TESTING OF PASSIVE SOLAR COMPONENTS USING IDENTIFICATION METHODS
An outdoor test procedure for passive solar components using shortterm monitoring data from tests cells is presented. The test cell is first calibrated with a highly insulated south wall. Different south wall components are then inserted in place of the calibration wall. The auxiliary heating is controlled by a pseudo random binary sequence which makes it possible to identify an energy balance model with shortterm measurements. Different energy balance models and identification methods are discussed and tested with measurements from two test cells. Measurement periods of 10 to 15 days give satisfactory results, provided there is sufficient variation in the solar radiation. Shorter periods, down to one day, can be used if no trends exist in the climatic variables. Time domain identification using the prediction error method has been used. A frequency domain solution has also been obtained using a finite Fourier transform and a subsequent minimization in the least squares sense. Both results are compatible. There is a systematic difference between the more sophisticated 7parameter model (2capacitance) and the 4 parameter model (1capacitance). The results from the 7 parameter model agree best with the calculated improvement in the description of the measured data.
Bibliographic Reference: EUR 11056 EN (1987) FS, 44 P., BFR 200, EUROFFICE, LUXEMBOURG, POB 1003
Availability: Can be ordered online
Record Number: 1989126002600 / Last updated on: 1988-02-01
Available languages: en