目录
摘要Ⅲ
关键词Ⅲ
AbstractⅣ
引言
引言1
1文献综述1
1.1碳交易价格影响因素方面的研究 1
1.2价格预测方面的研究 2
2理论分析2
2.1碳交易价格影响因素的理论分析 2
2.1.1气候环境3
2.1.2能源价格3
2.1.3政策制度3
2.1.4经济市场3
2.1.5国际碳交易价格3
2.2碳价预测的理论分析 4
3基于因子分析研究碳价影响因素 4
3.1因子分析法简介4
3.2变量的选择与数据处理4
3.2.1变量选择4
3.2.2数据处理5
3.3实证分析5
3.3.1因子分析适用性检测6
3.3.2判断公共因子个数6
3.3.3提取公共因子7
3.3.4因子旋转8
3.3.5因子得分9
4基于BP神经网络模型的碳价预测 10
4.1BP神经网络简介10
4.2数据选择与处理10
4.3实证分析11
4.3.1构建神经网络结构11
4.3.2BP神经网络训练11
4.3.3准确性分析12
5结论13
致谢14
参考文献15
附录A 月均碳交易价格及其影响因素数据16
附录B 公共因子与碳价数据17
附录C R语言源代码18
中国碳交易价格影响因素及预测
摘 要
低碳减排是可持续发展的重要组成部分，控制温室气体的排放不仅是人们所需要承担的社会责任，更是联合国对各个国家下达的任务。碳排放权交易市场因此产生，我国碳交易市场起步晚，空间差异大，且数据统计不完整，对我国来说创建整体碳交易市场、全面控制碳交易走向是十分有必要的。本文基于我国数据相对完整的五个碳交易试点，研究碳交易价格的影响因素以及对价格进行预测。根据查阅大量文献总结归纳 *51今日免费论文网|www.51jrft.com +Q: @351916072@ 
出影响我国碳交易价格的因素主要有气候环境、能源、政策、国际碳市场、经济市场五个方面，从中选取降水量、空气质量、最高气温、最低气温、煤炭价格指数、天然气价格、政策制度、居民消费价格指数、欧盟排放配额九个影响变量，运用因子分析法进行降维，归结成五个公共因子。在此基础上，建立BP神经网络模型对碳交易价格进行预测，利用五折交叉验证计算出碳价预测的平均相对误差为12.342%，证实了BP神经网络的适用性，能为今后的的相关研究提供方向，同时对碳交易市场也有一定的促进作用。
INFLUENCING FACTORS AND PREDICTION OF CARBON TRADING PRICE IN CHINA
ABSTRACT
Low carbon emission reduction is an important part of sustainable development. Controlling greenhouse gas emissions is not only the social responsibility that people need to bear, but also the task that the United Nations assigns to each country. Therefore, the carbon emission trading market also came into being. Chinas carbon trading market started late. At present, there are only eight carbon trading pilots, with large spatial differences and incomplete data statistics. It is very necessary for China to establish an overall carbon trading market and comprehensively control the trend of carbon trading. Based on five carbon trading pilot projects with relatively complete data in China, this paper studies the carbon trading price and its development situation from the overall perspective. According to a large number of literature review, it concludes that the carbon trading price in China is mainly affected by five aspects: energy price, climate and environment, policy and system, international carbon market and economic market, and selects precipitation, air quality, maximum temperature and maximum temperature Low temperature, coal price index, natural gas price, policy and system, consumer price index and EU emission quota are nine influencing variables. Factor analysis method is used to reduce dimensions, which can be summed up as five public factors. On this basis, this paper uses BP neural network model to predict carbon trading price, and uses 50% cross validation to calculate the average relative error of carbon price prediction is 12.342%, which proves the applicability of BP neural network, which can provide direction for future research, and also has a certain role in promoting carbon trading market.

原文链接：http://www.jxszl.com/jsj/sxtj/606718.html