分数阶时间导数方程和反常亚扩散过程——纪念茆诗松教授.pdf
《分数阶时间导数方程和反常亚扩散过程——纪念茆诗松教授.pdf》由会员分享,可在线阅读,更多相关《分数阶时间导数方程和反常亚扩散过程——纪念茆诗松教授.pdf(20页珍藏版)》请在咨信网上搜索。
1、应用概率统计第 40 卷第 2 期2024 年 4 月Chinese Journal of Applied Probability and StatisticsApr.,2024,Vol.40,No.2,pp.323-342doi:10.3969/j.issn.1001-4268.2024.02.007Time Fractional Equations and Anomalous Sub-Diffusions In Memory of Professor Shisong MaoCHEN Zhen-Qing(Department of Mathematics,University of Wash
2、ington,Seattle,WA 98195,USA)Abstract:In this paper,we survey some recent progress in the study of time fractional equationsand its interplay with anomalous sub-diffusions,with some improvements and extensions.Keywords:time fractional derivative;time fractional equation;subordinator;inverse subordina
3、-tor;strong and weak solution2020 Mathematics Subject Classification:primary 26A33,60H30;secondary 34K37Citation:CHEN Z-Q.Time fractional equations and anomalous sub-diffusionsJ.Chinese JAppl Probab Statist,2024,40(2):323342.1IntroductionSub-diffusions are random processes that can model the motions
4、 of particles that moveslower than Brownian motion(or the original underlying spatial motion),for example,dueto particle sticking and trapping.A prototype of anomalous sub-diffusions is modeled byYt=BLt,where B is a d-dimensional Brownian motion and Lt:=infr 0:Sr t,t 0,is the inverse of a-stable sub
5、ordinator S with 0 0,which grows sub-linearly in t.Here():=0t1etdt is the Gamma function.On the other hand,fractional calculus has attracted lots of attentions in several fieldsincluding mathematics,physics,chemistry,engineering,hydrology and even finance andE-mail:zqchenuw.edu.Received December 1,2
6、023.Revised January 23,2024.324Chinese Journal of Applied Probability and StatisticsVol.40social sciences;see,e.g.,3,57.The classical heat equation tu=u describes heatpropagation in homogeneous medium.The time-fractional diffusion equation tu=uwith 0 0 be a subordinator with S0=0.It is well known th
7、at there are aunique constant 0 and a unique L evy measure on(0,)satisfying0(1x)(dx)0,t 0,326Chinese Journal of Applied Probability and StatisticsVol.40Define for t 0,Lt=infs 0:Ss t,the inverse subordinator.Throughout thispaper,we assume the L evy measure of the subordinator S is infinite(which is e
8、quivalentto w(x):=(x,)being unbounded)on(0,)excluding compound Poisson processes.Under this assumption,almost surely,t 7 Stis strictly increasing and hence t 7 Ltiscontinuous.Let St:=St t.Clearly,S is a subordinator having L evy exponent0()=0(1 ex)(dx).For every a 0,by Fubini theorem,a0w(x)dx=a0(x,)
9、(d)dx=0(a0dx)(d)=0(a)(d)0.(5)Remark 1For a left continuous decreasing functionwon(0,)withlimxw(x)=0,it uniquely determines a Radon measureon(0,)so that(x,)=w(x)forx 0.By(4),is the L evy measure of a subordinator if and only ifwis locally integrableon0,),andis an infinite measure if and only ifwis un
10、bounded.For any locally bounded function f on 0,),we defined the Riemann-Liouville typeintegral byIwtf:=t0w(t s)f(s)ds=t0w(s)f(t s)ds,t 0.(6)In view of(5),the Laplace transform of Iwt(f)isLIwtf()=0()Lf()for 0.(7)Following 13,we define the generalized time fractional derivative with weight w bywtf(t)
