Nhảy đến nội dung

Hybrid Taguchi-cuckoo search algorithm for optimization of a compliant focus positioning platform


Thanh-Phong Dao*, Shyh-Chour Huang, and Pham Toan Thang

Source title: 
Applied Soft Computing, 57: 526-538, 2017 (ISI)
Academic year of acceptance: 

A long positioning range and a high first natural frequency are the two most important quality responses of a compliant focus positioning platform (CFPP). This paper aims to develop a hybrid Taguchi-cuckoo search (HTCS) algorithm to optimize overall the quality responses, simultaneously. The CFPP is designed via using flexure hinges. The length, width, and thickness of flexure hinges are considered as design variables. The Taguchi’s L16 orthogonal array is used to establish the experimental layout and the S/N ratios of each response are computed. The analysis of variance (ANOVA) is computed to investigate the effect of design parameters on the quality responses. Results of ANOVA are then utilized to limit the search space of design parameters that serves as initial population for the cuckoo search meta-heurist algorithm. The results showed that the HTCS algorithm is more effective than DE, GA, PSO, AEDE, and PSOGSA. The CFPP enables a long positioning range of 188.36 μm and a high frequency response of 284.06 Hz. The proposed HTCS approach can effectively optimize the multiple objectives for the CFPP and would be useful technique for related optimization problems.