Application Software Of Monte-Carlo Approach In Grounding Grids Corrosion Diagnosis

An application software to evaluate the grouding grid corrosion diagnosis algorithms is developed on the platform of MATLAB. A Monte-Carlo based approach is applied to form the testing samples by randomly changing the branch resistances and the initial values of the diagnosis algorithms. The averaged value and its standard deviation of relative accumulated error of branch resistances are used as the indexes for evaluation. Five grounding grid corrosion diagnosis algorithms are compared, such as, the linear least square algorithm, the nonlinear least square algorithm, the iteration least square algorithm, the Tabu search based algorithm and the Tabu-simplex hybrid algorithm.An application software to evaluate the testability of the branches of a grounding grid under a certain test scheme is developed on the platform of MATLAB. A Monte-Carlo based approach is applied to form the testing samples by randomly changing the branch resistances.The testability identification is based on the statistic analysis of the simulated corrosion diagnosis results of the branches. The branches with their averaged value of relative accumulated error less than the given threshold are determined as clear branches, the other branches are uncertain-branches. An improved method based on a minimum spanning tree algorithm is introduced, with which, not only the correctness of testability evaluation is remarkably increased, but also the calculation of testability evaluation is obviously reduced. Some instances are analyzed to show that the developed software is correct.In accordance with the touchable nodes deviation problems, the application software to analyze the influences of touchable nodes deviation on the accuracy of diagnosis and the testability of grounding grids corrosion diagnosis is developed. The application software is also based on a Monte-Carlo approach and is developed on the platform of MATLAB.