Vehicles stopping at signalized intersections during a red light is one of the main causes of traffic oscillations. Recently, deep reinforcement learning (DRL) methods have been applied to connected and automated vehicles (CAVs) traffic to reduce the traffic oscillation at signalized intersections. However, these methods do not perform well for mixed traffic flow, including human vehicles (HVs) and CAVs, especially when the CAV rate is low. We found that this was because they did not take into account the HVs stopping at a red light and causing oscillations. If this oscillation is ignored during the speed regulation, CAVs may conflict with the oscillation wave in the future, forcing a sudden and significant speed reduction and triggering the so-called “trajectory jerking” phenomenon. In order to address this problem, this study proposes a trajectory prediction-based DRL method. By introducing the prediction of the downstream vehicle trajectory into the design of the reward function, the leading CAV will adjust its speed in advance to avoid the future oscillation wave caused by HV’s stopping during the red phase. Simulation tests on various penetration rates of CAVs are conducted for the mixed traffic environment to evaluate the performance of the proposed method. The results show that the proposed method has two advantages. Even with low CAV penetration, the oscillations are suppressed remarkably well. And fuel consumption is also significantly reduced. This research provides a new idea to suppress traffic oscillations in a mixed traffic environment.

The vehicle rescue process for individuals in residential areas in disaster scenarios is a typical vehicle routing problem (VRP). However, most studies do not consider the factor of individual mobility. In residential areas, there are two types of individuals: individuals with high mobility and individuals with low mobility, such as the elderly. To improve the evacuation efficiency, besides ordinary vehicles, special vehicles equipped with wheelchairs and volunteers are also in great need. Thus, evacuation vehicles should consist of a heterogeneous fleet. Vehicles depart from parking lots, arrive at residential areas to pick up individuals, and then transport them to shelters. In other words, the origin and destination are different, but they are viewed as the same in classical VRP. Each residential area can be served directly by vehicles departing from parking lots or by vehicles that have already served others, which means demands can be split. All these make the VRP in emergency rescue more complicated than classical VRP. Therefore, we propose an integer liner program model – multi-parking lot and shelter heterogeneous vehicle routing problem with split pickup (MPSHVRPSP) model, which includes matching constraints of individuals and vehicles to satisfy the demands of different types of individuals, and considers the selectivity of parking lots and shelters too. We provide a Tabu Search (TS) algorithm with diversification strategy to solve the model and ensure the high quality of solution. A lot of experiments are carried out on various instances. Our results show that MPSHVRPSP can be applied to efficient evacuation of complicated scenarios that satisfies the demands of all individuals in residential areas. Besides, it is more reasonable compared with classical VRP, and TS can also obtain a satisfactory solution in less time. Furthermore, sensitivity analysis is conducted on factors that may affect the result of objective function.

In order to solve the problems that are not conducive to the automatic analysis and processing of related applications in the initial data of the road network, the main data is extracted from it to construct a traffic road network model. We designed an algorithm for extracting and constructing a traffic network model based on OSM files. The algorithm is mainly divided into preprocessing of OSM geographic data, road network extraction, road recognition, intersection recognition and merging, providing data basis for traffic simulation needs.