RBDO optimization method

UQ with UQLab Community Q&A and How To
,
1

Hello, everyone! First, I would like to express my sincere thanks to the UqLab team for such an amazing software. Recently, some questions have bothered me when I run with the UqLab and I really hope @moustapha can spend some time to help me.

  1. My program is a two-level RBDO problem. I have used the “CCMEAS” optimizer to explore the design space. But when I add the following commands

RBDOpts.Optim.CCMAES.TolFun = 1e-5;
RBDOpts.Optim.CCMAES.TolX = 1e-3;
RBDOpts.Optim.CCMAES.TolSigma= 1e-3;"

Some warnings came out:

Warning: The option “TolFun” is not a valid option for RBDO Optim.CCMAES options., it will not be used.
Warning: The option “TolX” is not a valid option for RBDO Optim.CCMAES options., it will not be used.
Warning: There were 2 invalid options found for RBDO Optim.CCMAES options.
In uq_options_remainder (line 52)
In uq_initialize_uq_rbdo (line 1172)
In uq_core_module/run_initialization_script (line 208)
In uq_core_analysis/add_module (line 96)
In uq_createAnalysis (line 83)
In A1_RBDO (line 84)

I don’t know why these warnings came out because I have strictly followed the RBDO manual. When I
changed the optimizer “CCMAES” to “HCCMAES”, the same warning problems came out.

  1. The surrogate-assisted RBDO problem can be solved in both two-level approach framework and single-loop approach framework? Which framework is better for solving a surrogate-assisted RBDO problem?

  2. If the cost function of a RBDO problem is not only dependent on design variables but also correlated with environmental variables, how to take the environmental variables into consideration in establishing the cost function model?