A state and fault prediction method based on RBF neural networks

Abstract

A state and fault prediction method based on RBF neural networks is proposed. The agricultural machinery is chosen as the experimental object of the method. There are 4 health level, such as failure, hazardous, sub-healthy and healthy. Some data of different provinces have been obtained, the health level can be acquired by RBF neural networks. The mathematical model of agricultural machinery is difficult to be proposed in this paper, so the traditional control algorithm can't be used in agricultural machinery. However, the RBF neural networks can solve this problem. At the same time, some vital factors should be considered, such as mileages, rotational speed, stubble height, water temperature, oil pressure of agricultural machinery. The rotational speed and stubble height have a big effect on fault prediction of agriculture. The experimental results verify the effectiveness of the proposed method.