基于数据挖掘的商业电力负荷预测及用电优化算法研究

发布时间：2018-03-17 11:34

本文选题：数据挖掘　切入点：用电优化　出处：《华侨大学》2015年硕士论文　论文类型：学位论文

【摘要】：随着经济的发展,社会用电量迅速增长,造成电网容量不足,能源利用率不高,环境和资源问题也面临重大挑战。因此,如何提高用电智能化水平,是提高社会用电效率,达到节能减排效果的重要措施。如今商用建筑能源消耗越来越大,对其用电进行智能化管理有重大意义。新兴商业建筑安装了电力能耗监控系统,所监测的数据为研究提供了有利条件。数据挖掘技术能够从海量数据中挖掘出有价值的内在信息,本文利用数据挖掘技术针对商业用户智能用电问题中的负荷预测和用电优化问题进行了研究。首先,在深度分析商业电力负荷特性的基础上,采用粗糙集对影响负荷预测的重要因素进行了约简,并将小波理论支持向量机相结合,建立了应用于商业电力负荷预测的小波支持向量机模型。实例表明,该模型在负荷突变处和波动较大的负荷序列中预测精度优于单一的支持向量机和神经网络预测模型。其次,针对商业用电主要负荷的特性,将其分为可控负荷和固定负荷,对可控负荷中的空调负荷建立用电成本最小化及舒适温度为目标的用电优化模型,对可控负荷中的可转移负荷建立了负荷转移调度优化模型,分别采用非支配遗传算法和自适应遗传算法进行寻优。结果表明,该方法在不影响人们用电舒适度的前提下降低了用电成本。最后,采用了C#与MATLAB混合编程技术开发了商业智能管理软件,并实现了商业电力负荷的预测和用电优化功能。
[Abstract]:With the development of economy, social electricity consumption is increasing rapidly, resulting in insufficient power grid capacity, low energy utilization ratio, and great challenges to environment and resources. Therefore, how to improve the intelligent level of electricity consumption is to improve the efficiency of power consumption in society. Important measures to achieve the effect of energy saving and emission reduction. Nowadays, commercial buildings are consuming more and more energy, so it is of great significance to intelligently manage their electricity consumption. The monitored data provide favourable conditions for research. Data mining techniques can extract valuable internal information from large amounts of data, In this paper, the data mining technology is used to study the load forecasting and power optimization problems in the intelligent power consumption problem of commercial users. Firstly, based on the in-depth analysis of the load characteristics of commercial electricity, The important factors affecting load forecasting are reduced by rough set, and the wavelet support vector machine is combined to establish the wavelet support vector machine model for commercial power load forecasting. The forecasting accuracy of this model is superior to that of single support vector machine and neural network model in load abrupt change and fluctuating load sequence. Secondly, it is divided into controllable load and fixed load according to the characteristics of main load of commercial electricity consumption. In this paper, the optimization model of power consumption with the aim of minimizing power cost and comfort temperature is established for the air conditioning load in controllable load, and the optimal model for load transfer scheduling is established for the transferable load in controllable load. The results show that the proposed method can reduce the power consumption cost without affecting the power comfort of people. The business intelligence management software is developed by using the mixed programming technology of C # and MATLAB, and the function of forecasting and optimizing the commercial power load is realized.
【学位授予单位】：华侨大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TP311.13;TM715

【参考文献】