtain optimal solutions with a practical limit of 25 units to be scheduled. This paper aims to improve the best mathematical programming model developed thus far in order to solve larger instances up to 40 units to optimality. The contribution of this paper is 4-fold: (i) larger instances can be solved to optimality by the off the shelf standard software; (ii) the new optimal solutions of the RTVP can be used to compare the results of heuristic procedures; (iii) the importance of modeling is demonstrated, as well as the huge impact that reformulation, redundant constraints and the elimination of symmetries have on the efficiency of MILPs is clearly established; finally (iv) a challenge to develop a customized optimization algorithm to rival the MILP solution efficiency for the RTVP is put forward.     

Effective learning hyper-heuristics for the course timetabling problem

Course timetabling is an important and recurring administrative activity in most educational institutions. This article combines a general modeling methodology with effective learning hyper-heuristics to solve this problem. The proposed hyper-heuristics are based on an iterated local search procedure that autonomously combines a set of move operators. Two types of learning for operator selection are contrasted: a static (offline) approach, with a clear distinction between training and execution phases; and a dynamic approach that learns on the fly. The resulting algorithms are tested over the set of real-world instances collected by the first and second International Timetabling competitions. The dynamic scheme statistically outperforms the static counterpart, and produces competitive results when compared to the state-of-the-art, even producing a new best-known solution. Importantly, our study illustrates that algorithms with increased autonomy and generality can outperform human designed problem-specific algorithms.     

Simulated annealing metaheuristics for the vehicle routing problem with time windows

This paper develops simulated annealing metaheuristics for the vehicle routing and scheduling problem with time window constraints. Two different neighborhood structures, the λ-interchange mechanism of Osman and thek-node interchange process of Christofides and Beasley, are implemented. The enhancement of the annealing process with a short-term memory function via a tabu list is examined as a basis for improving the metaheuristic approach. Computational results on test problems from the literature as well as large-scale real-world problem are reported. The metaheuristics achieve solutions that compare favorably with previously reported results.     

A genetic algorithm for the resource constrained multi-project scheduling problem

This paper presents a genetic algorithm for the resource constrained multi-project scheduling problem. The chromosome representation of the problem is based on random keys. The schedules are constructed using a heuristic that builds parameterized active schedules based on priorities, delay times, and release dates defined by the genetic algorithm. The approach is tested on a set of randomly generated problems. The computational results validate the effectiveness of the proposed algorithm.     

A new discrete particle swarm optimization approach for the single-machine total weighted tardiness scheduling problem with sequence-dependent setup times

In this paper we present a new Discrete Particle Swarm Optimization (DPSO) approach to face the NP-hard single machine total weighted tardiness scheduling problem in presence of sequence-dependent setup times. Differently from previous approaches the proposed DPSO uses a discrete model both for particle position and velocity and a coherent sequence metric. We tested the proposed DPSO mainly over a benchmark originally proposed by Cicirello in 2003 and available online. The results obtained show the competitiveness of our DPSO, which is able to outperform the best known results for the benchmark. In addition, we also tested the DPSO on a set of benchmark instances from ORLIB for the single machine total weighted tardiness problem, and we analysed the role of the DPSO swarm intelligence mechanisms as well as the local search intensification phase included in the algorithm.     

Local search methods for the flowshop scheduling problem with flowtime minimization

Flowshop scheduling is a very active research area. This problem still attracts a considerable amount of interest despite the sheer amount of available results. Total flowtime minimization of a flowshop has been actively studied and many effective algorithms have been proposed in the last few years. New best solutions have been found for common benchmarks at a rapid pace. However, these improvements many times come at the cost of sophisticated algorithms. Complex methods hinder potential applications and are difficult to extend to small problem variations. Replicability of results is also a challenge. In this paper, we examine simple and easy to implement methods that at the same time result in state-of-the-art performance. The first two proposed methods are based on the well known Iterated Local Search (ILS) and Iterated Greedy (IG) frameworks, which have been applied with great success to other flowshop problems. Additionally, we present extensions of these methods that work over populations, something that we refer to as population-based ILS (pILS) and population-based IG (pIGA), respectively. We calibrate the presented algorithms by means of the Design of Experiments (DOE) approach. Extensive comparative evaluations are carried out against the most recent techniques for the considered problem in the literature. The results of a comprehensive computational and statistical analysis show that the presented algorithms are very effective. Furthermore, we show that, despite their simplicity, the presented methods are able to improve 12 out of 120 best known solutions of Taillard’s flowshop benchmark with total flowtime criterion.     