11、:=ddtIwtf=ddtt0w(t s)f(s)f(0)ds,(8)whenever the right hand side is well defined.Its connection to the classical Caputo typefractional derivative,which is defined by the right hand side of(9),is given in the followinglemma.The advantage of our definition of the time fractional derivative wtf(t)is tha
12、twe do not need to assume a priori the existence of f(s)for almost every s (0,t)nor theabsolute convergence of the integralt0w(t s)f(s)ds.No.2CHEN Z.-Q.:Time Fractional Equations and Anomalous Sub-Diffusions327Lemma 2Iffis a local Lipschitz function on0,),thenwtf(t)exists for almosteveryt 0andwtf(t)
13、=t0w(t s)f(s)dsfor a.s.t 0.(9)ProofSince f is a local Lipschitz function on 0,),f(s)exists for a.e.s (0,)and for every t 0,|f(s)|is bounded over 0,t by a constant M(t)depending on t.Thust0w(t s)|f(s)|ds 0,which is a continuous function with G(0)=0;see 13;(2.2).Using the Fubini theorem and the integr
14、ation by parts formula,we have for everyT 0,T0g(t)dt=T0Tsw(t s)dtf(s)ds=T0G(T s)f(s)ds=G(T s)f(s)?T0T0G(T s)f(s)ds=G(T)f(0)+T0w(T s)f(s)ds=T0w(T s)f(s)f(0)ds.This establishes the lemma.?Remark 3WhenSis a-stable subordinator with0 0 to every t 0.Note that in this paper,w(x)is defined to be(x,)rather
15、thanits right continuous version(x,)as in 13.They differ only at possibly countablymany points on(0,).Proposition 4For everyt 0ands 0,P(Lt6 s)=P(Ss t)=s0Ew(t Sr)1tSrdr.(10)Consequently,0Ew(t Sr)1tSrdr=1for everyt 0.(11)ProofFor a 0,let f=1a,).Since S is a driftless subordinator with infiniteL evy me
16、asure and f is a right continuous non-decreasing function,we have by 20;328Chinese Journal of Applied Probability and StatisticsVol.40Corollary 2 thatEf(St)=f(0)+Et0(0,)f(Sr+z)f(Sr)(dz)dr.Hence we have by Fubinis theorem that for every a 0 and t 0,P(St a)=Et0(0,)1zaSr1Sr6a(dz)dr=Et0w(a Sr)1Sr6adr=t0
17、Ew(a Sr)1Sr6adr.(12)As almost surely r 7 Sris strictly increasing,we have for each fixed t 0,Lt6 r=Sr t a.s.and soP(Lt6 r)=P(Sr t)=t0Ew(t Sr)1Srtdr.This in particular implies that Lthas a density function r 7 Ew(tSr)1Sr 0.The following lemma is takenfrom 13;(2.5)and Corollary 2.1(ii).It together wit
18、h Proposition 4 played an importantrole in 13,18,19 for the probabilistic approach to time fractional equations.Proposition 5For everyt,s 0,t0w(t r)P(Ss r)dr=G(t)EG(t Ss)1tSs,0EG(t Sr)1Sr6tdr=tand0EG(t Sr)1tSrdr 6 t.3Time Fractional Parabolic EquationRecall that St;t 0 is a general subordinator with
19、 infinite L evy measure anddrift 0,whose Laplace exponent()is given by(3).Define w(x)=(x,)for x 0and 0():=0(1ex)(dx).Note that 0()is the Laplace exponent of the driftlesssubordinator St:=St t,t 0 having L evy measure.Clearly()=+0()and St=t+St.Since(0,)=,almost surely,t 7 Stis strictly increasing.Sup
