Teaching Optimization Algorithm and Simulation Analysis based on Self-learning Mechanism and Multi-class Interaction

Abstract

Teaching optimization algorithm is an intelligent optimization algorithm applied in the education. At the same time, it can solve complex optimization problems in other fields such as traffic flow optimization and logistics optimization. In response to the weak development ability, a teaching optimization algorithm based on self-learning mechanism is proposed by referring to general reverse learning methods. Meanwhile, a multi class interactive teaching optimization algorithm is proposed by combining clustering and partitioning methods based on Euclidean distance. By combining the two algorithms, a personalized and collaborative learning teaching environment is provided. When the function dimension is 30, the average function evaluations for the teaching optimization algorithm based on self-learning mechanism on unimodal function f1 is only 3859. On the multimodal function f2, the average function evaluations for this algorithm are only 4735, which is 2057 and 1367 less than the other two algorithms, respectively. Meanwhile, the success rates of this algorithm are all 100%. In addition, on the unconstrained function f6, the multi class interactive teaching optimization algorithm tends to converge when the function evaluations are 0.1×104. Traditional teaching optimization algorithms tend to converge only at 1.0×104. The two improved algorithms proposed in the study have better solution accuracy and stability, providing a reliable method reference for solving modern complex engineering problems.