为改善含高密度、多接入点建筑光伏并网发电系统的电网运行特性,从能量管理和区域电网的运行调度两方面分析了建筑光伏并网发电系统的发电量预测需求,探讨了区域建筑光伏发电量预测的关键技术,分析指出原理预报法建模困难。得出如下结论:采用神经网络法、支持向量机方法等人工智能方法是光伏发电量预测模型的主要发展方向;数值模式太阳能辐射预报的预测精度和空间尺度离电网调度运行的要求还有相当距离。最后初步设计了建筑光伏的发电量预测系统,提出采用综合多个点的太阳能辐射预报结果进行区域太阳能辐射预报。
In order to improve the power network operation characteristics of BIPV (Building Integrated Photovoltaic) grid-connected generation system including the high-density and multi-access point, the energy production forecasting demands of the BIPV grid-connected generation system are analyzed from two sides of energy management and regional power grid operation dispatch. And the key technologies of the regional BIPV energy production forecasting are discussed. It is illustrated by means of further analysis that the principle prediction method modeling is difficulty. The development direction for the energy production forecasting modeling of the BIPV grid-connected generation system is artificial intelligent methods such as neural network and support vector machines method. And the precision and spatial scale of the numerical solar radiation forecasting cannot meet the demand of the grid operation dispatch. At last the preliminary design of BIPV energy production forecasting system is put forward which adopts the regional solar radiation forecasting method based on multi-point forecast results.