ZIVARI-TLBO boosts optimization with zero-cost elite relay mechanism

ZIVARI-TLBO, introduced by Pezhman Zivari, is a zero-cost modification to the Teaching-Learning-Based Optimization (TLBO) algorithm. The core idea is a fixed ring relay: at scheduled events, each group in the population passes its best (elite) member to the next group. Since the elite has already been evaluated in the current generation, the relay copies its solution and stored fitness, incurring no additional objective-function calls. The recipient group only replaces its worst eligible learner if the incoming elite has a better fitness. This budget-consistent information sharing improves convergence without extra computational cost.

In extensive experiments on eight classical benchmark functions (dimensions 10, 30, 50, and 100) with 30 independent runs each, ZIVARI-TLBO recorded 728 wins, 11 ties, and 221 losses against the baseline grouped TLBO, with a rank-biserial effect size of 0.624. In a broader eight-method comparison (including WOA, DE, PSO, GWO, HHO, MCTLBO, and standard TLBO), ZIVARI-TLBO placed second overall (average rank 3.382), only behind Whale Optimization Algorithm (WOA, rank 2.914). It significantly outperformed TLBO, MCTLBO, DE, PSO, and GWO after Holm adjustment, but lost significantly to WOA and was statistically tied with HHO. However, results on five constrained engineering problems were mixed and sensitive to the static-penalty formulation used, so the method does not claim universal state-of-the-art performance on engineering problems or CEC benchmarks.