20、pose that Tt;t 0 is a strongly continuous semigroup with infinitesimal gener-ator(L,D(L)in some Banach space(B,)with the property that supt0Tt 0Tt 0Ttf 0 of a strong Markov process X=Xt,t 0;Px,x X on a Lusin space X that has a weak dual with respect to some-finite referencemeasure m on X.For every p
21、 1,Pt;t 0 is a strongly continuous semigroupin B:=Lp(E;m)with supt0Ptpp6 1.The infinitesimal generator(L,D(L)of Pt;t 0 in Lp(E;m)is called the Lp-generator of the Markov process X.(ii)Transition semigroup Pt;t 0 of a Feller process X=Xt,t 0;Px,x X ona locally compact separable metric space X.In this
22、 case,Pt;t 0 is a stronglycontinuous semigroup in the space(C(E),)of continuous functions on Xthat vanish at infinity equipped with uniform norm.The infinitesimal generator(L,D(L)of Pt;t 0 in B:=(C(X),)is called the Feller generator of X.(iii)Certain Feynman-Kac semigroups(can be non-local Feynman-K
23、ac semigroups or evengeneralized Feynman-Kac semigroups)in Lp-space or in C(X)of a Hunt processX;cf.2123.For 0,let G:=0etTtdt be the resolvent of the semigroup Tt;t 0 onthe Banach space B.Then by the resolvent equation,D(L)=G(B)=G1(B),whichis dense in the Banach space(B,).Recall that Lt:=infs 0:Ss t
24、,t 0,is theinverse of the subordinator S.Defineu(t)=ETLtf=0TsfdsP(Lt6 s)=0TsfdsP(Ss t).The following is essentially the main result of 13,Theorem 6 there,which gives theexistence and uniqueness of strong solutions to the time fractional parabolic equation(13).However,it contains an improvement(14)fo
25、r the positive drift case of the subordinatorS.This improvement was given in the Appendix of the arXiv version of 13,added afterits publication.For the readers convenience,we reproduce its proof here.Theorem 6(Strong solution)Suppose that(L,D(L)is the infinitesimal generatorof a uniformly bounded st
- 配套讲稿:
如PPT文件的首页显示word图标,表示该PPT已包含配套word讲稿。双击word图标可打开word文档。
- 特殊限制:
部分文档作品中含有的国旗、国徽等图片,仅作为作品整体效果示例展示,禁止商用。设计者仅对作品中独创性部分享有著作权。
- 关 键 词:
- 分数 时间 导数 方程 反常 扩散 过程 纪念 茆诗松 教授
1、咨信平台为文档C2C交易模式,即用户上传的文档直接被用户下载,收益归上传人(含作者)所有;本站仅是提供信息存储空间和展示预览,仅对用户上传内容的表现方式做保护处理,对上载内容不做任何修改或编辑。所展示的作品文档包括内容和图片全部来源于网络用户和作者上传投稿,我们不确定上传用户享有完全著作权,根据《信息网络传播权保护条例》,如果侵犯了您的版权、权益或隐私,请联系我们,核实后会尽快下架及时删除,并可随时和客服了解处理情况,尊重保护知识产权我们共同努力。
2、文档的总页数、文档格式和文档大小以系统显示为准(内容中显示的页数不一定正确),网站客服只以系统显示的页数、文件格式、文档大小作为仲裁依据,个别因单元格分列造成显示页码不一将协商解决,平台无法对文档的真实性、完整性、权威性、准确性、专业性及其观点立场做任何保证或承诺,下载前须认真查看,确认无误后再购买,务必慎重购买;若有违法违纪将进行移交司法处理,若涉侵权平台将进行基本处罚并下架。
3、本站所有内容均由用户上传,付费前请自行鉴别,如您付费,意味着您已接受本站规则且自行承担风险,本站不进行额外附加服务,虚拟产品一经售出概不退款(未进行购买下载可退充值款),文档一经付费(服务费)、不意味着购买了该文档的版权,仅供个人/单位学习、研究之用,不得用于商业用途,未经授权,严禁复制、发行、汇编、翻译或者网络传播等,侵权必究。
4、如你看到网页展示的文档有www.zixin.com.cn水印,是因预览和防盗链等技术需要对页面进行转换压缩成图而已,我们并不对上传的文档进行任何编辑或修改,文档下载后都不会有水印标识(原文档上传前个别存留的除外),下载后原文更清晰;试题试卷类文档,如果标题没有明确说明有答案则都视为没有答案,请知晓;PPT和DOC文档可被视为“模板”,允许上传人保留章节、目录结构的情况下删减部份的内容;PDF文档不管是原文档转换或图片扫描而得,本站不作要求视为允许,下载前自行私信或留言给上传者【自信****多点】。
5、本文档所展示的图片、画像、字体、音乐的版权可能需版权方额外授权,请谨慎使用;网站提供的党政主题相关内容(国旗、国徽、党徽--等)目的在于配合国家政策宣传,仅限个人学习分享使用,禁止用于任何广告和商用目的。
6、文档遇到问题,请及时私信或留言给本站上传会员【自信****多点】,需本站解决可联系【 微信客服】、【 QQ客服】,若有其他问题请点击或扫码反馈【 服务填表】;文档侵犯商业秘密、侵犯著作权、侵犯人身权等,请点击“【 版权申诉】”(推荐),意见反馈和侵权处理邮箱:1219186828@qq.com;也可以拔打客服电话:4008-655-100;投诉/维权电话:4009-655-100。