第76章 对火星轨道变化问题的最后解释(5/14)

投票推荐 加入书签 留言反馈

hestepsize.however,sincetheeccentricityofjupiter(~0.05)ismuchsmallerthanthatofmercury(~0.2),weneedsomecarewhenwecomparetheseintegrationssimplyintermsofstepsizes.
    intheintegrationoftheouterfiveplanets(f±),wefixedthestepsizeat400d.
    weadoptgauss'fandgfunctionsinthesymplecticmaptogetherwiththethird-orderhalleymethod(danby1992)asasolverforkeplerequations.thenumberofmaximumiterationswesetinhalley'smethodis15,buttheyneverreachedthemaximuminanyofourintegrations.
    theintervalofthedataoutputis200000d(~547yr)forthecalculationsofallnineplanets(n±1,2,3),andabout8000000d(~21903yr)fortheintegrationoftheouterfiveplanets(f±).
    althoughnooutputfilteringwasdonewhenthenumericalintegrationswereinprocess,weappliedalow-passfiltertotheraworbitaldataafterwehadcompletedallthecalculations.seesection4.1formoredetail.
    2.4errorestimation
    2.4.1relativeerrorsintotalenergyandangularmomentum
    accordingtooneofthebasicpropertiesofsymplecticintegrators,whichconservethephysicallyconservativequantitieswell(totalorbitalenergyandangularmomentum),ourlong-termnumericalintegrationsseemtohavebeenperformedwithverysmallerrors.theaveragedrelativeerrorsoftotalenergy(~10?9)andoftotalangularmomentum(~10?11)haveremainednearlyconstantthroughouttheintegrationperiod(fig.1).thespecialstartupprocedure,warmstart,wouldhavereducedtheaveragedrelativeerrorintotalenergybyaboutoneorderofmagnitudeormore.
    relativenumericalerrorofthetotalangularmomentumδaa0andthetotalenergyδee0inournumericalintegrationsn±1,2,3,whereδeandδaaretheabsolutechangeofthetotalenergyandtotalangularmomentum,respectively,ande0anda0aretheirinitialvalues.thehorizontalunitisgyr.
    notethatdifferentoperatingsystems,differentmathematicallibraries,anddifferenthardwarearchitecturesresultindifferentnumericalerrors,throughthevariationsinround-offerrorhandlingandnumericalalgorithms.intheupperpaneloffig.1,wecanrecognizethissituationinthesecularnumericalerrorinthetotalangularmomentum,whichshouldberigorouslypreserveduptoma

本章未完,点击下一页继续阅读


章节目录