In this paper, platoons of autonomous vehicles operating in urban road networks are considered. From a methodological point of view, the problem of interest consists of formally characterizing vehicle state trajectory...In this paper, platoons of autonomous vehicles operating in urban road networks are considered. From a methodological point of view, the problem of interest consists of formally characterizing vehicle state trajectory tubes by means of routing decisions complying with traffic congestion criteria. To this end, a novel distributed control architecture is conceived by taking advantage of two methodologies: deep reinforcement learning and model predictive control. On one hand, the routing decisions are obtained by using a distributed reinforcement learning algorithm that exploits available traffic data at each road junction. On the other hand, a bank of model predictive controllers is in charge of computing the more adequate control action for each involved vehicle. Such tasks are here combined into a single framework:the deep reinforcement learning output(action) is translated into a set-point to be tracked by the model predictive controller;conversely, the current vehicle position, resulting from the application of the control move, is exploited by the deep reinforcement learning unit for improving its reliability. The main novelty of the proposed solution lies in its hybrid nature: on one hand it fully exploits deep reinforcement learning capabilities for decisionmaking purposes;on the other hand, time-varying hard constraints are always satisfied during the dynamical platoon evolution imposed by the computed routing decisions. To efficiently evaluate the performance of the proposed control architecture, a co-design procedure, involving the SUMO and MATLAB platforms, is implemented so that complex operating environments can be used, and the information coming from road maps(links,junctions, obstacles, semaphores, etc.) and vehicle state trajectories can be shared and exchanged. Finally by considering as operating scenario a real entire city block and a platoon of eleven vehicles described by double-integrator models, several simulations have been performed with the aim to put in light the main f eatures of the proposed approach. Moreover, it is important to underline that in different operating scenarios the proposed reinforcement learning scheme is capable of significantly reducing traffic congestion phenomena when compared with well-reputed competitors.展开更多
Effective urban land-use re-planning and the strategic arrangement of drainage pipe networks can significantly enhance urban flood defense capacity.Aimed at reducing the potential risks of urban flooding,this paper pr...Effective urban land-use re-planning and the strategic arrangement of drainage pipe networks can significantly enhance urban flood defense capacity.Aimed at reducing the potential risks of urban flooding,this paper presents a straightforward and efficient approach to an urban distributed runoff model(UDRM).The model is developed to quantify the discharge and water depth within urban drainage pipe networks under varying rainfall intensities and land-use scenarios.The Nash efficiency coefficient of UDRM exceeds 0.9,which indicates its high computational efficiency and potential benefit in predicting urban flooding.The prediction of drainage conditions under both current and re-planned land-use types is achieved by adopting different flood recurrence intervals.The findings reveal that the re-planned land-use strategies could effectively diminish flood risk upstream of the drainage pipe network across 20-year and 50-year flood recurrence intervals.However,in the case of extreme rainfall events(a 100-year flood recurrence),the re-planned land-use approach fell short of fulfilling the requirements necessary for flood disaster mitigation.In these instances,the adoption of larger-diameter drainage pipes becomes an essential requisite to satisfy drainage needs.Accordingly,the proposed UDRM effectively combines land-use information with pipeline data to give practical suggestions for pipeline modification and land-use optimization to combat urban floods.Therefore,this methodology warrants further promotion in the field of urban re-planning.展开更多
Following a half century of popularity, central place theory experienced 20 years of neglect when the new urban system theory of network modeling gained attention at the beginning of the 1990s. However, central place ...Following a half century of popularity, central place theory experienced 20 years of neglect when the new urban system theory of network modeling gained attention at the beginning of the 1990s. However, central place theory remains valid, and it seems there has been a reemergence with it. Using the Greater Pearl River Delta (Greater PRD) as an experimental study region, this paper intends to present an empirical study that validates central place theory and shows that it can be integrated into an overall regional urban system. The study uses the compound Central Place Importance (CPI) to evaluate whether there is a hierarchy among the urban centers within the study area. The results indicate the existence of a hierarchy. Furthermore, empirical observation finds distinct complementarity relationships, rank-size distributions, and co-operative actions between the different cities, thus substantiating the claim that central place theory can be incorporated into an overall regional urban system. Besides, the presence of the densely distributed modern infrastructure system also appears to constitute a dimension of the overall urban system. There need further theoretical and empirical studies in order to support this proposition.展开更多
