G-FOLD: A Real-Time Implementable Fuel Optimal Large Divert Guidance Algorithm for Planetary Pinpoint

Introduction: Spacecraft accumulate large position and velocity errors during the atmospheric entry phase of a planetary mission due to atmospheric uncertainties and limited control authority. The powered descent phase, which is the last phase of Entry, Descent, and Landing (EDL), is when the lander makes a controlled maneuver to correct for these errors. This maneuver must be computed onboard in realtime because the state of the lander cannot be predicted at the start of powered descent phase. Current state-of-the-art onboard Powered Descent Guidance (PDG) algorithms used in this phase are inherited from the Apollo era. These algorithms do not explicitly optimize fuel usage or prevent landing mission constraints from being violated. As a result they cannot fully utilize the full spacecraft divert capability, and hence significantly limit the landing precision. We developed a new solution methodology and an associated onboard real-time implementable algorithm, called G-FOLD, for powered descent guidance for planetary landing, which autonomously generates fuel optimal landing trajectories in real-time. The algorithm is capable of generating any physically feasible large divert PDG trajectory that satisfies mission constraints. It has been used extensively at NASA JPL over the last seven years to perform future Mars lander mission analysis in an automated fashion. It has proved itself to be an essential analysis tool and the fact that it has been automated proves its numerical robustness. G-FOLD algorithm provides a key new technology required for planetary pinpoint landing. The National Research Council report on National Space Technology Roadmaps and Priorities identifies precision or pinpoint landing as a top priority for technology investment. Pinpoint landing capability is important to NASA because, as landing precision increases, robotic missions are able to access science targets, which are currently inaccessible or are very risky for a rover to reach from large distances. For crewed missions, the increased precision with minimal fuel requirements enables the landing of larger payloads in close proximity to predetermined targets. Description of G-FOLD Algorithm: We developed an optimal large divert PDG algorithm [1, 2, 3] that autonomously computes the fuel optimal path that takes the lander to a given surface target without violating any mission constraints. This powered descent Guidance algorithm for Fuel Optimal Large Diverts (G-FOLD) is needed for planetary pinpoint landing. It enables access to unreachable but scientifically valuable science targets for Mars sample return mission and to deliver large payloads necessary for human class planetary missions. The main purpose of Guidance, Navigation, and Control (GN&C) during the landing phase of planetary missions is to reduce the lander’s velocity from orbital or interplanetary velocities to a velocity near zero. In case of planets with atmospheres (such as Mars), this first involves an entry phase (see Fig. 1). This phase cancels most of the surface relative velocity. Once the lander slows down to supersonic speeds, a parachute is deployed. Then at a prescribed altitude and velocity, the parachute is released and the Powered Descent (PD) phase is initiated. Due to atmospheric uncertainties in winds and density and due to the passive deceleration during the parachute phase, the position and velocity relative to the target is dispersed significantly and they cannot be predetermined. In the example of Mars landing, the distance error can be in the order of 8-10 km with the velocity trigger (used in MSL) and 5-6 km with a range trigger for the start of parachute phase [4]. To achieve pinpoint landing (position error < 100 m at touchdown), an autonomous PDG algorithm is used to redirect the vehicle to the surface target in real-time to correct for these errors.