An iterated local search algorithm for the vehicle routing problem with convex time penalty functions

We propose an iterated local search algorithm for the vehicle routing problem with time window constraints. We treat the time window constraint for each customer as a penalty function, and assume that it is convex and piecewise linear. Given an order of customers each vehicle to visit, dynamic programming (DP) is used to determine the optimal start time to serve the customers so that the total time penalty is minimized. This DP algorithm is then incorporated in the iterated local search algorithm to efficiently evaluate solutions in various neighborhoods. The amortized time complexity of evaluating a solution in the neighborhoods is a logarithmic order of the input size (i.e., the total number of linear pieces that define the penalty functions). Computational comparisons on benchmark instances with up to 1000 customers show that the proposed method is quite effective, especially for large instances.     

An adaptive search for the response time variability problem

The Response Time Variability Problem (RTVP) is an NP-hard combinatorial scheduling problem, which has recently been reported and formalised in the literature. This problem has a wide range of real-world applications in mixed-model assembly lines, multi-threaded computer systems, broadcast of commercial videotapes and others. The RTVP arises whenever products, clients or jobs need to be sequenced in such a way that the variability in the time between the points at which they receive the necessary resources is minimised. We propose a greedy but adaptive heuristic that avoids being trapped into a poor solution by incorporating a look ahead strategy suitable for this particular scheduling problem. The proposed heuristic outperforms the best existing methods, while being much faster and easier to understand and to implement.     

Evolutionary search for difficult problem instances to support the design of job shop dispatching rules

Dispatching rules are simple scheduling heuristics that are widely applied in industrial practice. Their popularity can be attributed to their ability to flexibly react to shop floor disruptions that are prevalent in many real-world manufacturing environments. However, it is a challenging and time-consuming task to design local, decentralised dispatching rules that result in a good global performance of a complex shop.An evolutionary algorithm is developed to generate job shop problem instances for which an examined dispatching rule fails to achieve a good solution due to a single suboptimal decision. These instances can be easily analysed to reveal limitations of that rule which helps with the design of better rules. The method is applied to a job shop problem from the literature, resulting in new best dispatching rules for the mean flow time measure.     

An iterated local search algorithm for the time-dependent vehicle routing problem with time windows

We generalize the standard vehicle routing problem with time windows by allowing both traveling times and traveling costs to be time-dependent functions. In our algorithm, we use a local search to determine routes of the vehicles. When we evaluate a neighborhood solution, we must compute an optimal time schedule for each route. We show that this subproblem can be efficiently solved by dynamic programming, which is incorporated in the local search algorithm. The neighborhood of our local search consists of slight modifications of the standard neighborhoods called 2- opt^*, cross exchange and Or-opt. We propose an algorithm that evaluates solutions in these neighborhoods more efficiently than the ones computing the dynamic programming from scratch by utilizing the information from the past dynamic programming recursion used to evaluate the current solution. We further propose a filtering method that restricts the search space in the neighborhoods to avoid many solutions having no prospect of improvement. We then develop an iterated local search algorithm that incorporates all the above ingredients. Finally we report computational results of our iterated local search algorithm compared against existing methods, and confirm the effectiveness of the restriction of the neighborhoods and the benefits of the proposed generalization.     

Iterative approaches for a dynamic memory allocation problem in embedded systems

Memory allocation has a significant impact on energy consumption in embedded systems. In this paper, we are interested in dynamic memory allocation for embedded systems with a special emphasis on time performance. We propose two mid-term iterative approaches which are compared with existing long-term and short-term approaches, and with an ILP formulation as well. These approaches rely on solving a static version of the allocation problem and they take advantage of previous works for addressing the static problem. A statistic analysis is carried out for showing that the mid-term approach is the best one in terms of solution quality.     

NP-hardness of the single-variable-resource scheduling problem to minimize the total weighted completion time

Baker and Nuttle [K.R. Baker, H.L.W. Nuttle, Sequencing independent jobs with a single resource, Naval Research Logistics Quarterly 27 (1980) 499–510] studied the following single-variable-resource scheduling problem: sequencing n jobs for processing by a single resource to minimize a function of job completion times, when the availability of the resource varies over time. When the objective function to be minimized is the total weighted completion time, Baker and Nuttle conjectured that the problem is NP-hard. We show in this note that the conjecture is true.     

