ODYS Embedded MPC

ODYS specializes in developing Model Predictive Control (MPC) systems for next-gen controls in industrial production. Based on more than 25 years of scientific research, the expertise of the company covers advanced multivariable control design, efficient real-time optimization algorithms, efficient and robust software implementation, and consultancy on MPC design.

ODYS Embedded MPC is a software library for implementing Model Predictive Control solutions in real-time applications. It handles general multivariable linear time-varying and nonlinear prediction models. The software provides many advanced MPC design features, enabling extremely flexible and customized problem formulations in terms of cost function, constraints, preview, and degrees of freedom in the optimization. State estimation is performed by using state-of-the-art Kalman filtering techniques, optimized for computation efficiency and numerical robustness, also in single precision. The C-code is compliant with MISRA-C 2012 and supports single-precision arithmetic.

ODYS Embedded MPC is fast and robust in single precision, predictable in worst-case execution time, and requires very limited memory footprint. Ad-hoc sparse linear algebra allows efficient handling of sparsity in prediction models. The tight integration with ODYS’ MPC-dedicated QP solver means that MPC problems can be solved very efficiently (typically, twice as fast and with half the memory footprint than by using the general-purpose ODYS QP solver).

Being a standalone C-code library, the code can run on any industrial desktop PCs and any embedded platform that supports floating-point operations.

APIs to MATLAB^® and Simulink^® are available for testing the controller in simulation. The MPC controller can be defined in MATLAB code, rather than C code, and MATLAB Coder^® can be used to generate the corresponding C code. With no need for the user to write any C code during the initial phase of the design, it reduces the learning curve to get proficient with the software and the use of advanced MPC features.

ODYS Embedded MPC also has a plugin to run nonlinear MPC projects designed using the Model Predictive Control Toolbox^® for MATLAB^®. With a single click, users can port their existing project carried out with the Model Predictive Control Toolbox™ to fast and robust C code compatible with ODYS Embedded MPC.

In addition, ODYS Embedded MPC can be integrated with ODYS Deep Learning Toolset, a plugin for developing prediction models based on neural networks that can be exported to C-code for immediate use by ODYS’ MPC and state estimation functions.