1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
|
SUBROUTINE TB04BW( ORDER, P, M, MD, IGN, LDIGN, IGD, LDIGD, GN,
$ GD, D, LDD, INFO )
C
C SLICOT RELEASE 5.0.
C
C Copyright (c) 2002-2009 NICONET e.V.
C
C This program is free software: you can redistribute it and/or
C modify it under the terms of the GNU General Public License as
C published by the Free Software Foundation, either version 2 of
C the License, or (at your option) any later version.
C
C This program is distributed in the hope that it will be useful,
C but WITHOUT ANY WARRANTY; without even the implied warranty of
C MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
C GNU General Public License for more details.
C
C You should have received a copy of the GNU General Public License
C along with this program. If not, see
C <http://www.gnu.org/licenses/>.
C
C PURPOSE
C
C To compute the sum of an P-by-M rational matrix G and a real
C P-by-M matrix D.
C
C ARGUMENTS
C
C Mode Parameters
C
C ORDER CHARACTER*1
C Specifies the order in which the polynomial coefficients
C of the rational matrix are stored, as follows:
C = 'I': Increasing order of powers of the indeterminate;
C = 'D': Decreasing order of powers of the indeterminate.
C
C Input/Output Parameters
C
C P (input) INTEGER
C The number of the system outputs. P >= 0.
C
C M (input) INTEGER
C The number of the system inputs. M >= 0.
C
C MD (input) INTEGER
C The maximum degree of the polynomials in G, plus 1, i.e.,
C MD = MAX(IGN(I,J),IGD(I,J)) + 1.
C I,J
C
C IGN (input/output) INTEGER array, dimension (LDIGN,M)
C On entry, the leading P-by-M part of this array must
C contain the degrees of the numerator polynomials in G:
C the (i,j) element of IGN must contain the degree of the
C numerator polynomial of the polynomial ratio G(i,j).
C On exit, the leading P-by-M part of this array contains
C the degrees of the numerator polynomials in G + D.
C
C LDIGN INTEGER
C The leading dimension of array IGN. LDIGN >= max(1,P).
C
C IGD (input) INTEGER array, dimension (LDIGD,M)
C The leading P-by-M part of this array must contain the
C degrees of the denominator polynomials in G (and G + D):
C the (i,j) element of IGD contains the degree of the
C denominator polynomial of the polynomial ratio G(i,j).
C
C LDIGD INTEGER
C The leading dimension of array IGD. LDIGD >= max(1,P).
C
C GN (input/output) DOUBLE PRECISION array, dimension (P*M*MD)
C On entry, this array must contain the coefficients of the
C numerator polynomials, Num(i,j), of the rational matrix G.
C The polynomials are stored in a column-wise order, i.e.,
C Num(1,1), Num(2,1), ..., Num(P,1), Num(1,2), Num(2,2),
C ..., Num(P,2), ..., Num(1,M), Num(2,M), ..., Num(P,M);
C MD memory locations are reserved for each polynomial,
C hence, the (i,j) polynomial is stored starting from the
C location ((j-1)*P+i-1)*MD+1. The coefficients appear in
C increasing or decreasing order of the powers of the
C indeterminate, according to ORDER.
C On exit, this array contains the coefficients of the
C numerator polynomials of the rational matrix G + D,
C stored similarly.
C
C GD (input) DOUBLE PRECISION array, dimension (P*M*MD)
C This array must contain the coefficients of the
C denominator polynomials, Den(i,j), of the rational
C matrix G. The polynomials are stored as for the
C numerator polynomials.
C
C D (input) DOUBLE PRECISION array, dimension (LDD,M)
C The leading P-by-M part of this array must contain the
C matrix D.
C
C LDD INTEGER
C The leading dimension of array D. LDD >= max(1,P).
C
C Error Indicator
C
C INFO INTEGER
C = 0: successful exit;
C < 0: if INFO = -i, the i-th argument had an illegal
C value.
C
C METHOD
C
C The (i,j) entry of the real matrix D is added to the (i,j) entry
C of the matrix G, g(i,j), which is a ratio of two polynomials,
C for i = 1 : P, and for j = 1 : M. If g(i,j) = 0, it is assumed
C that its denominator is 1.
C
C NUMERICAL ASPECTS
C
C The algorithm is numerically stable.
C
C FURTHER COMMENTS
C
C Often, the rational matrix G is found from a state-space
C representation (A,B,C), and D corresponds to the direct
C feedthrough matrix of the system. The sum G + D gives the
C transfer function matrix of the system (A,B,C,D).
C For maximum efficiency of index calculations, GN and GD are
C implemented as one-dimensional arrays.
