Redundancy Allocation Test Problems
SOURCE: Coit, D.W. and A. Konak, (2006).
Multiple Weighted Objectives Heuristic for the Redundancy Allocation Problem, IEEE Transactions on Reliability, 55(3), 551-558.
In this paper, three new challenging test problems were proposed for the redundancy allocation problem in series-parallel systems. Each problem has 20 subsystems with four available component options for each subsystem. The maximum number of components allowed in each subsystem is 8. The problems were solved for 36 different combinations of cost (C) and weight constraints (W). Please refer to the paper for more information. The followings are AMPL input files and the best solution found for each problem instance. The format of the input files is easy to understand-NOTATION c: cost,
- Problem 2 (positively correlated)-AMPL input data file
- Problem 3 (negatively correlated)-AMPL input data file
- Problem 4 (negatively and positively correlated)-AMPL input data file
- Solutions of 36 Problem Instances
Network Design Problem with Relays Test Problems
SOURCE:Konak, A., (2012). Network Design Problem with Relays: A Genetic Algorithm with a Path-Based Crossover and a Set Covering Formulation, European Journal of Operational Research, 218(3), 829-837.
The following is the zipped AMPL input files for all test problems. Individual files are named based on their input parameters. For example, file 50N_5K_30L.dat is the input file for the test problem with 50 nodes, 5 commodities, and the relay constraint of 30 (lambda). There are two types of problems, type I where c_ij=d_ij and type II where c_ij=lambda-d_ij (please refer to the paper for more information). Type II problems are indicated using p3 in the file name (e.g., 50N_5K_30L_p3.dat is the type II version of 50N_5K_30L.dat). NOTATION nc: node relay cost, ac: arc cost, ad: arc distance, COM: set of commodities such that the first node is the source and the second node is the target.
AMPL input files were generated from the following text files. You may find them easier to read into your code. By columns, Node Data NOTATION – c1: node, c2: x coordinate, c3: y coordinate, c4: relay cost. By columns, Commodity Data NOTATION – c1:commodity index, c2:source node, and c3:destination node.
2-Edge Disjoint Network Design Problem with Relays Test Problems
SOURCE: Konak, A. (2014). “Two-Edge Disjoint Survivable Network Design Problem with Relays: A Hybrid Genetic Algorithm and Lagrangian Heuristic Approach,” Engineering Optimization, 46(1), 130-145.
There is only one type of problems, type I where c_ij=d_ij (please refer to the paper for more information). In this paper also, d_ij is also the Euclidean distance between nodes i and j. By columns, Node Data NOTATION – c1: node, c2: x coordinate, c3: y coordinate, c4: relay cost. By columns, Commodity Data NOTATION – c1:commodity index, c2:source node, and c3:destination node.
- Zipped node location files
- MS World file of Table 2 from the original paper . Note that the results in Table 1 are only for comparing various aspects of the algorithm).
Dynamic and Cyclic Facility Layout Test Problems
SOURCE: Kulturel-Konak S. and Konak, A. (2015). “A Large Scale Hybrid Simulated Annealing Algorithm for Cyclic Facility Layout Problem,” Engineering Optimization, 47(7), 963-973.
DFLP 12-3, DFLP 12-5, DFLP 20-3 are from Lacksonen (1997). In these problems, new departments are introduced in each planning period (two new departments per period in DFLP 12-3 and 12-5, and three departments per period in DFLP 20-3). In addition to rearrangement of the departments, a relayout cost is incurred when a new department replaces an existing department. Therefore, we considered the relayout cost of a department to be replaced as zero in the data file and then added this cost to the final objective function value in the paper results.