arXiv subjects

Alessandro Zavoli

Publications and source records attributed to Alessandro Zavoli.

11 recordsLinked to original sources

arXiv2022

Autonomous Upper Stage Guidance with Robust Splash-Down Constraint

This paper presents a novel algorithm, based on model predictive control (MPC), for the optimal guidance of a launch vehicle upper stage. The proposed strategy not only maximizes the performance of the vehicle and its robustness to external disturbances, but also robustly enforces the splash-down constraint. Indeed, uncertainty on the engine performance, and in particular on the burn time, could lead to a large footprint of possible impact points, which may pose a concern if the reentry points are close to inhabited regions. Thus, the proposed guidance strategy incorporates a neutral axis maneuver (NAM) that minimizes the sensitivity of the impact point to uncertain engine performance. Unlike traditional methods to design a NAM, which are particularly burdensome and require long validation and verification tasks, the presented MPC algorithm autonomously determines the neutral axis direction by repeatedly solving an optimal control problem (OCP) with two return phases, a nominal and a perturbed one, constrained to the same splash-down point. The OCP is transcribed as a sequence of convex problems that quickly converges to the optimal solution, thus allowing for high MPC update frequencies. Numerical results assess the robustness and performance of the proposed algorithm via extensive Monte Carlo campaigns.

math.OC↗

arXiv2022

Stochastic Control of Launch Vehicle Upper Stage with Minimum-Variance Splash-Down

This paper presents a novel synthesis method for designing an optimal and robust guidance law for a non-throttleable upper stage of a launch vehicle, using a convex approach. In the unperturbed scenario, a combination of lossless and successive convexification techniques is employed to formulate the guidance problem as a sequence of convex problems that yields the optimal trajectory, to be used as a reference for the design of a feedback controller, with little computational effort. Then, based on the reference state and control, a stochastic optimal control problem is defined to find a closed-loop control law that rejects random in-flight disturbance. The control is parameterized as a multiplicative feedback law; thus, only the control direction is regulated, while the magnitude corresponds to the nominal one, enabling its use for solid rocket motors. The objective of the optimization is to minimize the splash-down dispersion to ensure that the spent stage falls as close as possible to the nominal point. Thanks to an original convexification strategy, the stochastic optimal control problem can be solved in polynomial time since it reduces to a semidefinite programming problem. Numerical results assess the robustness of the stochastic controller and compare its performance with a model predictive control algorithm via extensive Monte Carlo campaigns.

math.OC↗

arXiv2021

Analysis of 3GM Callisto Gravity Experiment of the JUICE Mission

The ESA's JUICE mission will provide a thorough investigation of the Jupiter system and the Galilean moons during its nominal tour, comprising flybys of Europa and Callisto, and an orbital phase about Ganymede at the end of the mission. The 3GM experiment will exploit accurate Doppler and range measurements to determine the moons' orbits and gravity fields (both static and tidal) and infer their interior structure. This paper presents the attainable accuracies of the Callisto geodesy experiment and addresses the effect of different flybys mean anomaly distribution and geometry on the estimation of the tidal Love number k2.

astro-ph.EP↗

arXiv2020

Convex Optimization of Launch Vehicle Ascent Trajectory with Heat-Flux and Splash-Down Constraints

This paper presents a convex programming approach to the optimization of a multistage launch vehicle ascent trajectory, from the liftoff to the payload injection into the target orbit, taking into account multiple nonconvex constraints, such as the maximum heat flux after fairing jettisoning and the splash-down of the burned-out stages. Lossless and successive convexification are employed to convert the problem into a sequence of convex subproblems. Virtual controls and buffer zones are included to ensure the recursive feasibility of the process and a state-of-the-art method for updating the reference solution is implemented to filter out undesired phenomena that may hinder convergence. A hp pseudospectral discretization scheme is used to accurately capture the complex ascent and return dynamics with a limited computational effort. The convergence properties, computational efficiency, and robustness of the algorithm are discussed on the basis of numerical results. The ascent of the VEGA launch vehicle toward a polar orbit is used as case study to discuss the interaction between the heat flux and splash-down constraints. Finally, a sensitivity analysis of the launch vehicle carrying capacity to different splash-down locations is presented.

math.OC↗

arXiv2020

Reinforcement Learning for Low-Thrust Trajectory Design of Interplanetary Missions

This paper investigates the use of Reinforcement Learning for the robust design of low-thrust interplanetary trajectories in presence of severe disturbances, modeled alternatively as Gaussian additive process noise, observation noise, control actuation errors on thrust magnitude and direction, and possibly multiple missed thrust events. The optimal control problem is recast as a time-discrete Markov Decision Process to comply with the standard formulation of reinforcement learning. An open-source implementation of the state-of-the-art algorithm Proximal Policy Optimization is adopted to carry out the training process of a deep neural network, used to map the spacecraft (observed) states to the optimal control policy. The resulting Guidance and Control Network provides both a robust nominal trajectory and the associated closed-loop guidance law. Numerical results are presented for a typical Earth-Mars mission. First, in order to validate the proposed approach, the solution found in a (deterministic) unperturbed scenario is compared with the optimal one provided by an indirect technique. Then, the robustness and optimality of the obtained closed-loop guidance laws is assessed by means of Monte Carlo campaigns performed in the considered uncertain scenarios. These preliminary results open up new horizons for the use of reinforcement learning in the robust design of interplanetary missions.

