Integrated Network Capacity Expansion and Traffic Signal Optimization Problem: Robust Bi-level Dynamic Formulation

This paper presents a robust optimization formulation, with an exact solution method, that simultaneously solves continuous network capacity expansion, traffic signal optimization and dynamic traffic assignment when explicitly accounting for an appropriate robustness measure, the inherent bi-level nature of the problem and long-term O-D demand uncertainty. The adopted robustness measure is the weighted sum of expected total system travel time (TSTT) and squared up-side deviation from a fixed target. The model propagates traffic according to Daganzo’s cell transmission model. Furthermore, we formulate five additional, related models. We find that when evaluated in terms of robustness, the integrated robust model performs the best, and interestingly the sequential robust approach yields a worse solution compared to certain sequential and integrated approaches. Although the adopted objective of the integrated robust model does not directly optimize the variance of TSTT, our experimental results show that the robust solutions also yield the least-variance solutions.     

A Hybrid Route Choice Model for Dynamic Traffic Assignment

Network user equilibrium or user optimum is an ideal state that can hardly be achieved in real traffic. More often than not, every day traffic tends to be in disequilibrium rather than equilibrium, thanks to uncertainties in demand and supply of the network. In this paper we propose a hybrid route choice model for studying non-equilibrium traffic. It combines pre-trip route choice and en-route route choice to solve dynamic traffic assignment (DTA) in large-scale networks. Travelers are divided into two groups, habitual travelers and adaptive travelers. Habitual travelers strictly follow their pre-trip routes which can be generated in the way that major links, such as freeways or major arterial streets, are favored over minor links, while taking into account historical traffic information. Adaptive travelers are responsive to real-time information and willing to explore new routes from time to time. We apply the hybrid route choice model in a synthetic medium-scale network and a large-scale real network to assess its effect on the flow patterns and network performances, and compare them with those obtained from Predictive User Equilibrium (PUE) DTA. The results show that PUE-DTA usually produces considerably less congestion and less frequent queue spillback than the hybrid route choice model. The ratio between habitual and adaptive travelers is crucial in determining realistic flow and queuing patterns. Consistent with previous studies, we found that, in non-PUE DTA, supplying a medium sized group (usually less than 50%) of travelers real-time information is more beneficial to network performance than supplying the majority of travelers with real-time information. Finally, some suggestions are given on how to calibrate the hybrid route choice model in practice to produce realistic results.     

Stability issues for dynamic traffic assignment

This paper explores stability issues for operational route guidance control strategies for vehicular traffic networks equipped with advanced information systems, and develops a general procedure for the stability analysis of the associated dynamic traffic assignment (DTA) problems. The route guidance control strategies are modeled as dynamical systems, and the associated solution procedure enables computational tractability for real-time deployment. An important study insight is that the Lyapunov functions for the route guidance control models are their corresponding objective functions under DTA. This overcomes the key difficulty of constructing meaningful Lyapunov functions for DTA problems.     

SALA: A Self-Adaptive Learning Algorithm—Towards Efficient Dynamic Route Guidance in Urban Traffic Networks

In order to alleviate traffic congestion for vehicles in urban networks, most of current researches mainly focused on signal optimization models and traffic assignment models, or tried to recognize the interaction between signal control and traffic assignment. However, these methods may not be able to provide fast and accurate route guidance due to the lack of individual traffic demands, real-time traffic data and dynamic cooperation between vehicles. To solve these problems, this paper proposes a dynamic and real-time route selection model in urban traffic networks (DR2SM), which can supply a more accurate and personalized strategy for vehicles in urban traffic networks. Combining the preference for alternative routes with real-time traffic conditions, each vehicle in urban traffic networks updates its route selection before going through each intersection. Based on its historical experiences and estimation about route choices of the other vehicles, each vehicle uses a self-adaptive learning algorithm to play congestion game with each other to reach Nash equilibrium. In the route selection process, each vehicle selects the user-optimal route, which can maximize the utility of each driving vehicle. The results of the experiments on both synthetic and real-world road networks show that compared with non-cooperative route selection algorithms and three state-of-the-art equilibrium algorithms, DR2SM can effectively reduce the average traveling time in the dynamic and uncertain urban traffic networks.     

