基于时延特性建模的多断面短时交通流预测

针对现有交通流预测方法未充分考虑多断面车流演变规律，提出基于时延特性建模的时空相关性计算方法. 该方法采用对不同断面、不同时刻交通流的分布相似性度量，对输入的车辆到达数据序列进行切割构建时空相似度矩阵，得到相邻断面之间的时延参数. 基于时延特性建模，将多断面之间的流量信息进行融合，使用长短时记忆(LSTM)网络进行流量预测. 通过对实际路段数据的预测和结果分析，验证所提方法的有效性和实用性.     

基于多源数据的公交车能耗碳排放测算模型

公交车能耗碳排放强度与车辆、线路和驾驶员有显著相关关系，为精准刻画其能耗碳排放强度特征，整合OBD监测数据、加油(气)数据、运营排班数据等多源数据资源. OBD监测数据和加油(气)数据呈显著的线性关系，证明修正后的OBD监测数据可满足分析要求. 搭建“速度-能耗碳排放强度曲线”测算模型，幂函数关系的拟合优度R2 =0.972 6 为最高. 实证研究发现，平均速度在10~60 km/h 变化时，液化天然气(LNG)车比柴油车能耗碳排放强度高 3.3%~33.7%，双层车比铰接车高2.4%~13.3%；LNG铰接车在不同线路、相同速度下的强度相差9.6%；不同驾驶员在相同线路的能耗碳排放强度可相差24.2%. 模型为各城市基于多源数据开展公交能耗碳排放目标设定提供数据支撑.     

基于LSTM的舰船运动姿态短期预测

舰船的六自由度运动状态形成复杂的非线性过程,运动姿态会受到耦合作用、不定周期、噪声信号以及混沌特性等因素的干扰,因此很难得到精确的预测结果.为了提升舰船运动姿态的预测精度,利用舰船时间序列的特点,建立了基于长短期记忆单元(LSTM)模型,对其进行了舰船姿态预测仿真,将结果与时间序列分析法的结果进行对比.实例分析表明:基于LSTM模型的预测方法具有精确度高、易实现的特点.这为舰船运动短期预测提供了一个新的思路和方法.     

内河船舶波浪载荷研究

利用内河波浪散布图和修订波浪谱对内河系列船舶进行了波浪载荷的计算预报,首次分析了内河船舶波浪载荷的规律和特点,并与俄罗斯内河登记局(RRR)和法国-德国船级社(B-G)的内河规范载荷对应的等效模数要求进行了比较分析。结合实船测试结果表明：波浪载荷变化规律与国外同类规范一致,预报波浪载荷水平与测试结果及国外同类规范要求相当。内河船舶波浪载荷预报应考虑航速和水深的影响,载荷预报结果可以作为制定内河波浪载荷公式的基础。     

面向智能船舶的容错卡尔曼融合估计方法

针对智能船舶多传感器系统因未知海洋环境干扰和设备间干扰等因素导致的一个或数个传感器产生随机间歇性故障从而导致融合估计结果出现偏差甚至失真的问题,设计1种基于四分位滤波的容错方法,并针对该方法导致的观测时滞问题设计1种预报方法,提前预报观测值,进而抵消容错方法导致的时滞问题。此外,针对多传感器之间的互协方差难以准确估计的问题,采用CI融合估计方法进行融合估计。为验证算法的有效性和融合估计的精度,对带有间歇性故障的两传感器系统进行仿真试验,并与按矩阵、按对角阵和按标量3种分布式融合估计方法得到的结果进行对比。4种方法的均方误差系数大小对比结果显示,对于带间歇性故障的多传感器系统,设计的融合滤波不仅具有鲁棒性,而且具有较高的融合精度。     

提高TSP探测精度的途径

TSP作为目前最先进的隧道地质超前预报探测仪器，得到了广泛的应用。但是由于在现实中存在各种问题，从而导致该仪器的预测精度受到了极大限制，主要阐述如何提高其预测精度，更好地为隧道的建设服务。     

A methodology for evaluating the performance of model-based traffic prediction systems

Model-based traffic prediction systems (mbTPS) are a central component of the decision support and ICM (integrated corridor management) systems currently used in several large urban traffic management centers. These models are intended to generate real-time predictions of the system’s response to candidate operational interventions. They must therefore be kept calibrated and trustworthy. The methodologies currently available for tracking the validity of a mbTPS have been adapted from approaches originally designed for off-line operational planning models. These approaches are insensitive to the complexity of the network and to the amount and quality of the data available. They also require significant human intervention and are therefore not suitable for real-time monitoring. This paper outlines a set of criteria for designing tests that are appropriate for the mbTPS task. It also proposes a test that meets the criteria. The test compares the predictions of the mbTPS in question to those of a model-less alternative. A t-test is used to determine whether the predictions of the mbTPS are superior to those of the model-less predictor. The approach is applied to two different systems using data from the I-210 freeway in Southern California.     

基于三点模型的柴油机热工参数预测

准确预测热工参数劣化趋势是对柴油机实施预测维修的前提。已有预测技术均视不同时段信息所具有的建模价值相等,忽略了近期信息比远期信息更能反映设备当前运行态势这一客观事实,导致预测模型难以准确描述柴油机热工参数的真实变化规律。针对该问题,提出一种基于三点模型的柴油机热工参数预测方法,并采用新陈代谢法同步更新建模数据,以提高建模数据的利用效率及预测精度。与灰色系统GM(1,1)模型预测精度的对比结果表明,三点模型能够有效处理不同时段信息在反映设备运行态势能力上的差异,得到更为准确的预测结果。     

基于AIGA-WLSSVM的埋地管道腐蚀速率预测方法

为了降低埋地管道腐蚀影响因素之间的复杂相关性,提高腐蚀预测精度,文中提出一种基于自适应免疫遗传算法-加权最小二乘支持向量机(AIGA-WLSSVM)的埋地管道腐蚀速率预测建模方法,并采用AIGA优化模型参数,进一步提高模型的学习能力和稳定性。最后通过实例分析验证了AIGA-WLSSVM建模方法在埋地管道腐蚀速率预测中的可行性和有效性,为埋地管道的检修与更换提供参考。     

Real-time trip purpose prediction using online location-based search and discovery services

The use of smartphone technology is increasingly considered a state-of-the-art practice in travel data collection. Researchers have investigated various methods to automatically predict trip characteristics based upon locational and other smartphone sensing data. Of the trip characteristics being studied, trip purpose prediction has received relatively less attention. This research develops trip purpose prediction models based upon online location-based search and discovery services (specifically, Google Places API) and a limited set of trip data that are usually available upon the completion of the trip. The models have the potential to be integrated with smartphone technology to produce real-time trip purpose prediction. We use a recent, large-scale travel behavior survey that is augmented by downloaded Google Places information on each trip destination to develop and validate the models. Two statistical and machine learning prediction approaches are used, including nested logit and random forest methods. Both sets of models show that Google Places information is a useful predictor of trip purpose in situations where activity- and person-related information is uncollectable, missing, or unreliable. Even when activity- and person-related information is available, incorporating Google Places information provides incremental improvements in trip purpose prediction.