cs.LG↗

arXiv2020

EOS: a Parallel, Self-Adaptive, Multi-Population Evolutionary Algorithm for Constrained Global Optimization

This paper presents the main characteristics of the evolutionary optimization code named EOS, Evolutionary Optimization at Sapienza, and its successful application to challenging, real-world space trajectory optimization problems. EOS is a global optimization algorithm for constrained and unconstrained problems of real-valued variables. It implements a number of improvements to the well-known Differential Evolution (DE) algorithm, namely, a self-adaptation of the control parameters, an epidemic mechanism, a clustering technique, an $\varepsilon$-constrained method to deal with nonlinear constraints, and a synchronous island-model to handle multiple populations in parallel. The results reported prove that EOSis capable of achieving increased performance compared to state-of-the-art single-population self-adaptive DE algorithms when applied to high-dimensional or highly-constrained space trajectory optimization problems.

cs.NE↗

arXiv2019

Convex Optimization of Linear Impulsive Rendezvous

This note aims at providing a concise and self-contained document that describes a clear and easy-to-understand method, that could be useful for a reader that is approaching the linear-impulsive rendezvous topic for the first time, but that is also flexible enough to accommodate problem variations, such as additional constraints like bounded-magnitude DV, or direction constraints with minor modifications.

math.OC↗

arXiv2019

Integrated Optimization of Ascent Trajectory and SRM Design of Multistage Launch Vehicles

This paper presents a methodology for the concurrent first-stage preliminary design and ascent trajectory optimization, with application to a Vega-derived Light Launch Vehicle. The reuse as first stage of an existing upper-stage (Zefiro 40) requires a propellant grain geometry redesign, in order to account for the mutated operating conditions. An optimization code based on the parallel running of several Differential Evolution algorithms is used to find the optimal internal pressure law during Z40 operation, together with the optimal thrust direction and other relevant flight parameters of the entire ascent trajectory. Payload injected into a target orbit is maximized, while respecting multiple design constraints, either involving the alone solid rocket motor or dependent on the actual flight trajectory. Numerical results for SSO injection are presented.

math.OC↗

arXiv2019

GTOC X: Solution Approach of Team Sapienza-PoliTo

This paper summarizes the solution approach and the numerical methods developed by the joint team Sapienza University of Rome and Politecnico di Torino (Team Sapienza-PoliTo) in the context of the 10th Global Trajectory Optimization Competition. The proposed method is based on a preliminary partition of the galaxy into several small zones of interest, where partial settlement trees are developed, in order to match a (theoretical) optimal star distribution. A multi-settler stochastic Beam Best-First Search, that exploits a guided multi-star multi-vessel transition logic, is proposed for solving a coverage problem, where the number of stars to capture and their distribution within a zone is assigned. The star-to-star transfers were then optimized through an indirect procedure. A number of refinements, involving settle time re-optimization, explosion, and pruning, were also investigated. The submitted 1013-star solution, as well as an enhanced 1200-point rework, are presented.

math.OC↗

arXiv2019

A Time-Dependent TSP Formulation for the Design of an Active Debris Removal Mission using Simulated Annealing

This paper proposes a formulation of the Active Debris Removal (ADR) Mission Design problem as a modified Time-Dependent Traveling Salesman Problem (TDTSP). The TDTSP is a well-known combinatorial optimization problem, whose solution is the cheapest mono-cyclic tour connecting a number of non-stationary cities in a map. The problem is tackled with an optimization procedure based on Simulated Annealing, that efficiently exploits a natural encoding and a careful choice of mutation operators. The developed algorithm is used to simultaneously optimize the targets sequence and the rendezvous epochs of an impulsive ADR mission. Numerical results are presented for sets comprising up to 20 targets.

math.OC↗

arXiv2019

A Convex Optimization Approach for Finite-Thrust Time-Constrained Cooperative Rendezvous

This paper presents a convex approach to the optimization of a cooperative rendezvous, that is, the problem of two distant spacecraft that simultaneously operate to get closer. Convex programming guarantees convergence towards the optimal solution in a limited, short, time by using highly efficient numerical algorithms. A combination of lossless and successive convexification techniques is adopted to handle the nonconvexities of the original problem. Specifically, a convenient change of variables and a constraint relaxation are performed, while a successive linearization of the equations of motion is employed to handle the nonlinear dynamics. A filtering technique concerning the recursive update of the reference solution is proposed in order to enhance the algorithm robustness. Numerical results are presented and compared with those provided by an indirect method.

math.OC↗