基于深度强化学习的空间众包任务分配策略

针对动态在线任务分配策略难以有效利用历史数据进行学习、同时未考虑当前决策对未来收益的影响的问题,提出基于深度强化学习的空间众包任务分配策略.首先,以最大化长期累积收益为优化目标,基于马尔科夫决策过程从单个众包工作者的角度建模,将任务分配问题转化为对状态动作价值Q的求解及工作者与任务的一对一分配.然后采用改进的深度强化学...     

Online-SVR for short-term traffic flow prediction under typical and atypical traffic conditions

Most literature on short-term traffic flow forecasting focused mainly on normal, or non-incident, conditions and, hence, limited their applicability when traffic flow forecasting is most needed, i.e., incident and atypical conditions. Accurate prediction of short-term traffic flow under atypical conditions, such as vehicular crashes, inclement weather, work zone, and holidays, is crucial to effective and proactive traffic management systems in the context of intelligent transportation systems (ITS) and, more specifically, dynamic traffic assignment (DTA).To this end, this paper presents an application of a supervised statistical learning technique called Online Support Vector machine for Regression, or OL-SVR, for the prediction of short-term freeway traffic flow under both typical and atypical conditions. The OL-SVR model is compared with three well-known prediction models including Gaussian maximum likelihood (GML), Holt exponential smoothing, and artificial neural net models.The resultant performance comparisons suggest that GML, which relies heavily on the recurring characteristics of day-to-day traffic, performs slightly better than other models under typical traffic conditions, as demonstrated by previous studies. Yet OL-SVR is the best performer under non-recurring atypical traffic conditions. It appears that for deployed ITS systems that gear toward timely response to real-world atypical and incident situations, OL-SVR may be a better tool than GML.     

Simulation-based dynamic traffic assignment for short-term planning applications

Evaluation of Intelligent Transportation Systems (ITS) at the planning level requires the use of appropriate tools that can capture the dynamic and stochastic interactions between demand and supply. The objective of this paper is to present a methodological simulation-based framework for such applications and implement it in the context of dynamic traffic assignment. The framework consists of a mesoscopic supply simulator and a demand simulator that combines OD estimation capabilities with discrete travel behavior models. Simulation-based DTA systems are particularly suited to evaluate a wide range of Advanced Traffic Management Systems (ATMS) and Advanced Traveler Information Systems (ATIS). The simulation model performance is illustrated through two large-scale case studies in Irvine, California, and Lower Westchester County, NY.     

基于概率路由的出租车共乘调度算法

针对现有的拼车方案大多服务在线乘客请求,而忽略了离线乘客请求,导致出租车资源无法得到充分利用这一问题,提出了一种基于挖掘历史出行轨迹数据的概率路由拼车优化算法。该算法根据乘客请求的历史时空数据计算出区域概率转移矩阵,并使用该矩阵优化平台为出租车推荐拼车路径,提供了历史数据和拼车问题相融合的一种解决方案,可以有效提高出租车的载客量。在保障离线乘客接载率、在线乘客忍耐度的同时,使用松弛时间的度量指标,可以在O（n）内对整条路径的乘客忍耐时间进行评估预测,并用迪杰斯特拉算法对绕行区域进行路径规划,让出租车的绕行距离最短。使用滴滴GAIA真实数据集对算法有效性进行验证,结果显示,该算法在服务请求数量上高于基准算法12%。     

随机需求下双渠道供应链库存动态交互优化

针对双渠道供应链库存系统导致的缺货与库存积压等问题，在线上线下均为随机需求的条件下，考虑生产延迟和物流延迟，建立了双渠道库存的单独控制、集中控制和交叉补货控制这三种模式的动态优化模型。首先，以库存动态微分方程为基础，创新性地以控制理论为指导思想，以泰勒展开和拉普拉斯变换为手段，得到双渠道库存系统的反馈传递函数；其次，考虑了交叉补货的进销存过程中的周期间交互、上下游间交互以及渠道间交互，利用延迟控制、反馈控制和比例-积分-微分（PID）控制构造了双输入、双输出的复杂交互系统，以此寻求双渠道库存系统自身以及渠道间的动态供需双平衡，优化双渠道库存持有量，降低缺货次数和缺货量并使其保持动态稳定状态；最后，通过数值仿真实验，对比三种双渠道库存控制策略。仿真结果表明，在线上线下渠道为不同分布的随机需求时，交叉补货控制的剩余库存比独立库存控制降低了4.9%，交叉补货控制的缺货率与独立控制和集中控制相比分别下降了66.7%和60%。实验结果表明，在线上线下渠道为不同分布的随机需求的情况下使用双渠道交叉补货策略能很好地降低库存持有量，减少缺货次数和缺货量，从而节约库存成本。     

