The thesis addresses a new method of on-line condition monitoring and diagnostics for power converters. The proposed method requires an interleaved topology and it enables the determination of the branch in which additional power losses are generated due to degradation of components. First, the method is described, along with a detailed analytical analysis. Afterwards, analytical results are validated with a simulation model of a buck converter with realistic component models. Analytical expressions were derived in order to gain a better understanding of importance and cross-corelation of converter parameters and variables that determine the degradation of components. The effects of converter parameters on the measured variables representing degradation are further evaluated via graphical representation. The accuracy of analytical expressions was compared with the results of numerical simulations of a buck converter with realistic component models in Ansys Simplorer. The matching of obtained analytical and simulation results implies that the approximations, that have been made during analytical analysis, have little effect on the end result. This thesis shows that the proposed method can be used as an on-line condition monitoring tool, which does not require disconnecting the load from the power converter. Instead, the monitoring and diagnosis can be done without any interference to the nominal operation condition of the power converter. Consequently, there is no need to stop the industrial process, just to take the measurements for condition assesment.
|
