Summary Report for the Third International Competition On Computational Models of Argumentation

The Third International Competition on Computational Models of Argumentation (ICCMA’19) focused on reasoning tasks in abstract argumentation frameworks. Submitted solvers were tested on a selected collection of benchmark instances, including artificially generated argumentation frameworks and some frameworks formalizing real-world problems. This competition introduced two main novelties over the two previous editions: the first one is the use of the Docker platform for packaging the participating solvers into virtual “light” containers; the second novelty consists of a new track for dynamic frameworks.

A granular tabu search for the refrigerated vehicle routing problem with homogeneous fleet

The Refrigerated Capacitated Vehicle Routing Problem (RCVRP) considers a homogeneous fleet with a refrigerated system to decide the selection of routes to be performed according to customers' requirements. The aim is to keep the energy consumption of the routes as low as possible. We use a thermodynamic model to understand the unloading of products from trucks and the variables' efficiency, such as the temperature during the day influencing energy consumption. By considering various neighborhoods and a shaking procedure, this paper proposes a Granular Tabu Search scheme to solve the RCVRP. Computational tests using adapted benchmark instances from the literature demonstrate that the suggested method delivers high-quality solutions within short computing times, illustrating the refrigeration system's effect on routing decisions.

GRASP-Tabu Search Algorithms for the Route Planning Problem in Spatial Crowdsourcing

With the speedy progress of mobile devices, a lot of commercial enterprises have exploited crowdsourcing as a useful approach to gather information to develop their services. Thus, spatial crowdsourcing has appeared as a new platform in e-commerce and which implies procedures of requesters and workers. A requester submits spatial tasks request to the workers who choose and achieve them during a limited time. Thereafter, the requester pays only the worker for the well accomplished the task. In spatial crowdsourcing, each worker is required to physically move to the place to accomplish the spatial task and each task is linked with location and time. The objective of this article is to find an optimal route to the worker through maximizing her rewards with respecting some constraint, using an approach based on GRASP with Tabu. The proposed algorithm is used in the literature for benchmark instances. Computational results indicate that the proposed and the developed algorithm is competitive with other solution approaches.

Implications, conflicts, and reductions for Steiner trees

The Steiner tree problem in graphs (SPG) is one of the most studied problems in combinatorial optimization. In the past 10 years, there have been significant advances concerning approximation and complexity of the SPG. However, the state of the art in (practical) exact solution of the SPG has remained largely unchallenged for almost 20 years. While the DIMACS Challenge 2014 and the PACE Challenge 2018 brought renewed interest into Steiner tree problems, even the best new SPG solvers cannot match the state of the art on the vast majority of benchmark instances. The following article seeks to advance exact SPG solution once again. The article is based on a combination of three concepts: Implications, conflicts, and reductions. As a result, various new SPG techniques are conceived. Notably, several of the resulting techniques are (provably) stronger than well-known methods from the literature that are used in exact SPG algorithms. Finally, by integrating the new methods into a branch-and-cut framework, we obtain an exact SPG solver that is not only competitive with, but even outperforms the current state of the art on an extensive collection of benchmark sets. Furthermore, we can solve several instances for the first time to optimality.

Novel Metaheuristic Based on Iterated Constructive Stochastic Heuristic: Dhouib-Matrix-3 (DM3)

This paper presents a new metaheuristic named Dhouib-Matrix-3 (DM3) inspired by our recently developed constructive stochastic heuristic Dhouib-Matrix-TSP2 (DM-TSP2) and characterized by only one parameter: the number of iterations. The proposed metaheuristic DM3 is an iterative algorithm in which every iteration is based on two relay hybridization techniques. At first, the constructive stochastic heuristic DM-TSP2 starts by generating a different initial basic feasible solution and then each solution is intensified by the novel procedure Far-to-Near which exchanges far cities by closer ones using three perturbation techniques: insertion, exchange, and 2-opt. Experimental results carried out on the classical travelling salesman problem using the well-known TSP-LIB benchmark instances demonstrate that our approach DM3 outclasses the simulated annealing algorithm, the genetic algorithm, and the cellular genetic algorithm. Furthermore, the proposed DM3 is statistically concurrent to the hybrid simulated annealing cellular genetic algorithm. Nevertheless, DM3 is easier to implement and needs only one parameter to identify (the maximum number of iterations).

