- Page 1 and 2: Optimization Toolbox 3User’s Guid
- Page 3: Revision HistoryNovember 1990 First
- Page 7 and 8: Contents1Getting StartedWhat Is Opt
- Page 9 and 10: Other Examples That Use this Techni
- Page 11 and 12: Main Algorithm ....................
- Page 13 and 14: 7Functions — By CategoryMinimizat
- Page 15 and 16: 1Getting StartedWhat Is Optimizatio
- Page 17 and 18: What Is Optimization Toolbox?proble
- Page 19 and 20: Optimization Examplewhereis the nor
- Page 21 and 22: 2TutorialThe Tutorial provides info
- Page 23 and 24: IntroductionIntroductionOptimizatio
- Page 25 and 26: IntroductionEquation SolvingType No
- Page 27 and 28: Introductionresulting point where t
- Page 29 and 30: Examples That Use Standard Algorith
- Page 31 and 32: Examples That Use Standard Algorith
- Page 33 and 34: Examples That Use Standard Algorith
- Page 35 and 36: Examples That Use Standard Algorith
- Page 37 and 38: Examples That Use Standard Algorith
- Page 39 and 40: Examples That Use Standard Algorith
- Page 41 and 42: Examples That Use Standard Algorith
- Page 43 and 44: Examples That Use Standard Algorith
- Page 45 and 46: Examples That Use Standard Algorith
- Page 47 and 48: Examples That Use Standard Algorith
- Page 49 and 50: Examples That Use Standard Algorith
- Page 51 and 52: Examples That Use Standard Algorith
- Page 53 and 54: Examples That Use Standard Algorith
- Page 55 and 56:
Examples That Use Standard Algorith
- Page 57 and 58:
Examples That Use Standard Algorith
- Page 59 and 60:
Examples That Use Standard Algorith
- Page 61 and 62:
Large-Scale ExamplesGenerally speak
- Page 63 and 64:
Large-Scale ExamplesLarge-Scale Pro
- Page 65 and 66:
Large-Scale ExamplesNonlinear Equat
- Page 67 and 68:
Large-Scale Examples'LargeScale','o
- Page 69 and 70:
Large-Scale Examples3 16 0.0458181
- Page 71 and 72:
Large-Scale ExamplesF = 2 + 2*k-exp
- Page 74 and 75:
2 Tutoriali=1:(n-1); g = zeros(n,1)
- Page 76 and 77:
2 TutorialThe sparsity pattern of t
- Page 78 and 79:
2 Tutorialexitflag =3fval =270.4790
- Page 80 and 81:
2 Tutorial2.9310e+006Step 2: Call a
- Page 82 and 83:
2 Tutorialto 'on', fmincon knows to
- Page 84 and 85:
2 TutorialNorm of First-orderIterat
- Page 86 and 87:
2 Tutoriallight of this cost, one s
- Page 88 and 89:
2 Tutorial• Contains a nested fun
- Page 90 and 91:
2 Tutorialexitflag =3output =iterat
- Page 92 and 93:
2 Tutorial[fval,exitflag,output] =
- Page 94 and 95:
2 Tutorialload sc50bThis problem in
- Page 96 and 97:
2 TutorialBecause the iterative dis
- Page 98 and 99:
2 TutorialDefault Options Settings
- Page 100 and 101:
2 TutorialSetting More Than One Opt
- Page 102 and 103:
2 TutorialFunction-Specific Output
- Page 104 and 105:
2 Tutorialfzero and fminbndThe foll
- Page 106 and 107:
2 Tutorialfgoalattain,fmincon,fmini
- Page 108 and 109:
2 TutorialCalling an Output Functio
- Page 110 and 111:
2 TutorialThe arguments that the op
- Page 112 and 113:
2 TutorialRunning the ExampleTo run
- Page 114 and 115:
2 TutorialThe example displays a pl
- Page 116 and 117:
2 Tutorialx = fminbnd(fh, 3, 4)You
- Page 118 and 119:
2 TutorialTypicalProblemsandHowtoDe
- Page 120 and 121:
2 TutorialTroubleshooting (Continue
- Page 122 and 123:
2 TutorialSelected Bibliography[1]H
- Page 124 and 125:
3 Standard AlgorithmsLeast-Squares
- Page 126 and 127:
3 Standard AlgorithmsOptimization O
- Page 128 and 129:
3 Standard AlgorithmsUnconstrained
- Page 130 and 131:
3 Standard Algorithms(3-4)The optim
- Page 132 and 133:
3 Standard AlgorithmsThe line searc
- Page 134 and 135:
3 Standard AlgorithmsThe functions
- Page 136 and 137:
