论文部分内容阅读
本文提出了对动态系统参数分量序列分别建立其中一个参数关于其它参数的CAR模型,再进行参数的递推予报,最后进行状态递推予报的新方法。这种方法深入考虑了参数分量之间的影响,在某些情况下,可以提高予报精度,而参数的AR建模作为特例。本文还给出了这种方法用于油田产量和社会产值予报实例。
In this paper, we propose a new method to establish a CAR model with one parameter on the other parameters for the dynamic system parameter sequence, recursive the parameters and recursively report the state. This method takes into account in-depth the influence of the parameter components, and in some cases, can improve the accuracy of the report, while the parameter AR modeling is taken as a special case. This paper also gives an example of this method for oilfield production and social output.