为提高小样本轴承故障预测的精度,提出了灰色关联分析的神经网络预测模型.利用神经网络训练样本数据,并使用灰色关联分析不断调整其隐含层节点数,以寻找到最优的数据解,完成训练过程；基于已训练好的模型对未来时间点的运行状态进行分析,并根据设备的理论诊断标准实现故障预测.将提出的模型与同样可预测小样本波动数据的灰色马尔科夫预测模型进行实例应用对比,结果显示,灰色关联分析的神经网络预测模型效果明显优于灰色马尔科夫模型.%In order to improve the accuracy of fault prediction of bearing containing the data of small sample, the forecasting model is put forward based on neural network of grey relational analysis. The neural network is used to train the sample, and then the grey relational analysis is utilized to adjust constantly the number of node in hidden layers so that the optimal solution of question could be determined, finally the training process is completed. Based on the model trained to estimate the operation state of equipment in the future time, the fault prediction will be achieved according to the theoretical diagnosis standard. The proposed model is compared from an example with Gray-Markov method which can also predict the fluctuant data of less sample. Their result shows that the predictable effect based on neural network of grey relational analysis is better than the Grey-Markov's, which verify further the feasibility and effectiveness of the proposed method.
展开▼