3 Standard AlgorithmsCase 3.Case 4.
- Page 138 and 139:
3 Standard Algorithmsprocedure, the
- Page 140 and 141:
3 Standard AlgorithmsLeast-Squares
- Page 142 and 143:
3 Standard AlgorithmsGauss-Newton M
- Page 144 and 145:
3 Standard AlgorithmsNonlinear Leas
- Page 146 and 147:
3 Standard AlgorithmsThe implementa
- Page 148 and 149:
3 Standard Algorithmswhereis the n-
- Page 150 and 151:
3 Standard AlgorithmsThe step is co
- Page 152 and 153:
3 Standard Algorithmsbe canceled, L
- Page 154 and 155:
3 Standard AlgorithmsConsider Rosen
- Page 156 and 157:
3 Standard AlgorithmsThe functions
- Page 158 and 159:
3 Standard Algorithmsis called the
- Page 160 and 161:
3 Standard AlgorithmsLine Search an
- Page 162 and 163:
3 Standard AlgorithmsPhase 1. In ph
- Page 164 and 165:
3 Standard AlgorithmsWhen the probl
- Page 166 and 167:
3 Standard Algorithmsto characteriz
- Page 168 and 169:
3 Standard AlgorithmsWeighted Sum M
- Page 170 and 171:
3 Standard AlgorithmsFigure 3-11:Ge
- Page 172 and 173:
3 Standard AlgorithmsFigure 3-12:Ge
- Page 174 and 175:
3 Standard AlgorithmsTo overcome th
- Page 176 and 177:
3 Standard Algorithms[12] Fleming,
- Page 178 and 179:
3 Standard Algorithms[36] Powell, M
- Page 180 and 181:
4 Large-Scale AlgorithmsLarge-Scale
- Page 182 and 183:
4 Large-Scale AlgorithmsEquation 4-
- Page 184 and 185:
4 Large-Scale AlgorithmsDemos of La
- Page 186 and 187:
4 Large-Scale AlgorithmsIn a minimi
- Page 188 and 189:
4 Large-Scale AlgorithmsBox Constra
- Page 190 and 191:
4 Large-Scale AlgorithmsNonlinear L
- Page 192 and 193:
4 Large-Scale AlgorithmsLinear Leas
- Page 194 and 195:
4 Large-Scale Algorithms(4-16)which
- Page 196 and 197:
4 Large-Scale AlgorithmsThe algorit
- Page 198 and 199:
4 Large-Scale AlgorithmsSelected Bi
- Page 200 and 201:
4 Large-Scale Algorithms4-22
- Page 202 and 203:
5 Optimization ToolGetting Started
- Page 204 and 205:
5 Optimization ToolSteps for Using
- Page 206 and 207:
5 Optimization ToolSolverfminuncfse
- Page 208 and 209:
5 Optimization ToolProblem Setup (C
- Page 210 and 211:
5 Optimization Toolfgoalattain Prob
- Page 212 and 213:
5 Optimization ToolX2 (required) is
- Page 214 and 215:
5 Optimization Toolfminimax Problem
- Page 216 and 217:
5 Optimization Toolfminunc Problem
- Page 218 and 219:
5 Optimization ToolK1, K2, ..., Knt
- Page 220 and 221:
5 Optimization ToolX1 and X2 for wh
- Page 222 and 223:
5 Optimization Toollsqcurvefit Prob
- Page 224 and 225:
5 Optimization ToolConstraintsLinea
- Page 226 and 227:
5 Optimization ToolStart PointStart
- Page 228 and 229:
5 Optimization ToolRunning a Proble
- Page 230 and 231:
5 Optimization ToolSpecifying the O
- Page 232 and 233:
5 Optimization ToolFunction Value C
- Page 234 and 235:
5 Optimization ToolFor the medium-s
- Page 236 and 237:
5 Optimization ToolThe graphic abov
- Page 238 and 239:
5 Optimization Tool• Maximum iter
- Page 240 and 241:
5 Optimization ToolThegraphicaboves
- Page 242 and 243:
5 Optimization ToolGetting Help in
- Page 244 and 245:
5 Optimization ToolThe default valu
- Page 246 and 247:
5 Optimization ToolAlthough you can
- Page 248 and 249:
5 Optimization Tool-9*x(1)^2 - x(2)
- Page 250 and 251:
5 Optimization Tool• The Current
- Page 252 and 253:
5 Optimization Toolsubject to the c
- Page 254 and 255:
5 Optimization Tool• The Current
- Page 256 and 257:
6 Argument and Options ReferenceFun
