Now with a symbolic modeling engine with source transformation and automatic generation of first and second order derivatives!
The TOMLAB Base Module includes a large set of optimization solvers. Most of them were originally developed by the Applied Optimization and Modeling group (TOM). Since then they have been improved e.g. to handle Matlab sparse arrays and been further developed. The TOMLAB Base Module also includes the SOL solvers Tlsqr (LSQR), PDCO and PDSCO.
See the complete list of solvers in the TOMLAB Base Module >>
The TOMLAB Base Module can be extended with more powerful solvers. See our product list >>
The TOMLAB Base Module consists of more than 100 000 lines of Matlab m-file code with more than 80 optimization algorithms implemented.
- TOMLAB Base Module efficiently integrates five Fortran solvers: Tlsqr, LSEI, Tnnls, QLD and Tfzero. The other solvers are running part of the code in Fortran/C, thereby providing significant speed-ups.
- Box-bounded global optimization, solver glbSolve and the C-version glbDirect.
- Integer and nonlinearly constrained global optimization, solver glcSolve and the C-version glcDirect as well as glcCluster.
- Nonconvex quadratic programming, solver qpSolve. QLD for convex dense QP.
- Linear programming solvers milpSolve, QLD and DualSolve (dual LP problems).
- Mixed-Integer Linear Programming solvers mipSolve and cutplane. Either the QLD QP-solver or the DualSolve dual simplex solver may be used for subproblems, with warm starts.
- Constrained nonlinear minimization solvers conSolve and nlpSolve.
- Unconstrained nonlinear minimization solvers ucSolve and sTrustr.
- Robust solution of ill-conditioned nonlinear least squares with linear constraints, solver clsSolve.
- Sparse linear least squares using Tlsqr, or the nonlinear least squares solver clsSolve.
- Dense constrained linear least squares using LSEI.
- Nonnegative constrained linear least squares using Tnnls, a fast and robust replacement for the Matlab nnls routine.
- Solving non-linear systems of equations, solver clsSolve.
- Least squares problems with L1, L2 and Infinity norm using L1Solve, slsSolve and infSolve.
- Multistart optimization for finding multiple local minima with multiMin.
- Complete integration of automatic differentiation using the MAD toolbox.