As complex systems, the spatial structure of urban systems can be analyzed by entropy theory. This paper first calculates the interaction force between cities based on the gravity model, the spatial relationship matrix between cities is constructed using the method of network modeling, and the spatial network modeling of urban system can be calculated. Secondly, the Efficiency Entropy (EE), Quality Entropy (QE), and System Entropy (SE) of urban system network are calculated and analyzed by information entropy. Finally, taking the Huaihe River Basin (HRB) as a case study, model verification and empirical analysis are performed. It is found that the spatio–temporal pattern of the urban system network structure in the basin is uneven: in space, the urban system network in the HRB presents a layer-by-layer spatial distribution centered on the core city of Xuzhou; meanwhile, the overall urban system network in the basin presents an orderly development trend. This study has certain theoretical and practical value for the planning of urban and urban systems and the coordinated development of regions.

The intelligent transportation system (ITS) is one of the effective solutions to the problem of urban traffic congestion, and it is also one of the important topics of smart city construction. One particular application is the traffic monitoring and flow prediction. However, there are still challenges regarding both aspects. On the one hand, the current traffic monitoring relies heavily on the single object detection method that cannot achieve accurate statistics of moving target counting and, meanwhile, has limited speed advantage; on the other hand, the existing traffic flow prediction models rarely consider different weather conditions. Therefore, the present article attempts to propose a packaged solution, which combines a new target tracking and moving vehicle counting method and an improved long short-term memory (LSTM) network for traffic flow forecast with weather conditions. More specifically, the DCN V2 convolution kernel and MultiNetV3 framework are used to replace YOLOv4’s conventional convolution kernel and backbone network to realize multi-target tracking and counting, respectively. Subsequently, combined with the temporal characteristics of historical traffic flow, this article introduces weather conditions into the LSTM network and realizes the short-term prediction of traffic flow at the road junction level. This study carries out a series of experiments using the real traffic video data with a 2-month time span at a popular road junction in the downtown of Shenzhen, China. The results suggest that the proposed algorithms outperform the previous methods in terms of the 10% higher accuracy of target detection tracking and about a half reduction of traffic prediction error, when considering weather conditions.

In the intelligent transportation field, object recognition, detection, and location applications face significant real-time challenges. To address these issues, we propose an automatic sensor-based data loading and unloading optimization strategy for algorithm models. This strategy is designed for artificial intelligence (AI) application systems that leverage edge computing. It aims to solve resource allocation optimization and improve operational efficiency in edge computing environments. By doing so, it meets the real-time computing requirements of intelligent transportation business applications. By adopting node and sensor management mechanisms as well as efficient communication protocols, dynamic sensor-based data management of AI algorithm models was achieved, such as pedestrian object recognition, vehicle object detection, and ship object positioning. Experimental results show that while maintaining the same recall rate, the inference time is reduced to one tenth or even one twentieth of the original time. And this strategy can enhance privacy protection of sensor-based data. In the future research, we may consider integrating distributed computing under high load conditions to further optimize the response time of model loading and unloading for multi-service interaction, and enhance the balance and scalability of the system.