Article Preview
TopIntroduction
Assembling is one of the important steps in the electronics industry and related industries. Therein, small parts such as the miniaturized gear, pins, chips, male connectors must be inserted into female connectors (Ansel et al., 2002; Dechev et al., 2004; S. J. Hu & Camelio, 2016). In actual production conditions in most factories, the assembly process depends heavily on human skills. This is the cause of low production efficiency, unstable product quality, not competitive in terms of product prices, etc. In some factories, this process is replaced by industrial robots based on pneumatic and hydraulic systems (C. Chen et al., 2016; Zhong et al., 2019). However, these robots have limitations such as generating noise during operation, gripping with highly complex parts, small size components, difficult to process and assemble. Especially, such robots result in undesired errors during fabrication and assembly. This leads to difficulty in gripping and releasing objects. To overcome such limits, the compliant mechanism is alternatively used with the advantage of being monolithic, small, easy to fabricate, no noise (Howell, 2011). A compliant gripper (CG) is considered as an effective solution that can replace traditional robots. However, research, development, and application of the CG in the production process are still very limited. Based on the survey results, it can be seen that only one CG application study for the assembly system is performed (Ho et al., 2019a). From the problems mentioned above and the limited number of studies that have been done, research and development of CGs to overcome the above difficulties is considered essentially.
Regarding the analysis, design, and optimization of a CG, there are many approaches to implementation. However, there are two commonly used approaches: black-box approach (Nguyen et al., 2019) and kinematics-based design methods (Dsouza et al., 2018; Nandy et al., 2018; Qingsong, 2015; Tashakori et al., 2018; Xu, 2015). With a kinematics-based design method, the mathematical equations need to be established first, an optimization algorithm is then used to predict a set of optimal parameters. Through the establishment of mathematical models, it provides knowledge as well as a clear kinetic meaning. Notwithstanding, the limitation of this approach is dependent on the knowledge of designers. Also, some previous studies have shown that there have large errors between predicted results and experimental results (Ai & Xu, 2014). In contrast to the kinematics-based design methods, to avoid factors such as expert knowledge, inadequate description of deformed components, black-box approach allows the establishment of a virtual mathematical model that describes the relationship between input signals and output response. In recent years, this approach is used by many interested scientists (Dang et al., 2019; Ho et al., 2019a; K. Muriithi et al., 2017).