ارزيابي ابزارهاي مدل‌سازي و شبيه‌سازي مبتني بر عامل بر اساس استاندارد ايزو 25010

Evaluation of agent-based modeling and simulation tools based on ISO 25010

ﺑﺎ ﺗﻮﺟﻪ ﺑﻪ ﻧﻘﺶ ﮐﻠﯿﺪي و اﻫﻤﯿﺖ اﺳـﺘﻔﺎده از اﺑﺰارﻫـﺎي ﻣﺪلﺳـﺎزي و ﺷﺒﯿﻪﺳـﺎزي ﻣﺒﺘﻨـﯽ ﺑـﺮ ﻋﺎﻣـﻞ در ﮐﺎرﺑﺮدﻫﺎي ﻣﺨﺘﻠﻒ، ﻣﺎﻧﻨﺪ ﻣﺪﯾﺮﯾﺖ و ﺑﺮﻧﺎﻣﻪرﯾﺰي ﺷﻬﺮي، ﺷﺒﮑﻪﻫﺎي اﺟﺘﻤﺎﻋﯽ، ﺑﺎزارﻫﺎي ﻣﺎﻟﯽ، ﺟﺮﯾﺎنﻫﺎي ﺗﺮاﻓﯿﮑﯽ و ﻣﺪﯾﺮﯾﺖ ﺑﺤﺮان، ارزﯾﺎﺑﯽ و ﻣﻘﺎﯾﺴﻪ ﮐﻤّـﯽ اﯾـﻦ اﺑﺰارﻫـﺎ ﺑـﺮ اﺳـﺎس ﯾـﮏ ﭼـﺎرﭼﻮب اﺳـﺘﺎﻧﺪارد ﺧﻮشﺗﻌﺮﯾﻒ، ارزﺷﻤﻨﺪ اﺳﺖ. ﺗﺎﮐﻨﻮن ﭘﮋوﻫﺶﻫﺎي ﻣﺘﻌﺪدي ﺑﻪ ﺑﺮرﺳﯽ و ﻣﻘﺎﯾﺴﻪ اﺑﺰارﻫﺎي ﻣﺪلﺳﺎزي و ﺷﺒﯿﻪﺳﺎزي ﻣﺒﺘﻨﯽ ﺑﺮ ﻋﺎﻣﻞ از ﺟﻨﺒﻪﻫﺎي ﻣﺨﺘﻠﻒ ﭘﺮداﺧﺘﻪاﻧﺪ. ﺑﺎ اﯾﻦ وﺟﻮد، ﻫﯿﭻﯾﮏ از اﯾـﻦ ﺑﺮرﺳـﯽﻫﺎ ﺑـﺮ اﺳﺎس ﯾﮏ ﭼﺎرﭼﻮب اﺳﺘﺎﻧﺪارد ﺧﻮشﺗﻌﺮﯾﻒ و ﺑﻪ ﺻﻮرت ﮐﻤّﯽ ﻋﺪدي( اﻧﺠﺎم ﻧﺸﺪه اﺳﺖ. ﺑﻪﻣﻨﻈﻮر ﭘـﺮ ﮐﺮدن اﯾﻦ ﺷﮑﺎف ﺗﺤﻘﯿﻘﺎﺗﯽ، در اﯾﻦ ﻣﻘﺎﻟﻪ ﭘﻨﺞ ﻣﻮرد از ﭘﺮﮐﺎرﺑﺮدﺗﺮﯾﻦ اﺑﺰارﻫﺎي ﻣﺪلﺳﺎزي و ﺷﺒﯿﻪﺳﺎزي ﻣﺒﺘﻨﯽ ﺑﺮ ﻋﺎﻣﻞ ﺷﺎﻣﻞ GAMA ،Repast ،NetLogo ،AnyLogicو MASON، ﺑﺮ اﺳﺎس اﺳـﺘﺎﻧﺪارد ﺷﻨﺎﺧﺘﻪﺷﺪه اﯾﺰو 25010در ﺧﺼﻮص ﺳﻨﺠﺶ ﮐﯿﻔﯿﺖ ﻧﺮماﻓﺰار در ﻋﻤﻞ و ﺑﻪ روش ﻫﺪف-ﭘﺮﺳﺶ- ﻣﻌﯿﺎر ﺑﺎ ﯾﮑﺪﯾﮕﺮ ﻣﻘﺎﯾﺴﻪ و ﻣﻮرد ارزﯾﺎﺑﯽ ﻗﺮار ﮔﺮﻓﺘﻪاﻧﺪ. ﻧﺘﺎﯾﺞ ارزﯾﺎﺑﯽ ﻧﺸﺎن ﻣﯽدﻫﺪ ﮐﻪ اﺑـﺰار AnyLogic در ﻣﻘﺎﯾﺴﻪ ﺑﺎ دﯾﮕﺮان، از ﮐﯿﻔﯿﺖ ﺑﻬﺘﺮي ﺑﺮﺧﻮردار اﺳﺖ و اﯾﻦ ﻣﻮﺿﻮع، ﺑﯿﺶﺗﺮ ﺑﻪ ﺟﺬاﺑﯿﺖ اﺑـﺰار از دﯾـﺪﮔﺎه ﮐﺎرﺑﺮ و ﺗﻼش ﮐﻤﺘﺮ ﺑﺮاي ﺗﻮﺳﻌﻪ ﯾﮏ ﺑﺮﻧﺎﻣﻪ ﻣﺒﺘﻨﯽ ﺑﺮ ﻋﺎﻣﻞ در ﻣﻘﺎﯾﺴﻪ ﺑﺎ ﺳﺎﯾﺮ اﺑﺰارﻫﺎ، ﺑﺎز ﻣﯽﮔﺮدد.

Due to the key role and importance of using agent-based modeling and simulation tools in various applications, such as urban management and planning, social networks, financial markets, traffic flows, and crisis management, it is worthwhile to evaluate and quantitatively compare these tools based on a well-defined standard framework. So far, several studies have reviewed and compared agent-based modeling and simulation tools from different aspects. However, none of the studies are quantitatively performed based on a well-known standard framework. To fill this research gap, in this paper, five of the most widely used agent-based modeling and simulation tools including AnyLogic, NetLogo, Repast, GAMA, and MASON were compared and evaluated based on the well-known ISO 25010 standard, regarding the quality in use, by Goal-Question-Metric (GQM) method. The evaluation results show that AnyLogic tools are of better quality than others, and this is due to the attractiveness of the tool from the user's point of view and less effort to develop an agent-based application compared to other tools.