An MPC Approach to Transient Control of Liquid-Propellant Rocket Engines

Abstract

The current context of launchers reusability requires the improvement of control algorithms for their liquid-propellant rocket engines. Their transient phases are generally still performed in open loop. In this paper, it is aimed at enhancing the control performance and robustness during the fully continuous phase of the start-up transient of a generic gas-generator cycle. The main control goals concern end-state tracking in terms of combustion-chamber pressure and chambers mixture ratios, as well as the verification of a set of hard operational constraints. A controller based on a nonlinear preprocessor and on linear MPC (Model-Predictive Control) has been synthesised, making use of nonlinear state-space models of the engine. The former generates the full-state reference to be tracked while the latter achieves the aforementioned goals with sufficient accuracy and verifying constraints for the required pressure levels. Robustness considerations are included in the MPC algorithm via an epigraph formulation of the minimax robust optimisation problem, where a finite set of perturbation scenarios is considered.

A Reinforcement Learning Approach for Transient Control of Liquid Rocket Engines

“…5 shows an overshoot in the combustion chamber pressure as well as in the global mixing ratio with both the PID controllers and the RL agent. This characteristic behavior also occurs when an MPC approach is used [25]. Thus, the causes might be system-related.…”

“…The objective of our work is analogous to the investigation of Pérez-Roca et al [25], where a model predictive control (MPC) approach to control the start-up transient of a liquid rocket engine was studied. After the derivation of a suitable state-space model [26], a linear MPC controller was synthesized.…”

Trajectory Planning and Tracking via MPC for Transient Control of Liquid-Propellant Rocket Engines

“…The type of LPRE studied is a gas-generator-cycle engine, but the methodology is intended to be generally applicable to other cycles. As in [16], the control loop presented in this paper is based on Model Predictive Control (MPC), which is accompanied by a preprocessor for reference-trajectory generation. The MPC method has been chosen as the most adequate for this type of complex systems with hard operational constraints, as introduced in the next sections.…”

“…The modelling approach considered in this paper was first introduced in [15] and further explained and updated in [16]. Basically, a set of different models is used, characterised by diverse degrees of complexity according to their purposes.…”

A Reinforcement Learning Approach for Transient Control of Liquid Rocket Engines

“…The objective of our work is analogous to the investigation of Pérez-Roca et al [22], where a model predictive control (MPC) approach to control the start-up transient of a liquid rocket engine was studied. After the derivation of a suitable state-space model [23], a linear MPC controller was synthesized.…”