论文部分内容阅读
对相关参数预测的非线性时间序列进行分析 ,提出了用相关参数实现多参数、多步数预测的神经网络方法 ,建立了神经网络实时预测模型 ,并对预测参数进行了综合处理 ,不仅弥补了预测信息的不足 ,而且能够使故障状态提前得到多次警报 ,克服了传统预测方法所遇到的困难 ,提高了预测模型精度
The nonlinear time series predicted by the relevant parameters are analyzed. A neural network method to realize multi-parameter and multi-step number prediction with related parameters is proposed. The real-time neural network prediction model is established and the prediction parameters are comprehensively processed. Predict the lack of information, and can make the fault state in advance to get multiple alerts to overcome the difficulties encountered by the traditional prediction methods to improve the prediction model accuracy