基于优选组合预测技术的中长期负荷预测  

Medium and Long-term Load Forecasting Based on Optimized Combination Forecast Technology

作　　者：马星河[1] 闫炳耀[2] 唐云峰[3] 张均伟 

机构地区：[1]河南理工大学电气学院,焦作454003 [2]河南省电力公司焦作供电公司,焦作454001 [3]焦作煤业集团冯营电力有限责任公司,焦作454173 [4]河南新郑煤电有限公司,新郑451100

出　　处：《电力系统及其自动化学报》2015年第6期62-67,共6页Proceedings of the CSU-EPSA

摘　　要：中长期电力负荷预测是电力系统规划的基础和重要前提。该文提出优选组合预测技术分为模型筛选和组合模型筛选两大部分。单个模型的筛选方法选择改进的灰色关联度指标和冗余校验方法,改进的灰色关联度指标更加注重预测发展趋势预测的精准度,冗余校验是对已经筛选出来的模型进行冗余检查。选择5种组合预测方法,较大程度兼容组合预测的实用性和准确性。改进的预测误差指标体系验证了优选组合预测技术在中长期负荷预测中的合理性。Medium and long-term power load forecasting is the basis of power system planning and the main premise.The optimum combination of prediction model is divided into model selection and group model selection. A single model selection method is improved grey correlation degree index and redundancy check,the improved grey relation grade index is to predict the development trend forecast accuracy,selected model is redundancy check,which can weed out the combination of little help in predicting model. The article chooses five kinds of combination forecasting,which can be compatible with practicability and accuracy. Construction of improved prediction error index system can verify the rationality of optimum combined forecasting technology.

关 键 词：优选组合 模型筛选 改进的灰色关联度指标 冗余校验 改进的预测误差指标体系 

分 类 号：TM714[电气工程—电力系统及自动化]