基于负荷特性分析的中长期负荷预测研究

发布时间：2018-01-15 20:19

本文关键词：基于负荷特性分析的中长期负荷预测研究　出处：《湖南大学》2014年硕士论文　论文类型：学位论文

【摘要】：中长期负荷预测的结果是电力系统规划的基础和依据，是开展电量综合平衡、电力结构优化调整等工作的前提。开展中长期负荷预测有利于资源的优化配置、燃料计划的制定和电网运行方式的安排，是保证电网运行可靠性和合理性的重要前提，对电网规划部门有着重要的理论意义和实用价值。负荷特性分析作为电力系统中的一项重要工作，有助于预测人员准确地把握负荷曲线的变化规律，从而获得较为理想的预测结果。本文根据某省电网历史负荷数据和相关资料，在对地区电网中长期负荷特性指标统计和分析的基础上，深入探究了各特性指标与影响因素间的关系。之后，对负荷特性指标的特点和规律加以总结，为开展中长期负荷预测的研究工作做准备。相比于短期负荷，中长期负荷的影响因素繁杂、时间跨度长、不确定因素多。本文基于中长期负荷特性的分析结果，提出了电力负荷的主导因素辨识法，，并结合偏最小二乘回归法应用于年电量预测。该方法从负荷特性指标的角度出发，根据负荷分析的结果选出基准负荷指标，并结合Pearson相关分析法判别影响因素与基准负荷指标的关联程度，进而选出电力负荷的主导因素。通过算例分析，证实了此方法可大大减小预测模型中由于工作人员主观经验而带来的偏差，弱化了噪声因素的干扰，模型的实用性得到加强。同时，由于无关因素的剔除，模型对主成分的提取能力得到加强，预测结果更为精确。本文从电力企业长远发展和工作需要角度出发，开发了电力需求预测及负荷特性分析平台。系统基于Java平台和Oracle数据库，采用了多层体系的B/S（Browser/Server）结构，以电力需求预测和数据挖掘分析为核心，紧密结合计算机网络技术、通信技术、信息安全技术与智能技术。系统充分考虑了电力负荷的主要影响因素，密切联系了电网工作的实际需求，有效地提高了预测结果的准确性，为电网规划提供了重大的技术支撑。
[Abstract]:The result of medium and long term load forecasting is the foundation and basis of power system planning, and the premise of carrying out the work of comprehensive balance of electricity quantity and optimization and adjustment of power structure, etc. The development of medium and long term load forecasting is beneficial to the optimal allocation of resources. The formulation of fuel plan and the arrangement of power grid operation mode are the important premises to ensure the reliability and rationality of power grid operation, and have important theoretical significance and practical value to the power network planning department. As an important work in power system, load characteristic analysis is helpful for forecasters to accurately grasp the changing law of load curve. Based on the historical load data and related data of a province, the paper makes statistics and analysis of the medium and long-term load characteristics of regional power network. This paper probes into the relationship between the characteristic indexes and the influencing factors, and then summarizes the characteristics and rules of the load characteristic indexes in order to prepare for the research work of the medium- and long-term load forecasting. Compared with the short-term load, the medium and long term load has a complex influence factor, long time span and many uncertain factors. Based on the analysis results of the medium and long term load characteristics, the paper puts forward the main factor identification method of the power load. Combined with the partial least square regression method, the method is applied to the annual electricity forecasting. This method selects the benchmark load index according to the results of the load analysis from the point of view of the load characteristic index. Combined with the Pearson correlation analysis, the correlation degree between the influencing factors and the benchmark load index is determined, and then the dominant factors of the power load are selected. It is proved that this method can greatly reduce the deviation caused by the staff's subjective experience in the prediction model, weaken the interference of noise factors, and enhance the practicability of the model. At the same time, the irrelevant factors are eliminated. The ability of the model to extract principal components is enhanced, and the prediction results are more accurate. In this paper, the power demand forecasting and load characteristic analysis platform based on Java platform and Oracle database is developed from the perspective of the long-term development and work needs of electric power enterprises. A multilayer B / S browser / Server structure is adopted, which is based on power demand prediction and data mining analysis, and combines closely with computer network technology and communication technology. Information security technology and intelligent technology. The system fully considered the main factors of power load, closely linked to the actual needs of the grid work, and effectively improved the accuracy of the forecast results. It provides important technical support for power network planning.
【学位授予单位】：湖南大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TM715

【参考文献】