A three-phase matheuristic for capacitated multi-commodity fixed-cost network design with design-balance constraints
33
Citation
25
Reference
10
Related Paper
Citation Trend
Keywords:
Heuristics
Solver
Benchmark (surveying)
Neighbourhood (mathematics)
ABSTRACT Tabu search is a widely used heuristic search method. However, there are two drawbacks in traditional tabu search. First, tabu search, as well as other search methods, only provides the best solution obtained during the search process, and there is no way to know the quality of the obtained solutions. Second, tabu list helps tabu search avoid the problem of looping in a small cycle, but it cannot prevent tabu search from searching previously searched areas again or, worse, looping in a large cycle. The computation time of a search method can be reduced by implementing parallel processing. This study proposes a parallel deterministic simple tabu search, which computes more efficiently and overcomes the two drawbacks mentioned above. The results of our experiments show that it takes less computation effort for the proposed parallel tabu search to find a global optimal solution than for a conventional parallel tabu search.
Guided Local Search
Hill climbing
Local optimum
Cite
Citations (0)
Guided Local Search
Hill climbing
Cite
Citations (51)
The method using tabu search to identify system models is developed in this paper. By converting the system identification problem into an optimization problem in parameter space, the tabu search is used to seek for the global optimal solution as the optimal estimation of the parameters. Simulations on both the discrete and continuous systems are conducted to verify the feasibility. Compared with results obtained from genetic algorithm, the performance of the tabu search is found to be better in approaching the global optima. It is also proved that the method is effective in dealing with noises.
Guided Local Search
Identification
Local optimum
Hill climbing
Cite
Citations (0)
A hybrid tabu search algorithm was proposed in this paper for the reentrant scheduling problem on parallel machines with a remote service equipment.Concerning that only one start-point was used in the traditional tabu search algorithm which made it trapped in local optimum easily,a Restart method was established in the hybrid tabu search algorithm.When the traditional tabu search algorithm was trapped in local optimum,the Restart method was used to rebuild the initial solution and preceded with the tabu search algorithm.Thus,the traditional single start-point search was changed into multiple start-points search.Comparisons were made between the hybrid tabu search algorithm and a Coordinate Scheduling(CS) algorithm.The computational experiments show the effectiveness of the hybrid tabu search algorithm,whose optimization performance is superior to the CS algorithm.Moreover,the computation time is acceptable.
Guided Local Search
Hill climbing
Reentrancy
Cite
Citations (1)
Tabu Search is a metaheuristic that has proven to be very effective for solving various types of combinatorial optimization problems. To achieve the best results with a tabu search algorithm, significant benefits can sometimes be gained by determining preferred values for certain search parameters such as tabu tenures, move selection probabilities, the timing and structure of elite solution recovery for intensification, etc. In this paper, we present and implement some new ideas for fine‐tuning a tabu search algorithm using statistical tests. Although the focus of this work is to improve a particular tabu search algorithm developed for solving a telecommunications network design problem, the implications are quite general. The same ideas and procedures can easily be adapted and applied to other tabu search algorithms as well.
Guided Local Search
Hill climbing
Cite
Citations (74)
This article presents a review of a kind of evolutionary algorithms belonging to lamarckian learning group. It begins with providing background of evolutionary algorithms. Explanation on generic tabu search is given together with reviews of its variation. The reviews cover reactive tabu search (RTS), parallel tabu search (PTS), probabilistic tabu search (PrTS), hybrid tabu search (HTS) and adaptive tabu search (ATS). Finally, the article presents a meta-heuristic approach referred to as the management agent (MA) for search. With the MA, search performance can be improved dramatically as being confirmed by our results of surface optimization problems. In this article the most important patents are also discussed. Keywords: Evolutionary algorithms, tabu search, adaptive tabu search, management agent
Guided Local Search
Hill climbing
Cite
Citations (2)
This paper presents a Tabu Search based strategy which has been applied to optimize the objective function associated to the design of a solenoids system. For this kind of problem, Tabu Search strategies are able to find the global minimum with considerably better performance than that obtained using a Simulated Annealing algorithm. Algorithms used to perform the self-tuning of Tabu Search parameters have also been developed, achieving interesting improvements in terms of computing time and quality of the solutions. Such algorithms are an original development of Reactive Tabu Search and they are based on the past history of the search. Tabu Search strategies have also been applied in continuous optimization of the same problem with encouraging results.
Guided Local Search
Hill climbing
Cite
Citations (28)
A modified Tabu search algorithm is applied to power system reactive power optimization and a corresponding mathematical model is established in this paper in which the cost for line losses are considered to be minimun. On the basis of general Tabu search algorithm are analyzed and some corresponding improvements are made, thus, it is easily to stop the iteration of patial optimization and it is more probable to achieve the globe optimization by use of modified Tabu search algorithm. Comparing the optimized result for IEEE6 system by modified Tabu search algorithm with that by Tabu the result shows that modified Tabu search algorithm performs better in global optimization.
Guided Local Search
Hill climbing
Cite
Citations (16)
Tabu search is a widely used heuristic search method. However, there are two drawbacks in traditional tabu search. First, tabu search, as well as other search methods, only provides the best solution obtained during the search process, and there is no way to know the quality of the obtained solutions. Second, tabu list helps tabu search avoid the problem of looping in a small cycle, but it cannot prevent tabu search from searching previously searched areas again or, worse, looping in a large cycle.
The computation time of a search method can be reduced by implementing parallel processing. This study proposes a parallel deterministic simple tabu search, which computes more efficiently and overcomes the two drawbacks mentioned above. The results of our experiments show that it takes less computation effort for the proposed parallel tabu search to find a global optimal solution than for a conventional parallel tabu search.
Guided Local Search
Hill climbing
Local optimum
Cite
Citations (1)
In this article we propose a model for the topological design problem of multitechnology networks that includes the location of switches and their port configuration, the design of an access network (with single and double access links) and a backbone network. The model specifically takes into account different types of modular switches where each type is characterized by its cost, its capacity in terms of the number of slots, and by its switch fabric capacity. The mutitechnology qualifier stems from the fact that several technologies and rates can be used in the access network. The proposed model is of the integer-programming variety, and in order to find a good solution, we propose a starting heuristic that provides an initial solution and the tabu-search algorithm to improve the solution. Lower bounds are proposed and used to assess the performance of the tabu-based approach. Numerical results for randomly generated problems with up to 500 clients and 50 potential switch sites are presented. The tabu-search algorithm produced solutions that were, on average, within 2.11% of the best lower bound.
Port (circuit theory)
Branch and bound
Cite
Citations (19)