Search

This paper developed a hybrid convolution neural network for perceptual crowd counting, which could accurately predict density maps in extremely crowded scenes . It consisted of merely two components: the front-end was a dilated convolutional neural network to extract two-dimensional features ;the back-end deployed a fractionally stride convolution to lower the loss of image information caused by down-sampling. This paper designed the model structure based on the dataset Shanghai Tech, then in an attempt to acknowledge and analyze the performance of the algorithm, and afterwards made use of the evaluation indicators of the regression problem, the average absolute error ( MAE) and the mean-square error ( MSE) as the criteria. Additionally, testing the method on Shanghai Tech ( MAE= 100. 8), UCF CC_50 ( MAE = 305. 3) and WorldExpo' 10 datasets while the experiment results reveal that the proposed model can effectively reduce MAE and MSE when compared with previous methods . [ABSTRACT FROM AUTHOR]

Calibrating the microsimulation model is essential to enhance its ability to capture reality. The paper proposes a Bayesian neural network (BNN)-based method to calibrate parameters of microscopic traffic simulators, which reduces repeated running of simulations in the calibration and thus significantly improves the calibration efficiency. We use BNN with probability distributions on the weights to quickly predict the simulation results according to the inputs of the parameters to be calibrated. Based on the BNN model with the best performance, heuristic algorithms (HAs) are performed to seek the optimal values of the parameters to be calibrated with the minimum difference between the predicted results of BNN and the field-measured values. Three HAs are considered, including genetic algorithm (GA), evolutionary strategy (ES), and bat algorithm (BA). A TransModeler case of highway links in Shanghai, China, indicates the validity of the proposed calibration method in terms of error and efficiency. The results demonstrate that the BNN model is able to accurately predict the simulation and adequately capture the uncertainty of the simulation. We also find that the BNN-BA model produces the best calibration efficiency, while the BNN-ES model offers the best performance in calibration accuracy. [ABSTRACT FROM AUTHOR]

Railway transport becomes a more popular transportation in many countries due to its large transport capacity, low energy consumption, and benign environment. The passenger train service planning is the key of the rail operations system to balance the transport service and the passenger demand. In this paper, we propose a mixed binary linear programming formulation for the passenger train service planning to optimize the train route, frequency, stop schedule, and passenger assignment simultaneously. In addition, we analyze the computational complexities of the model and develop a Benders decomposition algorithm with valid inequalities to solve this problem. Finally, our model and algorithm are tested on a real-world instance of the Beijing-Shanghai high-speed railway line. The computational results show that our approach can solve these problems within reasonable solution time and small optimality gaps (less than 2.5%). [ABSTRACT FROM AUTHOR]

The problem of the optimal three-level location allocation of transfer center, processing factory and distribution center for supply chain network under uncertain transportation cost and customer demand are studied. We establish a two-stage fuzzy 0-1 mixed integer optimization model, by considering the uncertainty of the supply chain. Given the complexity of the model, this paper proposes a modified hybrid second order particle swarm optimization algorithm (MHSO-PSO) to solve the resulting model, yielding the optimal location and maximal expected return of supply chain simultaneously. A case study of clothing supply chain in Shanghai of China is then presented to investigate the specific influence of uncertainties on the transfer center, clothing factory and distribution center three-level location. Moreover, we compare the MHSO-PSO with hybrid particle swarm optimization algorithm and hybrid genetic algorithm, to validate the proposed algorithm based on the computational time and the convergence rate. [ABSTRACT FROM AUTHOR]

Disasters such as earthquake or tsunami can easily take the lives of thousands of people and millions worth of property in a fleeting moment. A successful emergency evacuation plan is critical in response to disasters. In this paper, we seek to investigate the multi-source, multi-destination evacuation problem. First, we construct a mixed integer linear programming model. Second, based on K shortest paths and user equilibrium, we propose a novel algorithm (hereafter KPUE), whose complexity is polynomial in the numbers of nodes and evacuees. Finally, we demonstrate the effectiveness of algorithm KPUE by a real evacuation network in Shanghai, China. The numerical examples show that the average computation time of the proposed algorithm is 95% less than that of IBM ILOG CPLEX solver and the optimality gap is no more than 5%. [ABSTRACT FROM AUTHOR]

