Robust Sensor Optimization for Liquid Propellant Rocket Engine Model Parameter Estimation

Parameter estimation can adjust the model as per the actual data, which is the key to reusable liquid propellant rocket engine health management. We introduce a nonlinear parameter estimation method, which contains estimability analysis and solving strategy. For certain parameters in the liquid propellant rocket engine (LPRE) model and certain processes, sensor networks determine the estimation accuracy. By considering sensor robustness to parameters and fault redundancy, we proposed a sensor optimization framework. A heuristic branch-and-bound solving strategy based on convex relaxation was developed. The effectiveness of the sensor optimization and parameter estimation methods was verified based on the case study of the space shuttle main engine. The proposed sensor optimization solving strategy has better performance than general-purpose solvers.