A state selection algorithm for the automated state model generator
Abstract
A powerful technique for the modeling and analysis of lumped-parameter power-electronic-based systems with changing topology has been previously developed. In this modeling approach, the state space representation of the overall system is generated automatically and updated dynamically based upon the topological state of the system. Although this method has been applied successfully to numerous power converter circuits, it fails to produce a state model for circuit topologies that include capacitive loops. In this thesis, a generalized state model generation algorithm is set forth that is applicable for all practically useful topologies of power electronic systems. In this method, a composite electrical network is viewed as inductive, capacitive, and algebraic subnetworks interconnected with each other. Decomposition of the global system into three elementary networks enables the identification of minimal sets of natural state variables; currents (fluxes) and voltages (charges), for the inductive and capacitive subsystems, respectively. A deterministic and robust state selection algorithm based on the theory of weighted graphs has been developed to enable network partitioning. The algorithm can also be used to ensure continuity of the state variables across topological boundaries. The network identification procedure for partitioning the global system has also been extended to a more general case of multiple networks. The algorithmically-generated state equations have a well-defined structure that is convenient for efficient computer implementation and modeling of networks with different types of parameters (constant, time-varying, linear, nonlinear, etc.) and switching characteristics.
Degree
Ph.D.
Advisors
Wasynczuk, Purdue University.
Subject Area
Electrical engineering|Computer science
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.