Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Scribd Logo
Upload

Welcome to Scribd!

  • 3 views

    Aduh 4

    Uploaded by

    Mas Fian

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd

    Aduh 4

    Uploaded by

    Mas Fian
    3 views9 pages

    Original Title

    aduh4

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    3 views9 pages

    Aduh 4

    Uploaded by

    Mas Fian

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    of 9

    The Reliability or Survivor function

    Let T be a continuous random variable


    representing a lifetime characteristic, say time to
    failure with PDF f(t) and CDF F(t) then the
    Reliability or Survivor function is given by
    R(t) = S(t) = Pr (T > t) = 1 – F(t)

    If t0 is given,
    S(t0) = Pr (T > t0) =
    is the reliability or survival at t0.
    S(t) = 1 – F(t)

    1,0
    F(t)

    0,5

    S(t)
    0,0

    Fig 1. CDF and Survivor function plots


    The p-th quantile of the distribution of T is the value tp such
    that
    Pr ( T≤ tp ) = F(tp) = p

    f(t)

    p 1-p

    tp t

    Fig 2. The p-th quantile

    Knowledge of tp in a lifetime distribution are


    relevant to, for example, the guaranteed lifetime of
    a consumer product.
    Example :

    A product which has reliability function


    2
    R(t) = S(t) = 2 + t 3

    where t is measured in years.


    a. when the company carries a six-month guarantee, find
    the probability that the product fails in the guarantee
    period.
    b. find the length of guarantee period necessary to give a
    failure probability 0,01 in the period.

    Answer :
    2
    a. 1 − R ( 0 ,5 ) = 1 − 3
    = 0,0588
    2 + ( 0 ,5 )

    2
    b. 0,99 = 3
    2+t
    1
    ⎛ 2 ⎞
    → t =⎜ −2 ⎟ 3 = 0,272 years
    ⎝ 0,99 ⎠
    In survival analysis the Mean Time To Failure (MTTF) is
    often of interest. This is given by :
    ∞
    μ = ∫t f ( t ) dt
    0
    ∞
    = [t F (t ) ]∞0 − ∫ F ( t ) dt
    0
    ∞
    = [ t { 1 − S (t ) } ] ∞
    0 − ∫ { 1 − S (t ) } dt
    0
    ∞
    = [ t { 1 − S (t ) } − t ] ∞
    0 + ∫ S ( t ) dt
    0
    ∞
    = [−t S (t ) ]
    ∞
    0 + ∫ S (t ) dt
    0
    ∞
    = ∫ S (t ) dt
    0
    Choosing the better Reliability or Survivor
    (1)

    0.8
    Choose S2
    0.6

    0.4
    S2(t)
    0.2 S1(t)

    0.5 1 1.5 2 2.5 3

    Fig 3. S2 has a larger MTTF than S1


    Choosing the better Reliability or Survivor
    (2)
    1

    0.8 S2(t) The choice as to wich


    was the better model
    0.6
    S1(t) is depend on the
    0.4
    variance. The smallest
    variance is the best.
    0.2

    1 2 3 4

    Fig 4. Two Reliability function with


    the same MTTF
    hazard function : h(t)
    h(t) is instantaneous failure rate
    h(t) = f(t) / S(t)
    The relationship between h(t), f(t), S(t) are :

    f(t)

    f (t )
    ∞

    - S’(t) ∞
    ∫ f ( x ) dx
    t
    ∫ f ( x ) dx
    t ⎧⎪ t ⎫⎪
    h (t ) Exp ⎨− ∫ h ( x ) dx ⎬
    ⎪⎩ 0 ⎪⎭
    − S '(t )
    S (t )
    S(t) h(t)
    ⎧⎪ t ⎫⎪
    Exp ⎨ − ∫ h ( x ) dx ⎬
    ⎪⎩ 0 ⎪⎭
    The hazard function may take a variety of
    forms :
    ¾ Constant : h(t) = λ
    Æ Exponential distribution
    ¾ h(t) is increasing function of t
    Æ Most common case
    ¾ h(t) is decreasing function of t
    Æ less common but may be true
    ¾ “bathtub” hazard :
    initial decreasing, followed by constant
    and finally increasing.

    You might also like