A Study of integrated intelligent marine monitoring system

摘要

This paper presents an integrated artificial intelligent approach for machine monitoring and fault detection, which merges artificial neural networks(ANN), expert systems(ES) and fuzzy technologies (FT) to overcome the problems in the conventional methods. In this paper we discuss the system control structure, management structure and the AI integration method. In the integrated method, first a multi-structure ANN with improved BP learning algorithm is proposed for reducing the dimensions and learning time of the ANN, then a fuzzy neural network(FNN) model with generalized probability sum & product operators is introduced for fault fuzzy reasoning, further an adaptive parameter correcting algorithm with single sample is presented for trend prediction.