MLDO tool design

Let us assume that we have a function

result runDOT(obj,g,x,lb,ub,c,params)

where:

double obj - the objective function f(x,c)
double[] g - array of constraints functions g[1](x,c), g[2](x,c), ... g[k](x,c)
double[] x - array of initial values of the design variables x[1],...,x[n]
double[] lb - array of lower bounds of x
double[] ub - array of upper bounds of x
double[] c - array of constants used in obj and g
int[] params - array of DOT parameters such as method, upper limit of iteration count, etc.
result - a data structure capturing the results of a DOT run, such as the value of the objective function, values of the design variables at optimum, number or iterations run, etc.

All At Once

result=runDOT(obj,g,x,lb,up,c,params)

design variables : z1, z2, z3, z4, z5, z6, z7 additional params: lambda, W

level 1: z1,    z3, z4, z5
level 2:     z2,        z5, z6, z7

level 1:

optimization: runDOT(z1,z3,z4,z5; lambda,W,r) --> z1*,z3*,z4*,t=z5*

objective:   obj = f(z1,z5; lambda,W,r)
constraints: g(1)= f(z3,z4,z5)
             g(2)= f(z1,z3,z4,z5)
             g(3)= -g(2)

level 2:

optimization: runDOT(z2,z5,z6,z7; lambda,W,t) --> z2*,r=z5* z6*,z7*

objective:   obj = f(z2,z5; lambda,W,t)
constraints: g(1)= f(z5,z6,z7)
             g(2)= f(z2,z5,z6,z7)
             g(3)= -g(2)