Design and Verification of Embedded Model Predictive Controller for Aero-engines

In this paper, an online computing model predictive controller is designed for aero-engines and implemented in a microprocessor. Due to the heavy burden of online optimization of model predictive control (MPC) and the high real-time requirements of aero-engines, it is very difficult to embed the MPC algorithm. In this paper, the MPC algorithm is implemented on low-end hardware devices. And this result is mainly achieved by implementing OSQP solver on low-frequency single-chip microcomputers. In order to verify the effectiveness of the controller, a hardware-in-the-loop (HIL) test platform is built for testing. When the controller program is running, the optimization solution can be completed online in a very short time, because the solution iteration process can always converge quickly. And the controller has a good control effect on the nonlinear aero-engine model. It is confirmed that the MPC algorithm can be embedded in the application field with complex model and high real-time performance such as aero-engines.