Constrained optimization of the magnetostrictive actuator with the use of penalty function method

• Lukasz Knypinski ^[1] ; Krzysztof Kowalski ^[1] ; Lech Nowak ^[1]
  1. [1] Institute of Electrical Engineering and Electronics, Poznan University of Technology, Polonia
• Localización: Compel: International journal for computation and mathematics in electrical and electronic engineering, ISSN 0332-1649, Vol. 37, Nº 5, 2018, págs. 1575-1584
• Idioma: inglés
• Enlaces
  • Texto completo
• Resumen

  • Purpose The purpose of this paper is to elaborate algorithm and software for the optimization of the actuator–capacitor system taking the dynamics parameters into account. The system is applied for driving the valve of plasma gun. Two optimization strategies are proposed and pondered. The penalty function approach has been expanded in detail.

    Design/methodology/approach The field-circuit mathematical model of the dynamics operation consists of the strongly coupled equations of the transient electromagnetic fields and the equations of the electric circuit. The numerical implementation is based on finite element method and step-by-step Cranck–Nicholson procedure. The genetic algorithm has been used in the optimization procedure. The sigmoidal transformation has been applied to adjust the classical external penalty function method to the genetic algorithm.

    Findings The modification consists in adaptation of the penalty function to the genetic algorithm. In the proposed approach, operations involving successive iterations of increasing penalty function and operations containing genetic iterations are intertwined with each other. The differences between these two procedures are getting blurred. The proposed approach is very effective. It is possible to achieve optimal solution even more than ten times faster than using the classical method.

    Originality/value The proposed approach can be successfully applied to designing and optimization of different electromagnetic devices, including functional constraints.