考虑气温因素影响的中长期负荷场景生成方法

Scenario Generation Method of Medium-and Long-term Load Series Considering Meteorological Factors

负荷场景生成是电力系统中长期规划的重要基础。基于历史数据建模生成的负荷场景难以充分表征负荷对气温的敏感性以及未来电力负荷发展趋势,针对此不足提出了一种考虑气温因素影响的中长期负荷场景生成方法。首先,对不同季节负荷与气象因素之间的相关性进行分析,将年度负荷序列按照季节特征和节假日特征划分为多个序列分别建模处理。对于与气温因素相关性较强的负荷,采用负荷分解理论将负荷分解为基准负荷和温变负荷两部分,并对温变负荷的峰值及曲线进行建模分析,在此基础上生成给定发展情景下的负荷序列;对于受气温因素影响不明显的负荷,通过建立日负荷特性指标联合概率分布模型和小时级负荷序列优化求解模型生成日负荷场景。最后,基于中国3个省级电网全年负荷数据进行测试,验证了所提方法的有效性和可行性。

Scenario generation of load series is an important basis for medium-and long-term power system planning.As the generated scenarios based on historical data could not represent the future development trend of power load adequately,this paper proposes a new scenario generation method of medium-and long-term load series considering the influence of meteorological factors.Firstly,the annual load series is divided into several subsequences based on the seasonal characteristics and correlation analysis.For meteorologically sensitive load,the total load is divided into base load and temperature-related load using the load decomposition theory.Peaks and curves of temperature-related load are modeled and then superimposed onto the base load to obtain the daily load series scenarios.For non-meteorologically sensitive load,the load scenarios are generated through a joint probability model of daily load characteristic indicators and an optimization model.The case study is conducted based on the annual load data of several provincial power grids in China which verifies the effectiveness and feasibility of the proposed method.