FOLIA OECONOMICA 285, 2013
[141]
Bronisław Ceranka
*, Małgorzata Graczyk
**CONSTRUCTION OF E–OPTIMAL SPRING BALANCE
WEIGHING DESIGNS FOR EVEN NUMBER OF OBJECTS
Abstract. The problem of the construction of spring balance weighing designs satisfying the
criterion of E-optimality is discussed. The incidence matrices of partially incomplete block designs are used to construction of the regular E-optimal spring balance weighing design.
Key words: E–optimal design, partially balanced incomplete block design, spring balance
weighing design.
I. INTRODUCTION
The study of weighing designs is supposed to be helpful in routine of weighing operations to determine unknown measurements of p objects using n measurement operations. These designs are applicable in a great variety of problems of measurements, for instance in metrology, dynamical system theory, computational mechanics and statistics. Results of experiment can be written as
,
e Xw
y where y is an n1 vector of observations, XΦnp
0,1 ,
0,1p n
Φ denotes the class of matrices X
xij , i1,2,...,n, j1,2,...,p, having entries xij 0 or 1 depending upon whether thej
th object is excluded or included in the ith measurement operation, w
w1,w2,...,wp
'is a vector representing unknown measurements of objects and e is an n1 vector of random errors. Our basic assumption is the following. There are not systematic errors and the errors are uncorrelated and have different variances, i.e. E
e 0nand Var
e 2G, where 0n is vector of zeros, G is then
n
positive definite diagonal matrix of known elements. If the design X is of full column rank, then all wj are estimable and the variance matrix of their best linear unbiased estimator is 2
X'G1X
1.* Professor, Department of Mathematical and Statistical Methods, Poznań University of Life
Sciences.
** Ph.D., Department of Mathematical and Statistical Methods, Poznań University of Life
In many problems concerning weighing experiments the E–optimal design is considered. For given variance matrix 2G, the regular E-optimal design there
is design for that the maximal eigenvalue of
X'G1X
1 attains the lower bound.Moreover, in any set of design matrices Φnp
0,1 , a regular E-optimal design may not exist, whereas E-optimal design exists always. The concept of E– optimality was considered in Raghavarao (1971), Banerjee (1975), Jacroux and Notz (1983), Pukelsheim (1993).The main purpose of this paper is to obtain a new construction method of regular E–optimal spring balance weighing designs.
II. E–OPTIMAL DESIGN
We consider the experiment in that using
h s s n n 1 measurement operations we determine unknown measurements of p objects. Without loss of generality we can assume that we have at our disposal h different installations with precision factors g1,g2,...,gh and ns measurements are taken with the precision
s
g
, s1,2,...,h, or thesen
s measurements are taken in different h conditions at the same installation. Thus h h h h h n h n n n n n n n n n n n n n n g g g I 0 0 0 0 0 0 I 0 0 0 0 0 0 I G 1 ' ' ' 1 2 ' ' ' 1 1 2 1 2 2 1 2 1 2 1 1 , gs0,s1,2,...,h. (1)
Now suppose that the design XΦnp
0,1 is divided into h matrices according to (1). Thus h X X X X 2 1 . (2)The theorem given in Katulska and Rychlińska (2010) presents the necessary conditions determining regular E-optimal design.
Theorem 1. Let
p
be an even number. Any nonsingular spring balance weighing design XΦnp
0,1 in the form (2) with the variance matrix of errors 2G, for G of (1), is the regular E– optimal if each row ofs X contains exactly 2 p ones and
' 1 1 1 ' 1 4 tr 2 1 4 tr p p p p p p p 1 1 G I G X G X . (3)In the special case G , Theorem 1 was presented in Jacroux and Notz In
(1983). Method of construction of the regular E–optimal spring balance weighing design for XΦnp
0,1 in the form (2) and G in (1), is given in Katulska and Rychlińska (2010). It is based on the incidence matrices of the balanced incomplete block designs.For convenience, from now on G is the same as defined in (1).
Here, we consider the case p is an even number. Under the assumption that the errors have different variances, we give a new construction method of regular E–optimal spring balance weighing design XΦnp
0,1 in the form (2). Therefore, we wide the class of possible optimal designs given in Jacroux and Notz (1983) and Katulska and Rychlińska (2010). Presented method is based on incidence matrices of two group divisible designs with the same association scheme.III. CONSTRUCTION OF THE DESIGN MATRIX
Now, we recall the definition of partially balanced incomplete block design with two associate classes given, for instance, in Clatworthy (1973).
An incomplete block design is said to be partially balanced with two associate classes if it satisfies the following requirements
(i) The experimental material is divided into b blocks of k units each, different treatments being applied to the units in the same block.
(ii) There are v
treatments each of which occurs in k r blocks.(iii) There can be established a relation of association between any two treatments satisfying the following requirements:
(a) Two treatments are either first associates or second associates. (b) Each treatment has exactly th associates, 1,2.
