11.07.2014 Views

Multiattribute acceptance sampling plans - Library(ISI Kolkata ...

Multiattribute acceptance sampling plans - Library(ISI Kolkata ...

Multiattribute acceptance sampling plans - Library(ISI Kolkata ...

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

...(1.2.7)<br />

To compare the relative changes in the OC function for changes in p i , we compute the<br />

P C ′ i obtained by differentiating the OC function as given in equation (1.2.7) with respect to<br />

m i . Thus,<br />

Slope j = −P C ′ j = g(c j , m j )<br />

r∏<br />

i=1:i≠j<br />

G(c i , m i ), j = 1, 2, ...r; c j ≥ 0, j = 1, 2, ...r<br />

...(1.2.8)<br />

Let m = m 1 + m 2 + .... + m r ; 0 ≤ m 1 ≤ m 2 ... ≤ m r ; ρ i = m i /m is fixed for i = 1, 2, ..., r<br />

and only m varies. In this case, 0 ≤ c 1 ≤ c 2 ... ≤ c r .<br />

Theorem 1.2.1 justifies to some extent the numerical results presented in Table 1.2.4.<br />

Theorem 1.2.1<br />

The function, H(m) = Slope i − Slope i+1 undergoes atmost one change of sign from positive<br />

to negative for m ≥ 0.<br />

Proof We need to show that the above statement is true for<br />

F (m) = g(c i , mρ i )G(c i+1 , mρ i+1 ) − G(c i , mρ i )g(c i+1 , mρ i+1 ).<br />

... (1.2.9)<br />

Without loss of generality we take i = 1, so that F (m)/ [e −m m c 1<br />

]=<br />

c 1 ∑<br />

i=0<br />

m [ c 2−i<br />

ρ c 2−i<br />

2 ρ c 1<br />

1 /((c 2 − i)!c 1 !)) − ρ c 1−i<br />

1 ρ c 2<br />

2 /((c 1 − i)!.c 2 !)) ] c 2 −c ∑ 1 −1<br />

+ ((mρ 2 ) x /x!)(ρ c 1<br />

1 /c 1 !)<br />

0 ≤ c 1 ≤ c 2 < ∞<br />

x=0<br />

...(1.2.10)<br />

We first note that all the coefficients of m x for x ≤ c 2 − c 1 − 1 are ≥ 0 in the second<br />

series function of the right hand side in (1.2.10).<br />

48

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!