Application of Tree Structured Transfer Function to an electromechanical system, its Comparison with Conventional Method and Application for Order Reduction
Lubna Moin, Vali Uddin
Abstract
Lubna Moin, Vali Uddin
Abstract
This research paper is built on recent work in synthesizing and analyzing tree-structured transfer functions which have the form of a continued fraction. A useful feature of this transfer function is the manner in which it allows state variables to be easily removed from a model. Two components: a means to readily synthesize tree-structured transfer functions and the possibility of using the same transfer function to examine the effects on system performance of eliminating state variables provide the foundation to a novel physically-based approach of model order reduction for physical system. In the research work this approach is developed and compared to conventional techniques. The research work addresses the problem of determining which physical state variables should be retained in a reduced order model that belongs to a class of lumped physical system models, the class is defined by SISO variable pair of systems that are passive and free of internal loops. The models synthesized in the work are intended primarily for controller design, and are expressed using bond graphs. An error term that describes the reduced order model frequency response uncertainty over a given frequency range of interest is also provided
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
This research paper is built on recent work in synthesizing and analyzing tree-structured transfer functions which have the form of a continued fraction. A useful feature of this transfer function is the manner in which it allows state variables to be easily removed from a model. Two components: a means to readily synthesize tree-structured transfer functions and the possibility of using the same transfer function to examine the effects on system performance of eliminating state variables provide the foundation to a novel physically-based approach of model order reduction for physical system. In the research work this approach is developed and compared to conventional techniques. The research work addresses the problem of determining which physical state variables should be retained in a reduced order model that belongs to a class of lumped physical system models, the class is defined by SISO variable pair of systems that are passive and free of internal loops. The models synthesized in the work are intended primarily for controller design, and are expressed using bond graphs. An error term that describes the reduced order model frequency response uncertainty over a given frequency range of interest is also provided
Key concepts: Transfer function, Tree (set theory), Reduction (mathematics), Computer science, State variable, Physical system, Function (biology), Model order reduction