(c) Given any two treatments which are th associates, the number of treatments common to the th associate of the first and the
th associate of the second isp
and is independent of the pair of treatments we start with. Also p
p , ,, 1,2.
(d) Two treatments which are th associates occur together in exactly
blocks, 1,2.
For a proper partially balanced incomplete block design 12. The numbers v, b, r, k, 1, 2 are called parameters of the first kind, whereas the numbers q, p, ,, 1,2 are called parameters of the second kind.
A group divisible design is a partially balanced incomplete block design with two associate classes for which the treatments may be divided into
m
groups of
t
distinct treatments each, such that treatments that belong to the same group are first associates and two treatments belonging to different groups are second associates. For group divisible design it is clear that vmt, q1 t1,
1
2 mt
q ,
t11t
m1
2r
k1
.Based on incidence matrices of two group divisible designs with the same association scheme, we construct regular E–optimal spring balance weighing design. For this purpose, we consider the design XΦnp
0,1 in (2) for
'2
1s s
s N N
X , (4)
where Nls is incidence matrix of the group divisible design with the same association scheme with parameters v, bls, rls, kls, 1ls, 2ls, s1,2,...,h,
. 2 , 1 l
Furthermore, let the condition
s s s s s 11 12 21 22 (5) be satisfied for each s. For XΦnp
0,1 in the form (2) and (4), we havev p and . 1 2 1
h s l ls b nTheorem 2 Suppose that
v
is an even number and that’s more N1s and s2
association scheme with parameters v, bls, rls, kls, 1ls, 2ls, l1,2, , ,..., 2 , 1 h
s for that Condition (5) is satisfied. If the conditions (i) b1sb2s2
r1sr2s
(ii) 4s
v1
v2
b1sb2s
are fulfilled simultaneously for each s, then XΦnp
0,1 in the form (2) and (4) with 2G, is the regular E–optimal spring balance weighing design.Proof. For given 2G, we choose
h
matrices in (4) and we form
0,1 p nΦ
X . According to above notation, we consider group divisible designs with the same association scheme for that (5) is satisfied. Hence we have
h s v v s s v s s s h s ls ls s l v v b b v v b b g g 1 ' 2 1 2 1 1 1 ' 1 2 1 1 ' 1 4 2 1 4 I 11 N N X G X . (6) Furthermore,
h s s s s v s v v h s l ls ls r r 1 ' 2 1 1 2 1 ' I 11 N N . Since Nls areincidence matrices of group divisible design satisfying Conditions (i) and (ii) then it is clear that
2 p
ks for each s. An easy computation shows that (6) is equivalent to (3) if and only if
1
4 2 2 1 v v b bs s s for each s. Thus we have
(ii). Taking into consideration Theorem 1 and the equality
s
s s s s r r v v b b 2 1 2 1 14 we obtain the Condition (i). Hence the result.
In given class XΦnp
0,1 we choose h matrices X in such a way that s
h s l ls b n 1 2 1. Let us mention, we don’t have to choose different h matrices X . s
The matrices X we form using the incidence matrices of group divisible s
designs with the same association scheme that parameters are given below. We present series of parameters of group divisible designs based on the book of Clatworthy (1973). The restrictions related to parameters r,k10 and u follow from assumption given in Clatworthy (1973).
Theorem 3. Let v4 and . 1 2 1
h s l ls b n Any nonsingular XΦn4
0,1 in (2) and (4), where N1s and N2s are incidence matrices of group divisible designs with the same association scheme with parameters(i) b1s 2
3t1
, r1s t3 1, k1s 2, 11s t1, 21st and
3 2
2 2 q b s , r2s q3 2, k2s 2, 12s q, 22s q1, t1,2,3, q0,1,2, (ii) b1s2
3t2
, r1s t3 2, k1s 2, 11s t2, 21s t and
3 4
2 2 q b s , r2s q3 4, k2s 2, 12s q, 22s q2, t1,2, q0,1,2, (iii) b1s u2
3
, r1s u3, k1s2, 11s u1, 21s1 andb
2s
4
u
, , 2 2 u r s k2s 2, 12s 0, 22s u, u1,2,3,4,5, (iv) b1s 16, r1s 8, k1s 2, 11s 0, 21s 4 and b2s 2
3u4
, , 4 3 2 u r s k2s2, 12s u4, 22s u, u1,2, (v)b
1s
18
,r
1s
9
,k
1s
2
,
11s
5
,
21s
2
and b2s u6
2
, ), 2 ( 3 2 u r s k2s 2, 12s u, 22s u3, u0,1,with 2G, is the regular E–optimal spring balance weighing design.
Proof. It is easily seen that parameters the group divisible designs satisfy Conditions (i) and (ii) of Theorem 2.
