Saïd Hanafi

An efficient tabu search approach for the 0-1 multidimensional knapsack problem

Abstract In this paper, we describe a new approach to tabu search (TS) based on strategic oscillation and surrogate constraint information that provides a balance between intensification and diversification strategies. New rules needed to control the oscillation process are given for the 0–1 multidimensional knapsack (0–1 MKP). Based on a portfolio of test problems from the literature, our method obtains solutions whose quality is at least as good as the best solutions obtained by previous methods, especially with large scale instances. These encouraging results confirm the efficiency of the tunneling concept coupled with surrogate information when resource constraints are present.

Genetic algorithm with iterated local search for solving a location-routing problem

Highlights? Hybrid genetic algorithm and iterated local search outperforms tabu search. ? Local and routing decisions are solved simultaneously. ? Iterated local search refines the genetic algorithm through successive iterations. This paper deals with a location routing problem with multiple capacitated depots and one uncapacitated vehicle per depot. We seek for new methods to make location and routing decisions simultaneously and efficiently. For that purpose, we describe a genetic algorithm (GA) combined with an iterative local search (ILS). The main idea behind our hybridization is to improve the solutions generated by the GA using a ILS to intensify the search space. Numerical experiments show that our hybrid algorithm improves, for all instances, the best known solutions previously obtained by the tabu search heuristic.

The Multidimensional 0-1 Knapsack Problem—Bounds and Computational Aspects

The multidimensional 0-1 knapsack problem (MKP) is a resource allocation model that is one of the most well-known integer programming problems. During the last few decades, an impressive amount of research on the 0-1 knapsack problem has been published in the literature, and efficient special-purpose methods have become available for solving very large-scale instances. However, the multidimensional case has received less attention from the operational research community. Although recent advances have made solving medium size instances possible, solving the NP-hard problem remains a very interesting challenge, especially when the number of constraints increases. This paper surveys the principal results published in the literature concerning both the problems theoretical properties and its approximate or exact solutions. The paper focuses on the more recent results—for example, those relevant to surrogate and composite duality, new preprocessing approaches creating enhanced versions of leading commercial software, and efficient metaheuristic-based methods.

Variable neighborhood search for location routing

In this paper we propose various neighborhood search heuristics (VNS) for solving the location routing problem with multiple capacitated depots and one uncapacitated vehicle per depot. The objective is to find depot locations and to design least cost routes for vehicles. We integrate a variable neighborhood descent as the local search in the general variable neighborhood heuristic framework to solve this problem. We propose five neighborhood structures which are either of routing or location type and use them in both shaking and local search steps. The proposed three VNS methods are tested on benchmark instances and successfully compared with other two state-of-the-art heuristics.

On the Convergence of Tabu Search

The Tabu Search (TS) meta-heuristic has proved highly successful for solving combinatorial and nonlinear problems. A key aspect of TS consists of using adaptive forms of memory to forbid the search process to revisit solutions already examined unless the trajectory to reach it is different. In Glover (ORSA Journal on Computing, 1990, 2, 4–32) a special memory design was proposed together with a choice rule for handling the situation where the method was compelled to revisit solutions already encountered. This proposal, which specified the exploration should resume from the earliest solution visited in the past, as accompanied by the conjecture that such a choice has implications for finiteness in the zero-one integer program and optimal set membership examples. Up to now numerous applications of TS in various areas of research are available, however, we are aware of only a few results concerning the convergence of TS. In this paper, we prove that Glovers conjecture is true if the neighborhood employed is strongly connected, yielding a “reversible” path from each solution to every other solution.

Variable neighbourhood decomposition search for 0-1 mixed integer programs

In this paper we propose a new hybrid heuristic for solving 0-1 mixed integer programs based on the principle of variable neighbourhood decomposition search. It combines variable neighbourhood search with a general-purpose CPLEX MIP solver. We perform systematic hard variable fixing (or diving) following the variable neighbourhood search rules. The variables to be fixed are chosen according to their distance from the corresponding linear relaxation solution values. If there is an improvement, variable neighbourhood descent branching is performed as the local search in the whole solution space. Numerical experiments have proven that exploiting boundary effects in this way considerably improves solution quality. With our approach, we have managed to improve the best known published results for 8 out of 29 instances from a well-known class of very difficult MIP problems. Moreover, computational results show that our method outperforms the CPLEX MIP solver, as well as three other recent most successful MIP solution methods.

New convergent heuristics for 0-1 mixed integer programming

Several hybrid methods have recently been proposed for solving 0-1 mixed integer programming problems. Some of these methods are based on the complete exploration of small neighborhoods. In this paper, we present several convergent algorithms that solve a series of small sub-problems generated by exploiting information obtained from a series of relaxations. These algorithms generate a sequence of upper bounds and a sequence of lower bounds around the optimal value. First, the principle of a linear programming-based algorithm is summarized, and several enhancements of this algorithm are presented. Next, new hybrid heuristics that use linear programming and/or mixed integer programming relaxations are proposed. The mixed integer programming (MIP) relaxation diversifies the search process and introduces new constraints in the problem. This MIP relaxation also helps to reduce the gap between the final upper bound and lower bound. Our algorithms improved 14 best-known solutions from a set of 108 available and correlated instances of the 0-1 multidimensional Knapsack problem. Other encouraging results obtained for 0-1 MIP problems are also presented.

New exact method for large asymmetric distance-constrained vehicle routing problem

In this paper we revise and modify an old branch-and-bound method for solving the asymmetric distance–constrained vehicle routing problem suggested by Laporte et al. in 1987. Our modification is based on reformulating distance–constrained vehicle routing problem into a travelling salesman problem, and on using assignment problem as a lower bounding procedure. In addition, our algorithm uses the best-first strategy and new tolerance based branching rules. Since our method is fast but memory consuming, it could stop before optimality is proven. Therefore, we introduce the randomness, in case of ties, in choosing the node of the search tree. If an optimal solution is not found, we restart our procedure. As far as we know, the instances that we have solved exactly (up to 1000 customers) are much larger than the instances considered for other vehicle routing problem models from the recent literature. So, despite of its simplicity, this proposed algorithm is capable of solving the largest instances ever solved in the literature. Moreover, this approach is general and may be used for solving other types of vehicle routing problems.

A dynamic programming algorithm for the bilevel knapsack problem

We propose an efficient dynamic programming algorithm for solving a bilevel program where the leader controls the capacity of a knapsack, and the follower solves the resulting knapsack problem. We propose new recursive rules and show how to solve the problem as a sequence of two standard knapsack problems.

Variable neighborhood search: basics and variants

Variable neighborhood search (VNS) is a framework for building heuristics, based upon systematic changes of neighborhoods both in a descent phase, to find a local minimum, and in a perturbation phase to escape from the corresponding valley. In this paper, we present some of VNS basic schemes as well as several VNS variants deduced from these basic schemes. In addition, the paper includes parallel implementations and hybrids with other metaheuristics.