Identification of Toroidal Continuously Variable Transmission Using Continuous Time System Identification Method

カテゴリ: 部門大会

論文No: TC4-3

グループ名: 【C】平成14年電気学会電子・情報・システム部門大会講演論文集

発行日: 2002/09/02

タイトル(英語): Identification of Toroidal Continuously Variable Transmission Using Continuous Time System Identification Method

著者名: Michel Mensler(Nissan Motor Co.,Ltd.),Shinichiro Joe(Nissan Motor Co.,Ltd.),Taketoshi Kawabe(Nissan Motor Co.,Ltd.)

著者名(英語): Michel Mensler(Nissan Motor Co.,Ltd.),Shinichiro Joe(Nissan Motor Co.,Ltd.),Taketoshi Kawabe(Nissan Motor Co.,Ltd.)

キーワード: identification|CVT|automotive

要約(日本語): This paper is focused on the estimation of the parameters of the Toroidal Continuous Variable Transmission (TCVT) model via two continuous-time system identification methods for several operating points in order to build a map of parameter values. Each operating point corresponds to one value of the TCVT gear ratio G and one value of the line pressure.

The TCVT model can be divided into two sub-models: one containing only static parameters and one containing a varying parameter. The parameter estimation phase is then done in two steps.

The first step performs the estimation of the parameters of the first model using the Poisson Moment Functional (PMF) method associated to an auxiliary model instrumental variable technique. The model used for the estimation is a two-inputs one-output model where the inputs are the actuator step number u and the TCVT power roller angle f and where the output is the trunnion offset y, and contains three unknown parameters.

The second step estimates the varying parameter using the Linear Integral Filter (LIF) method. The model used here is a single-input single-output model.

The estimated parameters are then compared to the available theoretical ones.

PDFファイルサイズ: 5,271 Kバイト