Emerging Physics-based CI Algorithms

Abstract

In this chapter, a set of (more specifically 22 in total) emerging physicsbased computational intelligence (CI) algorithms are introduced. We first, in Sect. 24.1, describe the organizational structure of this chapter. Then, from Sects. 24.2 to 24.23, each section is dedicated to a specific algorithm which falls within this category. The fundamentals of each algorithm and their corresponding performances compared with other CI algorithms can be found in each associated section. Finally, the conclusions drawn in Sect. 24.24 closes this chapter. IntroductionSeveral novel physics-based algorithms were detailed in previous chapters. In particular, Chap. 18 detailed the big bang-big crunch algorithm, Chap. 19 was dedicated to central force optimization algorithm, Chap. 20 discussed the charged system search algorithm, Chap. 21 introduced the electromagnetism-like mechanism algorithm, Chap. 22 was devoted to the gravitational search algorithm, and Chap. 23 described the intelligent water drops algorithm. Apart from this quasimature physics principles inspired CI methods, there are some emerging algorithms also fall within this category. This chapter collects 22 of them that are currently scattered in the literature and organizes them as follows: The effectiveness of these newly developed algorithms are validated through the testing on a wide range of benchmark functions and engineering design problems, and also a detailed comparison with various traditional performance leading CI algorithms, such as particle swarm optimization (PSO), genetic algorithm (GA), differential evolution (DE), evolutionary algorithm (EA), fuzzy system (FS), ant colony optimization (ACO), and simulated annealing (SA). Artificial Physics Optimization AlgorithmIn this section, we will introduce an emerging CI algorithm that is based on artificial physics or physicomimetics, a concept which was introduced in Spears et al. (2004a, b); Spears and Gordon (1999); Spears and Spears (2012). Fundamentals of Artificial Physics Optimization AlgorithmArtificial physics optimization (APO) algorithm was recently proposed in Xie et al. ( , b, 2010aXie et al. ( , b, 2011a, . Several APO applications and variants can also be found in the literature Popov 2012, 2013; Zeng 2009b, 2011;Mo and Zeng 2009; Wang and Zeng 2010a, b;Yang et al. 2010;Yin et al. 2010; Xie et al. 2011c, d;Wang et al. 2011). To implement the APO algorithm, the following steps need to be performed Biswas et al. 2013):• Initialization step: At this step, a swarm of individuals is randomly generated in the n-dimensional decision space. 376 24 Emerging Physics-based CI Algorithms• Force calculation step: At this step, according to the masses and distances between individual particles, the total force exerted on each particle is computed. In APO, the mass is defined via Eq. 24.1 Biswas et al. 2013):The force is then calculated through Eq. 24.2 Biswas et al. 2013):ð24:2ÞThe kth component of the total force F i;k exerted on individual i by all other individuals is acquired via Eq. 24.3...

Νοήμονες Υπολογιστικές Μέθοδοι Εμπνευσμένες Από Τον Φυσικό Κόσμο Για Την Βελτιστοποίηση Συστημάτων

“…Οι δημοσιεύσεις των Xing and Gao (2014), Biswas et al (2013), Can and Alatas (2015) και Siddique and Adeli (2016) έχουν χαρτογραφήσει εν μέρει την συγκεκριμένη κατηγορία αλγορίθμων, χωρίς ωστόσο σε καμία εξ αυτών να περιέχεται το σύνολο των μεθόδων. Στις υποενότητες που ακολουθούν γίνεται μια διεξοδική ανάλυση των μεθόδων Νοημοσύνης Εμπνευσμένης από Φυσικά Φαινόμενα και Νόμους των Επιστημών (ΝΕΦΦΝΕ).…”