NetLogo: An Open Source Plat­form for Study­ing Emer­gent Be­hav­iour

Here’s a brief note on Net Logo, which can be de­scribed as a multi-agent pro­gram­mable mod­el­ling en­vi­ron­ment. It is widely used by teach­ers, students and re­searchers to model and study com­plex sys­tems by cre­at­ing sim­u­la­tions and ‘play­ing’ with them.

OpenSource For You - - Contents - By: Hithu Anand The au­thor has a teach­ing back­ground and is cur­rently re­search­ing on smart grids. He can be reached at

Op­ti­mi­sa­tion is in­her­ent in na­ture. We have seen the key as­pects of op­ti­mi­sa­tion in pre­vi­ous edi­tions of OSFY. An­other in­ter­est­ing phe­nom­e­non is called emer­gent be­hav­iour, which could be nat­u­ral or ar­ti­fi­cial. Swarms of ants, schools of fish, flocks of birds, etc, are some of the nat­u­ral ex­am­ples of emer­gent be­hav­iour. The in­creased com­plex­ity of rout­ing air traf­fic is con­sid­ered to be ar­ti­fi­cial emer­gent be­hav­iour. Study­ing these com­plex be­hav­iours us­ing nu­mer­i­cal tech­niques is not easy. Hence, an agent based sim­u­la­tion plat­form called NetLogo is used to un­der­stand such phe­nom­ena.

In­tro­duc­ing NetLogo

NetLogo is GNU GPL li­censed free soft­ware, de­signed by Uri Wilen­sky, di­rec­tor of North­west­ern Univer­sity, in 1999. It is a JVM based cross-plat­form multi-agent sim­u­la­tion plat­form. NetLogo doesn’t re­quire cod­ing knowl­edge or any other pre­req­ui­sites. It is de­signed for a broad au­di­ence. So any­one

from sec­ondary school students to grad­u­ates can make use of this soft­ware to study their field of in­ter­est.


In­stal­la­tion of NetLogo in Ubuntu 14.04 is easy. First, down­load the NetLogo pack­age from http://ccl.north­west­ I have a 64-bit Linux com­puter, so I have down­loaded the pack­age ac­cord­ingly. For NetLogo ver­sion 6.0.1, mem­ory size of 197 MB is re­quired. The down­loaded file is com­pressed and must be ex­tracted in the home di­rec­tory (prefer­ably in the opt folder). Once ex­tracted, you are ready to use NetLogo on your com­puter. In the NetLogo di­rec­tory, dou­ble-click on the NetLogo ex­e­cutable to open the GUI of the soft­ware.

Par­ti­cle swarm op­ti­mi­sa­tion (PSO) in NetLogo

PSO is a na­ture in­spired swarm­ing the­ory based al­go­rithm in­vented by Kennedy and Eber­hart in the 1990s. Flocks of birds mi­grat­ing from place to place in search of food and shel­ter are the ba­sis for this al­go­rithm. For­tu­nately, NetLogo has in­built model li­braries where many mod­els like PSO are read­ily avail­able. The model demon­strates how the true op­ti­mum value 1 is found for var­i­ous pa­ram­e­ter set­tings like land­scape smooth­ness (struc­ture of valleys and hills), pop­u­la­tion size (num­ber of birds), etc.

The NetLogo in­ter­face is shown in Fig­ure 1. Fig­ure 3 shows the PSO model file avail­able in its li­brary. Fig­ure 2 shows the PSO find­ing its true value in 30 ticks. Fig­ure 4 shows the same at­tained in 22 ticks. Note that fewer ticks are ob­served for a pop­u­la­tion of 10. Fig­ures 5 and 6 show the above-men­tioned ini­tial and fi­nal re­sults.

Fig­ure 1: NetLogo in­ter­face 1

Fig­ure 3: NetLogo in­ter­face 3

Fig­ure 4: NetLogo in­ter­face 4

Fig­ure 2: NetLogo in­ter­face 2

Fig­ures 5: Sim­u­la­tion re­sults 1

Fig­ures 6: Sim­u­la­tion re­sults 2

Newspapers in English

Newspapers from India

© PressReader. All rights reserved.