C
C CONTRIBUTORS
C
C V. Sima, Research Institute for Informatics, Bucharest, May 2002.
C Based on the BIMASC Library routine TMCADD by A. Varga.
C
C REVISIONS
C
C V. Sima, Research Institute for Informatics, Bucharest, Feb. 2004.
C
C KEYWORDS
C
C State-space representation, transfer function.
C
C ******************************************************************
C
C .. Parameters ..
DOUBLE PRECISION ZERO
PARAMETER ( ZERO = 0.0D0 )
C .. Scalar Arguments ..
CHARACTER ORDER
INTEGER INFO, LDD, LDIGD, LDIGN, M, MD, P
C .. Array Arguments ..
DOUBLE PRECISION D(LDD,*), GD(*), GN(*)
INTEGER IGD(LDIGD,*), IGN(LDIGN,*)
C .. Local Scalars ..
LOGICAL ASCEND
INTEGER I, II, J, K, KK, KM, ND, NN
DOUBLE PRECISION DIJ
C .. External Functions ..
LOGICAL LSAME
EXTERNAL LSAME
C .. External Subroutines ..
EXTERNAL DAXPY, XERBLA
C .. Intrinsic Functions ..
INTRINSIC MAX, MIN
C ..
C .. Executable Statements ..
C
C Test the input scalar parameters.
C
INFO = 0
ASCEND = LSAME( ORDER, 'I' )
IF( .NOT.ASCEND .AND. .NOT.LSAME( ORDER, 'D' ) ) THEN
INFO = -1
ELSE IF( P.LT.0 ) THEN
INFO = -2
ELSE IF( M.LT.0 ) THEN
INFO = -3
ELSE IF( MD.LT.1 ) THEN
INFO = -4
ELSE IF( LDIGN.LT.MAX( 1, P ) ) THEN
INFO = -6
ELSE IF( LDIGD.LT.MAX( 1, P ) ) THEN
INFO = -8
ELSE IF( LDD.LT.MAX( 1, P ) ) THEN
INFO = -12
END IF
C
IF ( INFO.NE.0 ) THEN
C
C Error return.
C
CALL XERBLA( 'TB04BW', -INFO )
RETURN
END IF
C
C Quick return if possible.
C
IF( MIN( P, M ).EQ.0 )
$ RETURN
C
K = 1
C
IF ( ASCEND ) THEN
C
C Polynomial coefficients are stored in increasing order.
C
DO 30 J = 1, M
C
DO 20 I = 1, P
DIJ = D(I,J)
IF ( DIJ.NE.ZERO ) THEN
NN = IGN(I,J)
ND = IGD(I,J)
IF ( NN.EQ.0 .AND. ND.EQ.0 ) THEN
IF ( GN(K).EQ.ZERO ) THEN
GN(K) = DIJ
ELSE
GN(K) = GN(K) + DIJ*GD(K)
ENDIF
ELSE
KM = MIN( NN, ND ) + 1
CALL DAXPY( KM, DIJ, GD(K), 1, GN(K), 1 )
IF ( NN.LT.ND ) THEN
C
DO 10 II = K + KM, K + ND
GN(II) = DIJ*GD(II)
10 CONTINUE
C
IGN(I,J) = ND
ENDIF
ENDIF
ENDIF
K = K + MD
20 CONTINUE
C
30 CONTINUE
C
ELSE
C
C Polynomial coefficients are stored in decreasing order.
C
DO 60 J = 1, M
C
DO 50 I = 1, P
DIJ = D(I,J)
IF ( DIJ.NE.ZERO ) THEN
NN = IGN(I,J)
ND = IGD(I,J)
IF ( NN.EQ.0 .AND. ND.EQ.0 ) THEN
IF ( GN(K).EQ.ZERO ) THEN
GN(K) = DIJ
ELSE
GN(K) = GN(K) + DIJ*GD(K)
ENDIF
ELSE
KM = MIN( NN, ND ) + 1
IF ( NN.LT.ND ) THEN
KK = K + ND - NN
C
DO 35 II = K + NN, K, -1
GN(II+ND-NN) = GN(II)
35 CONTINUE
C
DO 40 II = K, KK - 1
GN(II) = DIJ*GD(II)
40 CONTINUE
C
IGN(I,J) = ND
CALL DAXPY( KM, DIJ, GD(KK), 1, GN(KK), 1 )
ELSE
KK = K + NN - ND
CALL DAXPY( KM, DIJ, GD(K), 1, GN(KK), 1 )
ENDIF
ENDIF
ENDIF
K = K + MD
50 CONTINUE
C
60 CONTINUE
C
ENDIF
C
RETURN
C *** Last line of TB04BW ***
END
|
