Domagoj Peharda

Short term hourly forecasting of gas consumption using neural networks

This paper presents a neural network based model for forecasting gas consumption for residential and commercial consumers. A feedforward neural network with sigmoid nodes and one hidden layer was trained by backpropagation. The model was validated on real data from a distribution area covering 7% of the total consumption in Croatia, consisting mostly of residential and commercial consumers.

Forecasting data for load flow

In order to achieve a more secure and economic operation of a power system its state should be known not only in the present but some time ahead. Transmission system operators do not have the same security margins like they had in the past. Most of the tools of security assessment operate on-line and provide valuable input to the system operators. Security assessment uses load flow calculation to provide results. If it could be possible to compute load flow several hours ahead it would also be possible to provide security assessment. In this paper a load flow calculation several hours in the future is discussed so developed and tested on-line methods for security assessment can be utilized.

Teaching Power System Modeling in an Object Oriented Environment

Electrical power system engineers often face a challenge of modeling power system components and performing complex calculations. This paper describes a teaching/learning process of modeling power systems networks and performing network calculations in an object-oriented environment.

Power system topology assessment and pre-estimation in an object oriented environment

Electrical power system engineers often face a challenge of software implementation of heuristic knowledge about power systems. This knowledge can be structured in a rule form. This is different from analysis and calculations knowledge which usually comes in form of algorithms and formulae. Object oriented environment enables power system engineers to transparently implement rules in objects, thus allowing more heuristic knowledge to be implemented without the use of complex models or algorithms. Technology has simplified the coding paradigm, thus allowing electrical engineers with little experience in programming to work on the development of the software.

Power Flows and Transmission Losses in the Croatian Network after the UCTE Reconnection

Consideration of losses of electric energy in the transmission network is important for power system operation and control. Development of the remote control system in the Croatian National Control Centre enabled an accurate calculation of losses and their comparison with the results measured in the terminal nodes. Additional value in proposed way of determination of losses lies in the fact that losses can be allocated by voltage level, lines and transformers. A new way of calculating the percentage of losses is proposed by the authors