Nhảy đến nội dung

A computational design of robotic grasper by intelligence-based topology optimization for microassembly and micromanipulation


Ngoc Thoai Tran, Minh Phung Dang, Alokesh Pramanik, Animesh Basak, S. Shankar, Dharam Buddhi, Thanh-Phong Dao

Source title: 
Robotics and Autonomous Systems, 156: 104209, 2022 (ISI)
Academic year of acceptance: 

In robotic microassembly and micromanipulation, robotic grasper plays a vital role in picking up and releasing the product. However, the design synthesis method for creating a new robotic grasper has not deeply considered yet. Therefore, this article presents an effective computation method for designing a new robotic grasper that can be used for microassembly and micromanipulation. Firstly, a new structural scheme of robotic grasper is made by using the topology procedure. The compliance is the objective with the stress constraint during the topology. Then, a new variant of the grasper is refined where compliant joints are needed to reach the better compliance and the elastic motions of grasping hands. In the second phase, the modeling establishment for predicting the behaviors of the grasper are built via using the intelligent computation, namely GENFIS#1-neuro-fuzzy inference system (ANFIS), GENFIS#2-ANFIS, and GENFIS#3-ANFIS. The numerical data of the grasper are collected via the design of experiment-based finite element method. The results indicated that the GENFIS#3-ANFIS type is the best solver for modeling the hand’s stroke, the resonant frequency, and the strain energy. Meanwhile, the GENFIS#2-ANFIS type was the best procedure for modeling the stress. Subsequently, the optimum geometrical dimensions of the grasper are searched by using the Bonobo optimizer to improve the four mentioned performances of the grasper. In the circumstance #1, the results found that the hand’s displacement is about 0.0922 mm, the resonant frequency is 67.6247 Hz, the elastic energy is 1.4550 mJ, and the stress is 6.7249 MPa. The circumstance #2 determined the resonant frequency of 67.6247 Hz, the hand’s displacement of 0.0883 mm, the elastic energy of 1.9914 mJ, and the stress of 6.7086 MPa. Finally, the circumstance #3 found the elastic energy of 2.0501 mJ, the hand’s displacement of 0.0884 mm, the resonant frequency of 80.0012 Hz, and the stress of 6.7046 MPa. Statistically compared with the other methods, the presented method is the simple and effective procedure for designing 3D printed robotic grasper.