Survey of modelling techniques used in optimisation of power electronic components

Abstract

Design optimisation techniques for power electronic converters have been the subject of numerous research studies for the past 15 years. Accurate modelling is the most important task in the optimisation, and, because the nature of optimisation problem demands evaluating an objective function numerous times, fast and simplified modelling techniques are required. The objective of this study is to categorise and analyse the previous studies related to modelling techniques incorporated in design optimisation of power electronic converters. Common trends in modelling and optimisation of power electronic converters are analysed and suggestions for future research in this field will be presented.

Design methodology for optimising a high insulation voltage insulated gate bipolar transistor gate driver signal transmission function

“…As described in [8], the optimisation process considers a mixed integer programming problem: some variables being continuous and other discrete (n 1 and n 2 ). In the latest references, many applications in the power electronics fields are addressed considering optimisation processes [12][13][14][15][16][17][18][19]. In the case of mixed integer programming problem, GAs are adapted as it was shown in [8,12,13].…”

Fundamental Operation of Marx Topology for High Boost Ratio DC-DC Converter

“…The heatsinks volume estimation for the switching devices and diodes in the proposed 3-stage MTBC and the ICRS are based on the semiconductor devices power loss and the cooling system performance index (CSPI) (23)- (25) . Besides, the simulation results of the conduction and switching losses for the both converters are considered for the heatsinks volume estimation.…”

“…On the other hand, the proposed 3-stage MTBC has three stage capacitors; meanwhile the ICSR has no stage capacitor. Fundamentally, between capacitor volume and stored energy in a capacitor has a linear relationship (23)- (25) . Thus the capacitor volume can be estimated based on stored energy in a capacitor.…”

Exploiting cloud computing in the multi-physics design and optimisation of electromagnetic devices

“…2) Winding Loss: The temperature and frequency dependent ac winding loss is determined from the ratio of effective ac to dc resistance, R ac /R dc , [19], [20]. The operating temperature of the inductor is emulated by scaling the resistivity, ρ, of the winding material with the well established linearised equation, (2), where T op , T 0 and α are the operating temperature, reference temperature and temperature coefficient of resistance respectively.…”

“…Table I details the set of electromagnetic FE analyses, along with the appropriate winding current, frequency and temperate cases, required to characterise the electromagnetic and loss behaviour over the temperature range of interest at a fixed frequency, f 0 , Sections II-A1 to II-A3. The electromagnetic model assumes pure sinusoidal excitation current, however, non-sinusoidal current can be approximated by decomposing the waveform into its Fourier components and evaluating the FE model at each significant harmonic along with the dc average and superimposing the result, [20], [25]. Hence, the number of FE analyses required, N EM , to characterise the electromagnetic and loss behaviour is given by, (6), where N h is the number of significant harmonics.…”