Scanning the Issue

Azim Eskandarian
2020 IEEE transactions on intelligent transportation systems (Print)  
A Mixed Path Size Logit (MPSL) model is proposed to analyze route choice behaviors in the process of taxi customersearching through considering spatio-temporal features of the route, including customer generation rate, path travel time, cumulative intersection delay, path distance, and path size. The GPS data were collected from about 36 000 taxi vehicles in Beijing city at 30-s intervals for six months. In the model application, Xidan district in the center of Beijing city is selected to
more » ... trate the effectiveness of the proposed model. The results indicated that the MPSL model could effectively analyze the route choice behavior in the customersearching process and express higher accuracy than the traditional Multinomial Logit model and basic PSL model. For the N-impulse transfer between two earth orbits, this article introduces N-1 intermediate orbits to describe the orbit transfer scheme. Based on the patched conic theory, candidate solutions can be analytically derived, constraints are removed from the optimization model, and the original problem is converted to a parameter optimization problem. The only difficulty lies in the initialization because the number of optimization variables increases linearly with N, which can be very large. This is settled by a hybrid optimization algorithm that comprises two searching methods. The problem is solved first by an improved particle swarm optimization method and then by an adaptive conjugate gradient method. The authors' method is adaptive to problems with any finite N and can calculate the optimal N in any transfer scenarios. Optimal Recourse Strategy for Battery Swapping Stations Considering Electric Vehicle Uncertainty W. Infante, J. Ma, X. Han, and A. Liebman To create comprehensive and resilient battery swapping stations, a two-stage optimization with recourse is proposed. In the planning stage, the investment for battery purchases is recommended even before the electric vehicle station visit uncertainties are made known. In the operation stage, the battery allocation decisions such as charging, discharging, and swapping are then coordinated. To apply the recourse strategy in creating representative scenarios, electric vehicle station visit distribution techniques are also proposed using a modified K-means clustering method. Aside from the sensitivity analysis made with swapping prices and charging intervals, the strategy comparisons with conventional strategies have also demonstrated the practicality of the proposed coordination to future electricity and transportation networks. This article proposes a global localization system in 3D LiDAR maps, which can localize the vehicle without any prior knowledge of its pose. Siamese neural network is developed to model the environments in mapping, and then similarities can be measured to achieve place recognition for global localization. Finally, Monte Carlo localization is used to localize the vehicle from scratch based on a Gaussian mixture model. The observability analysis is also presented as the theoretical foundations and practice guidelines to the localization system. The experimental results show that the proposed system can achieve global localization in 3D point clouds with effectiveness and high efficiency. Automated vehicles can change the society by improved safety, mobility, and fuel efficiency. However, due to the higher cost and change in the business model, over the coming decades, the highly automated vehicles will likely continue to interact with many human-driven vehicles. In the past, the control/design of the highly automated (robotic) vehicles mainly considers safety and efficiency but failed to address the "driving culture" of the surrounding human-driven vehicles. Thus, the robotic vehicles may demonstrate behaviors very different from other vehicles. The authors study this "driving etiquette" problem in this article. As the first step, they report the key behavior parameters of human-driven vehicles, derived from a large naturalistic driving database. The results can be used to guide future algorithm design of highly automated vehicles or to develop realistic human-driven vehicle behavior model in simulations. Background subtraction is an example of a moving object detection technique that uses machine vision systems. Conventional moving object detection methods need complicated thresholds for background modeling to address changes in illumination. This article proposes a novel background
doi:10.1109/tits.2020.2980087 fatcat:zncxp7cluzgybbw3zjx4nxefti