A hybrid teaching-learning slime mould algorithm for global optimization and reliability-based design optimization problems

作者：Zhong, Changting; Li, Gang^*; Meng, Zeng^*

来源：NEURAL COMPUTING & APPLICATIONS, 2022, 34(19): 16617-16642.

DOI：10.1007/s00521-022-07277-3

• Reliability-based design optimization
• Teaching-learning-based optimization
• Slime mould algorithm
• Adaptive chaos control method

Slime mould algorithm (SMA) is a novel metaheuristic algorithm with good performance for optimization problems, but it may encounter premature or low accuracy in complex optimization problems. This paper presents a hybrid SMA using teaching-learning based optimization (TLBO), called TLSMA, for solving global optimization and reliability-based design optimization (RBDO) problems. The key point of TLSMA is to combine the capacities of exploration and exploitation from SMA and TLBO, which can enhance the convergence ability of SMA. Moreover, the TLSMA is extended for solving RBDO problems under probabilistic constraints, which are handled by the adaptive chaos control method. The proposed algorithm is tested by a series of experiments with two parts. First, TLSMA is verified by 24 well-known benchmark optimization problems with unimodal and multimodal functions, and is compared with several state-of-the-art metaheuristic algorithms. The results of benchmark optimization problems show that TLSMA outperforms PSO, BBO, GWO, WOA, SSA, TLBO and SMA. Then, TLSMA-RBDO is tested by five RBDO problems, including a numerical and four engineering problems. The results illustrate that the proposed algorithm has high performance in the RBDO problems, which is significantly superior to the compared algorithms.