A diversified Tabu search approach for the open-pit mine production scheduling problem with metal uncertainty

This paper presents a metaheuristic solution approach based on Tabu search for the open-pit mine production scheduling problem with metal uncertainty. To search the feasible domain more extensively, two different diversification strategies are used to generate several initial solutions to be optimized by the Tabu search procedure. The first diversification strategy exploits a long-term memory of the search history. The second one relies on the variable neighborhood search method. Numerical results on realistic large-scale instances are provided to indicate the efficiency of the solution approach to produce very good solutions in relatively short computational times.     

A threshold accepting heuristic with intense local search for the solution of special instances of the traveling salesman problem

In real life scheduling, variations of the standard traveling salesman problem are very often encountered. The aim of this work is to present a new heuristic method for solving three such special instances with a common approach. The proposed algorithm uses a variant of the threshold accepting method, enhanced with intense local search, while the candidate solutions are produced through an insertion heuristic scheme. The main characteristic of the algorithm is that it does not require modifications and parameter tuning in order to cope with the three different problems. Computational results on a variety of real life and artificial problems are presented at the end of this work and prove the efficiency and the ascendancy of the proposed method over other algorithms found in the literature.     

Scheduling data transfers in a network and the set scheduling problem

In this paper we consider the online ftp problem. The goal is to service a sequence of file transfer requests given bandwidth constraints of the underlying communication network. The main result of the paper is a technique that leads to algorithms that optimize several natural metrics, such as max-stretch, total flow time, max flow time, and total completion time. In particular, we show how to achieve optimum total flow time and optimum max-stretch if we increase the capacity of the underlying network by a logarithmic factor. We show that the resource augmentation is necessary by proving polynomial lower bounds on the max-stretch and total flow time for the case where online and offline algorithms are using same-capacity edges. Moreover, we also give polylogarithmic lower bounds on the resource augmentation factor necessary in order to keep the total flow time and max-stretch within a constant factor of optimum.     

Metaheuristic algorithms hybridised with variable neighbourhood search for solving the response time variability problem

The response time variability problem (RTVP) is a scheduling problem with a wide range of real-world applications: mixed-model assembly lines, multi-threaded computer systems, network environments, broadcast of commercial videotapes and machine maintenance, among others. The RTVP arises whenever products, clients or jobs need to be sequenced in such a way that the variability in the time between the points at which they receive the necessary resources is minimised. Since the RTVP is NP-hard, several heuristic and metaheuristic techniques are needed to solve non-small instances. The published metaheuristic procedure that obtained the best solutions, on average, for non-small RTVP instances is an algorithm based on a variant of the variable neighbourhood search (VNS), called Reduced VNS (RVNS). We propose hybridising RVNS with three existing algorithms based on tabu search, multi-start and particle swarm optimisation. The aim is to combine the strengths of the metaheuristics. A computational experiment is carried out and it is shown that, on average, all proposed hybrid methods are able to improve the best published solutions.     

A branch and bound algorithm for the one-machine scheduling problem with minimum and maximum time lags

We consider the one-machine scheduling problem with minimum and maximum time lags while minimizing the makespan. This problem typically arises in a manufacturing environment where the next job has to be carried out within a specific time range after the completion of the immediately preceding job. We describe a branch and bound algorithm, based on the input and output of a clique and the relevant propositions, for finding the optimal waiting times. The computational experiments give promising results, showing whether a given instance is feasible or infeasible. With the proposed branch and bound algorithm we can either find an optimal schedule or establish the infeasibility within an acceptable run time.     

Solving the response time variability problem by means of a psychoclonal approach

The response time variability problem (RTVP) is a combinatorial scheduling problem that has recently appeared in the literature. This problem has a wide range of real life applications in fields such as manufacturing, hard real-time systems, operating systems and network environments. Originally, the RTVP occurs whenever products, clients or jobs need to be sequenced in such a way that the variability in the time between the instants at which they receive the necessary resources is minimized. Since the RTVP is hard to solve, heuristic techniques are needed for solving it. Three metaheuristic—multi-start, GRASP and PSO—algorithms proposed for solving the RTVP in two previous studies have been the most efficient to date in solving non-small instances of the RTVP. We propose solving the RTVP by means of a psychoclonal algorithm based approach. The psychoclonal algorithm inherits its attributes from Maslow’s hierarchy of needs theory and the artificial immune system (AIS) approach, specifically the clonal selection principle. In this paper, we compare the proposed psychoclonal algorithm with the three aforementioned metaheuristic algorithms and show that, on average, the psychoclonal algorithm strongly improves on the results obtained.