Optimum fixture locating layout for sheet metal part by integrating kriging with cuckoo search algorithm

• Published in: International journal of advanced manufacturing technology Vol. 91; no. 1-4; pp. 327 - 340
• Main Authors: Yang, Bo, Wang, Zhongqi, Yang, Yuan, Kang, Yonggang, Li, Xining
• Format: Journal Article
• Language: English
• Published: London Springer London 01.07.2017; Springer Nature B.V
• Subjects: Back propagation networks; CAE) and Design; Computer-Aided Engineering (CAD; Computing costs; Cost analysis; Deformation; Design optimization; Engineering; Finite element analysis; Finite element method; Hypercubes; Industrial and Production Engineering; Kriging; Latin hypercube sampling; Layouts; Mathematical analysis; Mechanical Engineering; Mechanical properties; Media Management; Metal sheets; Neural networks; Optimization; Original Article; Search algorithms; Training; Workpieces; Kriging surrogate model; Cuckoo search algorithm; Back propagation neural network; Fixture layout; Sheet metal
• ISSN: 0268-3768; 1433-3015
• DOI: 10.1007/s00170-016-9638-5

Fixture locating layout has a direct and influential impact on the sheet metal mechanical behavior and dimensional quality during the manufacturing process. The N-2-1 locating principle is adopted to design the fixture locating layout for the sheet metal part to determine the spatial location and restrain the excessive deformation. However, efficient optimal design of fixture layout is not an easy-to-implement and trivial task. The state-of-the-art evolutionary optimization of fixture layout aiming for workpiece deformation control often involves hundreds or even thousands of calls of finite element analysis and therefore is faced with uncomfortable and challenging computation cost and burden. In order to reduce the computational cost and improve the optimization efficiency, a new approach for optimum sheet metal fixture locating layout based on the N-2-1 principle is proposed in this paper. The training and test data sets are generated by running only a few times of finite element analysis on the design sites standing for different fixture locating layouts selected through Latin hypercube sampling. The kriging surrogate model is built based on the training sample set to approximate the implicit function relationship between the fixture locating layout and the concerned sheet metal deformation and meanwhile is compared with back propagation neural network in terms of prediction accuracy by the test sample set. The cuckoo search algorithm is applied to the kriging model to find the optimal fixture locating layout. Flat and curved sheet metal cases based on the “4-2-1” locating scheme are conducted, and the results indicate that the proposed approach is effective and efficient in the design and optimization of the sheet metal fixture locating layout.