Improved genetic algorithm for multi-objective reactive power dispatch problem

This paper presents an improved genetic algorithm (GA) approach for solving the multi-objective reactive power dispatch problem. Loss minimization and maximization of voltage stability margin are taken as the objectives. Maximum L-index of the system is used to specify the voltage stability level. Generator terminal voltages, reactive power generation of capacitor banks and tap changing transformer setting are taken as the optimization variables. In the proposed GA, voltage magnitudes are represented as floating point numbers and transformer tap-setting and reactive power generation of capacitor bank are represented as integers. This alleviates the problems associated with conventional binary-coded GAS to deal with real variables and integer variables with total number of permissible choices not equal to 2(5). Crossover and mutation operators which can deal with mixed variables are proposed. The proposed method has been tested on IEEE 30-bus system and is compared with conventional methods and binary-coded GA. The proposed method has produced the loss which is less than the value reported earlier and is well suitable for solving the mixed integer optimization problem. Copyright (C) 2007 John Wiley & Sons, Ltd.