QProblemB
Box-constrained QP solver (bounds only, no linear constraints)
Box-constrained QP solver (bounds only, no linear constraints)
Solves QPs with only variable bounds (no constraint matrix A):
Algorithm
Solves KKT conditions directly when active set is known. Uses Cholesky factorization of reduced Hessian.
$$\text{ s.t. }lb <= x <= ub$$
This is more efficient than QProblem when there are no general linear constraints. The active set consists only of bounds.
Key methods:
- init(): First QP solve (cold start)
- hotstart(): Subsequent solves with modified g, lb, ub
See Also
- QProblem for QPs with general linear constraints
Source
Header file: `include/qpOASES/QProblemB.hpp`