面向居民出行的空气质量可视分析系统研究与实现

发布时间：2018-09-14 16:42

【摘要】：随着科技的发展,数据被大量采集和存储,数据量呈爆炸式增长,如何从这些数据中挖掘出有价值的信息成为急需解决的问题。在面对庞大并且复杂的数据时,传统的数据挖掘和数据分析方法在探索数据时显得力不从心,因此我们需要一种行之有效的方法来解决这些问题。近些年来,作为以可视交互界面为基础的分析推理科学,可视分析为数据挖掘、数据分析提供了一种全新的手段,它以交互性、可视性等特点受到广大研究者的热烈欢迎,已渐渐成为研究热点。伴随着我国工业化进程的发展,以硫化物(SOx)、氮化物(NOx)、臭氧(O3)、碳化物(COx)、颗粒物(粒径小于等于10μm和2.5μm)为主的工业排泄物对空气质量造成的污染问题日益严重,对人们的日常出行和生活造成极大影响,据调查显示,当空气质量较差时,人们更愿意待在室内以减少非必须出行行为。因此针对空气质量与居民出行的可视化研究对于探究空气质量和居民出行之间的关系具有重要意义,它不仅可以为探索居民的日常出行行为提供重要参考,还可以引起交通、医疗等相关部门对空气质量的重视。因此探究空气质量和居民出行的可视化研究无论在理论上还是在实际应用中都具有非常重要的研究价值。本文为探究空气质量对居民出行的影响,设计并实现了面向居民出行的空气质量可视化分析系统,深度探索空气质量与居民出行的关系。系统以杭州市的空气质量、温度、POI分布和滴滴打车难易度数据为例,定义并计算POI带权活跃度、POI带权活跃度偏移率和基于k-means的POI聚类,并以可视化的方式提供可交互的分析方法。系统设计并实现了三个可视分析组件:Barbox、Stream和Smp。其中,Barbox为柱状-箱线图组件,提供空气质量和温度的可视化组件;Stream,堆积-流图组件,实现自定义类型POI带权活跃度双尺度导航视图;Smp,散点矩阵-日历热图-GeoMap组件,实现多变量分析和地理分布的视图。在各个组件的设计上,系统均融合了颜色、交互等特性,系统通过三个针对空气质量和居民出行的可视化组件,可以让用户在交互中探索空气质量对居民出行的影响以及空气质量和居民出行模式分析。
[Abstract]:With the development of science and technology, data is collected and stored in large quantities, and the amount of data increases explosively. How to mine valuable information from these data becomes a problem that needs to be solved urgently. In the face of large and complex data, the traditional methods of data mining and data analysis seem to be unable to explore the data, so we need an effective method to solve these problems. In recent years, as a science of analytical reasoning based on visual interaction interface, visual analysis provides a new method for data mining and data analysis, which is warmly welcomed by many researchers for its characteristics of interactivity and visibility. Has gradually become a research hotspot. With the development of industrialization in China, the pollution of air quality caused by industrial excreta, which is dominated by sulphide (SOx), nitride (NOx), ozone (O3) and carbide (COx), particles (particle size less than 10 渭 m and 2.5 渭 m), is becoming more and more serious. It has a great impact on people's daily travel and life. According to the survey, when the air quality is poor, people prefer to stay indoors in order to reduce the non-essential travel behavior. Therefore, the visualization research on air quality and resident travel is of great significance to explore the relationship between air quality and resident travel, which can not only provide an important reference for exploring residents' daily travel behavior, but also cause traffic. Health care and other related departments to pay attention to air quality. Therefore, it is of great value to explore the visualization of air quality and resident travel both theoretically and practically. In order to explore the effect of air quality on residents' travel, a visual analysis system for air quality is designed and implemented in this paper. The relationship between air quality and residents' travel is explored in depth. Taking the air quality, temperature distribution and Didi Taxi ease data of Hangzhou as an example, the paper defines and calculates the shift rate of POI weighted activity and POI clustering based on k-means. And provide interactive analysis methods in a visual way. The system has designed and implemented three visual analysis components: Barboxbus Stream and Smp.. The Barbox is a columnar box diagram component, a visualization component for air quality and temperature, a stack flow chart component, a custom type POI with weighted activity dual scale navigation view, a scatter point matrix, calendar heat map and a GeoMap module. Multivariate analysis and view of geographical distribution are implemented. In the design of each component, the system combines the characteristics of color, interaction and so on. The system adopts three visual components aimed at air quality and residents' travel. It allows users to explore the impact of air quality on residents' travel and the analysis of air quality and travel patterns.
【学位授予单位】：浙江工商大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP393.09

【相似文献】