FPGA-accelerated model predictive control for gas turbine aero-engines

Gas Turbine Engine (GTE) operability and safety constraints provide a challenging task for the control design. Model Predictive Control (MPC) provides a structured frame work for handling multivariable interactions and hard constraints through online optimisation. However, its deployment in embedded aerospace systems is constrained by computational complexity, timing requirements, and hardware resource limitations. This paper presents the design, implementation, and validation of a real-time MPC controller for a GTE using a Hardware Description Language (HDL) on an FPGA. The resulting quadratic programming (QP) problem is solved using an Alternating Direction Method of Multipliers (ADMM) algorithm. Unlike previous FPGA-based QP implementations that are primarily demonstrated on numerical benchmarks, this work presents a representative aerospace control application with system constraints derived from the GTE performance and safety requirements. A key aspect of the proposed architecture is its resource-efficient and reusable design, enabling deployment across different MPC problem sizes without requiring redesign of the underlying hardware. Closed-loop validation is performed using a hardware-in-the-loop (HIL) setup, where the FPGA-based QP solver operates in real-time with an industrial simulation platform executing the GTE model. The results demonstrate deterministic timing behaviour, real-time feasibility, and the suitability of resource-efficient FPGA-based embedded control for safety-critical applications.