Flight Readiness of ReFEx Guidance: Preparation for Integration Into Flight Software

Abstract

This paper presents the preparation steps to integrate the ReFEx Guidance Algorithms into the Flight Software (FSW) executed on the Guidance and Control Computer (GCC). The German Aerospace Center (DLR) "Reusability Flight Experiment" (ReFEx) aims to demonstrate the Guidance, Navigation, and Control (GNC) capabilities for an aerodynamically controlled Reusable Launch Vehicle (RLV) stage. The GNC subsystem is responsible for maintaining stability and steering the vehicle to achieve a target state, transitioning from the supersonic to subsonic flight regime. The Guidance task consists of planning and updating a feasible trajectory from separation from the launcher up to the target state. This paper describes two steps to increase the flight readiness of the ReFEx Guidance Algorithms. The first one is the process to go from an algorithm-development simulation environment to code suitable for integration in the FSW. The second one is the use of Software-in-the-Loop (SiL) tests to verify the consistency of the resulting code and of Embedded-in-the-Target (EiT) tests to verify the feasibility of the algorithms from a timing perspective. Both developments are significant steps towards achieving TRL 6 for these algorithms. The development of the Guidance algorithms was conducted in MATLAB/Simulink. The process to convert the algorithms to the target programming language for integration with the FSW (C++) is presented. The level of encapsulation, i.e. at which level to interface with the FSW, is driven by the executing frequency of the algorithms. Once encapsulated, Mathworks toolboxes are used to generate code. This automatic process has many advantages, including reducing the risk of human error and the deployment time of code. The resulting code was evaluated, assessing numerical and performance differences w.r.t. the original code and timing performance. SiL tests were used to assess: 1) numerical differences w.r.t. the original code, and 2) closed-loop performance of the generated code. EiT tests were used to: 1) evaluate the numerical effect of the compilation process and the hardware used, and 2) to verify that the execution time of the algorithms is within requirements.