Massive data broadcast by GPS-equipped vehicles provide unprecedented opportunities. One of the main tasks in order to optimize our transportation networks is to build data-driven real-time decision support systems. However, the dynamic environments where the networks operate disallow the traditional assumptions required to put in practice many off-the-shelf supervised learning algorithms, such as finite training sets or stationary distributions. In this paper, we propose BRIGHT: a drift-aware supervised learning framework to predict demand quantities. BRIGHT aims to provide accurate predictions for short-term horizons through a creative ensemble of time series analysis methods that handles distinct types of concept drift. By selecting neighborhoods dynamically, BRIGHT reduces the likelihood of overfitting. By ensuring diversity among the base learners, BRIGHT ensures a high reduction of variance while keeping bias stable. Experiments were conducted using three large-scale heterogeneous real-world transportation networks in Porto (Portugal), Shanghai (China), and Stockholm (Sweden), as well as with controlled experiments using synthetic data where multiple distinct drifts were artificially induced. The obtained results illustrate the advantages of BRIGHT in relation to state-of-the-art methods for this task. [ABSTRACT FROM AUTHOR]

Waste sorting is an imperative and significant issue in China, of which sorted-waste collection and transportation are indispensable parts. Despite its vital yet practical significance, few studies research mathematical models or algorithms of waste collection and transportation from the perspective of waste sorting. To address this issue, we extend a novel transportation model for the waste management system, namely, capacitated location routing problem with queuing time (CLRPQT) and design a cross-entropy and simulated-annealing based hyper-heuristic algorithm (CE-SAHH) for it. The main idea of this paper is three-fold: (1) As a particular property of this problem, source nodes cannot but need to be served by more than one vehicle that causes queuing time between a heterogeneous fleet of vehicles, which is novel in terms of the proposed model; (2) For the methodological contribution, a character encoding scheme, new decoding procedure, and local search strategy are designed embedded in the proposed method; (3) An integration of simulated annealing strategy and the cross-entropy-based hyper-heuristic algorithm is developed to overcome the combinatorial optimization problem with a more complex solution of this study. Finally, the results and analysis of three numeric experiments on benchmark datasets, new instances of CLRPQT, and simulation data in Shanghai, China, verify the effectiveness and universality of the proposed model and method. • A novel model with interactional routes for sorted-waste transportation is proposed. • A cross-entropy-based hyper-heuristic algorithm is proposed and performs well. • The novel model and method based on sorted waste enrich the waste management system. • Real-life data is used to verify the method and some managerial insights are drawn. [ABSTRACT FROM AUTHOR]

Time series prediction has been widely applied to the finance industry in applications such as stock market price and commodity price forecasting. Machine learning methods have been widely used in financial time series prediction in recent years. How to label financial time series data to determine the prediction accuracy of machine learning models and subsequently determine final investment returns is a hot topic. Existing labeling methods of financial time series mainly label data by comparing the current data with those of a short time period in the future. However, financial time series data are typically non-linear with obvious short-term randomness. Therefore, these labeling methods have not captured the continuous trend features of financial time series data, leading to a difference between their labeling results and real market trends. In this paper, a new labeling method called "continuous trend labeling" is proposed to address the above problem. In the feature preprocessing stage, this paper proposed a new method that can avoid the problem of look-ahead bias in traditional data standardization or normalization processes. Then, a detailed logical explanation was given, the definition of continuous trend labeling was proposed and also an automatic labeling algorithm was given to extract the continuous trend features of financial time series data. Experiments on the Shanghai Composite Index and Shenzhen Component Index and some stocks of China showed that our labeling method is a much better state-of-the-art labeling method in terms of classification accuracy and some other classification evaluation metrics. The results of the paper also proved that deep learning models such as LSTM and GRU are more suitable for dealing with the prediction of financial time series data. [ABSTRACT FROM AUTHOR]

