マルチアルゴリズムの可能な改良形GMDHの構成*
スポンサーリンク
概要
- 論文の詳細を見る
The GMDH (Group Method of Data Handling) is well known as a modeling method of a complex system. However, the models obtained have generally led to complex expressions because partial polynomials with two variables are used in the method. In this paper, a revised GMDH is proposed from a point of view of practical use. The revised points are following : 1) partial polynomials consist of multi variables and can be selected arbitrarily in each selection layer ; 2) optimal partial polynomials can be generated automatically ; 3) the AIC (Akaike’s Information Criterion) is used as a criterion for selecting intermediate variables, generating optimal partial polynomials and stopping the multi-layered calculation ; 4) it is possible to constitute a modeling algorithm according to the intention of a model maker. Two different algorithms constituted under the concept of this revised GMDH are applied to a simple illustrative example, and these results are compared and discussed.
- 北見工業大学の論文
著者
関連論文
- むだ時間を持つ連続値形制御対象の線形離散値系への変換
- マルチアルゴリズムの可能な改良形GMDHの構成*
- 構造未知システムのモデリング手法*
- モデルの多項関数応答に追従する離散値形線形制御系の構成*
- ハイブリッド位相進み補償器*
- 離散値形完全モデル追従コントロ-ラの構成 (現代制御理論によるサ-ボ系設計論とその応用)