基于logistic组合模型的城市电力饱和负荷预测研究

发布时间：2019-04-19 07:44

【摘要】：电力负荷增长与经济社会的发展具有相同的趋势,在不同的经济社会发展阶段呈现不同的特点。当经济社会发展到一定阶段,受区域能源结构、土地资源、人口规模及环境压力等条件的限制,电力负荷的增长速度将会放缓甚至停止增长,逐步进入到电力负荷的饱和发展阶段。电力饱和负荷可以确定城市电网规划中电网发展的最终规模,为电源点、输电网和配电网的远期规划提供基础数据,进而指导近期城市电网的建设和改造,避免不必要的改扩建工程,为城市未来电力的发展预留空间,确保城市规划与电网发展相协调。本文通过对典型发达地区和国家的电力负荷发展规律的分析,总结出电力负荷的发展历程和在电力饱和负荷阶段的社会经济特征,进而完善了判定电力负荷进入饱和发展阶段的量化指标体系,同时介绍并总结了常用饱和负荷预测模型的方法和特点。通过分析与电力发展息息相关的经济、社会、人口、资源、政策、环境等影响因素,建立了电力负荷影响因素的意识模型,运用解释结构模型分析了各影响因素之间的结构关系,将影响电力负荷变化的因素划分为表层原因、浅层原因和深层原因三个层次。考虑保留传统logistic模型特征,运用灰色模型的精确极差格式建立极差格式的灰色logistic模型,避免了传统预测模型中参数主观化。考虑对电力负荷变化产生直接影响的表层因素,运用PSO优化的神经网络模型将logistic预测模型中的速度增长因子函数化,建立logistic拓展模型。然后采用方差-协方差和偏离度两种方式建立logistic组合预测模型分析并预测城市饱和负荷,避免单一预测的缺点,增强模型的实用性和灵活性,提高预测精度。最后分析了北京市经济、人口和电力发展状况,结合历史数据对北京市未来用电量需求量进行了预测,结果表明:从2026-2027年开始,北京市用电需求量年增长率小于2%,用电需求总量达到1331.68-1350.91亿千瓦时,进入饱和发展阶段。
[Abstract]:The growth of electric power load has the same trend with the development of economy and society, and it has different characteristics in different stages of economic and social development. When economic and social development reaches a certain stage, under the constraints of regional energy structure, land resources, population size and environmental pressure, the growth rate of electric power load will slow down or even stop growing. Gradually enter the saturated development stage of power load. The power saturation load can determine the final scale of power grid development in urban power grid planning, provide the basic data for the long-term planning of power supply point, transmission grid and distribution network, and then guide the construction and transformation of urban power grid in the near future. Avoid unnecessary reconstruction and expansion projects, reserve space for the future development of urban electricity, and ensure coordination between urban planning and power grid development. Based on the analysis of the law of power load development in typical developed regions and countries, this paper summarizes the development process of power load and its socio-economic characteristics in the stage of saturation load. Furthermore, the quantitative index system for judging the power load entering the saturation stage is improved, and the methods and characteristics of the commonly used saturated load forecasting models are introduced and summarized at the same time. Based on the analysis of economic, social, population, resource, policy, environment and other factors which are closely related to the development of electric power, the consciousness model of influencing factors of electric power load is established. The structural relationship among the factors is analyzed by using the interpretive structural model. The factors influencing the change of power load are divided into three levels: surface layer, shallow layer and deep layer. Considering the characteristics of the traditional logistic model, the grey logistic model of the extreme difference scheme is established by using the exact range scheme of the grey model, which avoids the subjectivity of the parameters in the traditional prediction model. Considering the surface factors that have a direct effect on the change of power load, the speed growth factor in logistic forecasting model is functioned by the neural network model optimized by PSO, and the logistic expansion model is established. Then the logistic combination forecasting model is established by variance-covariance and deviation to avoid the shortcomings of single prediction and to enhance the practicability and flexibility of the model and to improve the prediction accuracy. Finally, the development of economy, population and electricity in Beijing is analyzed. Based on historical data, the future demand for electricity consumption in Beijing is predicted. The results show that from 2026 to 2027, the annual growth rate of electricity demand in Beijing will be less than 2%. Total electricity demand reached 1331.68- one hundred and thirty five billion ninety one million kilowatt-hours, entering the stage of saturation development.
【学位授予单位】：华北电力大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TM715

【参考文献】