Machine learning applications in activity-travel behaviour research: a review

ABSTRACT

This paper reviews the activity-travel behaviour literature that employs Machine Learning (ML) techniques for empirical analysis and modelling. Machine Learning algorithms, which attempt to build intelligence utilizing the availability of large amounts of data, have emerged as powerful tools in the fields of pattern recognition and big data analysis. These techniques have been applied in activity-travel behaviour studies since the early ’90s when Artificial Neural Networks (ANN) were employed to model mode choice decisions. AMOS, an activity-based modelling system developed in the mid-’90s, has ANN at its core to model and predict individual responses to travel demand management measures. In the dawn of 2000, ALBATROSS, a comprehensive activity-based travel demand modelling system, was proposed by Arentze and Timmermans using Decision Trees. Since then researchers have been exploring ML techniques like Support Vector Machines (SVM), Decision Trees (DT), Neural Networks (NN), Bayes Classifiers, and more recently, Ensemble Learners to model and predict activity-travel behaviour. A large number of publications over the years and an upward trend in the number of published articles over time indicate that Machine Learning is a promising tool for activity-travel behaviour analysis and prediction. This article, first of its kind in the literature, reviews these studies and explores the trends in activity-travel behaviour research that apply ML techniques. The review finds that mode choice decisions have received wide attention in the literature on ML applications. It was observed that most of the studies identify the lack of interpretability as a serious shortcoming in ML techniques. However, very few studies have attempted to improve the interpretability of the models. Further, some studies report the importance of feature engineering in ML-based studies, but very few studies adopt feature engineering before model development. Spatiotemporal transferability of models is another issue that has received minimal attention in the literature. In the end, the paper discusses possible directions for future research in the area of activity-travel behaviour modelling using ML techniques.     

牵引供电系统优化设计研究

本文论述电气化铁道供电系统优化设计的新课题。在建立牵引仿真(或牵引计算)数据库基础上,应用数学规划方法提出了寻优设计的两个数学模型:最小能损模型和最少工程费用模型,并详细讨论了以最小能损为目标函数的数学模型及算法。给出的算例表明,按寻优设计方案能有效地减少整个牵引供电系统中的功率损失(能量损失)。     

基于地质横剖面的隧道工程三维地质建模方法研究

为直观准确地表达地质与隧道工程之间的相互关系,对三维地质与隧道的建模方法进行研究。首先,提出一种基于地质横剖面轮廓线的归一化三维地质体模型重构方法; 然后,以已建立的三维地质模型为基础,对基于隧道横断面的三维建模及隧道与地质模型合并过程中的模型切割、分层进行研究; 最后,开发基于断面的三维隧道地质建模及可视化系统。通过建模实例显示: 采用本文的方法能够取得良好的三维地质及隧道建模效果,直观展示隧道与地层间的相互关系,较为丰富地展现隧道的内部细节。     

Simulating Demand-responsive Transportation: A Review of Agent-based Approaches

Abstract

In light of the need to make better use of existing transport infrastructure, demand-responsive transportation (DRT) systems are gaining traction internationally. However, many systems fail due to poor implementation, planning, and marketing. Being able to realistically simulate a system to evaluate its viability before implementation is important. This review investigates the application of agent-based simulation for studying DRT. We identify that existing simulations are strongly focused on the optimisation of trips, usually in favour of the operator, and rarely consider individual preferences and needs. Agent-based simulations, however, permit incorporation of the latter, as well as capture the interactions between operators and customers. Several areas of future research are identified in order to unify future research efforts.     

Road Safety Trends at National Level in Europe: A Review of Time-series Analysis Performed during the Period 2000–12

Abstract

This paper presents a review of time-series analysis of road safety trends, aggregated at a national level, which has been performed in the period 2000–12 and applied to European national data sets covering long time periods. It provides a guideline and set of best practices in the area of time-series modelling and identifies the latest methods and applications of national road safety trend analysis in Europe. The paper begins with the methodological framework adopted for aggregate time-series modelling that will be considered, and then discusses a number of relevant applications to long-period data aggregated at the national level, whether for countries alone, or for groups of countries. Some analyses, which were performed at the disaggregated level, are also provided, as they are being used more and more. Finally, the paper summarizes and discusses the significant changes in aggregate road safety trend analysis which occurred during the period and provides recommendations for continuing these research efforts.     

A tool for calculating aircraft emissions and its application to Greek airspace

This paper quantifies the impact of aircraft emissions on local air quality and climate change. Aircraft emissions during the cruise cycle and the landing/take-off cycle are considered. A tool is developed that computes emission values using real-time air traffic data derived from various databases. Emissions include carbon dioxide, hydrocarbons, carbon monoxide and nitrogen oxides. The overall output is a detailed ‘emissions map’ of a given territory that enables the identification of critical emission spots including routes, airports, season, aircraft type and flight category. The method can be used for real-time monitoring of airline emissions and for policy analysis. The proposed tool and resulting outputs are illustrated in the case of the Greek airport system using domestic, international and overflights. Demand volatility driven mainly by tourism and its impact on emissions is assessed.     

