Engineering-Research-Activity2015｜SHINSHU UNIV.

29/75

Research on “Multi-Objective Optimization” using “Evolutionary Computation” and its application to solve various real-world problems.Tanaka, Hernanand Akimoto Lab. has been working on “Multi-Objective Optimization” using the “Evolutionary Computation”, inspired from the natural evolution of living things, and the development of its applied technology. “Multi-Objective Optimization” must involve several conflicting objective functions such as quality and cost, and this is the key issue in industrial applications and decision making.Our laboratory develops new “Multi-Objective Evolutionary Algorithms (MOEA)” that effectively and efficiently solve complex optimization problems including many objective functions and design variables, and we are pursuing the collaborative research with other research institutes and companies to apply the technology to various real-world problems that require “Multi-Objective Optimization.”“Multi-Objective Evolutionary Algorithms” evaluate many individuals in a given population using multiple objective functions , and choose superior individuals as parents, who then produce their offspring by crossing over and mutating genetic information. By repeating the evolution of the entire population, we can obtain a set of compromised solutions (many options to be chosen) called POS. We hope to apply this technology to numerous applications such as smart grid, cloud computing and derivation of design variables (set of parameters to express complex shapes) in manufacturing.The alumni of our laboratory are working as research and development engineers at the manufacturers in a wide range of areas such as electrical, electricity, communication, information processing, machinery, electric device and medical system. Some of them are faculties at universities including Shinshu University.Professor Kiyoshi TanakaProf. Tanaka has worked at the Faculty of Engineering since 1995. Currently, he is the associate dean of the Faculty of Engineering. He has put great effort into industry liaisons with companies and international exchange with universities and research centers abroad. His laboratory has energetically conducted research and education globally by receiving many international students and researchers.The concept of problem-solving using Evolutionary Computation for large scale/many-objective optimization problem; derivation of the set of optimal solutions in the management and the operation of complicated system.Shinshu Univ. has signed the agreement for academic cooperation and exchange with CINVESTAV which is the most prestigious graduate university in Mexico. (From the left: President of CINVESTAV Prof. Palacio, Prof. Coello, Prof. Tanaka.)TanakaLabElectrical& ElectronicEngineeringIn the FutureAfter GraduationHernanLabElectrical& ElectronicEngineeringIn the FutureAfter GraduationEvolutionarycomputationisusedtofindoptimalsolutionstomulti-objective,largescale,complexoptimizationproblemsEvolutionarycomputationsimulatesevolutionanditisappliedtosolveavarietyofdesigninnovationandsustainabilityproblemsSmartpowergrids,intelligentmobilityandtransportationsystems,andintelligentwatergridsarethreekeysystemstowhichevolutionarycomputationcanbeusedtooptimizetheirdesignandimprovetheirsustainabilitytomeettheneedsofthefuture.Additionally,designinnovationinkeyindustries,suchasautomobileandspaceexploration,areareasinwhichevolutionarycomputationwillplayanimportantrole.IntheLaboratorywelearnaboutreal-worldproblem-solvingandoptimizationbycomputationalmeans.Thisrequiresaclearunderstandingoftheproblem,itsmodeling,andthecreationofaprogramtosolveit.Thereisalsoopportunitiestocollaborateinjointresearchwithindustry.Theseactivitiescomplementstheeducationreceivedintheundergraduatecurseandbroadenstheskillsofengineers.AGUIRREHERNANAssociate ProfessorEcuadorian Engineer. PhD from ShinshuUniversity in 2003. Collaborates actively with industry, promotes international exchange, and conducts joint research with national and international institutions. Evolutionarycomputationusescomputationalmodelsofnaturalevolutiontosolvecomplexproblemsinscienceandengineering.Thecreationofsustainablesystemstomeetourneedswithoutcompromisingfuturegenerationsisasocialchallenge.Itdemandsthedevelopmentofnewtechnologiesandtheredesignofourinfrastructure,balancingenvironmental,economic,andsocietalneeds.Sustainabilityproblemsareintrinsicallycomplex,dynamic,large-scale,spanseveraldisciplines,andrequiremultidisciplinaryeffortsandmethodstosolvethem.Weusethepowerofevolutionarycomputationfordesigninnovationandoptimizationofsolutionstocomplexsustainabilityproblems.Evolutionary Multi-objective Optimization and Sustainability27

※このページを正しく表示するにはFlashPlayer10.2以上が必要です