Comparative Analysis of Backtrack Search Optimization Algorithm (BSA) with other Evolutionary Algorithms for Global Continuous Optimization.

Abstract- In the real world scenario we come across the problem of optimization a number of times. Finding the best solution among the available set of solutions becomes mandatory. A number of numerical techniques are already present in literature which aims at optimizing the result however, they are not feasible to be used in each type of problem. Hence we are tending towards evolutionary algorithms which are more powerful tools to fetch the best results without using any set formulae. A Number of algorithms are already available in literature however they have a problem of getting stuck in local minima or their time of convergence is too high. In this paper I have implemented Backtracking Search Optimization Algorithm (BSA). BSA uses two set of populations i.e. old and new which prevents it from getting stuck into local minima. Its selection, crossover and mutation processes are different from the other methods and it yields the most optimized solution in lesser time. The claim is supported by the results of its comparison with different techniques and BSA is proved to give better results and in lesser time