A novel methodology for maximization of Reactive Reserve using Teaching-Learning-Based Algorithm

Abstract

This paper presents a new algorithm to optimize the reactive reserve. As the amount of reactive reserves at generating station is a measure of voltage stability. A new approach for scheduling of reactive power control variables for voltage stability enhancement using teaching-learning-based optimization (TLBO) has been developed. An objective function selected for maximization of reactive reserve for maintaining the voltage stability. A Sensitivity inequality relation analysis based proximity indicator has been selected for obtaining desired stability margin whose value along with reactive power reserve maximization assures desired static voltage stability margin. TLBO has been selected because this is an efficient optimization method for large scale non linear optimization problems for finding global solutions. It is basically based on the influence of a teacher on an output of learners in a class. Developed algorithm has been implemented on 6-bus standard test systems. The problem of reactive power optimization has played an important role in the reliable and optimum operation of power system. Reactive power reserve optimization (RPRO) has complex and non-linear characteristic with large number of equality and inequality constraints (1). The reactive power optimization problem is a nonlinear combinational and computational optimization problem and during last two decades there are many efforts has been devoted to the development are done by using different mathematical methods known as optimization techniques for solving the reactive power optimization problems. Primarily the conventional optimization techniques such as linear programming and non linear programming are in practice with the advantage of computational speed and convergence with the objective function of continuous and differentiable value. These are so named as conventional optimization techniques because they cannot handle the large and discrete-continuous problems such as reactive power optimization. So recently, computational intelligence based optimization techniques are in practice and have been proposed in the application of reactive power optimization such as genetic algorithm (GA), Tabu search, simulation annealing, particle swarm optimization(PSO), differential evolution(DE) and most recent one is teaching- learning-based-optimization (TLBO).these all are consider as most practical, user friendly and powerful scheme to obtain the global optimum solution for different optimization problems(2).