No abstract

The dynamic reliability of configuration transformation is the key to ensure the normal operation of metamorphic mechanisms with multiple configurations in engineering applications. In this work, a controllable metamorphic palletizing robot for handling the workpieces was investigated. The failure function was established using the dynamic reliability model of its configuration transformation, where the uncertain variables were considered as probabilistic random and non-probabilistic interval variables. Next, a reliability analysis method used to determine and solve this function was proposed. The upper and lower failure function bounds were considered as probabilistic random variables enabling the calculation of the corresponding reliability. Regarding the case study, the reliability results of various configuration transformation types calculated via the presented method are close to those gained by the neural network-based Monte-Carlo method. Furthermore, in the experimental study, the measured data were obtained by the orthogonal design and fitted using a neural network. The reliability results show that the relative errors between the experimental results and those obtained through the proposed method are small, which suggests that the proposed method is effective. Therefore, this study provides an appropriate safety evaluation reference for the configuration transformation of metamorphic mechanisms.