Explosive Simulations in Dynare for OLG model

Dear Professor Pfeifer and Dynare Users,

I am currently working on my thesis paper and modeled a two-period discrete OLG model with public sector and endogenous mortality: government finances its debt with issued bonds. In order to calculate my initial guesses for steady state I use the separate .m file. Then I manually transfer results in .mod file.

I also want to go through values of mu and build a graph of lifetime utility (u_lifetime) first theoretical moments. I have a working loop, but the problem is with my model. At some point (when mu is around 0.093) loop terminates, since there can be no steady state found or the shock is explosive. In some cases, when I change initial guesses, I have the problem of B-K conditions being not satisfied (also probably due to the explosiveness of a shock, when per labor debt b is positive.). Really do not know, what is the problem. I am really struggling for any help. Please, any advice would be welcome.

You may also find files attached:

OLG_dip2_amend.mod (4.2 KB) ss_test_omega.m (3.2 KB)

Here is the loop file with parameters.

param.m (1.5 KB)