DC–DC converter

  With increasing complications and varieties of satellites, the requirement for attitude control becomes higher. As a significant actuator, a reaction flywheel has high output precision of angular momentum, strong anti-interference ability, and fast response, whose output torque accuracy directly affects attitude control precision. Thus, how to improve the stabilization and control precision for angular momentum is a key issue and urgent to be solved. The reaction flywheel can be regarded as a motor with a biggish moment of inertia. For the drive motor of reaction flywheel domestic and overseas, the brushless DC motor (BLDCM) is utilized widely with a relatively wide range of rotational speed, and the rotational speed regulation requires that the motor’s line voltage changes with the rotational speed. Therefore, a voltage-mode controlled buck converter is mainly applied as a typical DC–DC converter.

  As is known, the DC–DC converter is a highly nonlinear plant with abundant dynamical behaviors, such as the Hopf bifurcation,1 period doubling bifurcation,2 border collision bifurcation,3 tangent bifurcation,4 coexisting attractors,5

etc. The major difficulty in controlling the DC–DC converter is the exact modeling.

  At present, the modeling approaches for DC–DC converters are generally divided into two types: one based on small
signal analysis and the other using large signal analysis. The small signal analysis consists of the state-space averaging approach 6 and the averaged equivalent circuit approach.7 However, the large signal analysis is complicated and also
classified into two categories: one is based on the state-space averaging or circuit-average method, such as the phase plane method,8 the switching signal flow graph approach,9 etc.; the other uses state variables ripple determination, like the extended ripple analysis,10 the nth harmonic three terminal model, 11 the Krylov Bogoliubov Mitropolskii (KBM) method,12 the Volterra series determination,13 the describing function method,14 etc. Anyway, most control strategies are still based on the state-space averaging model or the linearization small signal model.