Problems with simscape solvers
7 views (last 30 days)
I'm working on problem were I feed my simscape (pneumatics) model with long time series, which is quite noisy and unstable. I know that such a signal gives hard time for differential euqation solvers (for implemented in simscape too). But the hardest time for solvers is initial few time steps. It freezes for a few timesteps and then after everything goes quite smoothly. It seems for me quite strange that at the very beginning of simulation it struggles while everthing else goes smoothly and the behaviour of my time series is quite ugly all the time: at the begining and further. Does anybody have some thoughts about it and how those long initial calculations could somehow alleviated?
Steven on 20 Oct 2015
Instead of changing the simulink solver parameters, try clicking on the solver configuration block. I needed more accuracy and I started changing the simulink solver ( I changed the timestep and tried different solvers). I had the same result, my simscape simulation locked up. I set the simulink solver back to auto and variable timestep, and found the Solver Configuration block ( you know, the one that every simscape simulation has to have. ) There are 2 solvers and you can play with the time step. It appears that simulink and simscape have two different ODE solvers that can talk to each other through PS->S and S->PS blocks