It is convenient to remark that the proofs of Theorems 4-8 are similar to analysis given in the proof of Theorem 3. We leave to the reader to verify relations between the parameters of group divisible designs.
Theorem 4. Let v6 and .
1 2 1
h s l ls b n Any nonsingular XΦn6
0,1 in (2) and (4), where N1s and N2s are incidence matrices of group divisible design with the same association scheme with parameters(i) b1s 4t, r1s 2t, k1s 3, 11s 0, 21s t and b2s 6t, r2s 3t, , 3 2s k 12s 2t, 22s t, t1,2,3, (ii) b1s2
2t5
, r1s t2 5, k1s 3, 11s t1, 21s t2 and , 6 2 t b s r2s 3t, k2s 3, 12s t1, 22s t, t1,2, (iii) b1s 12, r1s 6, k1s 3, 11s 4, 21s 2 and b2s 2
5t4
, , 4 5 2 t r s k2s 3, 12s 2t, 22s t2 1, t0,1, (iv) b1s 16, r1s 8, k1s 3, 11s 4, 21s 3 and b2s 2
5t2
, , 2 5 2 t r s k2s 3, 12s t2, 22s t2 1, t0,1,Theorem 5. Let v8 and h s l bls n 1 2 1 . Any nonsingular
X
Φ
n8
0
,
1
in (2) and (4), where
N
1s andN
2s are incidence matrices of group divisible design with the same association scheme with parameters(i) b1s t4( 1), r1s t2( 1), k1s 4, 11s 0, 21s t1 and
t
b2s 64 , r2s 62
t
, k2s 4, 12s 6, 22s5t, t1,2,3, (ii) b1s 2(3t2), r1s t3 2, k1s 4, 11s t1,
21s t
2
and
t
b2s 46 , r2s 43
t
, k2s 4, 12s 4t, 22s5t, t1,2, with 2G, is the regular E–optimal spring balance weighing design.Theorem 6. Let v10 and .
1 2 1
h s l ls b n Any nonsingular XΦn10
0,1 in (2) and (4), where N1s and N2s are incidence matrices of group divisible design with the same association scheme with parameters b1s 8u, r1s 4u,, 5 1s k 11s0, 21s 2u and b2s10u, r2s 5u, k2s 5, 12s 4u, , 2 22s u
u1,2, with 2G, is the regular E–optimal spring balance weighing design.
Theorem 7. If N1 and N2 are incidence matrices of group divisible designs with the same association scheme with parameters v2(2u1), b14u,
, 2 1 u r k1 u2 1, 110, 21u and vb2 2(2u1), r 2 k2 2u1, , 2 12 u
22 u, u1,2,3,4, then any spring balance weighing design
4 1 22 1
0,1 2 Φ h u u X in the form X1 X* h for
, ' 2 1 * N N X with , 2G is the regular E–optimal.
Theorem 8. If N1 and
N
2 are incidence matrices of group divisible designwith the same association scheme with parameters v u4( 1), b12(2u1), , 1 2 1 u r k1 u2( 1), 11 u2 1, 21u and vb24(u1), ), 1 ( 2 2 2k u
r 120, 22 u1, u1,2,3,4, then any spring balance weighing design XΦ2h4u3 4u1
0,1 in the formX
1
X
*h for
'2 1
* N N
REFERENCES
Banerjee K.S. (1975), Weighing Designs for Chemistry, Medicine, Economics, Operations Research, Statistics. Marcel Dekker Inc., New York.
Clatworthy W.H. (1973), Tables of Two-Associate-Class Partially Balanced Design. NBS Applied Mathematics Series 63.
Jacroux M., Notz W. (1983), On the optimality of spring balance weighing designs. The Annals of Statistics 11, 970–978.
Katulska K., Rychlińska E. (2010), On regular E-optimality of spring balance weighing designs. Colloquium Biometricum 40, 165–176.
Pukelsheim F. (1993), Optimal design of experiment. John Wiley and Sons, New York. Raghavarao D. (1971), Constructions and Combinatorial Problems in Designs of Experiments, John Wiley Inc., New York.
Raghavarao D., Padgett L.V. (2005), Block Designs, Analysis, Combinatorics and Applications. Series of Applied Mathematics 17, Word Scientific Publishing Co. Pte. Ltd.
Bronisław Ceranka, Małgorzata Graczyk
KONSTRUKCJA REGULARNEGO E–OPTYMALNEGO SPRĘŻYNOWEGO UKŁADU WAGOWEGO DLA PARZYSTEJ LICZBY OBIEKTÓW
W pracy przedstawiono zagadnienie konstrukcji sprężynowego układu wagowego spełniającego kryterium E–optymalności. Do konstrukcji macierzy układu wykorzystano macierze incydencji częściowo zrównoważonych układów bloków.