Page 99 - Design of Solar Thermal Power Plants
P. 99

2.6 SOLAR IRRADIANCE PREDICTION METHODS        87

           TABLE 2.14  Comparison of Estimation Result Errors for Monthly Mean Global Solar
                       Radiation Resources
                                               Model
           Month       Model Expressions       RMSE      MABE      MBE
           January     H/H 0 ¼ 0.157 þ 0.680 S/S 0  6.18 E-02  1.12 E-01   3.50 E-03
           February    H/H 0 ¼ 0.120 þ 0.721 S/S 0  5.43 E-02  9.10 E-02   3.30 E-03
           March       H/H 0 ¼ 0.116 þ 0.709 S/S 0  4.80 E-02  8.35 E-02   1.55 E-09
           April       H/H 0 ¼ 0.111 þ 0.678 S/S 0  4.33 E-02  7.75 E-02   1.30 E-03
           May         H/H 0 ¼ 0.132 þ 0.624 S/S 0  4.70 E-02  8.16 E-02   5.50 E-03
           June        H/H 0 ¼0.127 þ 0.621 S/S 0  4.83 E-02  8.90 E-02   3.82 E-02
           July        H/H 0 ¼0.193 þ 0.478 S/S 0  5.45 E-02  9.35 E-02   1.40 E-03
           August      H/H 0 ¼ 0.192 þ 0.487 S/S 0  5.36 E-02  9.38 E-02   4.20 E-03
           September   H/H 0 ¼ 0.165 þ 0.559 S/S 0  5.11 E-02  8.30 E-02   3.80 E-03
           October     H/H 0 ¼ 0.102 þ 0.685 S/S 0  4.52 E-02  7.66 E-02   1.80 E-03
           November    H/H 0 ¼ 0.078 þ 0.754 S/S 0  5.33 E-02  8.77 E-02   1.50 E-03
           December    H/H 0 ¼ 0.160 þ 0.656 S/S 0  7.22 E-02  1.23 E-01   9.01 E-05
           Mean value  H/H 0 ¼ 0.139 þ 0.637 S/S 0  5.27 E-02  9.10 E-02   5.40 E-03




           conclusion can be made that the smaller the RMSE of the comparison
           result, the higher the precision.
              As shown in Table 2.14, the mean RMSE for various months of the full-
           site single-month estimation model is 5.27%. Under clear weather con-
           ditions, atmospheric transmission coefficients can be reflected through
           sum of regression coefficients a and b. Monthly variation of sum of
           regression coefficients corresponding to the estimation model developed
           in the book is shown in Fig. 2.9.
              According to Fig. 2.9, on average in China, the sum of these coefficients
           gradually decreases with the beginning of the new year, bottoming in July
           before enjoying a gradual increase thereafter, with atmospheric trans-
           parency reaching a maximum in spring and winter and a minimum in
           summer. This is caused by several factors: in most areas in China, the
           second and third quarters are the main precipitation seasons and
           respective weather conditions are comparatively humid; whereas during
           spring and winter, the weather is comparatively dry; it is not easy for
           precipitation to form. Humidity is form easily on cloudy days under
   94   95   96   97   98   99   100   101   102   103   104