| 一种改进GMDH的新方法 |
| |
| 引用本文： | 劳兰Xun.一种改进GMDH的新方法[J].汕头大学学报(自然科学版),1997,12(1):9-14. |
| |
| 作者姓名： | 劳兰Xun |
| |
| 作者单位： | 汕头大学数学系 |
| |
| 摘 要： | 本文提出在ＧＭＤＨ中利用“新息贡献”准则选择关键变量（包括初始变量和中间变量）、自动构成最优部分多项式，并且给出了递推算法以避免重复运算，从而优化了模型结构，提高了精度，简化了建模过程。最后用仿真例子验证了新方法的有效性。
|
| 关 键 词： | GMDH，新息，非线性系统，系统辨识，关键变量 |
A new method for improving on GMDH |
| |
| Lao Lanjun.A new method for improving on GMDH[J].Journal of Shantou University(Natural Science Edition),1997,12(1):9-14. |
| |
| Authors: | Lao Lanjun |
| |
| Abstract: | In this paper the "innovation-contribution" criterion is presented to select es- sential variables (including initial and intermediate variables) and to determine the opti- mal partial polynomials in GMDH , with a recurrence algorithm proposed to avoid dupli- . cate computation , thereby optimizing the model structure , ataining remarkable model precision , and simplifying the modelling process. Identification of a simulated system demonstrates the effectiveness of this new method. |
| |
| Keywords: | group method of data handling (GMDH) : innovation nonlinear system t sys- tens identification : essential variables |
| 本文献已被 CNKI 维普 等数据库收录！ |
