Sérgio Ricardo de Souza

An adaptive genetic algorithm to the single machine scheduling problem with earliness and tardiness penalties

This paper deals with the Single Machine Scheduling Problem with Earliness and Tardiness Penalties, considering distinct due windows and sequence-dependent setup time. Due to its complexity, an adaptive genetic algorithm is proposed for solving it. Five search operators are used to explore the solution space and the choice probability for each operator depends on the success in a previous search. The initial population is generated by the combination between construct methods based on greedy, random and GRASP techniques. For each job sequence generated, a polynomial time algorithm are used for determining the processing initial optimal date to each job. During the evolutive process, a group with the best five individuals generated by each crossover operator is built. Then, periodically, a Path Relinking module is applied taking as base individual the best one so far generated by the algorithm and as guide individual each one of the five best individuals generated by each crossover operator. Three variations of this algorithm were submitted to computational experiments. The results shows the effectiveness of the proposal.

An immune-inspired algorithm for an unrelated parallel machines' scheduling problem with sequence and machine dependent setup-times for makespan minimisation

This paper proposes an immune-inspired algorithm to the problem of minimising the makespan on unrelated parallel machines, with sequence dependent setup times. The initial population is generated through the construction phase of the Greedy Randomised Adaptive Search Procedure (GRASP). An evaluation function is proposed to help the algorithm escape from local optima. A Variable Neighbourhood Descent (VND) local search heuristic, which makes significant use of the characteristics of the problem, is proposed as a somatic hypermutation operator to accelerate the convergence of the algorithm. A population re-selection operator, which strategically keeps good quality solutions with a high level of dispersion in the search space, is also proposed. The experiments performed show that the proposed algorithm enables better results than those reported in recent literature studies.

Biased random-key genetic algorithm for fiber installation in Optical Network Optimization

The problem of Fiber Installation in Optical Network Optimization consists in routing a set of lightpaths (all-optical connections), such that the cost of the optical components necessary to operate the network is minimized. We propose a genetic algorithm with random keys that extends the best heuristic in the literature by embedding it into an evolutionary framework. Computational results showed that the new heuristic improves the best heuristic in the literature.

An adaptive genetic algorithm for solving the single machine scheduling problem with earliness and tardiness penalties

This paper deals with the Single Machine Scheduling Problem with Earliness and Tardiness Penalties, considering distinct time windows and sequence-dependent setup time. Due to the complexity of this problem, an adaptive genetic algorithm is proposed for solving it. Many search operators are used to explore the solution space where the choice probability for each operator depends on the success in a previous search. The initial population is generated by applying GRASP to five dispatch rules. For each individual generated, a polynomial time algorithm is used to determine the initial optimal processing date for each job. During the evaluation process, the best individuals produced by each crossover operator, in each generation undergo refinement in order to improve quality of individuals. Computational results show the effectiveness of the proposed algorithm.

A multiagent architecture for solving combinatorial optimization problems through metaheuristics

This article introduces MAM - multiagent architecture for metaheuristics, whose objective is to combine metaheuristics, through the multiagent approach, for solving combinatorial optimization problems. In this architecture, each metaheuristic is developed in the form of an autonomous agent, cooperatively interacting in an environment. This interaction between one or more agents is carried out through information exchange in the search space of the problem, seeking to improve the same objective. MAM is a flexible architecture, which can be used for solving different optimization problems, without the need to rewrite algorithms. In this paper, the MAM architecture is specialized for genetic algorithm (GA), iterated local search (ILS) and variable neighborhood search (VNS) metaheuristics in order to solve the vehicle routing problem with time windows (VRPTW). Computational tests were performed and results are presented, showing the effectiveness of the proposed architecture.

An adaptive genetic algorithm to solve the Single Machine Scheduling Problem with Earliness and Tardiness Penalties

This paper deals with the Single Machine Scheduling Problem with Earliness and Tardiness Penalties, considering distinct time windows and sequence-dependent setup time. Due to the complexity of this problem, an adaptive genetic algorithm is proposed for solving it. Many search operators are used to explore the solution space where the choice probability for each operator depends on the success in a previous search. The initial population is generated by applying GRASP to five dispatch rules. For each individual generated, a polynomial time algorithm is used to determine the initial optimal processing date for each job. During the evaluation process, the best individuals produced by each crossover operator, in each generation undergo refinement in order to improve quality of individuals. Computational results show the effectiveness of the proposed algorithm.

A Multi-agent Metaheuristic Optimization Framework with Cooperation

This article address a Multiagent Metaheuristic Optimization Framework. In this proposal, each agent acts independently in the search space of a combinatorial optimization problem. The Framework allows the simultaneous execution of various agents, in a cooperative way. The coalition concept of cooperation is adopted. The agents have auto-learning abilities, based on reinforcement learning. The ability of cooperation and its influence on the quality of solutions provided by the agents are confirmed by performed experiments. In addition, experiments show that influence is greater when the number of agents is increased.

A Clonal Selection Algorithm for Makespan Minimization on Unrelated Parallel Machines with Sequence Dependent Setup Times

This work presents a proposal to implement the clonal selection algorithm (CLONALG) for the make span minimization problem on unrelated parallel machines with sequence-dependent setup times. The initial population is created through the GRASP algorithm, whereas an evaluation function based on make span values and on a historical of recent improvements helps the algorithm to escape from local optima. A local search heuristic, based in the VND method, was used as the hyper mutation somatic operator, in order to accelerate the convergence of the algorithm. The performed experiments show that the proposed operators improve the performance of standard CLONALG algorithm as well as allow to obtain better results than those presented in the most recent literature.

Iterated Local Search for Fiber Installation in Optical Network Optimization

Abstract The problem of Fiber Installation in Optical Network Optimization consists in routing a set of lightpaths (all-optical connections), such that the cost of the optical components necessary to operate the network is minimized. We propose a novel Iterated Local Search heuristic. Computational results showed that the new heuristic is better than the best heuristic in the literature.

Variable Neighborhood Descent applied to Multi-way Number Partitioning Problem

Abstract This paper presents an algorithm for the optimization version of the Multi-Way Number Partitioning Problem (MWNPP). This problem consists in distributing the elements of a given sequence into k disjoint subsets so that the sums of each subset elements fit in the shortest interval. The metaheuristic Variable Neighborhood Descent (VND), a deterministic variant of Variable Neighborhood Search (VNS), adapted for solving the MWNPP, has a good performance over instances less than six subsets. It is carried out a comparative study with two algorithms, Karmarkar-Karp Heuristic and Longest Processing Time, using randomly generated instances and objective functions values. The statistical tests show that results of the VND proposed are significantly better than literature constructive methods and its improvements.