roadmap.md

Roadmap

ocp_nlp

• Gauss-Newton SQP
• exact Hessian SQP
• partial tightening
• HPNMPC
• blockSQP

ocp_qp

• qpOASES v1.0
• qpOASES v2.0 (no dynamic memory allocation, same structure as rest of QP solvers)
• block condensing (should maybe get its own category)
• OOQP
• qpDUNES
• HPMPC
• FORCES
• OSQP

sim

• explicit Runge-Kutta
• lifted IRK
• collocation integrators GL
• collocation integrators Radau
• discrete-time systems
• second order sensitivities

general

• allow for models with varying state and control dimensions
• unit tests in Python (unit_test or nosetests) or MATLAB (built-in framework)
• 'code generation' of driver files such that users don't have to code them manually
• short Python doctests to exemplify usage of interface functions
• installation via conda and/or pip
• RTI implementation similar to ACADO
• Provide Simulink blocks as an interface (self-written or Legacy Code Tool)
• HPMPC and HPIPM for MSVC / dSpace
• Proper installation instructions/Troubleshooting on the Wiki/Readme
• Installation instructions for MacOS
• Provide separate documentation page