A new methodology for multi-step-ahead forecasting was proposed herein which combined the wavelet transform(WT), artificial neural network(ANN) and forecasting strategies based on the changing characteristics of available parking spaces(APS). First, several APS time series were decomposed and reconstituted by the wavelet transform. Then, using an artificial neural network, the following five strategies for multi-step-ahead time series forecasting were used to forecast the reconstructed time series: recursive strategy, direct strategy, multi-input multi-output(MIMO) strategy, DIRMO strategy(a combination of the direct and MIMO strategies), and newly proposed recursive multi-input multi-output(RECMO) strategy which is a combination of the recursive and MIMO strategies. Finally, integrating the predicted results with the reconstructed time series produced the final forecasted available parking spaces. Three findings appear to be consistently supported by the experimental results. First, applying the wavelet transform to multi-step ahead available parking spaces forecasting can effectively improve the forecasting accuracy. Second, the forecasting resulted from the DIRMO and RECMO strategies is more accurate than that of the other strategies. Finally, the RECMO strategy requires less model training time than the DIRMO strategy and consumes the least amount of training time among five forecasting strategies.
The impacts of four different car-following types on rear-end crash risks at a freeway weaving section were evaluated using trajectory data, in which Type 1 represents car following car, Type 2 represents car following truck, Type 3represents truck following car and Type 4 represents truck following truck. The time to collision( TTC) was introduced as the surrogate safety measure to determine the rear-end crash risks. Then, the trajectory data at a freeway weaving section was used for the case-controlled analysis. Three logistic regression models were developed with different TTC thresholds to quantify the impacts of different car-following types. The explanatory factors were also analyzed to investigate possible reasons for the results of logistic regressions. Results showthat the rear-end crash risk of Type3 is 3. 167 times higher than that of Type 1 when the TTC threshold is 2 s. However, the odds ratios of Type 2 and Type4 are both smaller than 1, which indicates a safer condition.The analysis of explanatory factors also shows that Type 3 has the largest speed differences and the smallest net gaps. This is consistent with vehicle operation features at a weaving section and is also the reason for the larger rear-end crash risks. The results of this study reflect the mechanism of rear-end crash risks of different car-following types at the freeway weaving section.
As an essential component of bus dwelling time, passenger boarding time has a significant impact on bus running reliability and service quality. In order to understand the passengers’ boarding process and mitigate passenger boarding time, a regression analysis framework is proposed to capture the difference and influential factors of boarding time for adult and elderly passengers based on smart card data from Changzhou. Boarding gap, the time difference between two consecutive smart card tapping records, is calculated to approximate passenger boarding time. Analysis of variance is applied to identify whether the difference in boarding time between adults and seniors is statistically significant. The multivariate regression modeling approach is implemented to analyze the influences of passenger types, marginal effects of each additional boarding passenger and bus floor types on the total boarding time at each stop. Results show that a constant difference exists in boarding time between adults and seniors even without considering the specific bus characteristics. The average passenger boarding time decreases when the number of passenger increases. The existence of two entrance steps delays the boarding process, especially for elderly passengers.
为获取居民公交出行的换乘信息,设计了一套基于多分类支持向量机(multi-class support vector machine)的公交换乘识别方法.通过融合GPS数据和公交IC卡数据获取训练样本,利用多分类支持向量机进行样本训练,选取最佳训练样本量,并采用网格搜索法结合粒子优化算法对模型参数进行标定,以获取最优SVM分类模型.测试结果显示模型分类精度可达90%.以佛山市公交车GPS数据和IC卡数据对算法进行验证,并获取公交换乘量、公交换乘比例等基本换乘数据.结果表明:算法可在少样本条件下完成公交换乘识别,且分类识别精度高,尤其适用于公交线网复杂的大城市公交换乘识别,有助于在公交前期规划时进行线路布设和枢纽选址.
