Large Tanker Motion Model Identification Using Generalized Ellipsoidal Basis Function-Based Fuzzy Neural Networks

In this paper, the motion dynamics of a large tanker is modeled by the generalized ellipsoidal function-based fuzzy neural network (GEBF-FNN). The reference model of tanker motion dynamics in the form of nonlinear difference equations is established to generate training data samples for the GEBF-FNN algorithm which begins with no hidden neuron. In the sequel, fuzzy rules associated with the GEBF-FNN-based model can be online self-constructed by generation criteria and parameter estimation, and can dynamically capture essential motion dynamics of the large tanker with high prediction accuracy. Simulation studies and comprehensive comparisons are conducted on typical zig-zag maneuvers with moderate and extreme steering, and demonstrate that the GEBF-FNN-based model of tanker motion dynamics achieves superior performance in terms of both approximation and prediction.

I. Introduction

In the entire guidance, navigation, and control system [1], the vessel maneuvering dynamics plays a fundamental role, and have attracted various achievements on vessel motion models of Abkowitz [2], MMG [3], and response type [4], which possess distinct features different from each other. In the Abkowitz model [2], accurate hydrodynamic derivatives can be pursued while the physical concepts of variables are lost, and thereby resulting in difficulties for control system design. As the simplifications of the Abkowitz model, the MMG and response models with lower accuracy incorporate with analysis and synthesis of model-based control systems. In this respect, various methods [5]–[9] have been proposed to identify hydrodynamic derivatives and/or input–output nonlinearities of vessel systems. Unfortunately, the resulting models are involved in complicated mathematical formulation of vessel maneuvering which is strongly associated with the presence of hydrodynamic nonlinearities pertaining to the vessel dynamics. Apparently, traditional methods would inevitably lead to a dilemma between the accuracy and complexity of a vessel motion model.