Particle Swarm Optimization to solve Multiple Dipole Modelling problems in space applications

An advanced modeling algorithm based on Particle Swarm Optimization (PSO) has been developed to solve Multiple Dipole Modelling (MDM) problems in space applications. Multiple Dipoles Modelling is a technique to represent spacecraft units as a set of equivalent magnetic dipoles able to reconstruct, in the far-field distance, the same original magnetostatic field. This procedure allows preparing a magnetic model of the spacecraft during design and development phases and foreseeing the magnetostatic state of the spacecraft during operation in the final orbit. This latter aspect plays an important role in mission with equipment susceptible to magnetic fields since the spacecraft behaviour with changing environment can be predicted and taken into account during design and development [1]. During the last decades, the MDM problem has been addressed in different ways by many authors for several applications. A main difference resides in the mathematical approach for implementation of the optimisation technique used as solver, which can be of deterministic [2]–[3] or stochastic [4]–[9] nature. For space applications mainly deterministic methods have been applied; nevertheless, due to the highly nonlinear nature of the problem, classic deterministic methods are not always the best choice for this application (problem of local minima and need of suitable initial guesses.). Therefore our research has been driven towards the investigation of an advanced stochastic method.