With the boom of big data, the Internet contains more and more personal behavior information, but it is difficult to extract effectively. A model involving multivariate processing capability must be constructed to deal with these time series with complex characteristics. In this paper, a novel hybrid model embedding Baidu Search Index is therefore proposed to implement multi-step ahead subway passenger flow forecasting. Firstly, we collect data from informative Baidu Search Index, reduce dimensionality, and screen out the powerful predictors by statistical analysis. Secondly, we extract matching common modes at similar time scales between the subway passenger flow and screened Baidu Search Index via multivariate mode decomposition being optimized by multi-objective algorithm. Furthermore, to eliminate pseudo statistical causality, we select the optimal combination of modal components between subway passenger flow and its corresponding Baidu Search Index at each time scale by an innovative multi–modal analysis strategy. Thirdly, we reconstruct the forecasting values of each selected optimal combination as the final results. The empirical results of Beijing, Shanghai and Guangzhou show that the proposed model can significantly outperform six benchmark models in both the level and directional accuracy. So introducing Baidu Search Index creates a sound opportunity to enhance the subway passenger flow forecasting ability. [ABSTRACT FROM AUTHOR]

Vehicle-based sensors can be used for traffic monitoring. These sensors are usually set with long sampling intervals to save communication costs and to avoid network congestion. In this paper, we are interested in understanding the traffic-monitoring performance that we can expect from such vehicle-based mobile sensor networks, despite the incomplete information provided. This is a fundamental problem to be addressed. A performance evaluation has been carried out in Shanghai, China, by utilizing the vehicle-based sensors installed in about 4000 taxies. Two types of traffic status-estimation algorithms, i.e., the link-based and the vehicle-based, are introduced and analyzed. The results show that estimations of the traffic status based on these imperfect data are reasonably accurate. Therefore, the feasibility of such an application is demonstrated. [ABSTRACT FROM AUTHOR]

Purpose - Taking into account the long-term influences of the non-linear behavior of the material as well as the large deformation and contact conditions, the limiting factors of the computer simulation are the computer runtime and the memory requirement during solution of seismic response analysis for immersed tunnel. This research aims to overcome these problems. Design/methodology/approach - This research deals with parallel explicit finite element simulation with domain decomposition for seismic response analysis of immersed tunnel, which is the non-linear and time-dependent behavior of complex structures in engineering. A domain decomposition method based on parallel contact algorithm and dynamic-explicit time integration procedure are used, and the latter is used for the solution of the semi-discrete equations of motion, which is very suited for parallel processing. Using the high performance computer SGI Onyx3800, the seismic response analysis of the immersed tunnel in Shanghai is processed with more than 1.2 million nodes and more than 1 million elements in final finite element model. Findings - The results show numerical scalability of this algorithm and reveal the dangerous joints in this immersed tunnel under Tangshan seismic acceleration, and it could also provide references for the antiseismic design of the immersed tunnel. Originality/value - With the increasing demands in the scale, accuracy and speed of numerical simulation in geotechnical engineering, parallel computing has its great application in this area. This paper fulfils an identified method need, and it is believed more and more research work will be devoted to this research field in the near future. [ABSTRACT FROM AUTHOR]

Reinforcement learning has potential in the area of intelligent transportation due to its generality and real-time feature. The Q-learning algorithm, which is an early proposed algorithm, has its own merits to solve the train timetable rescheduling (TTR) problem. However, it has shortage in two aspects: Dimensional limits of action and a slow convergence rate. In this paper, a deep deterministic policy gradient (DDPG) algorithm is applied to solve the energy-aimed train timetable rescheduling (ETTR) problem. This algorithm belongs to reinforcement learning, which fulfills real-time requirements of the ETTR problem, and has adaptability on random disturbances. Superior to the Q-learning, DDPG has a continuous state space and action space. After enough training, the learning agent based on DDPG takes proper action by adjusting the cruising speed and the dwelling time continuously for each train in a metro network when random disturbances happen. Although training needs an iteration for thousands of episodes, the policy decision during each testing episode takes a very short time. Models for the metro network, based on a real case of the Shanghai Metro Line 1, are established as a training and testing environment. To validate the energy-saving effect and the real-time feature of the proposed algorithm, four experiments are designed and conducted. Compared with the no action strategy, results show that the proposed algorithm has real-time performance, and saves a significant percentage of energy under random disturbances. [ABSTRACT FROM AUTHOR]