随机需求下双渠道供应链库存动态交互优化

针对双渠道供应链库存系统导致的缺货与库存积压等问题,在线上线下均为随机需求的条件下,考虑生产延迟和物流延迟,建立了双渠道库存的单独控制、集中控制和交叉补货控制这三种模式的动态优化模型。首先,以库存动态微分方程为基础,创新性地以控制理论为指导思想,以泰勒展开和拉普拉斯变换为手段,得到双渠道库存系统的反馈传递函数;其次,考虑了交叉补货的进销存过程中的周期间交互、上下游间交互以及渠道间交互,利用延迟控制、反馈控制和比例-积分-微分（PID）控制构造了双输入、双输出的复杂交互系统,以此寻求双渠道库存系统自身以及渠道间的动态供需双平衡,优化双渠道库存持有量,降低缺货次数和缺货量并使其保持动态稳定状态;最后,通过数值仿真实验,对比三种双渠道库存控制策略。仿真结果表明,在线上线下渠道为不同分布的随机需求时,交叉补货控制的剩余库存比独立库存控制降低了4.9%,交叉补货控制的缺货率与独立控制和集中控制相比分别下降了66.7%和60%。实验结果表明,在线上线下渠道为不同分布的随机需求的情况下使用双渠道交叉补货策略能很好地降低库存持有量,减少缺货次数和缺货量,从而节约库存成本。     

A-RESCUE: An Agent based Regional Evacuation Simulator Coupled with User Enriched Behavior

Household behavior and dynamic traffic flows are the two most important aspects of hurricane evacuations. However, current evacuation models largely overlook the complexity of household behavior leading to oversimplified traffic assignments and, as a result, inaccurate evacuation clearance times in the network. In this paper, we present a high fidelity multi-agent simulation model called A-RESCUE (Agent-based Regional Evacuation Simulator Coupled with User Enriched behavior) that integrates the rich activity behavior of the evacuating households with the network level assignment to predict and evaluate evacuation clearance times. The simulator can generate evacuation demand on the fly, truly capturing the dynamic nature of a hurricane evacuation. The simulator consists of two major components: household decision-making module and traffic flow module. In the simulation, each household is an agent making various evacuation related decisions based on advanced behavioral models. From household decisions, a number of vehicles are generated and entered in the evacuation transportation network at different time intervals. An adaptive routing strategy that can achieve efficient network-wide traffic measurements is proposed. Computational results are presented based on simulations over the Miami-Dade network with detailed representation of the road network geometry. The simulation results demonstrate the evolution of traffic congestion as a function of the household decision-making, the variance of the congestion across different areas relative to the storm path and the most congested O-D pairs in the network. The simulation tool can be used as a planning tool to make decisions related to how traffic information should be communicated and in the design of traffic management policies such as contra-flow strategies during evacuations.     

一种基于网格和密度的数据流聚类算法

在"数据流分析"这一数据挖掘的应用领域中,常规的算法显得很不适用.主要是因为这些算法的挖掘过程不能适应数据流的动态环境,其挖掘模型、挖掘结果不能满足实际应用中用户的需求.针对这一问题,本文提出了一种基于网格和密度的聚类方法,来有效地完成对数据流的分析任务.该方法打破传统聚类方法的束缚,把整个挖掘过程分为离线和在线两步,最终通过基于网格和密度的聚类方法实现数据流聚类.     