Novel Formulations and Logic-Based Benders Decomposition for the Integrated Parallel Machine Scheduling and Location Problem

We investigate the discrete parallel machine scheduling and location problem, which consists of locating multiple machines to a set of candidate locations, assigning jobs from different locations to the located machines, and sequencing the assigned jobs. The objective is to minimize the maximum completion time of all jobs, that is, the makespan. Though the problem is of theoretical significance with a wide range of practical applications, it has not been well studied as reported in the literature. For this problem, we first propose three new mixed-integer linear programs that outperform state-of-the-art formulations. Then, we develop a new logic-based Benders decomposition algorithm for practical-sized instances, which splits the problem into a master problem that determines machine locations and job assignments to machines and a subproblem that sequences jobs on each machine. The master problem is solved by a branch-and-cut procedure that operates on a single search tree. Once an incumbent solution to the master problem is found, the subproblem is solved to generate cuts that are dynamically added to the master problem. A generic no-good cut is first proposed, which is later improved by some strengthening techniques. Two optimality cuts are also developed based on optimality conditions of the subproblem and improved by strengthening techniques. Numerical results on small-sized instances show that the proposed formulations outperform state-of-the-art ones. Computational results on 1,400 benchmark instances with up to 300 jobs, 50 machines, and 300 locations demonstrate the effectiveness and efficiency of the algorithm compared with current approaches. Summary of Contribution: This paper employs operations research methods and computing techniques to address an NP-hard combinatorial optimization problem: the parallel discrete machine scheduling and location problem. The problem is of practical significance but has not been well studied in the literature. For the problem, we formulate three novel mixed-integer linear programs that outperform state-of-the-art formulations and develop a new logic-based Benders decomposition algorithm. Extensive computational experiments on 1,400 benchmark instances with up to 300 jobs, 50 machines, and 300 locations are conducted to evaluate the performance of the proposed models and algorithms.

A Reduced Variable Neighborhood Search Approach to the Heterogeneous Vector Bin Packing Problem

The two-dimensional heterogeneous vector bin packing problem (2DHet-VBPP) consists of packing the set of items into the set of various type bins, respecting their two resource limits. The problem is to minimize the total cost of all bins. The problem, known to be NP-hard, can be formulated as a pure integer linear program, but optimal solutions can be obtained by the CPLEX Optimizer engine only for small instances. This paper proposes a metaheuristic approach to the 2DHet-VBPP, based on Reduced variable neighborhood search (RVNS). All RVNS elements are adapted to the considered problem and many procedures are designed to improve efficiency of the method. As the Two-dimensional Homogeneous-VBPP (2DHom-VBPP) is more often treated, we considered also a special version of the RVNS algorithm to solve the 2DHom-VBPP. The results obtained and compared to both CPLEX results and results on benchmark instances from literature, justify the use of the RVNS algorithm to solve large instances of these optimization problems.

Hybrid pointer networks for traveling salesman problems optimization

In this work, we proposed a hybrid pointer network (HPN), an end-to-end deep reinforcement learning architecture is provided to tackle the travelling salesman problem (TSP). HPN builds upon graph pointer networks, an extension of pointer networks with an additional graph embedding layer. HPN combines the graph embedding layer with the transformer’s encoder to produce multiple embeddings for the feature context. We conducted extensive experimental work to compare HPN and Graph pointer network (GPN). For the sack of fairness, we used the same setting as proposed in GPN paper. The experimental results show that our network significantly outperforms the original graph pointer network for small and large-scale problems. For example, it reduced the cost for travelling salesman problems with 50 cities/nodes (TSP50) from 5.959 to 5.706 without utilizing 2opt. Moreover, we solved benchmark instances of variable sizes using HPN and GPN. The cost of the solutions and the testing times are compared using Linear mixed effect models. We found that our model yields statistically significant better solutions in terms of the total trip cost. We make our data, models, and code publicly available https://github.com/AhmedStohy/Hybrid-Pointer-Networks.

A large neighborhood search approach for the paint shop scheduling problem

Minimizing the setup costs caused by color changes is one of the main concerns for paint shop scheduling in the automotive industry. Yet, finding an optimized color sequence is a very challenging task, as a large number of exterior systems for car manufacturing need to be painted in a variety of different colors. Therefore, there is a strong need for efficient automated scheduling solutions in this area. Previously, exact and metaheuristic approaches for creating efficient paint shop schedules in the automotive supply industry have been proposed and evaluated on a publicly available set of real-life benchmark instances. However, optimal solutions are still unknown for many of the benchmark instances, and there is still a potential of reducing color change costs for large instances. In this paper, we propose a novel large neighborhood search approach for the paint shop scheduling problem. We introduce innovative exact and heuristic solution methods that are utilized within the large neighborhood search and show that our approach leads to improved results for large real-life problem instances compared to existing techniques. Furthermore, we provide previously unknown upper bounds for 14 benchmark instances using the proposed method.

A two-level based genetic algorithm for solving the soft-clustered vehicle routing problem

The soft-clustered vehicle routing problem (Soft-CluVRP) is a relaxation of the clustered vehicle routing problem (CluVRP), which in turn is a variant of the generalized vehicle routing problem (GVRP). The aim of the Soft-CluVRP is to look for a minimum cost group of routes starting and ending at a given depot to a set of customers partitioned into a priori defined, mutually exclusive and exhaustive clusters, satisfying the capacity constraints of the vehicles and with the supplementary property that all the customers from the same cluster have to be supplied by the same vehicle. The considered optimization problem is NP-hard, that is why we proposed a two-level based genetic algorithm in order to solve it. The computational results reported on a set of existing benchmark instances from the literature, prove that our novel solution approach provides high-quality solutions within acceptable running times.