On-orbit service mission optimization for geosynchronous satellites based on metaheuristic algorithms

This thesis proposes a comprehensive and efficient on-orbit service (OOS) project design scheme. The multi-objective optimization of the optimal selection of mission objects, mission allocations, and execution sequences of OOS vehicles and the formation of low-thrust rendezvous orbits are all concurrently examined using metaheuristic algorithms.

First, we investigate advanced metaheuristic algorithms for global optimization problems. To overcome the present bottlenecks that limit the accuracy of metaheuristics in conjunction with broadening the area of optimization problems, we propose a whale army optimization algorithm and a proportional, integral, and derivative differential evolution algorithm. The integrated high-accuracy, rapid-convergence, and stable optimizers are then developed using comprehensive information utilization principles and flexible parameter determination methods. We evaluate the performance of the proposed algorithms using simple benchmark functions, CEC-2014 real-parameter numerical optimization problems, and real-world constrained engineering design problems. Test results illustrate that the proposed algorithms can provide a faster local convergence rate, higher convergence accuracy, and lower computational complexity than numerous typical approaches, including the top algorithms in CEC competitions.

Second, we propose novel low-thrust shape approximation models for rendezvous orbits in space. To address the general low-thrust orbit in space, a transformed high-order polynomial model based on the finite Fourier series is presented to improve the precision of finite Fourier series shape-based function. Furthermore, a two-stage shape-approximate model of the low-thrust rendezvous orbit for geosynchronous satellites is developed, accounting for the trajectory safety. The proposed algorithms investigate the three objectives of minimum flight time, fuel consumption, and engine thrust in the multi-objective rendezvous orbit optimization problem, subjected to constraints which include angular momentum, trajectory safety, and fuel limitation. Simulation results show that this approach has both the high precision of numerical methods and the low processing complexity offered by shape approximation techniques. Additionally, the Pareto front of the multi-objective optimization orbit is introduced to analyze a complete optimal solution set. Furthermore, we demonstrate that our proposed approach is appropriate for generating the preliminary orbit for the pseudo-spectral method.

An OOS project in space is established using one or more OOS vehicles and several fuel-filling stations providing a real-time service to satellites in geosynchronous orbits. The unique ideas of satellite cooperative relationships and diverse service requirements of a satellite are developed and introduced into the OOS project. The optimization of the OOS project primarily comprises three steps: selecting several client satellites from the potential service targets to develop the optimal customer group, determining the service objects and flight sequences of the service spacecraft while considering the fuel-filling operation, and optimizing the service spacecraft’s low-thrust rendezvous orbits. The Pareto front of an OOS implementation plan is analyzed in simulation experiments, considering economic value, risk factor, satellite collaboration, fuel consumption, and other practical goals under the constraints of execution time bounds, fuel-filling station usage limitations, and maximum engine thrust, among other things. Furthermore, by making use of low-thrust orbit can save half the fuel compared to the three-impulse orbit, and fuel economy can be further improved through the use of fuel-filling stations.