论文部分内容阅读
以终端区交通流为研究对象,基于航迹谱聚类结果定义并提取交通流特征,分析了特征间关系与交通流相态演化规律,发掘了实测数据下交通流的自由态、平稳态与拥堵态,以此为先验经验进一步设计因子分析与遗传期望最大化模糊聚类算法相结合的终端区交通流态势识别方法,实现对交通流状态影响因素与交通流隐性特征的提取,选取典型繁忙终端区的实测数据进行验证。分析结果表明:基于客观数据挖掘的交通流态势识别方法具有良好的适应性与准确性,自由态、平稳态与拥堵态的模型识别数量分别为6、36、37,管制员判别数量分别为7、40、32,误差率分别为14.3%、10.0%、15.6%,模型识别率均在84%以上;提取的交通流相态及时空特征可从局部细节构建终端区整体运行态势,为终端区流量时空分布调配与进离场程序优化提供支撑。
Taking the terminal traffic flow as the research object, the traffic flow characteristics are defined and extracted based on the results of the cluster analysis. The evolution of the relationship between the characteristics and the phase flow of the traffic flow is analyzed, and the free and steady states of the traffic flow are explored. And congestion state as a priori experience to further design factor analysis and genetic algorithm to maximize the combination of fuzzy clustering terminal traffic flow pattern recognition method to achieve the influence of traffic flow state and traffic flow characteristics of the recessive extraction, Select the typical busy terminal area of the measured data to verify. The analysis results show that the recognition method of traffic flow potential based on objective data mining has good adaptability and accuracy. The number of free-form, steady-state and congested states are 6, 36 and 37, respectively, and the number of controllers is 7,40 and 32 respectively, the error rates are 14.3%, 10.0% and 15.6% respectively, and the model recognition rates are above 84%. The extracted traffic flow phase and space-time features can build the overall operation trend of the terminal area from the local details, Regional distribution of space-time distribution and departure procedures to optimize the support.