An Optimal Approach for Determining Microgrid State of Connection to Utility with Local Information

For correct performanceof protection and control systems in microgrids, the islanding conditions should be identified as soon as possible.In this paper, an optimal method is proposed to determine the input parameters of classification method by using local information. The goal of optimization is to minimizie the time and maximize the percision of microgrid connection state detection. Also, in classification method, three states, i.e. islanding, reconnecting to the utility and other events are considered. The objective function of this problem is defined as summation of weighted time factor and detection error. The inputs of the islanding detection problem can be a large number of network parameters such as frequency, voltage, current, power, their sequence components or the rate of change. Consequently, the variables of the classification problem are the number and type of input parameters of the classification problem.The genetic algorithm is used to solve this optimal problem, and support vector machine is used for the classification. In order to assess the validity of the proposed method, different situations are considered for the operation of the grid, and in each case, different events are modeled. The results are compared with the methods of other papers, and the advantage of the suggested method is shown.