الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 136 |  |  |
| | | from | 136 | | |
Abstract
Asset management is now one of the most important issues in the electrical energy industry since it is one of the requirements of smart grid. The conventional way of forming the hazard rate function of transformers applies the best modeling based on the history data using Hazard Plotting Approach (HPA) under different functions: Normal Distribution, Lognormal Distribution, Weibull Distribution, and Smallest Extreme Value Distribution.
In this thesis, a new method is proposed to develop the best modeling using Artificial Neural Network- based polynomial model with minimum error to represent the hazard rate function of the power transformers. The procedure of applying the proposed methodology is simple. The quality of the obtained results ensures the adequacy of applying this methodology for expecting the failure age of the transformer.
The developed polynomial equation is used to estimate the Probability Density Function (PDF), and Cumulative Distribution Function (CDF). The Life Cycle Cost (LCC) is modeled in the form of expectation function based on the hazard rate and the expected penalty cost, which is a function of the Expected Energy Not Supply (EENS).
Three substation configurations have been assumed to check the validity of the proposed algorithm: two standalone transformers, two transformers with bus coupler, and three transformers with two bus couplers. The studied substation is assumed to supply a commercial load, an industrial load or a residential load, one at a time. The LCC is evaluated for the considered case studies after calculating the penalty cost corresponding to each case study which is calculated from the energy not served. The results obtained from the study propose the optimum age of renewing the transformers of the substations which yields the minimum value of LCC.