2. License Variants

Each problem type requires a dedicated solver method in order to be solved quickly and efficiently. FORCESPRO is available in different variants in order to adapt to each user’s needs. When receiving a FORCESPRO license on the portal(https://my.embotech.com) a user can select one of the available variants which is best suited for the problem to be solved. At any point, a user can decide to upgrade to a larger variant in order to include additional solver methods in their available tool set for FORCESPRO.

The available variants are (smaller variants are included in larger ones):

• S (Variant S)

• M (Variant M)

• L (Variant L)

2.1. Variant Summary

In the tables below you can find a summary of the components provided with each variant of FORCESPRO.

Table 2.1 Problem types supported for each variant

	S	M	L
Problem Type
LP	✓	✓	✓
QP	✓	✓	✓
QCQP	✓	✓	✓
BI-QP		✓	✓
NLP (SQP)		✓	✓
NLP (IP)			✓
MINLP			✓

Table 2.2 Interfaces provided for each variant

	S	M	L
Interface
MATLAB Low-Level	✓*	✓	✓
Python Low-Level	✓*	✓	✓
MATLAB Y2F	✓	✓	✓
MathWorks MPC Toolbox™ (Linear MPC)	✓	✓	✓
MATLAB High-Level		✓**	✓
Python High-Level		✓**	✓
MathWorks MPC Toolbox™ (Nonlinear MPC)		✓**	✓
* No Binary Constraints
** Only with SQP method

2.2. Variant S

This variant is used for generation of convex solvers. This variant should be used for solving:

• LP problems

• QP problems

• QCQP problems

This variant is delivered with the following interfaces:

• MATLAB Low-level Interface (Low-level interface)

• Python Low-level Interface (Low-level interface)

• MATLAB Y2F Interface (Y2F Interface)

• MathWorks Model Predictive Control Toolbox™ - Linear MPC (MathWorks Linear MPC Plugin)

2.3. Variant M

This variant further enables the generation of SQP solvers for NLPs and the solution of Binary-Integer QPs. This variant should be used for solving:

• Binary-Integer QP problems (Binary constraints)

• NLP Problems using SQP methods (Sequential quadratic programming algorithm)

This variant is delivered with the following interfaces:

• MATLAB High-level Interface (High-level Interface) with codeoptions.solvemethod = ‘SQP_NLP’;

• Python High-level Interface (High-level Interface) with codeoptions.solvemethod = ‘SQP_NLP’

• MathWorks Model Predictive Control Toolbox™ - Nonlinear MPC (MathWorks Nonlinear MPC Plugin) with options.SolverType = ‘SQP’;

2.4. Variant L

This variant provides the full experience of FORCESPRO and enables all its features. This variant further enables the solution of:

• NLP problems with Interior-Point Methods and SQP

• MINLP problems (Mixed-integer nonlinear solver)

This variant is delivered with the following interfaces:

• MATLAB High-level Interface (High-level Interface) with full support

• Python High-level Interface (High-level Interface) with full support

• MathWorks Model Predictive Control Toolbox™ - Nonlinear MPC (MathWorks Nonlinear MPC Plugin) with full support