Trajectories, representing the movements of objects in the real world, carry significant stop/move semantics. The detection of trajectory stops poses a critical problem in the study of moving objects and becomes even more challenging due to the inevitable noise recorded along with true data. To extract stops with a variety of shapes and sizes from single trajectories with noise, this paper presents a sequence oriented clustering approach, in which noise points within the sequence of a stop can be identified and classified as a part of the stop. In our method, two key concepts are first introduced: (1) a core sequence that defines sequence density based not only on proximity in space but also continuity in time as well as the duration over time; and (2) an Eps-reachability sequence that aggregates core sequences that overlap or meet over time. Then, three criteria are presented to merge Eps-reachability sequences interrupted by noise. Further, an algorithm, called SOC (Sequence Oriented Clustering), is developed to automatically extract stops from a single trajectory. In addition, a reachability graph is designed that visually illustrates the spatio-temporal clustering structure and levels of a trajectory. Finally, the proposed algorithm is evaluated against two baseline methods through extensive experiments based on real world trajectories, some with serious noise, and the results show that our approach is fairly effective in recognizing trajectory stops.

Small to medium sized transportation and logistics companies are usually constrained by limited computing and IT professional resources on implementing an efficient parallel metaheuristic algorithm for planning or management solutions. In this paper we extend the standard meta-description for genetic algorithms (GA) with a simple non-trivial parallel implementation. Our parallel GA framework is chiefly concerned with the development of a straightforward way for engineers to modify existing genetic algorithm implementations for real transportation and logistics problems to make use of commonly available hardware resources without completely reworking complex, useful and usable codes. The framework presented at its parallel base is a modification of the primitive parallelization concept, but if implemented as described it may be gradually extended to fit the qualities of any underlying problem better (via the adaptation of the merging and communications functions).We present our framework and computational results for a classical transportation related combinatorial optimization problem – the traveling salesman problem with a standard sequential genetic algorithm implementation. Our empirical analysis shows that this simple extension can lead to considerable solution improvements. We also tested our assumptions that the framework is easily implemented by an engineer not initially familiar with genetic algorithms to implement the framework for another minimum multiprocessor scheduling problem. These case studies verify that our framework is better than primitive parallelization because it gives empirically better results under equitable conditions. It also outperforms fine grained parallelization as it is easier and faster to implement.

With the development of artificial intelligence and the improvement of hardware computing power, deep learning models have become widely used in the Internet of Things (IoT) field, especially for analyzing spatiotemporal data collected by wireless sensors. Recurrent Neural Networks (RNN) such as Long Short-Term Memory (LSTM) network are generally used for these time series data. Hyperparameter settings of model training are regarded as essential factors for the performance of deep learning models. Manually optimizing hyperparameters not only cost more resources but also be more likely to set hyperparameters that follow stereotypes, resulting in unreasonable hyperparameter settings and poor model performance. As one of the most important fields in Automated Machine Learning (AutoML) research, automated hyper-parameter optimization (HPO) mainly includes grid search, hyperparameter search based on genetic algorithm, etc. Whereas these methods have their own drawbacks. In this paper, an automated hyper-parameter optimization method based on parallel genetic algorithm (PGA) is proposed. According to the process of PGA, this paper divided hyperparameter optimization into several stages, including population initialization, fitness function, tournament selection, crossover operators, mutation operators, subgroup exchange, and end of evolution. Then the proposed hyperparameter optimization method is implemented in LSTM models and tested on two different time series datasets collected by real-world IoT sensors. By comparing our proposed method with other mainstream HPO methods in different datasets, it is proved that our HPO method based on PGA shows better performance on both time costs and prediction results.

The long production cycle and huge cost of collecting road network data often leave the data lagging behind the latest real conditions. However, this situation is rapidly changing as the positioning techniques ubiquitously used in mobile devices are gradually being implemented in road network research and applications. Currently, the predominant approaches infer road networks from mobile location information (e.g., GPS trajectory data) directly using various extracting algorithms, which leads to expensive consumption of computational resources in the case of large-scale areas. For this reason, we propose an alternative that renews road networks with a novel spiral strategy, including a hidden Markov model (HMM) for detecting potential problems in existing road network data and a method to update the data, on the local scale, by generating new road segments from trajectory data. The proposed approach reduces computation costs on roads with completed or updated information by updating problem road segments in the minimum range of the road network. We evaluated the performance of our proposals using GPS traces collected from taxies and OpenStreetMap (OSM) road networks covering urban areas of Wuhan City.

Exploring the heterogeneous characteristics of the urban expansion process is essential for understanding the dynamics of the urban spatial structure. Many studies focused on depicting the spatio-temporal characteristics based on urban expansion patches. However, measuring heterogeneous characteristics of urban expansion from agglomeration areas comprising the expanded urban construction land patches have not been adequately explored. This study presents a novel approach and two improved indices for characterizing the heterogeneity of urban spatial agglomeration areas during urban expansion. Firstly, we proposed a Gaussian mixture model considering multiple constrains and density-based spatial clustering of applications with noise (DBSCAN) integration method to identify and extract the urban agglomeration areas automatically. Secondly, the gradient analysis and the compact index using the inverse “S” function are introduced to explore the spatio-temporal characteristics from a macrocosmic perspective. Finally, the compactness index (NCI) and normalized dispersion index (NDIS) are improved based on agglomeration area data. The microcosmic heterogeneous characteristics are measured by these two improved indices and the positional offset characteristics indices (POCIS). The method was implemented in the urban area of Changsha, Hunan Province, China in 2005, 2010, and 2015. The results show that (1) compared to that in the Changsha City Master Plan (2003–2020), the recognition rate was higher in the agglomeration areas than others. (2) The overall expansion trend in Changsha transitioned toward decentralization, making Changsha a polycentric city. (3) The agglomeration of urban expansion in the east-west direction became compact; that in the north-south direction became looser; most clusters expanded to the west and a new sub-center would appear. The proposed method can effectively characterize their heterogeneity, which can provide valuable references for urban planning and policymaking.

Electric vehicles (EVs) have attracted considerable attention because of their clean and high-energy efficiency. Reasonably planning a charging station network has become a vital issue for the popularization of EVs. Current research on optimizing charging station networks focuses on the role of stations in a local scope. However, spatial features between charging stations are not considered. This paper proposes a charging station selection method based on the graph convolutional network (GCN) and establishes a charging station selection method considering traffic information and investment cost. The method uses the GCN to extract charging stations. The charging demand of each candidate station is calculated through the traffic flow information to optimize the location of charging stations. Finally, the cost of the charging station network is evaluated. A case study on charging station selection shows that the method can solve the EV charging station location problem.