Page 183 - Design of Solar Thermal Power Plants

168 3. GENERAL DESIGN OF A SOLAR THERMAL POWER PLANT

There are two basic methods to establish the mathematical model. One
is the mechanism analysis method, which is to establish a model by
utilizing modeling information or prerequisite provided by modeling
assumptions on the basis of the analysis on the internal mechanism of
matters; it is normally referred to as the white box, such as the energy
balance equation and thermal and mass transfer theory. The other one is
the system identification method, which is to establish a model by uti-
lizing modeling assumptions or actual input and output information of
matter system offered to system test data while being absolutely ignorant
of the internal mechanism of system; it is normally referred to as the black
box. According to different principles, model identification methods can
be summarized into four categories: (1) the least-square method,
including the least-square method, the extended least-square method,
auxiliary variable method and the generalized least-square method; (2)
gradient correction parameter identification method, such as stochastic
approximation method; (3) probability density approximation parameter
identification method, such as the maximum likelihood method; (4) new
methods that have been recently developed, such as the blur identifica-
tion method, neural network identification method, wavelet identification
method and inheritance identification method.
To sum up, thermal performance test model of solar collector is to
establish the mechanism model according to the energy balance theory
and heat-transfer principle, and identify the undetermined parameters of
the model by applying the least-square method; it belongs to a grey box
model. Such test model based on empirical methods is deemed to be able
to generate highly precise results with the respective range of applied
parameters.

3.3.3.2 Assumed Conditions of Dynamic Test Model
Dynamic test model of parabolic trough solar collector shall satisfy the
following assumptions.
1. Based on the requirement of ASHRAE 93 standard on volume flow
of heat-transfer fluid being within 2% of the set value during the
test, according to the dynamic test model assumption, volume flow
variation of synthetic oil passing through the parabolic trough
collector within the test period does not exceed 2% of the mean
value of the period. However, considering that density variation of
synthetic oil within the application temperature normally exceeds
10%, such as the density variation of synthetic oil applied under
Experimental Conditions of this research, which is shown in
Fig. 3.26. Therefore, mass flow of synthetic oil passing through the
parabolic trough collector within the test period is assumed to be
within 20% of the mean value without any intensive fluctuation of