Improvement of the Gravitational Search Algorithm by means of Low-Discrepancy Sobol Quasi Random-Number Sequence Based Initialization

Altinoz, O. Tolga
Weber, Gerhard Wilhelm
Nature-inspired optimization algorithms can obtain the optima by updating the position of each member in the population. At the beginning of the algorithm, the particles of the population are spread into the search space. The initial distribution of particles corresponds to the beginning points of the search process. Hence, the aim is to alter the position for each particle beginning with this initial position until the optimum solution will be found with respect to the pre-determined conditions like maximum iteration, and specific error value for the fitness function. Therefore, initial positions of the population have a direct effect on both accuracy of the optima and the computational cost. If any member in the population is close enough to the optima, this eases the achievement of the exact solution. On the contrary, individuals grouped far away from the optima might yield pointless efforts. In this study, low-discrepancy quasi-random number sequence is preferred for the localization of the population at the initialization phase. By this way, the population is distributed into the search space in a more uniform manner at the initialization phase. The technique is applied to the Gravitational Search Algorithm and compared via the performance on benchmark function solutions.


Random Sequences in Vehicle Routing Problem
Gülşen, Mehmet Emin; Yayla, Oğuz; Department of Cryptography (2022-9-1)
Vehicle Routing Problem (VRP) is a classical combinatorial optimization and integer programming problem. The goal of VRP is to find the optimal set of routes to given set of destination points with a fleet of vehicles. In this thesis, we have focused on the a variant of VRP which is Capacitated Vehicle Routing Problem (CVRP) and present two different combination of heuristic algorithms with random projection clustering technique and also provide comparison of random number generators on Monte Carlo Simulati...
Optimizations of antennas using heuristic algorithms supported by the multilevel fast multipole algorithm
Önol, Can; Ergül, Özgür Salih; Department of Electrical and Electronics Engineering (2015)
In this study, an optimization environment based on heuristic algorithms supported by the multilevel fast multipole algorithm (MLFMA) is presented for different antenna problems involving either excitation or geometry optimizations. The heuristic algorithms are implemented in-house by aiming more effective interactions between electromagnetic solvers and optimization algorithms, instead of black box interactions. Excitation optimizations of various array geometries for desired radiation characteristics are ...
Optimization of nonuniform array geometry for DOA estimation with the constraint on gross error probability
Birinci, Toygar; Tanık, Yalçın (2007-10-01)
In this work, a novel method is proposed to optimize the array geometry for DOA estimation. The method is based on minimization of fine error variances with the constraint that the gross error probability is below a certain threshold. For this purpose, a metric function that reflects the gross and fine error characteristics of the array is proposed. Theoretical analyses show that the minimization of this metric function leads to small DOA estimation error variance and small gross error probability. Analyses...
Rigorous Solutions of Large-Scale Scattering Problems Discretized with Hundreds of Millions of Unknowns
Guerel, L.; Ergül, Özgür Salih (2009-09-18)
We present fast and accurate solutions of large-scale scattering problems using a parallel implementation of the multilevel fast multipole algorithm (MLFMA). By employing a hierarchical partitioning strategy, MLFMA can be parallelized efficiently on distributed-memory architectures. This way, it becomes possible to solve very large problems discretized with hundreds of millions of unknowns. Effectiveness of the developed simulation environment is demonstrated on various scattering problems involving canonic...
Tuning of handover parameters in LTE-A heterogeneous networks
Tunçel, Emrah; Şener, Cevat; Department of Computer Engineering (2014)
In this work, a new handover parameter optimization algorithm for the LTE-A networks is presented. Most of the previous solutions that are proposed to solve the handover parameter optimization problem, optimize the handover parameters with considering all of the neighbor cells together. But with the introduction of heterogeneous networks to the LTE-A standards, the deployments are no longer homogeneous and the cell borders have different inter-cell interference conditions that changes according to the diffe...
Citation Formats
O. T. Altinoz, A. E. YILMAZ, and G. W. Weber, “Improvement of the Gravitational Search Algorithm by means of Low-Discrepancy Sobol Quasi Random-Number Sequence Based Initialization,” ADVANCES IN ELECTRICAL AND COMPUTER ENGINEERING, pp. 55–62, 2014, Accessed: 00, 2020. [Online]. Available: