Traffic information prediction is one of the most essential studies for traffic research, operation and management. The successful prediction of traffic speed is increasingly significant for the benefits of both road users and traffic authorities. However, accurate prediction is challenging, due to the stochastic feature of traffic flow and shallow model structure. Furthermore, environmental factors, such as rainfall influence, should also be incorporated to improve accuracy. Inspired by deep learning, this paper investigates the performance of deep belief network (DBN) and long short-term memory (LSTM) to conduct short-term traffic speed prediction with the consideration of rainfall impact as a non-traffic input. The deep learning models have the ability to learn complex features of traffic flow pattern under various rainfall conditions. To validate the performance of rainfall-integrated DBN and LSTM, the traffic detector data from an arterial in Beijing are utilised for model training and testing. The experiment results indicate that with the combination input of speed and additional rainfall data, deep learning models have better prediction accuracy over other existing models, and also yields improvements over the models without rainfall input. Furthermore, the LSTM can outperform the DBN to capture the time-series characteristics of traffic speed data.

In this study, social force model (SFM) is extended by using a discretisation grid to permit pedestrians to change their desired speed directions dynamically. In reality, other pedestrians may obscure the visions of the behind pedestrians, so the behind pedestrians will be blocked if they insist to walk in the final desired directions calculated by the SFM. So, a dynamic destination choice model is established to provide pedestrians a series of available intermediate destinations. Based on the dynamic destination choice model, the authors use a discretisation grid to represent all the potential moving directions of pedestrians, and model the weight of every potential moving direction. The direction with the maximum weight is selected as the optimal route at that time step. Besides, pedestrians prefer to pass near to crowds with a low relative velocity and choose the low space occupancy route when several routes have the same pedestrian density. So, pedestrian speeds, space un-occupancy, route length, crowd density and object pedestrian ratio are used to develop the weight model. The modified SFM guarantees pedestrians to obtain an available and optimal route. Compared to other models, the proposed models can be used to reproduce the behavior of bidirectional pedestrians more really.

Protection of vulnerable road users has been an increasingly important requirement for today's autonomous vehicles. Analysis of drivers’ response to potential collision with pedestrians can provide guidance for vehicle evasive manoeuvre and data support for traffic collision modelling. Field traffic data were collected by video recording and image processing at urban midblock crosswalks in Beijing, China and Munich, Germany. The drivers’ response to vehicle–pedestrian conflict is discussed by deceleration and acceleration phases considering the lane-based kinematics of vehicles. Some basic characteristics describing the driver behaviour in an evasive process are statistically analysed, including the yielding decision, space history of deceleration, deceleration rate choice, and accelerating behaviour. Results show significant differences in drivers’ yielding decisions. In addition, the onset of braking and the trend of the space history of average deceleration rate during the drivers’ approaching process have been quantitatively proposed. The contributing factors to driver's yielding behaviour are analysed by the Binary Logit Model. The behavioural differences address the needs to make intercultural adaptation when introducing new autonomous vehicle technologies to developing countries. This study provides a basis for establishing evasive decision corrections for drivers or driverless vehicles.

Hard shoulder running (HSR) and queue warning are two active traffic management (ATM) strategies and are commonly used to alleviate highway traffic congestion. This study proposes an optimisation model for HSR operation in coordination with queue warning service during non-recurring traffic accident condition using an updated cell transmission model (CTM). The CTM in this study is updated in two aspects: first, the capacity reduction caused by traffic accident is considered in the CTM; second, a purposive desired lane-changing rate is updated in the CTM considering a driver's movement from one lane to another depending on the lane blocking after receiving queue warning. The HSR operation combined with queue warning is optimised to minimise the total time delay during traffic accident. The model is tested on the A12 highway in the Netherlands under two typical accident scenarios and presents a certain application value.

A new dual-mode power-split device is introduced in this study for application in the transmission system of hybrid electric vehicles. The new system provides two modes of operation and a supervisory control strategy is responsible to determine the vehicle's present operating mode. Based on the selected mode, which is done according to the driving conditions, maximum efficiency and minimum fuel consumption are achieved. Pontryagin's minimum principle has been applied for designing an optimal control strategy during which the Hamiltonian is minimised. The Hamiltonian is calculated by using Pareto maps which provide best operating points of the engine according to the power demand. The simulation results show improvements in fuel consumption for the new system in comparison to the Toyota Hybrid System as the first commercialised and the most accepted and popular power-split powertrain system.

