Search

Design and optimization of high-frequency inductive components is a complex task because of the huge number of variables to manipulate, the strong interdependence and the interaction between variables, the nonlinear variation of some design variables as well as the problem nonlinearity. This paper proposes a multi-objective design methodology of a 200-W flyback transformer in continuous conduction mode using genetic algorithms and Pareto optimality concept. The objective is to minimize loss, volume and cost of the transformer. Design variables such as the duty cycle, the winding configuration and the core shape, which have great effects on the former objectives but were neglected in previous works, are considered in this paper. The optimization is performed in discrete research space at different switching frequencies. In total, 24 magnetic materials, 6 core shapes and 2 winding configurations are considered in the database. Accurate volume and cost models are also developed to deal with the optimization in the discrete research space. The bi-objective (loss–volume) and tri-objective (loss–volume–cost) optimization results are presented, and the variations of the design variables are analyzed for the case of 60 kHz. An example of a design (30 kHz) is experimentally verified. The registered efficiency is 88% at full load. [ABSTRACT FROM AUTHOR]

Power system stabilizers (PSSs) associated with generators are mandatory requirements for damping low-frequency oscillations of a multimachine power system. Nevertheless, PSSs work well at particular network configuration and steady-state conditions for which they were designed. Therefore, the aim of this paper is the design of a robust PSS able to ensure the stability of the system for a wide range of loading conditions and various faults with higher performance. The main motivation for this design is to simultaneously shift as much as could be possible the lightly damped and undamped system electromechanical modes, at different loading conditions and system configurations, into pre-specified zone in the s-plane called D-shape sector. Hence, the problem of robustly tuning the PSSs parameters is formulated as a multiobjective optimization problem (MOP) with an eigenvalue-based objective functions. An improved version of non-dominated sorting genetic algorithms (NSGAII) is proposed to solve this MOP. The performance of the proposed NSGAII-based PSS (NSGAII-PSS) under different loading conditions, system configurations and disturbances is tested and examined for different multimachine power systems. Eigenvalue analysis and nonlinear simulations show the effectiveness and robustness of the proposed controllers NSGAII-PSSs and their ability to provide efficient damping of low-frequency oscillations. [ABSTRACT FROM AUTHOR]