内蒙电网负荷预测及交易计划决策系统的研究与设计-软件工程专业论文 word格式.docxVIP

摘要摘要内蒙古电力公司为了适应电力系统运行模式的转变，及时开展了电力系统控 制手段的研究，由内蒙古电力调通中心与清华大学合作开发了电力交易市场所需 的计划决策系统，它包括年、月、日交易计划决策系统和年、月、日负荷预测系 统，投入运行后，取得了较好的效果。日负荷预测系统为不同的预测情况设计了不同的算法，实践证明这样的做法 是有效的，在负荷变化不大的情况下，使用自回归方法即可得出比较满意的预测 结果，而在负荷波动比较大的情况下，则使用人工神经网络方法可得到相对满意 的结果，在节假日的时候（例如春节），只有使用外推插值方法才能够得出比较符 合实际情况的预测结果。年、月负荷预测系统用线性增长法和自回归方法进行年电量预测，基本能够 满足要求。年、月负荷预测系统分别采用自回归方法、平均增长率法和最小二乘 法进行月电量的预测，能够满足实际运行的要求。关键词:负荷预测决策系统模型和算法ABSTRACTABSTRACTIn order to adapt to the transfer of the power system operation mode, Inner Mongolia Power Co., Ltd. started to launch the study on the control measures for the power system. Jointly with Tsinghua University, Inner Mongolia Power Dispatching Center developed the planning decision-making system for the power trading market, which consists of the yearly, monthly and daily trading plan decision-making system and the yearly, monthly and daily load forecast system. Good results were achieved after operation.The different ways of calculations are designed for the different situations of daily load forecast system and proved to be effective. Under the relatively stable condition of the load, the more satisfactory result can be obtained by the way of self-resettling. If the load changes greatly, the man-made nerves network is used to achieve the better result. However, during the holidays like Spring Festival, the relatively practical forecast can be done only by the way of extrapolation.The linear increase for the yearly, monthly and daily load forecast system and self-resetting for annual power amount forecast can meet the demand. The self-resetting, average increase rate and min two-multiple methods for the yearly, monthly and daily load forecast system can fulfill the requirement of practical operation.Key words: load forecast decision-making model and calculation method目录目录摘要 ................................................................................................................................................ IABSTRACT............................................