Many recent applications of intelligent transportation systems require both real-time and network-wide traffic flow data as input. However, as the detection time and network size increase, the data volume may become very large in terms of both dimension and scale. To address this concern, various traffic flow data compression methods have been proposed, which archive the low-dimensional subspace rather than the original data. Many studies have shown the traffic flow data consist of different components, i.e. low-dimensional intra-day trend, Gaussian type fluctuation and burst components. Existing compression methods cannot compress the burst components well and provide very limited choices of compression ratio (CR). A better compression method should have the ability to archive all the dominant information in different components of traffic flow data. In this study, the authors compare the influence of different data reformatting, archive the bursts defined before in descending order with respect to the absolute value of the burst points and propose a flexible compression framework to balance between burst components and low-dimensional intra-day trend. Experimental results show that the proposed framework promotes the reconstruction accuracy significantly. Moreover, the proposed framework provides more flexible choices with respect to CR, which can benefit a variety of applications.

A vehicle detection method is developed based on three-axis anisotropic magneto resistive (AMR) sensors along travel lane markings. The method integrates multiple algorithms and uses different geomagnetic waveforms which are disturbed by vehicles passing a single AMR sensor. This method comprehensively analyses X-, Y-, and Z-axis information and applies a double-window algorithm to extract a single vehicle waveform. The vehicle mixed algorithm (VMA) is developed to differentiate vehicles driving by the AMR sensor simultaneously and determine vehicle flow rates. In addition, the vehicle motion-state discrimination algorithm (VMSDA) is developed to distinguish the vehicle operating status (i.e. driving on the left lane, the lane line, or the right lane). The field experimental tests verified the effectiveness of the both algorithms. Results indicate that the average accuracy rates of VMA and VMSDA can, respectively, be up to 98.0 and 96.4%.

In this research, a real-time positioning method, which utilises crowdsourced positioning data obtained from smartphone GPS is developed. Such vehicle location information obtained from crowdsourcing and smartphones in public transport could replace traditional automatic vehicle location systems. However, the location information from smartphone GPS is more erroneous. The proposed methodology serves as an alternative to existing positioning methods to improve the vehicle positioning accuracy. The developed enhanced particle filter algorithm takes smartphone GPS positioning data [from multiple passengers in a single transit vehicle (e.g. bus)] as input data. This ‘crowdsourced’ data can then be utilised to calculate the vehicles’ positioning information with better accuracy using the developed enhanced particle filter algorithm. The developed algorithm was tested using data collected on 14 different bus routes in urban and suburban areas of Mumbai, India, and it was identified that the algorithm is effective in reducing the average error up to 21.3% from a regular smartphone GPS and 10% from extended Kalman filter algorithm and was able to curtail positioning error within 8.672 m (average over 14 routes).

Signal lights are essential for maintaining the operational efficiency and safety in urban road networks. Operational efficiency and safety at intersection have been two important topics in transportation science. In this study, the authors propose a car-following model to investigate the impacts of signal light on driving behaviour, fuel consumption and emissions during the whole process that each vehicle runs across the intersection. In particular, the proposed model has explicitly considered the behaviours at an intersection with countdown device that provides instantaneous information to drivers. The proposed model is tested by numerical analysis and the results indicate that the model can enhance the operational efficiency and the traffic safety near the intersection, and also reduce the average fuel consumption of the vehicles. Sensitivity analysis indicates that the vehicles’ initial time headway at the road origin may have major influences on the flow capacity and the total fuel consumption.

In single vehicle probe data measured on German freeways the authors have revealed free flow→synchronised flow→free flow transitions that occur before traffic breakdown at a bottleneck occurs. Thus resulting in the formation of a congested pattern. The empirical findings of these phase transitions confirm a recent microscopic stochastic theory of traffic breakdown developed by Kerner. Only because of the recently introduced possibility of gathering larger amounts of anonymised vehicle data – including a sequence of positions of each car – the authors are able to show the phenomenon of these phase transitions in measured floating car data. This contribution reveals empirical findings in microscopic data which support and prove some of the recent theoretical findings of the nature of traffic breakdown.