Open Collections

Implementation of a full frequency-dependent transmission line model within the framework of the (OVNI) real-time power system simulator

University of British Columbia

This thesis presents the methodology to implement a full frequency-dependent transmission line model in UBC's OVNI (Object Virtual Network Integrator) real-time power system simulator. OVNI utilizes an object-oriented approach to represent and implement its element models. Object-oriented programming permits a flexible, reliable, and expandable solution to the simulation program. In particular, OVNI represents each element model as a class inheriting common characteristics and properties from a parent class of basic elements. Each element object interacts with the network solver, the core, only by exchanging such parameters as external node voltages, external node names, the equivalent conductance matrix and the equivalent history source vector through some well-defined member functions. Model developers only needs to correctly and efficiently implement these member functions to successfully incorporate the model with the core. This clean-cut abstraction between element models and the core provides model developers with complete freedom and independence while designing, implementing, and upgrading models of different electric properties and complexities. The transmission line model implemented in this thesis is a phase-domain full frequency-dependent model (zLine). This model was developed in Ph.D projects by Castellanos [15] and Yu [16]. The model is suitable for time-domain EMTP (Electromagnetic Transient Program) simulations within OVNFs real-time framework. Z-line is accurate, efficient, numerically stable and strongly suited for multi-circuit asymmetrical line configurations. The model divides the line length into a number of small segments and separates the wave propagating in each segment into a constant idealline section and a frequency-dependent loss section. A numerically stable curve fitting routine was modified to synthesize elements of the loss matrix so that the equivalent time-domain model for the loss section can be formulated with an integration rule. The implementation of zLine in this thesis work is divided into three major tasks: preprocessing and initialization of model parameters, computation of equivalent conductance matrix, and update of equivalent history sources. The pre-processing task involves three subtasks. First of all, it requires the execution of mtLine, a program that generates line parameter matrices from geometrical conductor configurations. Moreover, it requires the execution of a modified fitting routine that synthesizes the frequency-dependent loss matrix with a series of rational functions. Finally a series of I/O routines are required to organize and generate the input files necessary to run the programs mentioned above. An ANSI C compatible function is chosen to integrate the execution of all those routines under the control of the simulator. The update of history sources and computation of conductance matrices take advantage of the abstraction offered by object-oriented programming between the line and the modelling segments. Each history source update and matrix computation is performed first at the segment level for each zLine segment and then accumulated at the line level to form the overall history vector and conductance matrix of the line. Internal nodes are hidden away by applying the node-hiding technique to reduce network complexity and further streamline solution efficiency. A comprehensive set of test cases are run and compared with proven results and exact models in Microtran to ensure the correctness and accuracy of the implementation methodology. zLine's accuracy and absolute numerical stability for any asymmetrical line configurations render it an ideal candidate as the first frequency-dependent line model to be included with OVNI.

Thesis/Dissertation

application/pdf

2009-12-03

For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use.

http://hdl.handle.net/2429/16252

Electrical and Computer Engineering

University of British Columbia

UBCV

DSpace