Design optimization of 3D-manufactured monolithic stage for biomedical polishing applications by neural artificial network and teaching learning studying based optimization algorithm

Abstract

Ultra-precision polishing engineering is a current trend in ensuring the good surface quality for biomedical applications. In precision engineering, compliant mechanism, so-called monolithic mechanism, has an increasing trend in alternating conventional rigid counterparts which can be utilized for biomedical polishing application. Therefore, this article proposes an optimal design for the monolithic stage. The stage is employed in driving the polishing head. The stage is operated within from several millimeters to hundreds of millimeters. In this article, the stage was tended to guide the head to polish the biomedical samples. Hypercube sampling design is combined with finite element method to collect the data. The teaching learning-based optimization algorithm is firstly coupled with the artificial intelligence in modeling the behaviors of the stage (resonant frequency, displacement, strain energy, and stress). Next, the teaching learning studying based optimization algorithm is extended to search the optimal parameters of the stage. Three single objective optimization scenarios and two multiobjective optimization scenarios of the monolithic stage are conducted to show the effectiveness of the offered method. The predictors were well-formed by using the teaching learning-based optimization-coupled artificial neural network. The results found that the performance indicators (R and R2) are almost equal to one with the mean square error (MSE) and root MSE values are very small. The optimal results that the stage can work with a highest frequency of 760.7903 Hz, a largest stroke of 0.7325 mm (about 732.5 μm), and a strain energy of 2.5999 mJ. Meanwhile, the von Mises stress is almost much lower than the yield strength of AL 7075-T73. The designed method can effectively use in designing the stage. The stage can be fabricated by using selective laser melting additive manufacturing technology, and it can be employed to guide the polishing head for biomedical samples.