Comparing a Parameterless Technique Jaya with Parameters-Based Evolutionary Algorithms

Every day a new meta-heuristic technique, especially evolutionary algorithms (EAs), is added to the pool of existing optimization techniques. This emerging pool of EAs only increases the dilemma about the selection of the right EA for a design problem. One of the solutions for this problem can be comparing various EAs on similar grounds to find the best-suited EA for a design situation. The present work addresses the same by depicting a detailed comparison of Jaya, a new parameterless optimization technique with seven other EAs. Two well-known benchmark networks are considered to test the efficiency of Jaya over other techniques. The predefined stopping criteria of 1000, 10,000, and 40,000 function evaluations are used for each benchmark network. Statistical analysis performed establishes the superiority of Jaya over other techniques. Quality checks for the optimal results obtained from Jaya for minimum pressure required constraint are also presented as part of this work.