太阳位置算法的计算误差对辐射预测的影响

以往太阳辐射预测计算的研究多采用柯伯方程等误差较大的公式,往往忽略了太阳方位参数计算误差对辐射预测计算模型的影响。忽略其他因素影响,仅计算不同太阳方位参数误差所对应的太阳辐照度,计算结果与天文年历标准值的计算结果进行对比,发现太阳方位参数的误差不仅对太阳辐照度计算造成7%以上的误差（时角大于70°时）,而且在日出和日落之间临界点时也会产生极大的数值计算误差。针对以上问题,本文提出了引入儒略日为时间变量和采用大数据量最小二乘法拟合改进的数值模拟法,在保持计算公式简易的同时,显著降低了临界点的计算误差,而且辐射计算结果的相对误差小于0.2%。

Effect of Calculation Error of Solar Position Algorithm on Radiation Prediction Calculation

The influence of the calculation error of the solar azimuth parameter on the radiation prediction model was often neglected in the past studies, formulas with large error such as the Cooper equation was used. By eliminating the influence of other factors, only the solar irradiance calculation results corresponding to the different solar azimuth parameter errors were calculated and compared with the standard astronomical calendar calculation results. It was found that the error of the solar azimuth parameter not only caused an error of more than 7% for the solar irradiance calculation (when the time angle was greater than 70 degrees), but also caused a large calculation error at the transition point between sunrise and sunset. In response to the above questions, this paper improved the numerical simulation method using the Julian Day as the time variable and using the large data quantity least squares method to fit it. Not only greatly reducing the calculation error of the critical point, but also the relative error of calculated result was less than 0.2%.