【摘要】 光伏系统日发电量预测对于提高可远程监控的离网光伏设备的电能管理具有重要的意义。针对这一需求,通过分析影响光伏系统日发电量的因素,提出了用加权平均总云量以量化日天气类型,并以此改进选取相似日过程中气象特征向量的构造。通过计算分析相似度筛选出与预测日特征相似的历史数据,作为预测模型的训练样本,利用支持向量机回归(SVR)对光伏系统的日发电量进行预测,并通过某地太阳能LED路灯的实测数据对模型进行验证,计算分析了预测误差。结果表明,该方法具有较高的预测精度,所提模型具有有效性和实用性。更多还原

【Abstract】 The daily generation forecast of photovoltaic( PV) system is significant to improve the energy management of off-grid PV equipment which can be monitored remotely. In response to the demand,by analyzing the factors that affect the daily generation of PV system,a weighted average total cloud is proposed to quantify the daily weather patterns and to improve the construction of meteorological eigenvectors in the process of selecting similar days. By calculating and analyzing similarity degree,the historical data similar to the features of the forecast day are selected and considered as the training samples,the daily power generation of PV system is forecasted by support vector machine regression( SVR). The method is validated by the actual data of solar LED street lights and the forecast error is calculated and analyzed. The results show that the proposed method has high accuracy and the model is effective and practicable.更多还原