Integration control for traffic corridors considering guidance information

The problem of designing integration traffic strategies for traffic corridors with the use of ramp metering, speed limit, and route guidance is considered in this paper. As an improvement to the previous work, the presented approach has the following five features： 1） modeling traffic flow to analyze traffic characteristics under the influence of variable speed limit, on-ramp metering and guidance information; 2） building a hierarchy model to realize the integration design of traffic control and route guidance in traffic corridors; 3） devising a multi-class analytical dynamic traffic assignment （DTA） model for traffic corridors, where not only the route choice process will be different for each user-class, but also the traffic flow operations are user-class specific because the travel time characteristic for each user-class is considered; 4） predicting route choice probabilities adaptively with real-time traffic conditions and route choice behaviors corresponding to variant users, rather than assuming as pre-determined; and 5） suggesting a numerical solution algorithm of the hierarchy model presented in this paper based on the modified algorithm of iterative optimization assignment （IOA）. Preliminary numerical test demonstrates the potential of the developed model and algorithm for integration corridor control.     

DENFIS: dynamic evolving neural-fuzzy inference system and itsapplication for time-series prediction

This paper introduces a new type of fuzzy inference systems, denoted as dynamic evolving neural-fuzzy inference system (DENFIS), for adaptive online and offline learning, and their application for dynamic time series prediction. DENFIS evolve through incremental, hybrid (supervised/unsupervised), learning, and accommodate new input data, including new features, new classes, etc., through local element tuning. New fuzzy rules are created and updated during the operation of the system. At each time moment, the output of DENFIS is calculated through a fuzzy inference system based on m-most activated fuzzy rules which are dynamically chosen from a fuzzy rule set. Two approaches are proposed: (1) dynamic creation of a first-order Takagi-Sugeno-type fuzzy rule set for a DENFIS online model; and (2) creation of a first-order Takagi-Sugeno-type fuzzy rule set, or an expanded high-order one, for a DENFIS offline model. A set of fuzzy rules can be inserted into DENFIS before or during its learning process. Fuzzy rules can also be extracted during or after the learning process. An evolving clustering method (ECM), which is employed in both online and offline DENFIS models, is also introduced. It is demonstrated that DENFIS can effectively learn complex temporal sequences in an adaptive way and outperform some well-known, existing models     

移动边缘计算中的动态用户分配方法

在边缘计算环境中,为用户匹配合适的服务器是一个关键问题,可以有效提升服务质量.文中将边缘用户分配问题转换为一个受距离和服务器资源约束的二分图匹配问题,并将其建模为一个0-1整数规划问题进行优化.在离线状态下,基于精确式算法的优化模型可以求得最优分配策略,但其求解时间过长,无法处理规模较大的数据,不适用于现实服务环境.因...     

Dynamic Traffic Assignment with More Flexible Modelling within Links

Traffic network models tend to become very large even for medium-size static assignment problems. Adding a time dimension, together with time-varying flows and travel times within links and queues, greatly increases the scale and complexity of the problem. In view of this, to retain tractability in dynamic traffic assignment (DTA) formulations, especially in mathematical programming formulations, additional assumptions are normally introduced. In particular, the time varying flows and travel times within links are formulated as so-called whole-link models. We consider the most commonly used of these whole-link models and some of their limitations.In current whole-link travel-time models, a vehicle's travel time on a link is treated as a function only of the number of vehicles on the link at the moment the vehicle enters. We first relax this by letting a vehicle's travel time depend on the inflow rate when it enters and the outflow rate when it exits. We further relax the dynamic assignment formulation by stating it as a bi-level program, consisting of a network model and a set of link travel time sub-models, one for each link. The former (the network model) takes the link travel times as bounded and assigns flows to links and routes. The latter (the set of link models) does the reverse, that is, takes link inflows as given and finds bounds on link travel times. We solve this combined model by iterating between the network model and link sub-models until a consistent solution is found. This decomposition allows a much wider range of link flow or travel time models to be used. In particular, the link travel time models need not be whole-link models and can be detailed models of flow, speed and density varying along the link. In our numerical examples, algorithms designed to solve this bi-level program converged quickly, but much remains to be done in exploring this approach further. The algorithms for solving the bi-level formulation may be interpreted as traveller learning behaviour, hence as a day-to-day traffic dynamics. Thus, even though in our experiments the algorithms always converged, their behaviour is still of interest even if they cycled rather than converged. Directions for further research are noted. The bi-level model can be extended to handle issues and features similar to those addressed by other DTA models.     

