ABC-BPNN Based Enhanced Hybrid Strategy for the Prediction Model of 6-DoF FKP
Huizhi Zhu, Jing Wu, Wenxia Xu, Jian Huang, Yanan Li, Baocheng Yu · 2022 International Conference on Advanced Robotics and Mechatronics (ICARM) · 2022
Ususally, Back Propagation Neural Network (BPNN) was used to solve the problem of complex forward kinematics for Gough-Stewart platform. However, BPNN will falls short of the neural network requirements in a hybrid strategy. Then, to improve the prediction ability of neural network model, the ABC-BPNN, a new neural network for employment in hybrid algorithm was proposed in this paper. This paper firstly introduced kinematics model of 6-DoF. Subsequently, improving the prediction ability of neural network model to the pose of the machine by ABC optimization. Lastly, we designed three models and proved that the network model with one hidden layer has the best prediction effect and the lowest metrics by qualitative comparisons.