【作者】 刘歆；

【导师】 王洪礼；

【作者基本信息】 天津大学 ， 管理科学与工程， 2008， 博士

【摘要】 电力工业是国民经济的基础支撑产业,电力供应能否满足日益增长的国民经济发展和人民生活水平提高的需要,是关系到国计民生的重要问题。高质量的、准确的电量需求预测和供应规划,是电力工业与国民经济协调发展的前提。本文的主要研究内容及成果包括以下几个方面:首先在分析天津市社会经济发展状况的基础上,阐述了天津市电力供需状况、电网及电源概况、电网装备水平以及存在的问题。并运用支持向量机理论对天津市短期(半年)的电力需求状况进行了预测研究。对中长期(四年)的用电需求分别采用了四种方法进行预测:首次采用随机梯度回归方法对居民用电量进行预测;并首次采用多元自适应样条方法对第三产业用电量进行预测;采用模糊人工神经网络对第一产业的用电量进行预测,采用逐步回归法对第二产业用电量进行预测,最后得到了天津市中长期用电总量的科学预测。本文还针对供电企业目前在资产管理方面存在的诸多问题,首次重新定义了供电企业资产最小统计单位,并引入二级模糊综合评判和多目标决策法等科学的投资决策方法来改善目前供电企业的资产投资管理,在研究和分析供电分公司业务及生产特点的基础上,采用4GL语言与面向对象技术结合,开发出了供电企业资产管理信息系统。最后从电力行业的发展环境分析入手,阐述了国内电力行业的宏观发展现状,并对天津市电力行业的发展现状进行了分析,在此基础上,提出了天津电力公司的“十一五”发展策略。更多还原

【Abstract】 Power industry is the elementary industry of the national economical system, and whether the electricity supply can meet the demand of the economic development and the progress of the people’s living standard is a key problem to sustainability of the whole world. A high quality and accurate of the power demand prediction is the prerequisite of the sustainable development between the power industry and national economy. The detailed research work is as follows:First, based on the analysis of the society and economical status in Tianjin, the thesis discussed the power supply and demand, electric network and source of power, status-of-equipment of electric network and the open questions. Next using the support vector machine theory, we predicted the short period(half year)demand of the power in Tianjin.The thesis used four different methods to predict the mid and long term power demand: we predicted the power demand of the residents first with the stochastic gradient regression method; and predicted the power demand of the third industry with multivariate adaptive spline regression; then we predicted the power demand of the first industry with fuzzy artificial neutral network, and the power demand of the second industry with stepwise regression method. Last, we got the mid-long term power demand of Tianjin.To the many problems exist in the assets management of the power supply company, the thesis first redefined the minimum statistical unit of the power company’s assets, and improved the management of the investment of the power company with introduction the scientific investment decision method such as second-level fuzzy integrated diagnosis and multi-objective decision. Then based on the analysis of the characters of the professional work in power company, we development the management system of the power company assets with 4GL language and object-oriented technology.By analyzing the development environment of the power industry, the thesis discussed macro development status of the national and Tianjin power industry, and then put forward the development strategy of the Tianjin Electronic Power Company.更多还原