Black Box Modeling of Power Electronic Converters

In Distributed Energy System (DES), different converters are interconnected. For converters to be interconnected, behavior of individual converter should be known to avoid instabilities. Now to find the behavior of the converter, either we can go for white box modeling in which the internal details

Project Title

Black Box Modeling of Power Electronic Converters

Project Area of Specialization

Electrical/Electronic Engineering

Project Summary

In Distributed Energy System (DES), different converters are interconnected. For converters to be interconnected, behavior of individual converter should be known to avoid instabilities. Now to find the behavior of the converter, either we can go for white box modeling in which the internal details of the converter are given which is not the case for commercial converters as the data is confidential. The second method for finding the behavior of the converter is Black Box modeling in which no prior data will be required for finding the behavior of the converter except access to input and output terminals of the converter. With access to input and output terminals we find certain parameters which are used to develop Black box model of the converter. Four parameters are required for DC-DC converter, nine parameters are required for AC-DC/DC-AC converter and sixteen parameters are required for AC-AC converter. For all converters, input voltage and output current are the input parameters while output voltage and input current are the output parameters. We apply step at input parameters and observe response of output parameters i.e. output voltage and input current.

There are two different techniques with the help of which black box model is developed; one is time domain measurements and the other one is frequency domain measurements. We are using time domain measurements in which step is applied at input parameters at a certain time to observe the response of output parameters.

Project Objectives

The conventional way of generation, transmission and distribution of electricity is strongly affected by increasing demand of electric power due to technological development. In order to increase the efficiency and reliability of the system, certain revisions are required to be made to it. In order to design more reliable, economic and environment friendly electric power distribution systems for certain future applications i.e. more electric aircraft, hybrid electric vehicles, all electric ships, commercial and residential services etc. readymade power modules are available. To integrate solar system and wind system into the existing grid, commercial modules are used but the problem is that we cannot interconnect different converters/systems without knowing their behavior and in most cases, manufacturers do not provide internal details of the converter.

The objective here is to adopt such modeling techniques which will not use any prior information about internal details of the converter known as black box modeling. Black box modeling is based on certain parameters which will be measured with input-output terminals only, for finding the behavior of the converter and will not require any prior information about converter internal structure.

Project Implementation Method

To interconnect different converters, behavior of each individual converter should be known to avoid instabilities. It is difficult to have the behavior of converter when internal details of the converter are not given while in case of commercial converters manufacturer do not provide internal details of the converter. Black box modeling is a technique with the help of which behavior of the converter is determined with access to input-output terminals only, without having internal details of the converter. For implementation of any circuit, first step is to simulate that circuit and then go for a prototype.

For DC-DC converter, g-parameter model is developed. It is also known as two port model. When step is applied at output current, we measure output voltage and input current known as Output Impedance (Zo=Vo/Io) and Back Current Gain (Hi=Iin/Io) respectively. When we apply step at input voltage, output voltage and input current is again measured known as Audio-Susceptibility (Go=Vo/Vin) and Input Admittance (Yi=Iin/Vin) respectively.

Verification is a step in which the same step is applied to actual model and Black box model to see how closely they are matching. Validation is another step in which different step is applied to see the response of actual and black box model. We have performed the above steps for buck converter, boost converter and then we have moved to interconnected converters. We design black box model for a specific operating point. When step exceeds those limits then response of the black box model and actual converter does not match. Then we go for non-linear behavior of the converter. Furthermore, a decoupling procedure is also followed to remove the effect of source and load from the model which is known as Unterminated response.

Currently, we are working on Distributed Energy System (DES). We are having three sources; one is from conventional grid which is converted to DC, second source is PV and the third one is battery. We are using various loads. We are using MATLAB software for Black Box Modeling of Power Electronic Converters and System Identification Toolbox for measurement for g-parameters as transfer functions. Transfer functions are measured from the actual model.

Benefits of the Project

Design Engineers are facing problems in interconnecting different converters when prior data is not available. For converters to be interconnected, behavior of each converter should be known. The methodology we are using will help design Engineers to interconnect various converters even if prior data is not available. Later, it can be extended to systems instead of converters. Currently, we are working on the behavior of individual converters, later we can have a black box model of a complete system.

Internal details of converter are not required for converter response and we can have behavior of a converter with access to input and output terminals only.

System stability can be observed under Nyquist stability criteria which can be determined in case of interconnected converters. It can be determined with the help of Output Impedance (Zo) of one converter and Input Admittance (Yi) of another converter. And we have already measured these transfer functions for Black box modeling.

Another benefit is that no need to remove converter from the system as we can have the behavior of the converter by removing the effect of load and source from the model known as decoupling procedure (Unterminated Response).

Technical Details of Final Deliverable

As discussed earlier, to interconnect different converters, behavior of each individual converter should be known. To find behavior of a converter internal data should be available while in case of commercial converters manufacturers do not provide internal details, so it is difficult for design Engineers to interconnect different converters without having their behavior.

Black box modeling is a technique with the help of which converter behavior can be determined while having access to input and output terminals only, without having any internal data of the converter

==> Black box model with the help of which response of the converter can be observed.

==> Black box model of Distributed Energy System (DES).

Final Deliverable of the Project

HW/SW integrated system

Core Industry

Energy

Other Industries

Energy

Core Technology

Clean Tech

Other Technologies

Clean Tech

Sustainable Development Goals

Affordable and Clean Energy

Required Resources

If you need this project, please contact me on contact@adikhanofficial.com