Be aware also the impact of the DB2_WORKLOAD registry variable on these settings.
Study carefully again the doc pages for dft_degree, intra_parallel, max_querydegree.
They seem clear to me. Perhaps translations into your native language are available from IBM.
I suggest you use: dft_degree and max_querydegree both as -1 (notice the minus) and intra_parallel=1 and also that you prove the effect of these via benchmarking your workload.
Normally it is best to let the optimizer work out what is the best parallelism for each query. You can limit the max degree of parallelism via max_querydegree if needed. Additionally if you prove that a specifc workload/statement has a measurable benefit from a specific degree of parallelism then your programmers can use the special register CURRENT DEGREE for that workload/statement which can be different from dft_querydegree.