On-line testing of passive solar components
A generally applicable method for on-line testing of passive solar components is presented. The method is based on a state-space formulation of difference equations describing major heat flows in an outdoor test cell. This formulation makes it possible to use powerful statistical tools for the identification of the model, including on-line estimation of parameters using an extended Kalman filter, and estimation of the accuracy of the obtained estimates. The method is designed to allow for: (1) estimation of parameters also in the case of autocorrelated noise by an extended matrix method, (2) tracking of time varying parameters by a technique for discounting old data, and (3) testing of sensor failure. An application of the method is also presented using data from the Passive Solar Test Facility of the J.R.C in Ispra. The application shows that thermal parameters may be estimated which agree well with the results from two other identification methods. Moreover, a time variation in the effective solar aperture could be followed using a rectangular window. This variation could also be detected by tests based on the obtained residuals in the recursions assuming that no variation had occurred.
Bibliographic Reference: EUR 11823 EN (1988)
Record Number: 198910078 / Last updated on: 1994-12-01
Original language: en
Available languages: en