No abstract

The efficient protection of electric power distribution networks against lightning discharges is a crucial problem for distribution electric utilities. To solve this problem, the great challenge is to find a solution for the installation of surge arresters at specific points in the electrical grid and in a sufficient quantity that can ensure an adequate level of equipment protection and be within the utility’s budget. As a solution to this problem of using ATP (Alternative Transient Program), this paper presents a methodology for optimized surge arrester allocation based on genetic algorithm (GA), with a fitness function that maximizes the number of protected equipment according to the financial availability for investment in surge arresters. As ATP may demand too much processing time when running large distribution grids, an innovative procedure is implemented to obtain an overvoltage severity description of the grid and select only the most critical electric nodes for the incidence of lightning discharges, in the GA allocation procedure. The results obtained for the IEEE-123 bus electric feeder indicate a great reduction of flashover occurrence, thus increasing the equipment protection level.

No abstract

Data analysis plays an important role in our InformationEra; however, most of statistical and machine learning algorithms were not developed to tackle the ubiquitous issue of missing values. In pattern classification, several strategies have been proposed to handle this problem, where missing data imputation is the most used one, which can be viewed as an optimization problem where the goal is to reduce the bias imposed by the absence of information. Although most imputation methods are restricted to one type of variable only (categorical or numerical), they usually ignore information within incomplete instances. To fill these gaps, we propose an evolutionary missing data imputation method for pattern classification, based on a genetic algorithm, which is suitable for mixed-attribute datasets and takes into account information from incomplete instances and model building -more specifically, the classification accuracy. To assess the performance of our method, we used three algorithms in order to represent the three groups of classification methods: 1) rule induction learning, 2) approximate models and 3) lazy learning. Experiments have shown that the proposed method outperforms some well-established missing value treatment methods.

No abstract