Rosen-Suzuki Problem

The Rosen-Suzuki problem is a function of four variables with three nonlinear constraints on the variables. It is taken from problem 43 of Hock and Schittkowski (1981). The objective function is

$\displaystyle  f(x) = x_1^2 + x_2^2 + 2x_3^2 + x_4^2 - 5x_1 - 5x_2 - 21x_3 + 7x_4  $

The nonlinear constraints are

$\displaystyle  0  $
$\displaystyle  \leq  $
$\displaystyle  8 - x_1^2 - x_2^2 - x_3^2 - x_4^2 - x_1 + x_2 - x_3 + x_4  $
$\displaystyle 0  $
$\displaystyle  \leq  $
$\displaystyle  10 - x_1^2 - 2x_2^2 - x_3^2 - 2x_4^2 + x_1 + x_4  $
$\displaystyle 0  $
$\displaystyle  \leq  $
$\displaystyle  5 - 2x_1^2 - x_2^2 - x_3^2 - 2x_1 + x_2 + x_4  $

Since this problem has nonlinear constraints, only the NLPQN and NLPNMS subroutines are available to perform the optimization. The following code solves the problem with the NLPQN subroutine:

The F_HS43 module specifies the objective function, and the C_HS43 module specifies the nonlinear constraints. The OPTN vector is passed to the subroutine as the OPT input argument. See the section Options Vector for more information. The value of OPTN[10] represents the total number of nonlinear constraints, and the value of OPTN[11] represents the number of equality constraints. In the preceding code, OPTN[10]=3 and OPTN[11]=0, which indicate that there are three constraints, all of which are inequality constraints. In the subroutine calls, instead of separating missing input arguments with commas, you can specify optional arguments with keywords, as in the CALL NLPQN statement in the preceding code. For details about the CALL NLPQN statement, see the section NLPQN Call.

The initial point for the optimization procedure is $x=(1,1,1,1)$, and the optimal point is $x^*=(0,1,2,-1)$, with an optimal function value of $f(x^*) = -44$. Part of the output produced is shown in Figure 14.4.

Figure 14.4: Solution to the Rosen-Suzuki Problem by the NLPQN Subroutine

Optimization Start
Parameter Estimates
N Parameter Estimate Gradient
Objective
Function
Gradient
Lagrange
Function
1 X1 1.000000 -3.000000 -3.000000
2 X2 1.000000 -3.000000 -3.000000
3 X3 1.000000 -17.000000 -17.000000
4 X4 1.000000 9.000000 9.000000

Parameter Estimates 4
Nonlinear Constraints 3

Optimization Start
Objective Function -19 Maximum Constraint Violation 0
Maximum Gradient of the Lagran Func 17    

Iteration   Restarts Function
Calls
Objective
Function
Maximum
Constraint
Violation
Predicted
Function
Reduction
Step
Size
Maximum
Gradient
Element
of the
Lagrange
Function
1   0 2 -41.88007 1.8988 13.6803 1.000 5.647
2   0 3 -48.83264 3.0280 9.5464 1.000 5.041
3   0 4 -45.33515 0.5452 2.6179 1.000 1.061
4   0 5 -44.08667 0.0427 0.1732 1.000 0.0297
5   0 6 -44.00011 0.000099 0.000218 1.000 0.00906
6   0 7 -44.00001 2.573E-6 0.000014 1.000 0.00219
7   0 8 -44.00000 9.118E-8 5.098E-7 1.000 0.00022

Optimization Results
Iterations 7 Function Calls 9
Gradient Calls 9 Active Constraints 2
Objective Function -44.00000026 Maximum Constraint Violation 9.1181918E-8
Maximum Projected Gradient 0.0002260958 Value Lagrange Function -44
Maximum Gradient of the Lagran Func 0.0002211456 Slope of Search Direction -5.098303E-7

Optimization Results
Parameter Estimates
N Parameter Estimate Gradient
Objective
Function
Gradient
Lagrange
Function
1 X1 -0.000001247 -5.000003 -0.000012904
2 X2 1.000027 -2.999945 0.000221
3 X3 1.999993 -13.000027 -0.000053971
4 X4 -1.000003 4.999994 -0.000020534


In addition to the standard iteration history, the NLPQN subroutine includes the following information for problems with nonlinear constraints:

  • CONMAX is the maximum value of all constraint violations.

  • PRED is the value of the predicted function reduction used with the GTOL and FTOL2 termination criteria.

  • ALFA is the step size $\alpha $ of the quasi-Newton step.

  • LFGMAX is the maximum element of the gradient of the Lagrange function.