An efficient evolutionary algorithm for channel resource managementin cellular mobile systems

Modern cellular mobile communications systems are characterized by a high degree of capacity. Consequently, they have to serve the maximum possible number of calls while the number of channels per cell is limited. The objective of channel allocation is to assign a required number of channels to each cell such that both efficient frequency spectrum utilization is provided and interference effects are minimized. Channel assignment is therefore an important operation of resource management and its efficient implementation increases the fidelity, capacity, and quality of service of cellular systems. Most channel allocation strategies are based on deterministic methods, however, which result in implementation complexity that is prohibitive for the traffic demand envisaged for the next generation of mobile systems. An efficient heuristic technique capable of handling channel allocation problems is introduced as an alternative. The method is called a combinatorial evolution strategy (CES) and belongs to the general heuristic optimization techniques known as evolutionary algorithms (EAs). Three alternative allocation schemes operating deterministically, namely the dynamic channel assignment (DCA), the hybrid channel assignment (HCA), and the borrowing channel assignment (BCA), are formulated as combinatorial optimization problems for which CES is applicable. Simulations for representative cellular models show the ability of this heuristic to yield sufficient solutions. These results will encourage the use of this method for the development of a heuristic channel allocation controller capable of coping with the traffic and spectrum management demands for the proper operation of the next generation of cellular systems     

GIS and ITS Traffic Assignment: Issues in Dynamic User-Optimal Assignments

Dynamic traffic assignment (DTA) is at the heart of much ITS research. Assigning traffic, whether for planning purposes or for real time route guidance, is a difficult problem. Recent advances in user-optimal dynamic traffic assignment have built on the methods developed for static user-optimal assignments. Since assignment models are complex, they will not use many of the network analysis functions found in commercial GIS packages. Custom software will have to be developed. In this paper we explore the problems faced in solving static and dynamic assignments and relate those problems to information that is likely to be based in a traffic control centers GIS database. Because of the size of the problem and the need for faster-than-real-time analysis, how and when data is transferred between a GIS to analysis modules is important. Further, many approaches for GIS software design and spatial data handling, such as OOP and dynamic segmentation, may impose too much overhead to be of much use in time-sensitive ITS applications.     

Comparison of Methods for Path Flow Reassignment for Dynamic User Equilibrium

Models to describe or predict of time-varying traffic flows and travel times on road networks are usually referred to as dynamic traffic assignment (DTA) models or dynamic user equilibrium (DUE) models. The most common form of algorithms for DUE consists of iterating between two components namely dynamic network loading (DNL) and path inflow reassignment or route choice. The DNL components in these algorithms have been investigated in many papers but in comparison the path inflow reassignment component has been relatively neglected. In view of that, we investigate various methods for path inflow reassignment that have been used in the literature. We compare them numerically by embedding them in a DUE algorithm and applying the algorithm to solve DUE problems for various simple network scenarios. We find that the choice of inflow reassignment method makes a huge difference to the speed of convergence of the algorithms and, in particular, find that ??travel time responsive?? reassignment methods converge much faster than the other methods. We also investigate how speed of convergence is affected by the extent of congestion on the network, by higher demand or lower capacity. There appears to be much scope for further improving path inflow reassignment methods.     

Calibration and validation of a simulation-based dynamic traffic assignment model for a large-scale congested network

Deployment of a simulation-based dynamic traffic assignment (DTA) model requires calibration of a large number of demand and supply input parameters. In this paper, we present a data-intensive framework for deployment, calibration, and validation of a simulation-based DTA model of Melbourne, Australia as a large-scale congested network. The model consists of 55,719 links and 24,502 nodes and simulates almost 2.1 million commuters in a 4-h morning peak period. We propose a machine learning based technique to classify and calibrate the traffic flow fundamental diagrams against empirical data obtained from a large number of freeway loop detectors across the network. An optimization framework for estimating the time-dependent origin-destination (TDOD) demand is also presented. To enhance the quality of the TDOD demand estimation, we apply a departure time profiling technique to account for the spatial differences between OD pairs. The paper also demonstrates the impact of adaptive driving on the quality of the DTA calibration. A comparison between the simulated and observed link volumes over 1,250 locations across the entire network shows that the calibration procedure generates an approximately 30% improvement in the root mean square error (RMSE). The impact of pedestrians and cyclists on the vehicular traffic is also implicitly considered in the central business district (CBD) area to improve underestimation of the simulated travel times. Validation results suggest that the calibrated DTA model successfully replicates traffic patterns in the network and support the future applications of the model for various transportation operations and planning purposes.