Estimating multi-class dynamic origin-destination demand through a forward-backward algorithm on computational graphs

Ma, Wei; Pi, Xidong; Qian, Sean

doi:10.1016/j.trc.2020.102747

Cited by 41 publications

(24 citation statements)

References 63 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To accurately capture spatial patterns of on-road traffic, we use the output of a link-level, origin-destination by vehicle class traffic model of Pittsburgh (Ma et al, 2020). This traffic model simulates traffic counts and speed by hour of day using observations from Pennsylvania Department of Transportation sites throughout Pittsburgh.…”

Section: Model Applicationmentioning

confidence: 99%

“…Improvements in the resolution of emission inventories have been focused on traffic as this source exhibits significant variability at high resolutions. Recent approaches to building high-resolution traffic inventories include origin-destination by vehicle class (Ma et al, 2020), synthetic population mobility (Elessa Etuman and Coll, 2018), and fuel sales combined with traffic counts (McDonald and McBride, 2014). Other sectors such as biomass burning for residential heating and commercial cooking have been identified as very uncertain in current inventories (Day et al, 2019).…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Source-resolved variability of fine particulate matter and human exposure in an urban area

et al. 2022

View full text Add to dashboard Cite

Abstract. Increasing the resolution of chemical transport model (CTM) predictions in urban areas is important to capture sharp spatial gradients in atmospheric pollutant concentrations and better inform air quality and emissions controls policies that protect public health. The chemical transport model PMCAMx (Particulate Matter Comprehensive Air quality Model with Extensions) was used to assess the impact of increasing model resolution on the ability to predict the source-resolved variability and population exposure to PM2.5 at 36×36, 12×12, 4×4, and 1×1 km resolutions over the city of Pittsburgh during typical winter and summer periods (February and July 2017). At the coarse resolution, county-level differences can be observed, while increasing the resolution to 12×12 km resolves the urban–rural gradient. Increasing resolution to 4×4 km resolves large stationary sources such as power plants, and the 1×1 km resolution reveals intra-urban variations and individual roadways within the simulation domain. Regional pollutants that exhibit low spatial variability such as PM2.5 nitrate show modest changes when increasing the resolution beyond 12×12 km. Predominantly local pollutants such as elemental carbon and primary organic aerosol have gradients that can only be resolved at the 1×1 km scale. Contributions from some local sources are enhanced by weighting the average contribution from each source by the population in each grid cell. The average population-weighted PM2.5 concentration does not change significantly with resolution, suggesting that extremely high resolution PM2.5 predictions may not be necessary for effective urban epidemiological analysis at the county level.

show abstract

Section: Model Applicationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Source-resolved variability of fine particulate matter and human exposure in an urban area

et al. 2022

View full text Add to dashboard Cite

show abstract

“…With the underlying mathematical tool performing AD, CG is an effective modeling platform which is able to deal with large-scale data and capture nonlinear relationship. As CG-based NL model is just one part of the DACPE framework, interested readers are referred to papers by Wu et al, 2018a , Ma et al, 2020 on the underlying computational graph building blocks. With the connection across main elements of the NL model (namely Eqs.…”

Section: Problem Statement and Methodologymentioning

confidence: 99%

Dynamic activity chain pattern estimation under mobility demand changes during COVID-19

Liu

Zhu

et al. 2021

Transportation Research Part C: Emerging Technologies

View full text Add to dashboard Cite

During the coronavirus disease 2019 pandemic, the activity engagement and travel behavior of city residents have been impacted by government restrictions, such as temporary city-wide lockdowns, the closure of public areas and public transport suspension. Based on multiple heterogeneous data sources, which include aggregated mobility change reports and household survey data, this paper proposes a machine learning approach for dynamic activity chain pattern estimation with improved interpretability for examining behavioral pattern adjustments. Based on historical household survey samples, we first establish a computational graph-based discrete choice model to estimate the baseline travel tour parameters before the pandemic. To further capture structural deviations of activity chain patterns from day-by-day time series, we define the activity-oriented deviation parameters within an interpretable utility-based nested logit model framework, which are further estimated through a constrained optimization problem. By incorporating the long short-term memory method as the explainable module to capture the complex periodic and trend information before and after interventions, we predict day-to-day activity chain patterns with more accuracy. The performance of our model is examined based on publicly available datasets such as the 2017 National Household Travel Survey in the United States and the Google Global Mobility Dataset throughout the epidemic period. Our model could shed more light on transportation planning, policy adaptation and management decisions during the pandemic and post-pandemic phases.

show abstract

“…Yet, the method failed to integrate car-sharing modes. In the literature, 40 the authors designed a forward-backward algorithm as a solution to overcome MCDODE formulation on computational graphs. The authors also devised a fresh concept, that is, tree-based cumulative curves to determine multi-class dynamic assignment ratio (DAR) matrix.…”

Section: Literature Reviewmentioning

confidence: 99%

Intelligent optimization for charging scheduling of electric vehicle using exponential Harris Hawks technique

Devendiran¹,

Kasinathan²,

Ramachandaramurthy

et al. 2021

Int J Intell Syst

View full text Add to dashboard Cite

The coordination of modern transportation system depends heavily on intelligent techniques, information assortment, and its analysis. Sensors play a crucial role in information assortment in charging scheduling of electric vehicles (EVs). EVs are destined to become inevitable due to their innate economic contribution, climate improvement, and social attributes as per United Nation's sustainable development goals. Innovation in EV has gained the interest of many researchers since it is one of the novel green transportation sectors. Moreover, EVs are essential to preserve conventional fuels and to maximize the utilization of renewable sources. Nevertheless, EVs have short driving ranges due to their battery limitation, which hinders the reliability. The charging stations (CS) for EVs are also unevenly distributed. This paper presents a novel strategy to schedule the charging points in EV CSs. The goal is to determine the convenient CS for EVs through Vehicular Ad-hoc Network (VANET) model. In this model, the CSs are determined and prioritized using four phases, such as driving, charge planning, charging scheduling, and

show abstract

Estimating multi-class dynamic origin-destination demand through a forward-backward algorithm on computational graphs

Cited by 41 publications

References 63 publications

Source-resolved variability of fine particulate matter and human exposure in an urban area

Source-resolved variability of fine particulate matter and human exposure in an urban area

Dynamic activity chain pattern estimation under mobility demand changes during COVID-19

Intelligent optimization for charging scheduling of electric vehicle using exponential Harris Hawks technique

Contact Info

Product

Resources

About