摘要
人工神经元网络的研究技术在理论和实际应用上已经比较成熟,在信号处理系统中也采用该技术进行非线性时间序列信号的预测分析。但是由于该理论黑箱模型的特点,无法引入先验知识,从而预测精度难以提高。针对该问题,文中提出了智能神经网络的动态预测模型,引入智能神经元,建立区别于传统神经网络的预测模型,达到了较为理想的预测效果。并以工业生产参数的时间序列预测——某油井生产过程中MinCurrent参数值,作为实验模型,对该方法进行了验证,结果表明了该模型预测精度较高、计算速度快。
The artificial neuron network technology has been developed in theory and practical application, so it is also used in signal pro- cessing system to predict nonlinear time series. But because of its character of the black box and being unable to introduce prior knowl- edge, it is hard to improve prediction accuracy. To solve this problem, introduce the dynamic intelligent neural networks, a new predic- tion model building up by dynamic predietion and introducing intelligent neuron, which achieves the better prediction result. Then, take the prediction of industrial parameters- Mineurrent-a parameter of a oil for example to validate the method and reveal its practical appli- cation value： high prediction accuracy and fast calculation speed.
出处
《计算机技术与发展》
2012年第3期74-76,80,共4页
Computer Technology and Development
基金
陕西省教育科研计划项目(11JK0984)
关键词
智能神经元
神经网络
BP算法
时间序列
预测模型
intelligent neuron
neural network
BP algorithm
time series
prediction model
