The implementation of model predictive controllers on low-cast hardware such as micro-controllers has been traditionally hampered by the high computing and associated memory demands of the algorithm. This paper describes a completely automatic way to implement an MPC controller on embedded hardware starting from a dynamic model in ml. The resultant controller runs stand-alone on the embedded hardware, is extremely fast, exhibits a modest memory footprint and best of all, requires no particular embedded programming experience from the user.
展开▼
