Position accuracy improvement of linear motor container transfer system using DR-FNNs

摘要

We introduced a multi-step prediction control of linear motor-based container transport system for high precision using dynamically constructed recurrent fuzzy neural networks (DR-FNN). Linear motor-based container transport system (LMCTS) is the horizontal transfer system for yard automation, which is proposed to take the place of automated guided vehicle in the maritime container terminal. LMCTS is the system that can suddenly changed its model by loading and unloading container. The proposed control system is using two DR-FNN for multi-step prediction. Consequently, the system has the ability to adapt for a huge rolling friction, cogging force ripple, and sudden changes by loading and unloading.