Tampere University of Technology

TUTCRIS Research Portal

Evaluation of Dead-Time Effect of Grid-Connected Inverters Using Broadband Methods

Research output: Chapter in Book/Report/Conference proceedingConference contributionScientificpeer-review


Original languageEnglish
Title of host publication2018 IFAC Symposium on System Identification
Number of pages6
Publication statusPublished - 2018
Publication typeA4 Article in a conference publication
EventIFAC Symposium on System Identification -
Duration: 9 Jul 201811 Jul 2018

Publication series

ISSN (Electronic)2405-8963


ConferenceIFAC Symposium on System Identification


Power electronic inverters are devices that are used to interface renewables, such as photovoltaic panels or wind turbines, with the electrical distribution system. The inverter should control its output current to follow a sinusoidal reference to avoid distorting the voltage waveform of the power system. However, in real inverters dead-time is used in the control signals which causes harmonics in inverter output current, thus, creating undesired power quality problems. The amount of distortion is difficult to predict due to nonlinear nature of the deadtime. Moreover, the harmonic distortion depends on many variables, such as control parameters of the inverter and impedance of the power system. This paper shows that the frequency response from inverter control signal (duty ratio) to grid current can be used to evaluate the effect of dead time on power quality. The frequency response is identified by perturbing the inverter control signal with a maximum length binary sequence (MLBS). The MLBS employs several advantageous properties, such ease of digital implementation and low peak factor. The magnitude of the identified frequency response is shown to follow the same trend as the amount of current distortion. The magnitude is affected by the power system and control parameters, which is in line with the observations made based on time-domain waveforms. Therefore, the effect of dead-time on power quality could potentially be modeled using a linearized model.

Publication forum classification