An efficient dynamical model of reluctance actuators with flux fringing and magnetic hysteresis

Rauch-Tung-Striebel Smoother for Position Estimation of Short-Stroke Reluctance Actuators

“…The continuous-time model presented in Section II-A includes the most important electromagnetic phenomena. As shown in [23], it matches real measurements with high accuracy. Thus it is considered the reference, emulating the real system, for analysis purposes.…”

“…Due to its complexity, magnetic hysteresis is most commonly ignored in this class q = 2 (motion) of actuators, even if its effect is not negligible. Still, there are some proposals for integrating magnetic hysteresis models into the dynamic system of reluctance actuators [22], [23].…”

“…Following the approach from [23], the differential permeability µ ∆ = dB c /dH c -which provides the relation between the core magnetic flux density B c and field strength H c -can be expressed as a piecewise function of φ, H c and φ (with different behavior for increasing and decreasing magnetic flux),…”

“…This is a compromise between measuring only electrical signals and measuring directly the position. On the one hand, if only the current and voltage are measured, the estimator must be based on a more complex and accurate model, such as the ones presented in [22] or [23]. This makes the model identification and estimator implementation much more challenging and expensive.…”

On the Stability of Electromechanical Switching Devices: An Analytic Study of Hysteretic Switching Behaviors

“…The first one is the magnetic equivalent circuit (MEC) approach, which results in low-order dynamical models well suited for control or estimation, as well as for computationally inexpensive simulations [5], [6], [12]. Although these models only capture the dynamics of a small number of variables, they can be made very accurate by including phenomena such as eddy currents or magnetic saturation and hysteresis [13], [14]. The second approach is the use of high-order numerical models, e.g., those based on the finite element method [4], [15], which provide much more detailed results, but are generally computationally expensive and not appropriate for some classes of analysis.…”