Application of genetic algorithm for the improved performance of boost converters Kumaraguru Prabakar School of electrical, computer and energy engineering The University of Tennessee Knoxville TN 37916, USA kprabaka@utk.edu Fangxing Li School of electrical, computer, and energy engineering The University of Tennessee Knoxville TN 37916, USA fli6@utk.edu Abstract—DC-DC Boost converters play a major role in DC-DC and DC-AC conversion applications like electric vehicle, integration of photovoltaic (PV) and distributed energy resources into the grid. When these converters operate they might introduce electromagnetic interference (EMI) and harmonics into the system. In order to avoid this, passive filters are prominently used. The methodology proposed here makes use of evolutionary computing algorithm in order to generate an optimized filter design for a boost converter. The proposed generic filter design procedure involves three steps, first selecting a suita-ble filter, second identifying the transfer functions involved, and finally using evolutionary algorithms to optimize the selected filter design. This methodology was implemented for LC filter design for a DC-DC boost converter using genetic algorithm. Index Terms—Genetic algorithm, Boost converter, power electronics, LC filters, filter design optimization. I. INTRODUCTION In the past recent years, there has been significant development in the high speed power electronics control and power electronics switches which can switch at a higher speed[1]. In spite of these developments, switching nature of the electromagnetic energy processing in power electronics generates electromagnetic interference (EMI) [2]. Electromagnetic disturbances thus caused travel by conduction on wiring, radiation in space, and inter-circuit capacitive or inductive coupling[3]. The mode of EMI travel typically depends on the switching frequency. EMI spreads through conduction for switching frequencies less than 10 MHz, while EMI spreads through radiation for higher switching frequencies. It becomes a necessity that a filter needs to be added to the circuit, in order to smooth the interference. The work presented here introduces an evolutionary algorithm based approach towards the filter design. II. FILTER DESIGN FOR POWER ELECTRONIC CONVERTERS Power electronic converters also tend to generate current- related interference at their input, injecting noise to the power grid, and voltage-related interference at their outputs, which may disturb operation of communication and control systems in proximity to the converter. In order to avoid such interferences active and passive filters are used[4]. Active filters are dynamic and adjustable solutions to power quality problems. These filters are able to compensate current and voltage harmonics, reactive power, and to improve voltage balance in three-phase systems. Active filters are developed with pulse width modulated converters. Based on the topology of the converters, typical active power filters available are shunt, series, or hybrid. Due to the complexity involved in the operation of active filters, passive filters such as filters with LCR has been traditionally used for mitigation of harmonic distortion. The design procedure proposed here is for passive filters in DC-DC boost converters. A. Low-pass LC filters topologies Low-pass filters are common in EMI mitigation. They are often called as radio-frequency (RF) filters. These filters are typically installed on the input and output sides of power converters. Passive filters are usually capacitive or inductive- capacitive circuits. Basic topologies of low-pass filters are indicated in Figure 1. Passive EMI filters are preferred than the active filters instead of their effective filtering due to passive filters highly reduced complexity and cost. The inverted Gamma LC filter topology is considered here for the LC filter design methodology. When these filters are used with the converters, the input impedance of the converter varies significantly and this causes difficulty in optimizing the EMI filter design. The proposed methodology makes use of this impedance variation and tries to find an optimized value for the inverted-gamma LC filter. This work made use of Engineering Research Center shared facilities supported by the Engineering Research Center Program of the National Science Foundation and the Department of Energy under NSF Award Number EEC-1041877 and the CURENT Industry Partnership Program