A Novel Tree-Seed Teaching Learning based optimization Algorithm for constrained optimization problems

Nature-inspired algorithms have proven invaluable in addressing complex, non-linear and non-convex optimization problems. Combining two or more strategies can improve accuracy, adaptability, and efficiency in tackling complex problems while efficiently dealing with uncertainties and nonlinearity. In this research, we present a hybrid approach known as Tree-Seed Teaching Learning based Optimization algorithm (TSTLBO), in which the teaching phase of Teaching Learning based Optimization (TLBO) algorithm is modified by the seeds generation approach of the Tree seed algorithm (TSA) to increase problem-solving skills. TLBO's exploration capabilities are limited. Keeping this in mind, we combined TLBO and the TSA algorithm to take advantage of both in a single variant, resulting in improved performance when addressing complex optimization problems. The proposed hybrid algorithm has better capability to escape from local optima with faster convergence than the standard TLBO and TSA. The performance of the proposed hybrid approach was evaluated against 30 IEEE CEC-2017 (IEEE Congress on Evolutionary Computation 2017) benchmark functions of varied complexity and structural design problems. The computed results are compared to numerous existing state-of-the-art studies. Experimental findings reveal that the proposed TSTLBO outperforms other existing algorithms. Among the 30 benchmark functions, TSTLBO achieves superior results on 25 and gives same values for 2 functions for dimension 10, whereas for dimension 30, the proposed algorithm performs better on 18 and gives equal values for 4 functions. Furthermore, the TSTLBO results are evaluated using the statistical analysis test. Among four real-life problems, TSTLBO ranked first.