| Budget Amount *help |
¥1,800,000 (Direct Cost: ¥1,800,000)
Fiscal Year 1993: ¥400,000 (Direct Cost: ¥400,000)
Fiscal Year 1992: ¥700,000 (Direct Cost: ¥700,000)
Fiscal Year 1991: ¥700,000 (Direct Cost: ¥700,000)
|
|---|
| Research Abstract |
On sysytem identification with multivariate time series data in the reactor noise, it is necessary to examine properties of innovation models systematically. Coefficient matrices of a data-oriented innovation model have been determined by using the singular value decomposition of Hankel matrix of which the elements are correlation functions, and properties of its conservation quantities such as poles and zeros have been examined. Furthermore, effects of observation noise to zeros of innovation model have been examined, and summarized as rules.
Since the nuclear reactor is a system with thermal and hydro-dynamical feedback mechanics, it is necessary to examine relations between closed loop transfer functions and open loop transfer functions describing feedback loops. Since the relationship between both transfer functions is a nonlinear transformation, poles and zeros of the fitted model are not conserved in general. It has been reported that there is a possibility of pole-zero cancellati
… More
on in a fitted time series model.
Along the line of above studies it has been recognized that there exists a stable particular solution of Riccati equation, which appears in the formulation of the innovation model. The investigation of existence of the stable particular solution brings us to know that the condition of atable solution is satisfied if zeros of feedback system are invertible. If a reactor is of minimun phase (of which zeros are invertible), the Riccati equation has a stable solution under the condition, and open loop transfer functions can be identified theoretically. Otherwise, there appears an additional equivalent loop in the feedback system, and then the additional loop makes difficulty in system identification of open loop transfer functions.
Although the key model in the formulation is an equivalent innovation model, its representation is not unique. The transformation from one representation to the other has been studied. Corresponding Riccati equations are also examined, and properties and relationship between Riccati equations have been made clear. Less
|
