节点文献

基于PCA数据降维和神经网络的能源审计对标评价方法的研究

Standard Evaluation Method Research Based on PCA and ANN for Energy Audit

【作者】 党银宁

【导师】 周乃君;

【作者基本信息】 中南大学 , 动力工程, 2011, 硕士

【摘要】 随着我国社会经济的飞速发展,资源的约束越来越突出,在这种情况下,为了保证经济“又好又快”的发展,我们国家经济结构要面临转型,即从过去那种“高投入、高能耗、高污染、低产出”的模式向“低投入、低能耗、低污染、高产出”转变。但,目前我国仍然沿袭了以消耗大量资源为特征的传统发展模式,导致能源浪费现象十分严重。这不仅不利于企业绩效的提高和企业的发展,而且破坏了环境,影响了国民经济的可持续发展。世界各国普遍采用能耗评价体系来评价企业能耗水平,但现行企业能耗评价体系在指标设定上着重经济指标,对于环保指标设置的不够全面,且指标冗余性较大,不利于企业能源审计方法的推广,这将导致能源审计对国家能源与环境可持续性发展的导向性变差。因此,加强现有能源审计体系的环保指标,优化指标体系的内部结构,并发展新的能源审计工具,对推进和加快能源审计方法在企业的应用具有重要的现实意义。本文以现有企业能源审计评价体系和国家可持续发展的最新战略部署为基础,结合笔者多年从事能源审计工作的切身经验,从体系指标在使用中存在的问题出发,提出了以下改进:1)为降低现行能源审计指标评价体系的指标冗余度,提出用PCA主成分分析的方法降低数据维度,所处理得到的主元保留了原数据的绝大部分特征,降低了能源审计数据处理的难度。2)根据国家政策和对标管理理念,提出了基于BP神经网络的人工智能能源审计方法,并对6家钢铁企业的同年能源审计数据进行了应用。结果表明,该模型能够达到较好的模式识别效果,能有效降低企业对能源审计机构的业务依赖性。3)为了全面监督企业低碳化耗能的水平,引入了能源审计新指标-对标低碳系数,并对企业对标低碳系数进行了分级。本文的研究成果为完善能源审计体系和审计数据分析提供了新的思路。

【Abstract】 With China’s rapid social and economic development, resource constraints become increasingly prominent, in this case, in order to ensure economics’ "good and fast" development, China’s economic structure should be a transition. That is, from the past that the "high input, high energy consumption, high pollution and low output "model to the" low input, low energy consumption, low pollution, high-yield" change. However, China is still consuming a large amount of resources which characterized by the traditional development model, resulting in energy waste seriously. This is not conducive to the improvement of enterprise performance and enterprise development, but also damaged the environment, affecting the sustainable development of the national economy.Nowadays, world widely used energy evaluation system to evaluate enterprise energy consumption level has been used, current enterprise energy evaluation system in the index set, focusing on economic indicators for environmental protection index set is not comprehensive, and index redundancy bigger, unfavorable to enterprise energy audit method of promotion, this will lead to energy audit of national energy and environmental sustainability guideness becomes poor. Therefore, strengthening existing energy audit system, optimizing the environmental protection index of the indicator system of internal structure, and developing new energy audit tools, have important practical significance to advancing and accelerating energy audit methods in enterprise applications.Based on the existing enterprise energy audit evaluation system and the latest strategy of sustainable development, combining the work experience of the author in energy audit, the following improvement from system parameters angle were put forward:1) In order to reduce the index redundancy of the current energy audit evaluation index system, the PCA (principal component analysis) method to reduce data dimensions were proposed, which can get the most characteristics of original data, reducing the difficulty of energy audit data processing. 2) According to the national policies and the standard management concept, a new energy audit method based on BP neural network of artificial intelligence was introduced and, this new method was applied in 6 companies iron & steel enterprises’energy audit data processing.3) In order to fully supervise enterprises’low energy consumption level, a new index of energy audit, low carbon coefficient via standard was introduced and the index for enterprises was graded already.The result of this dissertation provides new ideas for perfecting the energy audit system and audit data analysis methods system.

  • 【网络出版投稿人】 中南大学
  • 【网络出版年期】2012年 01期
  • 【分类号】F239.6;F206;TP183
  • 【被引频次】4
  • 【下载频次】328
  • 攻读期成果
节点文献中: 

本文链接的文献网络图示:

本文的引文网络