Campus parking supply impacts on transportation mode choice

Parking demand is a significant land-use problem in campus planning. The parking policies of universities and large corporations with facilities located in small urban areas shape the character of their campuses. These facilities will benefit from a simplified methodology to study the effects of parking availability on transportation mode mix and impacts on recruitment and staffing policies. This paper, based on a case study of North Dakota State University in the United States, introduces an analytical framework to provide planners with insights about how parking supply and demand affects campus transportation mode choice. The methodology relies only on aggregate mode choice data for the special generator zone and the average aggregate volume/capacity ratio projections for all external routes that access the zone. This reduced data requirement significantly lowers analysis cost and obviates the need for specialized modelling software and spatial network analysis tools. Results illustrate that the framework is effective for analysing mode choice changes under different scenarios of parking supply and population growth.     

The asymmetric effects of income and fuel price on air transport demand

Forecasts of passenger demand are an important parameter for aviation planners. Air transport demand models typically assume a perfectly reversible impact of the demand drivers. However, there are reasons to believe that the impacts of some of the demand drivers such as fuel price or income on air transport demand may not be perfectly reversible. Two types of imperfect reversibility, namely asymmetry and hysteresis, are possible. Asymmetry refers to the differences in the demand impacts of a rising price or income from that of a falling price or income. Hysteresis refers to the dependence of the impacts of changing price or income on previous history, especially on previous maximum price or income. We use US time series data and decompose each of fuel price and income into three component series to develop an econometric model for air transport demand that is capable of capturing the potential imperfectly reversible relationships and test for the presence or absence of reversibility. We find statistical evidence of asymmetry and hysteresis – for both, prices and income – in air transport demand. Implications for policy and practice are then discussed.     

Use of Bayesian inference method to model vehicular air pollution in local urban areas

Traffic Related Air Pollution (TRAP) studies are usually investigated using different categories such as air pollution exposure for health impacts, urban transportation network design to mitigate pollution, environmental impacts of pollution, etc. All of these subfields often rely on a robust air pollution model, which also necessitates an accurate prediction of future pollutants. As is widely accepted by the heath authorities, TRAP is considered to be the major health issue in urban areas, and it is difficult to keep pollution at harmless levels if the time sequenced dynamic pollution and traffic parameters are not identified and modelled efficiently. In our work here, artificial intelligence techniques, such as Bayesian Networks with an optimized configuration, are used to deliver a probabilistic traffic data analysis and predictive modelling for air pollution (SO₂, NO₂ and CO) at very local scale of an urban region with up to 85% accuracy. The main challenge for traditional data analysis is a lack of capability to reveal the hidden links between distant data attributes (e.g. pollution sources, dynamic traffic parameters, etc.), whereas some subtle effects of these parameters or events may play an important role in pollution on a long-term basis. This study focuses on the optimisation of Bayesian Networks to unveil hidden links and to increase the prediction accuracy of TRAP considering its further association with a predictive GIS system.     

Assessing the importance of transportation activity data for urban emission inventories

The aim of this research is the implementation of a GPS-based modelling approach for improving the characterization of vehicle speed spatial variation within urban areas, and a comparison of the resulting emissions with a widely used approach to emission inventory compiling. The ultimate goal of this study is to evaluate and understand the importance of activity data for improving the road transport emission inventory in urban areas. For this purpose, three numerical tools, namely, (i) the microsimulation traffic model (VISSIM); (ii) the mesoscopic emissions model (TREM); and (iii) the air quality model (URBAIR), were linked and applied to a medium-sized European city (Aveiro, Portugal). As an alternative, traffic emissions based on a widely used approach are calculated by assuming a vehicle speed value according to driving mode. The detailed GPS-based modelling approach results in lower total road traffic emissions for the urban area (7.9, 5.4, 4.6 and 3.2% of the total PM10, NO_x, CO and VOC daily emissions, respectively). Moreover, an important variation of emissions was observed for all pollutants when analysing the magnitude of the 5th and 95th percentile emission values for the entire urban area, ranging from −15 to 49% for CO, −14 to 31% for VOC, −19 to 46% for NO_x and −22 to 52% for PM10. The proposed GPS-based approach reveals the benefits of addressing the spatial and temporal variability of the vehicle speed within urban areas in comparison with vehicle speed data aggregated by a driving mode, demonstrating its usefulness in quantifying and reducing the uncertainty of road transport inventories.