基于神经网络的PM2_5质量浓度预测研究_付彦丽.docx
《基于神经网络的PM2_5质量浓度预测研究_付彦丽.docx》由会员分享,可在线阅读,更多相关《基于神经网络的PM2_5质量浓度预测研究_付彦丽.docx(72页珍藏版)》请在淘文阁 - 分享文档赚钱的网站上搜索。
1、 论文题目: 基于神经网络的 PM2.5 质量浓度预测研究 学科门类:工学 一级学科:计算机科学与技术 培养单位:电气与信息工程学院 硕士生:付彦丽 导 师:杨云教授 2016 年 6 月 The Research of Prediction PM2.5 Mass Concentration Based on Neural Network A Thesis Submitted to Shaanxi University of Science and Technology in Partial Fulfillment of the Requirements for the Degree of M
2、aster of Engineering Science By Yanli Fu Supervisor: Prof. Yun Yang June 2016 基于神经网络的 PM2.5 质量浓度预测研究 摘要 目前我国己经进入了后工业化的时代,但与之相随而来的是环境与发展 的极度不和谐,环境在发展过程中受到了巨大的破坏,尤其表现在我们赖以 生存的空气、土壤、水等方面。近些年,空气质量急剧恶化,雾霾以越来越 高的频率、越来越大的范围出现在全国各地,通过呼吸系统对人体健康造成 严重的危害。而随着生活水平的提高,人们对于环境的关注越来越高,对于 自己生活区域的环境质量的要求也随之升高,空气质量成为人们
3、关注的焦点 问题。本文以此为出发点,对我国现有累积的大量历史空气质量数据进行挖 掘,寻找 PM2.5 与其他空气污染物间的非线性关系,针对 PM2.5 小区域范 围内质量浓度差距大、民众难以及时得到预报信息等问题,对 PM2.5 每小 时的质量浓度进行预测。本文主要完成的工作如下: (1) 根据 PM2.5 的形成原因,分析可以引起其质量浓度变化的因素, 建立数学模型。可入肺颗粒物 PM2.5 的成因复杂,组成成分包含直接排放 的一次粒子以及由光化学反应形成的二次粒子,主要包括有机碳、元素碳、 土壤尘、硫酸铵或亚硫酸铵、硝酸铵、铵盐、半挥发性有机物等。结合监测 站所监测空气污染物,最终选取 C
4、O、 N02、 03-1、 03-8、 S02、 PM10 六种 污染物,作为 PM2.5 质量浓度的影响因子。 (2) 获取空气环境污染物历史监测数据,并对数据进行预处理。在监 测数据中,会出现偶然的异常数据条,比如监测值全为零的空数据。在使用 数据前需要对异常数据进行剔除,以免影响之后的预测结果。对剔除后数据 的进行归一化处理,让不同数量级的数据在一个范围内取值,避免由于数量 级的差距产生的预测误差。对样本数据进行划分,以合适的比例将其分为训 练数据集和测试数据集。选择合适的预测分析工具,由于本研究中需要对大 量数据进行高速、高效的矩阵运算处理,故选择 Matlab 作 为主要工具。 (3
5、) 研究神经网络,分析神经网络的原理、执行流程、参数设置、运 算过程,使用反向传播神经网络作为基础进行预测。通过经验公式以及试错 法确定该网络的最佳隐含层神经元个数,设计最佳网络结构;根据函数的适 用范围,选择合适的传递函数、训练函数、学习函数。使用训练数据对网络 进行训练,在网络训练结束后,使用 sim ()函数以及测试集对训练好的网 络进行测试预测。最后,将网络的预测结果进行统计,计算其预测结果的可 接受度、相对误差,分析网络性能、优缺点。 (4) 针对 BP 神经网络的缺点,提出改进的神经网络,使用模糊系统、 遗传算法对神经网络进行优化。模糊系统将神经网络进行模糊化处理,即模 糊其网络输
6、入及连接权值,明晰网络的推理过程,解决神经网络求解问题时 的黑箱特性。将神经网络的输入 /输出作为模糊系统的输入 /输出,用神经网 络的隐含节点表示隶属度函数和模糊规则。遗传算法对神经网络的初始连接 权进行优化,提高网络的全局搜索能力、收敛速度,解决神经网络容易陷入 局部最小的问题。该算法需要根据网络的进化目标,选择与其相匹配的个体 适应度函数以及进行遗传操作的方法,最终,将最优的初 始连接权赋值给神 经网络。 (5) 在 Matlab 下编写完整的优化后的神经网络的 .m 程序,将预处理 后的数据输入网络进行训练、测试。分别对拟合结果以及测试结果进行分析, 在保证网络没有出现过拟合的情况下,
7、对三种方法的预测结果进行统计、对 比分析。数据结果表明,使用遗传算法优化的神经网络在 PM2.5 质量浓度 预测上的表现最佳,提高了预测结果的精确度,降低了其误差率。 关键词: PM2.5 质量浓度,预测,神经网络, T-S 模糊模型,遗传算法 THE RESEARCH OF PREDICTION PM2.5 MASS CONCENTRATION BASED ON NEURAL NETWORK ABSTRACT At present our country has been entered into the post-industrial era, but at same time extre
8、me discord happened between environment and development. The environment has been huge destroyed in the development process, especially in the air, soil, water and so on several aspects. In recent years, air quality has been rapidly exacerbated, haze appears in countries at higher frequencies and la
9、rger ranges, and it cause serious harm to human health through respiratory system. With the improvement of human living standards, people start to pay more and more attention on the environment and the requirements of environmental quality of their own living regional also increase, air quality prob
10、lems become the focus issue. This article use it as a starting point to mining air quality monitoring data, and find nonlinear relationship between PM2.5 and other air pollutants. Aim at problems such as,PM2.5 mass concentration has big gap between different areas, people are difficult to get forcas
