重庆长寿电网负荷特性分析与负荷预测研究

发布时间：2018-02-20 23:14

本文关键词： 负荷预测负荷预测系统网络技术灰色模型负荷特性分析　出处：《湖北工业大学》2016年硕士论文　论文类型：学位论文

【摘要】：长久以来,电力负荷预测都是研究电力市场的至关重要课题,尤其是近年来,电网的规模不断扩增,研究电网负荷时需要考虑的因素也不断增多。冬夏时节空调的负荷增大,电网使用量达到最大值,电网进入丰谷期;随着近几年工业的不断发展,工业用电量也在不断增加;一些不可预知且不断出现的自然灾害等因素,对测量电网负荷造成诸多无法预料的困难;并且从近几年电网数据统计发现,电网负荷率和利用小时都在不同程度的降低,让我们的电网安全和经济运行上的困难进一步加大。想要控制好我们国家在电网项目中的投资,最大限度的提高发电设备和燃料的使用效率,对水电、火电发电之间按照一定的比例进行重新划分,达到节约能源减少污染排放的目标,那么将把负荷分析预测的工作做好就显得非常关键;这对于我们的大型用电企业来讲,好处也是非常之多的,一方面在供电设备上面的投资减少,另一方面还可以实现电网公司的削峰平谷,而且由于峰谷电价的存在,使得大型用电企业既可以合理安排生产工作,也可以把生产成本降低;最后对于普通的民众来讲,电力负荷预测能保证用户在用电高峰使用上高质量的电能而且保证他们的用电的需求量,进而居民家里的电气设备的使用寿命也得到提高。电力系统负荷的科学准确预测是电力部门进行正确决策的依据和保障之一,它将有利于发电部门制定经济合理的系统发电的相关计划,有效的控制成本;根据电网数据预测结果,对相关部门制定强有力的电力规划方案,提供可靠数据来源;对计划用电管理非常有利,获得需求分析情况,推进电力工业的市场化;有利于调度部门对电力提前进行调度安排,不仅有利于电力收入,而且对社会公共效益提升也有很大的帮助;对保证电网的稳定运行也是十分有力的。本文里提出了一种电网负荷的预测方法,这是一种在灰色模型的基础上,通过不断改进得到的,这种预测办法首选对指数进行加权,打破原始负荷的排列形式,再研究受到日类型以及气象条件因素影响较大的母线,依据灰色关联度的理论建立日特征向量,在此基础上找到一些最优相似日,最后历史负荷的样本即上述选取日的母线负荷。根据远的小、近的大的规律,在选取负荷样本时,选取那些影响相对较大的母线以就近原则,反之也成立。这种方法既充分利用了样本里面数据的有用信息,又使得样本中数据的随机性减小,还能够减弱异常值的影响。本文通过研究重庆市长寿供电分公司的计算实例,充分证明了对电力负荷分类处理的实用性和科学性,并且也在一定程度上,体现出本文里提出的负荷预测方法,有很高的实践效益。
[Abstract]:For a long time, power load forecasting has been a very important topic in the study of power market. Especially in recent years, the scale of power grid has been expanded, and the factors to be considered in the study of power grid load have been increasing, and the load of air conditioning is increasing in winter and summer. With the development of industry in recent years, the power consumption of industry is also increasing, and some unpredictable and recurring natural disasters and other factors, It is difficult to measure the load of the power network, and it is found that the load rate and the hours of utilization are decreasing in different degree from the statistics of the power network data in recent years. We want to control our country's investment in power grid projects, maximize the efficiency of power generation equipment and fuel, and make use of hydropower. If thermal power generation is reclassified according to a certain proportion to achieve the goal of saving energy and reducing pollution emissions, it will be very critical to do a good job of load analysis and forecasting; this is very important for our large power enterprises. The benefits are also very large. On the one hand, the investment in power supply equipment is reduced, on the other hand, it is possible to realize the peak-cutting and leveling valley of power grid companies. Moreover, due to the existence of peak and valley electricity prices, large power enterprises can not only reasonably arrange production work, It can also reduce production costs; finally, for ordinary people, power load forecasting can ensure that consumers use high quality electricity at peak consumption and ensure their demand for electricity. Furthermore, the service life of the electrical equipment in the household has also been improved. The scientific and accurate prediction of the power system load is one of the bases and guarantees for the power department to make the correct decision. It will be helpful for the power generation department to make the relevant plan of economic and reasonable system power generation, effectively control the cost, according to the forecast result of the power network data, make the powerful power planning plan for the relevant department, provide the reliable data source; It is very beneficial to the management of planned electricity, to obtain the situation of demand analysis, to promote the marketization of the electric power industry, to facilitate the dispatch department to arrange ahead of time for the electricity, and not only to benefit the power revenue, It is also very helpful to the promotion of social and public benefits, and it is also very powerful to ensure the stable operation of the power grid. In this paper, a forecasting method of power grid load is proposed, which is based on the grey model. Through continuous improvement, this forecasting method is preferred to weight the index, break the arrangement of the original load, and then study the busbar which is greatly affected by the day type and meteorological conditions. Based on the theory of grey correlation degree, the daily eigenvector is established, and on this basis, some optimal similar days are found. The sample of the last historical load is the bus load of the selected day above. According to the law of far small and near large, the load sample is selected when the load sample is selected. This method not only makes full use of the useful information of the data in the sample, but also reduces the randomness of the data in the sample. By studying the example of Changshou Power supply Branch in Chongqing, this paper fully proves the practicability and scientific nature of the classification of electric power load, and to some extent, The method of load forecasting proposed in this paper has high practical benefit.
【学位授予单位】：湖北工业大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TM714;TM715

【参考文献】