مشخصات پژوهش

صفحه نخست /Deep Bernoulli optimisation ...
عنوان Deep Bernoulli optimisation for solving 2D/3D ψ tempered fractional optimal control problems
نوع پژوهش مقاله چاپ‌شده
کلیدواژه‌ها ψ-temperedfractional optimalcontrolproblems; deepneuralnetwork method;fractional-order Bernoullifunctions;error estimate;collocationmethod
چکیده Thisstudyintroducestwo-andthree-dimensionaloptimalcontrolproblemscharacterisedbythe ψ-temperedfractionalderivativeandproposesanovelnumericalmethodologybasedondeep fractional-orderBernoullioptimisationtoachievetheirefficientsolution.Tothisend,theproblems underconsiderationarefirstreformulatedasequivalentvariationalproblems.Subsequently,adeep neuralnetworkemployingthefractional-orderBernoullifunctionsandsinhasactivationfunctionsis utilisedtoapproximatethestatevariable.Tofacilitatetheeffectiveimplementationoftheproposed method,weconstructseveral integraloperatorsofbothintegerandψ-temperedfractionalorders basedonthebasisfunctionsderivedfromthedeepneuralnetwork.Theseoperatorsarediscretised viatheFejérquadratureruleadaptedtotheψ-temperedfractionalcalculus,ensuringstabilityand high-precisionintegration.Theproposedapproachconvertsthe2D/3Dψ-temperedfractionalopti malcontrolproblemintoanalgebraicsystemusingdeepneuralnetworkswithintegraloperators andGauss–Legendreintegration,whichisefficientlysolvedviaNewton’smethod.Theproposed methodcombinessimpleimplementation,lowcomputationalcost,andhighaccuracy.Itseffective nessandrobustnessarevalidatedthroughseveralrepresentative2Dand3Dexamples,confirming themethod’sapplicabilitytocomplexfractionaloptimalcontrolproblems.
پژوهشگران پریسا رحیم خانی (نفر اول)، ثابت عبدالجواد (نفر دوم)