• cubesats
• optimization
• satellite design optimization
• satellite system design
• satellite task scheduling
• simulation
• small satellites

The demand for small satellites, and in particular CubeSats in the microsatellite range, has been growing steadily in the last years. The University of Pisa is working on one of such satellites, the EXCITE CubeSat, on behalf of the Italian Space Agency (ASI). As such, this work was born with the objective of determining a framework of techniques needed to generate an optimized design for satellites and then to optimize the usage of onboard energy to execute mission tasks, in order to facilitate future work on EXCITE. This objective will be accomplished by building a tool capable of generating an optimized design from a database of available components provided by the users, by solving a multidimensional knapsack problem; this will allow to establish all the parameters and performance capabilities of the satellites, which will be used to determine an optimized task scheduling after simulating its base behavior over a given period of time. As such, the tool will be made of three sections: the first one will be the design optimization section, the second one will be the orbital simulation section, and the third one will be the task scheduling optimization section. Many user-defined parameters will be needed as inputs, together with the aforementioned database, to solve such problems in any conditions the engineers may require; indeed, such parameters will grant high versatility and customization options to the tool. This will allow the generation and operation of very different satellites or scenarios. The resolution will need the definition of an engineering and mathematical model capable of describing the aforementioned problems, introducing constraints needed to represent the physical and engineering conditions and limitations. These problems will be defined as integer linear programming problems, which will allow to obtain an exact solution to these problems while being able to solve them in a reasonably low amount of time with commercial hardware. These optimization problems will be solved through the use of the Branch and Bound method. Lastly, special care was taken in defining the constraints of the optimization problems so to include all the needed engineering boundaries while guaranteeing a reasonable runtime.