Kozlov V. Synthesis of active power filter systems for electric load with quasi-stationary work regimes

Українська версія

Thesis for the degree of Candidate of Sciences (CSc)

State registration number

0416U002036

Applicant for

Specialization

  • 05.09.03 - Електротехнічні комплекси та системи

30-04-2016

Specialized Academic Board

Д. 08.080.07

Essay

In the thesis the actual problem of improving of the active power filters efficiency is solved. The main idea of modernization of active filter control system is to use the found extremum relation between changing of power inverter parameters and the process of power quality correction. The area of applying of modernized active power filter is network with slowly changing loads. The application of proposed extremum seeking control based on found physical relation can improve the quality of correction of electric energy flow and simplify the active power filter measurement system. The tasks which can be solved with help of active power filter based on proposed new control system are: reactive power compensation, THD level reduction, balancing of load, technological control of electric parameters. The thesis is also involves the questions of efficiency determining of active filters which work at industrial objects. The particular attention is given to using of electric meter systems for evaluation of correction process efficiency. The operability of active power filter with proposed new control system is estimated experimentally with help of physical modeling: it was conducted the experiment of reducing voltage unbalance (reduced from 8.8% to 3.2%). Also it was conducted the mathematical modeling of real industry object (excavator EKG5). It would be noted the next results: the reactive energy by cycle is reduced by 56 times, voltage THD level changes in borders from 5.5% to 21%, current THD level changes from 18% to 400% during most part of cycle. Digital control system of proposed active power filter is adopted for tasks of technological control of electric parameters.

Files

Similar theses