- Page 258 and 259:
6 Argument and Options ReferenceInp
- Page 260 and 261:
6 Argument and Options ReferenceOut
- Page 262 and 263:
6 Argument and Options ReferenceOpt
- Page 264 and 265:
6 Argument and Options ReferenceOpt
- Page 266 and 267:
6 Argument and Options ReferenceOpt
- Page 268 and 269:
6 Argument and Options ReferenceOpt
- Page 270 and 271:
6 Argument and Options ReferenceOpt
- Page 272 and 273:
6 Argument and Options ReferenceCor
- Page 274 and 275:
6 Argument and Options Referenceopt
- Page 276 and 277:
6 Argument and Options Referenceopt
- Page 278 and 279:
6 Argument and Options ReferenceSta
- Page 280 and 281:
6 Argument and Options Reference6-2
- Page 282 and 283:
7 Functions — By CategoryMinimiza
- Page 284 and 285:
7 Functions — By Category7-4
- Page 286 and 287:
intprogPurposeEquationSolve binary
- Page 288 and 289:
intprogexitflagoutputInteger identi
- Page 290 and 291:
intprogNodeSearchStrategyStrategy t
- Page 292 and 293:
intprog• If the algorithm finds a
- Page 294 and 295:
intprogSee Alsolinprog, optimset, o
- Page 296 and 297:
fgoalattainPurposeEquationSolve mul
- Page 298 and 299:
fgoalattain“Avoiding Global Varia
- Page 300 and 301:
fgoalattainnonlconThe function that
- Page 302 and 303:
fgoalattainweightA weighting vector
- Page 304 and 305:
fgoalattainoutputupper Upper bounds
- Page 306 and 307:
fgoalattainPlotFcnsRelLineSrchBndRe
- Page 308 and 309:
fgoalattainC = [1 0 0; 0 0 1];K0 =
- Page 310 and 311:
fgoalattainmethods assume the funct
- Page 312 and 313:
fminbndPurposeEquationFind minimum
- Page 314 and 315:
fminbndoutput-1 Algorithm was termi
- Page 316 and 317:
fminbndThevalueattheminimumisy = f(
- Page 318 and 319:
fminconPurposeEquationFind minimum
- Page 320 and 321:
fmincon[x,fval,exitflag,output,lamb
- Page 322 and 323:
fminconIftheHessianmatrixcanalsobec
- Page 324 and 325:
fminconOutputArguments“Function A
- Page 326 and 327:
fminconstepsizefirstorderoptFinal s
- Page 328 and 329:
fminconGradObjMaxFunEvalsMaxIterOut
- Page 330 and 331:
fminconY is a matrix that has the s
- Page 332 and 333:
fminconMedium-Scale Algorithm OnlyT
- Page 334 and 335:
fmincon• Specify the feasible reg
- Page 336 and 337:
fminconsubproblem at each iteration
- Page 338 and 339:
fminimaxPurposeEquationSolve minima
- Page 340 and 341:
fminimaxInputArguments“Function A
- Page 342 and 343:
fminimaxnonlconThe gradient consist
- Page 344 and 345:
fminimaxexitflagInteger identifying
- Page 346 and 347:
fminimaxDiffMaxChangeDiffMinChangeD
- Page 348 and 349:
fminimaxRelLineSrchBndDurationTolCo
- Page 350 and 351:
fminimaxset to 'iter'). The depende
- Page 352 and 353:
fminsearchPurposeEquationFind minim
- Page 354 and 355:
fminsearchoutput0 Number of iterati
- Page 356 and 357:
fminsearch8.1777e-010This indicates
- Page 358 and 359:
fminsearchNotesfminsearch is not th
- Page 360 and 361:
fminunc[x,fval,exitflag,output,grad
- Page 362 and 363:
fminuncThe gradient is the partial
- Page 364 and 365:
fminunccgiterationsstepsizeNumber o
- Page 366 and 367:
fminuncPlotFcnsTolFunPlots various
- Page 368 and 369:
fminuncY is a matrix that has the s
- Page 370 and 371:
fminuncHessUpdateMethod for choosin
- Page 372 and 373:
fminuncx =1.0e-015 *0.1110 -0.8882f
- Page 374 and 375:
fminunc[3] Coleman, T.F. and Y. Li,
- Page 376 and 377:
fseminfDescriptionfseminf finds a m
- Page 378 and 379:
fseminffunThefunctiontobeminimized.
