OBJECT-ORIENTED MODELING AND IMPLEMENTATION FOR STEADY STATE ANALYSIS OF POWER SYSTEMS

A THESIS

to be submitted by

M. P. SELVAN

for the award of the degree

of

DOCTOR OF PHILOSOPHY

DEPARTMENT OF ELECTRICAL ENGINEERING INDIAN INSTITUTE OF TECHNOLOGY, MADRAS NOVEMBER 2005

1

1. INTRODUCTION Digital simulation of power system gathers importance in various engineering studies for power system design, planning and training, ranging from a simple load flow to complex contingency, security and transient stability analysis (Pandit et al., 2000). Power system computation programs were traditionally developed using function-oriented methodology and implemented in Fortran with its associated advantages and limitations. Due to escalating power demand and expanding power system network, the operation and control of contemporary power system become more complex and cumbersome. The requirements on the software are increasing day by day and the application software being used in energy management system (EMS) and distribution management system (DMS) has to be upgraded to meet the requirements. Simulation and operation of the deregulated power system requires sophisticated software tools. Hence, the power industry computer applications are undergoing a revolution to significant phenomena such as the replacement of mainframe computer solutions with networked microcomputers and workstations, as well as the movement toward object-oriented (OO) software development methodologies. Modern control centers make use of the open distributed systems with client server architecture to distribute their functions among different computers (Dy-Liacco, 1994). REAL WORLD MODEL MATHEMATICAL MODEL SOFTWARE MODEL PROGRAMMING IMPLEMENTATION STEADY STATE ANALYSIS & TIME DOMAIN SIMULATION

Fig. 1 Simulation software development procedure. Fig.1 shows the steps involved in the development of steady-state analysis and time domain simulation software. The development process begins with the real world model, which is the physical system. The behavior of the physical system is mathematically modeled using linear and/or nonlinear, algebraic and/or differential equations. The mathematical equations and the data corresponding to the physical components are translated into the software model, which describes the data specifications on which all the operations are performed. Software modeling plays a significant role in representing the mathematical model of the system into the computer memory. The developed software model is then implemented in any programming language to obtain a full-fledged simulation module. 2

In practice, developing a better software model, which supports various analyses of the problem domain, is a challenging task. In traditional function-oriented programming, where the problem is decomposed into various functions performing a particular task, the behavior of the system is completely decoupled from the characteristics or attributes of the system. The data of the problem domain is kept global and is passed as arguments into the functions for performing desired computations. The relationships between different physical components are not replicated in the software. This reduces the ‘one to one’ matching between the physical system and the software model. Consequently, the software model turns out to be very difficult to realize when the physical system becomes more complex. This is where the object-oriented modeling can provide a better solution. Object-oriented modeling decomposes the problem domain into various objects and performs a particular task by the interaction between objects (Booch, 1994; Ambler 2001). Researchers exploited the advantages of object-oriented methodology (OOM) to address the issues involved in the development of power system analysis applications since early ’90s (Neyer et al., 1990; Zhou, 1996).

2. MOTIVATION

Over the last decade, a large number of power system problems,...