Error Status 3 after multiple sqp iterations

User Questions
1

Hi guys,

I have quite an interesting issue.

At the moment I am trying to transfer my time-dependent FC-Truck model to a spatial model. Therefore I reduced it to its basics: FC, Battery and longitudinal vehicle dynamics including slopes.

When optimizing the fuel consumption of this model I get the following solver output:

iter	res_stat	res_eq		res_ineq	res_comp	qp_stat	qp_iter
0	5.132551e+06	1.132428e+02	2.742500e+05	0.000000e+00	0	0
1	3.903288e+06	5.748785e+02	2.010342e+03	5.595437e-04	0	31
2	0.000000e+00	0.000000e+00	0.000000e+00	0.000000e+00	3	1

How can I interpret these outputs? The infeasibility after one successful sqp iteration irritates me.

I already tried reducing the number of constraints and improving the initialization. Both didn’t change the ‘stat’-output.

Thanks in advance,
Maximilian

2

Hi,

One more thin you can try is to use the option nlp_solver_ext_qp_res (set it to 1) and print the statistics again.
This will show you the residuals of the QP.

It can happen that the QP after some SQP iterations become infeasible.

It seems weird to me that changing the initialization does not change the solver statistics.

Best,
Jonathan