11、t information timely and so on, to forcast PM2.5 mass concentration hourly, the paper main work is as follows: (1) According to the formation of PM2.5, the reason for the change of mass concentration was analyzed and the mathematical model was established. The formation of PM2.5 is complex, which co
12、ntains a direct emission of a particle and two particles formed by the photochemical reaction. Mainly including organic carbon, elemental carbon, soil dust, ammonium sulfate or ammonium sulfite, ammonium nitrate, ammonium salt, semi volatile organic compounds, etc. Combined with the monitoring stati
13、ons to monitor air pollutants, the final selection of CO, NO2, O3-I, O3-B, SO2, PM 10 six kinds of pollutants, as the impact factor of PM2.5 mass concentration. (2) Getting environmental monitoring data of air pollutants, and preconditioning of the data. There are some occasional abnormal data in mo
14、nitoring data, such as empty data which all values are zero. Before use them we need to remove abnormal data to avoid their affect in prediction results. To normalize after excluding, normalized can make the different magnitude data in in the same range, avoid the prediction error caused by the gap
15、of magnitude. The sample data are divided into training data set and test data set with appropriate proportion. Choosing Matlab as the main tool for predictive analysis, because of its high speed, efficiency and low cost in vast matrix operation (3) Research of neural networks, analysis its principl
16、e, executing process, parameter setting, calculating process and etc. Using back-propagation neural networks as predicting basis. Then through empirical formula and trial-and-error method to determine the best hidden layer neuron number, design the best network structure and select the appropriate t
17、ransfer function, training function, learning function.Using the training function to train the network, after network training using sim( ) with a trained network to predict test data. Statistic the network forcast results, calculating its forecast acceptability, relative error, analyze network per
18、formance, advantages and disadvantages. (4) Aim at the disadvantage of back-propagation neural natwork, proposed a improved neural networks, using fuzzy system and genetic algorithm to optimize neural networks. Fuzzy system fuzzing the neural network input and connection weight, clear the reasoning
19、process of neural network, solve the black-box nature of neural network. Using neural networks input/output as flizzy systems input/output,using neural networks hidden nodes to express membership functions and fuzzy rules. Genetic algorithm optimized the initial connection weights of neural network
20、and improve the global searching ability, convergence speed of the network, solve the problem of neural network is easy to fall into local minimum. Accroding to the evolution target to select individual fitness function and genetic operation methods. Finally, assign the optimal initial connection we
- 配套讲稿:
如PPT文件的首页显示word图标,表示该PPT已包含配套word讲稿。双击word图标可打开word文档。
- 特殊限制:
部分文档作品中含有的国旗、国徽等图片,仅作为作品整体效果示例展示,禁止商用。设计者仅对作品中独创性部分享有著作权。
- 关 键 词:
- 基于 神经网络 PM2_5 质量 浓度 预测 研究 付彦丽
限制150内