LOAD FLOW STUDIES USING PARALLEL PROCESSORS

Abstract

The work presented in this dissertation deals with application of sparsity techniques to load flow studies. As the size of the system increases, it becomes increasingly difficult to accomodate the problem within the core memory of the computer. Row Wise packing has been used to store admittance and Jacobian matrix. The aim was to reduce the execution time. Thus this technique finds application for on-line studies. A full chapter is included in this work for sparsity techniques. Chapter-IV on parallel processing deals with implementation algorithm. A- test matrix has been taken to show the validity of algorithm. The various problems involved in the use of the parallel processor has been shown. Decoupled N-R method was used to carry out load flow solution. Due to the some reasons the same algorithm could not be used to solve power flow equations. This technique is fully sparse oriented. The coputational work has been carried out on "TATA ELXSI Power Series-3220" Computer System of the Roorkee University.