Weighted superposition attraction algorithm for binary optimization problems
- PDF / 2,173,534 Bytes
- 27 Pages / 439.37 x 666.142 pts Page_size
- 46 Downloads / 178 Views
Weighted superposition attraction algorithm for binary optimization problems Adil Baykasoğlu1 · Fehmi Burcin Ozsoydan1 · M. Emre Senol1 Received: 27 March 2018 / Revised: 27 August 2018 / Accepted: 10 September 2018 © Springer-Verlag GmbH Germany, part of Springer Nature 2018
Abstract Weighted superposition attraction algorithm (WSA) is a new generation populationbased metaheuristic algorithm, which has been recently proposed to solve various optimization problems. Inspired by the superposition of particles principle in physics, individuals of WSA generate a superposition, which leads other agents (solution vectors). Alternatively, based on the quality of the generated superposition, individuals occasionally tend to perform random walks. Although WSA is proven to be successful in both real-valued and some dynamic optimization problems, the performance of this new algorithm needs to be examined also in stationary binary optimization problems, which is the main motivation of the present study. Accordingly, WSA is first designed for stationary binary spaces. In this modification, WSA does not require any transfer functions to convert real numbers to binary, whereas such functions are commonly used in numerous approximation algorithms. Moreover, a step sizing function, which encourages population diversity at earlier iterations while intensifying the search towards the end, is adopted in the proposed WSA. Thus, premature convergence and local optima problems are attempted to be avoided. In this context, the contribution of the present study is twofold: first, WSA is modified for stationary binary optimization problems, secondarily, it is further enhanced by the proposed step sizing function. The performance of the modified WSA is examined by using three well-known binary optimization problems, including uncapacitated facility location problem, 0–1 knapsack problem and a natural extension of it, the set union knapsack problem. As demonstrated by the comprehensive experimental study, results point out the efficiency of the proposed WSA modification in binary optimization problems. Keywords Weighted superposition attraction algorithm · Binary optimization · Uncapacitated facility location problem · 0–1 Knapsack problem · Set union knapsack problem * Adil Baykasoğlu [email protected]; [email protected] http://web.deu.edu.tr/baykasoglu Extended author information available on the last page of the article
13
Vol.:(0123456789)
A. Baykasoğlu et al.
1 Introduction As recently introduced by Baykasoğlu and Akpinar (2015, 2017), Weighted Superposition Attraction algorithm (WSA) is a novel swarm intelligence-based metaheuristic algorithm, proposed to solve real-valued constrained and unconstrained optimization problems. WSA draws inspiration from the superposition of particles principle in physics. Solution vectors in WSA generate a superposition, which is followed by some of the agents (solution vectors), depending on the quality of the generated superposition. As mentioned by Baykasoğlu and Akpinar (2015,
Data Loading...
