Open AccessCalibration of erroneous branch parameters utilising learning automata theory
Author(s) -
Helmzadeh Amin,
Kouhsari Shahram Montaser
Publication year - 2016
Publication title -
iet generation, transmission and distribution
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.92
H-Index - 110
eISSN - 1751-8695
pISSN - 1751-8687
DOI - 10.1049/iet-gtd.2015.1046
Subject(s) - computer science , dependency (uml) , sensitivity (control systems) , algorithm , probabilistic logic , function (biology) , calibration , selection (genetic algorithm) , sign (mathematics) , probabilistic automaton , limit (mathematics) , mathematical optimization , mathematics , artificial intelligence , statistics , engineering , mathematical analysis , evolutionary biology , electronic engineering , biology
Incorrectness of simulation model parameters can lead to erroneous results in power system operation and planning studies. Therefore, model parameters should be verified according to measurements obtained from the actual system. Mismatches between simulation results and corresponding field measurements can be considered as a sign for the necessity of model verification. Inaccurate branch parameters in simulation models can lead to misguided results in protection, operation and planning studies. An efficient algorithm is proposed in this study for detection and correction of erroneous branch parameters. First, suspicious parameters are selected utilising a sensitivity‐based probabilistic method. Applying learning automata theory, selection probabilities will be modified according to simulation responses. A novel formulation is proposed for calculating selection probabilities according to simulation responses. Afterwards, selected parameters are modified by utilising an iterative Newton–Raphson scheme. Unlike state estimation‐based methods, required sensitivities are calculated numerically. Therefore, the explicit mathematical formulation of measurements dependency on model parameters is unnecessary. The procedure of erroneous parameter detection and correction will be continued until the value of sum of squared errors (SSE), the objective function in optimisation procedure, becomes sufficiently small. Simulation results demonstrate the high efficiency of the proposed method.