Genetic Algorithm Based Parameters Identification for Power Transformer Thermal Overload Protection
Recent studies by various authors have shown as the IEEE Transformer Loading Guide model and the more recent modified equations, proposed by the Working Group K3 of the IEEE “Power System Relaying Committee”, are lacking in accuracy in prediction the wind
- PDF / 851,537 Bytes
- 4 Pages / 595.276 x 790.866 pts Page_size
- 29 Downloads / 149 Views
Genetic Algorithm Based Parameters Identification for Power Transformer Thermal Overload Protection V. Galdi, L. Ippolito, A. Piccolo, A. Vaccaro * 'Department of Electronic & Electrical Engineering, University of Salerno, Fisciano (SA), 1-84084, Italy
Abstract Recent studies by various authors have shown as the IEEE Transformer Loading Guide model and the more recent modified equations, proposed by the Working Group K3 of the IEEE "Power System Relaying Committee", are lacking in accuracy in prediction the winding hottest spot temperature of a power transformer in presence of overload conditions. This is mainly due to the deviation of the parameters of the thermal model of the power transformer in presence of overload conditions. In the paper a novel technique to identify the thermal parameters to be used for the estimation of the hot spot temperature is presented. The proposed method is based on a Genetic Algorithm (GA) which, working on the load current and on the measured hot spot temperature pattern, pennits to identify a corrected set of parameters for the thermal model of the power transformer. Thanks to data obtained from experimental tests, the GA based method is tested to evaluate the performance of the proposed method in terms of accuracy.
1 Introduction Monitoring and protection of mineral-oil-filled power transformers are of critical importance in power systems, since they can cause widespread power outages of the distribution power systems. Today, protection of power transformer is of critical importance considering that the utilities, in order to increase system operation margins in presence of overload conditions, are compelled to adopt a new approach aimed to load power transformers beyond nameplate ratings, for a short or a long time. This approach, although it allows the exploitation of the full capacity of the transformer, requires an accurate monitoring of the transformer thermal state and in particular of the evolution of the hot spot temperature of the windings [1,2,3]. Undoubtedly a direct measurement of the windings hottest spot temperature can guarantee a very accurate transformer thermal monitoring, but it needs of a optical fibre based measurement station, which could be expensive for medium power electrical transformers. An alternative technique frequently adopted for the thermal transformer protection is based on the estimation of the windings hottest spot temperature by analytical models descriptive of the heating equations driving the internal heat flow exchanges. V. Krková et al. (eds.), Artificial Neural Nets and Genetic Algorithms © Springer-Verlag Wien 2001
In particular, amongst the possible approaches, the procedure reported in the IEEE load guide [4], with the improvement proposed in [5], seems to be handy and therefore particularly suitable for the implementation on a programmable unit. As recent studies have shown [3,6,7,8], the main limitation of this approach is that the model accuracy decays drastically in presence of overload conditions due to the deviation of the cha
Data Loading...
