A library for model predictive control using a structure exploiting active-set method with support for low-rank updates.
See doc/usage.md for basic usage.
When compiling for embedded devices, it may be beneficial to use 32-bit floats instead of the default 64-bit floats inside the solver. This reduces the accuracy but can improve the computation time on devices without support for 64-bit floats in hardware. This can be done by disabling the default f64 feature. Note that this also changes the default values of the tolerance settings.