摘要
为了科学准确地预测近期公交客流量,提出了应用灰色马尔可夫模型进行预测的方法。利用历史数据建立灰色GM(1,1)模型,通过确定系数可获得公交客流量的时间响应序列及还原值的表达式,从而可获得未来年公交客流量的发展序列值,结合马尔可夫链过程将随机序列状态划分为3类,通过确定状态转移矩阵可获得序列处于各状态的概率值及与各状态对应的预测中值,最终求得各序列的修正值。通过铜陵市公交客流的历史数据,预测了近3年的公交客流。实例证明该预测模型具有较高的精度,能够指导公交经营管理者近期的决策行为。
In order to forecast the short-term public transit volume accurately, a forecasting model applying Gray-Markov is developed. Based on historical data, this method developed gray GM （ 1, 1 ） model, determined coefficient and gained the expression of time response series and reduction value, obtained short-term transit volume sequences, then divided series into three states, ascertained transition matrix and attained states probability and median value, finally acquired modified value. A forecast example is given by using historical data of public transit volume of Tonglin, which proves that this method is accurate and can help operators and managers to make decisions.
出处
《公路交通科技》
CAS
CSCD
北大核心
2007年第9期120-123,共4页
Journal of Highway and Transportation Research and Development
基金
教育部科学技术研究资助项目(105086)
江苏省自然科学基金重点资助项目(BK2005205)
关键词
交通工程
客流预测
灰色马尔可夫
公交
traffic engineering
volume forecasting
Gray-Markov
public transit
