Jarmo Partanen

Using fuzzy sets to model the uncertainty in the fault location process of distribution networks

In the computerized fault diagnosis of distribution networks the heuristic knowledge of the control center operators can be combined with the information obtained from the network database and SCADA system. However, the nature of the heuristic knowledge is inexact and uncertain. Also the information obtained from the remote control system contains uncertainty and may be incorrect, conflicting or inadequate. This paper proposes a method based on fuzzy set theory to deal with the uncertainty involved in the process of locating faults in distribution networks. The method is implemented in a prototype version of the distribution network operation support system. >

A modified dynamic programming algorithm for sizing, locating and timing of feeder reinforcements

The design of reinforcements for existing distribution feeders is a time-dependent problem. A new optimization algorithm for solving this problem is presented. The optimization method is based on modified dynamic programming, the voltage drop limit is taken into account during the optimization. The problem is not therefore a fixed net-problem, and the basic principle of dynamic programming has to be modified. Experience with the method indicates that the algorithm is valid and effective in practical design work. >

A practical approach for estimating future outage costs in power distribution networks

Quantitative reliability analysis is integrated as a part of a distribution network information and design system. The outage rates, outage durations and costs of outages are evaluated using the data included in the network database together with the reliability data for network components and the per unit values for the energy not supplied. Using the reliability calculation program together with other programs, optimum development plans for each network can be studied. In these studies, both the technical restrictions and the total costs, including the costs of investments, losses, and outages, must be included. >

The interface between a conventional database and an expert system application

Abstract Efficient fault locating and network reconfigurating in medium-voltage distribution networks call for flexible connection with the SCADA (system central and data acquisition) and network database systems. Using the information from these systems together with the heuristic knowledge of the control centre operator a high-level knowledge-based assistance system can be built for fault diagnosis of distribution networks. Such a system is being developed by the Power Engineering Group at Tampere University of Technology. This paper gives a detailed description of the dynamic frame-based model of the distribution network and outlines the use of the interface between the network database system and the expert system for fault diagnosis in a fault situation.