- Page 380 and 381:
fseminfThe vectors or matrices K1,
- Page 382 and 383:
fseminfOptionsOptimization options
- Page 384 and 385:
fseminfvalues other than the recomm
- Page 386 and 387:
fseminf0.4023The function value and
- Page 388 and 389:
fseminfs = [2 2];end% Sampling setw
- Page 390 and 391:
fseminfAlgorithmLimitationsSee Also
- Page 392 and 393:
fsolveInputArguments“Function Arg
- Page 394 and 395:
fsolveOutputArguments“Function Ar
- Page 396 and 397:
fsolveMedium-Scale and Large-Scale
- Page 398 and 399:
fsolveJacobMultFunction handle for
- Page 400 and 401:
fsolveNonlEqnAlgorithmSpecify one o
- Page 402 and 403:
fsolvestarting at the point x= [1,1
- Page 404 and 405:
fsolvedogleg method described in [8
- Page 406 and 407:
fsolve[6] Moré, J. J., “The Leve
- Page 408 and 409:
fzeroNote For the purposes of this
- Page 410 and 411:
fzerooutput-4 Complex function valu
- Page 412 and 413:
fzeroAlgorithmLimitationsReferences
- Page 414 and 415:
gangstrPurposeSyntaxDescriptionSee
- Page 416 and 417:
linprogx = linprog(f,A,b,Aeq,beq,lb
- Page 418 and 419:
linprogLargeScaleUse large-scale al
- Page 420 and 421:
linprogNonzero elements of the vect
- Page 422 and 423:
linprogNote The preprocessing steps
- Page 424 and 425:
linprogReferences[1] Dantzig, G.B.,
- Page 426 and 427:
lsqcurvefitx = lsqcurvefit(fun,x0,x
- Page 428 and 429:
lsqcurvefitOutputArguments“Functi
- Page 430 and 431:
lsqcurvefitmedium-scale algorithm.
- Page 432 and 433:
lsqcurvefitJacobMultFunction handle
- Page 434 and 435:
lsqcurvefitLevenbergMarquardt Choos
- Page 436 and 437:
lsqcurvefitcubic polynomial method
- Page 438 and 439:
lsqcurvefitSee Also@ (function_hand
- Page 440 and 441:
lsqlinx = lsqlin(C,d,A,b,Aeq,beq,lb
- Page 442 and 443:
lsqlincgiterationsfirstorderoptNumb
- Page 444 and 445:
lsqlinJacobMultFunction handle for
- Page 446 and 447:
lsqlin0.6721];lb = -0.1*ones(4,1);u
- Page 448 and 449:
lsqlinDiagnosticsLarge-Scale Optimi
- Page 450 and 451:
lsqnonlinPurposeEquationSolve nonli
- Page 452 and 453:
lsqnonlinfunThe function whose sum
- Page 454 and 455:
lsqnonlinoutputupperUpper bounds ub
- Page 456 and 457:
lsqnonlinJacobianMaxFunEvalsMaxIter
- Page 458 and 459:
lsqnonlinMaxPCGIterMaximum number o
- Page 460 and 461:
lsqnonlinGauss-Newton method, which
- Page 462 and 463:
lsqnonlin[5] Marquardt, D., “An A
- Page 464 and 465:
lsqnonneg[x,resnorm,residual,exitfl
- Page 466 and 467:
lsqnonnegAlgorithmNoteslsqnonneg us
- Page 468 and 469:
optimsetPurposeSyntaxDescriptionOpt
- Page 470 and 471:
optimsetTolFunTolXTypicalXPositive
- Page 472 and 473:
optimsetThis statement makes a copy
- Page 474 and 475:
optimtoolSee Alsooptimset8-190
- Page 476 and 477:
quadprog[x,fval] = quadprog(...) re
- Page 478 and 479:
quadprogalgorithmcgiterationsfirsto
- Page 480 and 481:
quadprogHessMultFunction handle for
- Page 482 and 483:
quadprogNext, invoke a quadratic pr
- Page 484 and 485:
quadprogAlgorithmLarge-Scale Optimi
- Page 486 and 487:
quadprog[2] Gill, P. E. and W. Murr
- Page 488 and 489:
Indexlarge-scale 4-9equation solvin
- Page 490 and 491:
Indexmerit function 3-38minimax exa
- Page 492:
Indexdetermination of 4-4subspace,