在信息物理高度融合背景下,快速、准确地检测虚假数据注入攻击(false data injection attacks,FDIAs)是城市配电网安全稳定运行的关键。提出一种基于对抗性自动编码器的城市配电网FDIAs检测方法,将自动编码器和生成对抗网络结合,能够提...在信息物理高度融合背景下,快速、准确地检测虚假数据注入攻击(false data injection attacks,FDIAs)是城市配电网安全稳定运行的关键。提出一种基于对抗性自动编码器的城市配电网FDIAs检测方法,将自动编码器和生成对抗网络结合,能够提取数据特征,发现FDIAs引起的配电网数据异常,并在少量标记数据的基础上完成网络训练,避免出现高昂标记成本的同时,还可减少FDIAs检测对网络拓扑结构的依赖。通过典型配电网案例仿真和结果分析,验证所提方法与现有FDIAs检测方法相比,在检测精度与效率方面都有一定优势,适用于规模日益庞大的城市配电网。展开更多
文摘In this paper, platoons of autonomous vehicles operating in urban road networks are considered. From a methodological point of view, the problem of interest consists of formally characterizing vehicle state trajectory tubes by means of routing decisions complying with traffic congestion criteria. To this end, a novel distributed control architecture is conceived by taking advantage of two methodologies: deep reinforcement learning and model predictive control. On one hand, the routing decisions are obtained by using a distributed reinforcement learning algorithm that exploits available traffic data at each road junction. On the other hand, a bank of model predictive controllers is in charge of computing the more adequate control action for each involved vehicle. Such tasks are here combined into a single framework:the deep reinforcement learning output(action) is translated into a set-point to be tracked by the model predictive controller;conversely, the current vehicle position, resulting from the application of the control move, is exploited by the deep reinforcement learning unit for improving its reliability. The main novelty of the proposed solution lies in its hybrid nature: on one hand it fully exploits deep reinforcement learning capabilities for decisionmaking purposes;on the other hand, time-varying hard constraints are always satisfied during the dynamical platoon evolution imposed by the computed routing decisions. To efficiently evaluate the performance of the proposed control architecture, a co-design procedure, involving the SUMO and MATLAB platforms, is implemented so that complex operating environments can be used, and the information coming from road maps(links,junctions, obstacles, semaphores, etc.) and vehicle state trajectories can be shared and exchanged. Finally by considering as operating scenario a real entire city block and a platoon of eleven vehicles described by double-integrator models, several simulations have been performed with the aim to put in light the main f eatures of the proposed approach. Moreover, it is important to underline that in different operating scenarios the proposed reinforcement learning scheme is capable of significantly reducing traffic congestion phenomena when compared with well-reputed competitors.
基金supported by the National Key Research and Development Program of China(Grant No.2022YFC3202002)the National Natural Science Foundation of China(Grant Nos.52025092,52209087 and 52379065).
文摘Effective urban land-use re-planning and the strategic arrangement of drainage pipe networks can significantly enhance urban flood defense capacity.Aimed at reducing the potential risks of urban flooding,this paper presents a straightforward and efficient approach to an urban distributed runoff model(UDRM).The model is developed to quantify the discharge and water depth within urban drainage pipe networks under varying rainfall intensities and land-use scenarios.The Nash efficiency coefficient of UDRM exceeds 0.9,which indicates its high computational efficiency and potential benefit in predicting urban flooding.The prediction of drainage conditions under both current and re-planned land-use types is achieved by adopting different flood recurrence intervals.The findings reveal that the re-planned land-use strategies could effectively diminish flood risk upstream of the drainage pipe network across 20-year and 50-year flood recurrence intervals.However,in the case of extreme rainfall events(a 100-year flood recurrence),the re-planned land-use approach fell short of fulfilling the requirements necessary for flood disaster mitigation.In these instances,the adoption of larger-diameter drainage pipes becomes an essential requisite to satisfy drainage needs.Accordingly,the proposed UDRM effectively combines land-use information with pipeline data to give practical suggestions for pipeline modification and land-use optimization to combat urban floods.Therefore,this methodology warrants further promotion in the field of urban re-planning.
文摘Following a half century of popularity, central place theory experienced 20 years of neglect when the new urban system theory of network modeling gained attention at the beginning of the 1990s. However, central place theory remains valid, and it seems there has been a reemergence with it. Using the Greater Pearl River Delta (Greater PRD) as an experimental study region, this paper intends to present an empirical study that validates central place theory and shows that it can be integrated into an overall regional urban system. The study uses the compound Central Place Importance (CPI) to evaluate whether there is a hierarchy among the urban centers within the study area. The results indicate the existence of a hierarchy. Furthermore, empirical observation finds distinct complementarity relationships, rank-size distributions, and co-operative actions between the different cities, thus substantiating the claim that central place theory can be incorporated into an overall regional urban system. Besides, the presence of the densely distributed modern infrastructure system also appears to constitute a dimension of the overall urban system. There need further theoretical and empirical studies in order to support this proposition.
文摘在信息物理高度融合背景下,快速、准确地检测虚假数据注入攻击(false data injection attacks,FDIAs)是城市配电网安全稳定运行的关键。提出一种基于对抗性自动编码器的城市配电网FDIAs检测方法,将自动编码器和生成对抗网络结合,能够提取数据特征,发现FDIAs引起的配电网数据异常,并在少量标记数据的基础上完成网络训练,避免出现高昂标记成本的同时,还可减少FDIAs检测对网络拓扑结构的依赖。通过典型配电网案例仿真和结果分析,验证所提方法与现有FDIAs检测方法相比,在检测精度与效率方面都有一定优势,适用于规模日益庞大的城市配电网。
