假设{εi;-∞〈i∞}是一列独立同分布(i.i.d.)随机变量,满足Eε1=0,Eε1^2〈∞〈i〈∞}是一列绝对可和的实数列,关于滑动平均过程Xk=+∞ ∑ i=-∞ ai+kεi,k≥1,已经得到矩形式完全收敛的精确渐近结果:假设E|ε1|^3〈∞,则对1〈p〈2,r〉1+p/2,若E|ε1|^r〈∞,那么lim ε→0 ε^2(r-p)/(2-p)-1 ^∞ ∑n=1 n^r/-p-2-1/ p E{|Sn|-εn^1/p}+=p(2-p)/(r-p)(2r-p-2) E|Z|^2(r-p)/(2-p),本文将以上定理中E|ε1|^3〈∞的条件去掉,得到相同结论,并且在Eε1^2〈∞的条件下得到:假设0≤δ1,α为正实数,并且满足1/2-1/α〈δ〈1-1/α,则lim ε→0 ε^2δ+2/α-1 ^∞∑n=2 ((log n)^(δ-1/2)α/n^3/2) E{|Sn|-ε√n(log n)^α}+ =α/(δα+)(2δα+2-α) E|Z|^2δ+2/α,其中Z服从均值为0,方差为0,方差为τ^2=σ^2(^+∞ ∑ i=-∞ ai)^2 的正态分布.
The precise asyrnptotics in the complete convergence of moving-average processes Xk=+∞ ∑ i=-∞ ai+kεi,is discussed, where {εi;-∞〈i∞}is a doubly infinite sequence of i. i, d. random variables with mean zeros and finite variances,{α i;-∞〈i∞} is an absolutely summable sequence of real numbers. Set Sn=^n ∑k=1 Xk,n≥1,the following precise asymptotics of moving-average processes is proved without the existence of E|ε1|^3,For 1〈p〈2 ,r〉l+p12, it:holds lim ε→0 ε^2(r-p)/(2-p)-1 ^∞ ∑n=1 n^r/-p-2-1/ p E{|Sn|-εn^1/p}+=p(2-p)/(r-p)(2r-p-2) E|Z|^2(r-p)/(2-p), In addition, 0≤δ1,α be a positive number and 1/2-1/α〈δ〈1-1/α, then lim ε→0 ε^2δ+2/α-1 ^∞∑n=2 ((log n)^(δ-1/2)α/n^3/2) E{|Sn|-ε√n(log n)^α}+ =α/(δα+)(2δα+2-α) E|Z|^2δ+2/α where Z has a normal distribution with mean 0 and variance τ^2=σ^2(^+∞ ∑ i=-∞ ai)^2 .