Electric vehicles have been gaining attention as a cleaner means of transportation that is low-carbon and environmentally friendly and can reduce greenhouse gas emissions and air pollution. Despite EVs’ many advantages, widespread adoption will negatively affect the electric grid due to their random and volatile nature. Consequently, predicting the charging demand for electric vehicles is becoming a priority to maintain a steady supply of electric energy. Time series methodologies are applied to predict the charging demand: traditional and deep learning. RNN, LSTM, and transformers represent deep learning approaches, while ARIMA and SARIMA are traditional techniques. This research represents one of the first attempts to use the Transformer model for predicting EV charging demand. Predictions for 3-time steps are considered: 7 days, 30 days, and 90 days to address both short-term and long-term forecasting of EV charging load. RMSE and MAE were used to compare the model’s performance. According to the results, the Transformer outperforms the other mentioned models in terms of short-term and long-term predictions, demonstrating its ability to address time series problems, especially EV charging predictions. The proposed Transformers framework and the obtained results can be used to manage electricity grids efficiently and smoothly.
Electric vehicles (EVs) contribute to reducing fossil fuel dependence and environmental pollution problems. However, due to complex charging behaviors and the high demand for charging, EVs have imposed significant burdens on power systems. By providing reliable forecasts of electric vehicle charging loads to power systems, these issues can be addressed efficiently to dispatch energy. Machine learning techniques have been demonstrated to be effective in forecasting loads. This research applies six machine learning methods to predict the charging demand for EVs: RNN, LSTM, Bi-LSTM, GRU, CNN, and transformers. A dataset containing five years of charging events collected from 25 public charging stations in Boulder, Colorado, USA, is used to validate this approach. Compared to other highly applied machine learning models, the transformer method outperforms others in predicting charging demand, demonstrating its ability for time series forecasting problems.
15 September 2019 in Transportation Research Record: Journal of the Transportation Research Board
The concepts of connectivity, accessibility, and mobility (CAM) are key measures of transport network performance that have been discussed extensively in the literature. However, there has been little work that discussed the relationships among these concepts. A clear discourse on these concepts and their interrelationships can help agencies carry out more objective evaluations of projects that seek to improve at least one of these measures of transportation performance. This paper presents three alternative perspectives (models) of the CAM relationship: the nested, snowman, and three-way overlapping models. The paper also presents, for project appraisal purposes, two alternative ways of classifying the three CAM concepts. The first is based on the three concepts in their basic forms; the second considers some variation of these concepts in addition to aspects of the network topology, operational performance, road condition, and socioeconomic characteristics of the project’s area of influence. The conceptual framework outlined in this paper contributes towards a holistic approach to the appraisal of low-volume road projects, programs, or existing networks based on their impact on overall CAM.
5 January 2020 in Computer-Aided Civil and Infrastructure Engineering
Construction projects are often associated with partial or full road closures, which result in user costs and community disruptions in terms of reduced business productivity. A number of studies have addressed the problem of scheduling construction projects based on a variety of stakeholder objectives. Yet still, there seems to exist a few gaps regarding (1) possible tradeoffs between road user cost reduction and business cost reduction associated with optimal scheduling, (2) role of the project type (rehabilitation and capacity expansion) on the solution methodology, and (3) lack of solution algorithm to address the problem complexity by deriving the optimal solution. In addressing these gaps, this article adopts a novel approach for developing an optimal project schedule for multiple road projects within a construction horizon. The goal is to minimize the overall cost of the projects to road users and adjacent businesses over the construction horizon. The project scheduling problem is formulated as a mixed‐integer nonlinear program. We solve the problem using a local decomposition method. The methodology is demonstrated using the Sioux Falls city network with two project types: capacity expansion and rehabilitation. The results of the numerical experiment suggest that (1) the solution algorithm converges to optimal solution in finite iterations and (2) a network‐wide scheduling of urban road projects using explicit optimization can yield a significant reduction in business disruption costs while incurring a relatively smaller increase in system travel time, and overall, is superior to a schedule developed only considering the total system travel time.
For the purposes of long-term planning and budgeting, infrastructure user cost allocation, and financial need forecasts, infrastructure agencies seek knowledge of the annual expenditure levels for maintaining their assets. Often, this information is expressed in dollars per unit dimension of the infrastructure and is estimated using observed data from historical records. This paper presents an artificial neural network (ANN) approach for purposes of estimating annual expenditures on infrastructure maintenance and demonstrates the application of the approach using a case study involving rural interstate highway pavements. The results of this exploratory study demonstrate that not only is it feasible to use ANN to derive reliable predictions of annual maintenance expenditures (AMEX) at aggregate level, but also it is possible to identify the influential factors of such expenditures and to quantify the sensitivity of AMEX to such factors.
21 April 2017 in Transportmetrica A: Transport Science
Lane and shoulder widths are important highway safety factors because wider lanes and shoulders generally help to reduce crashes. In practice however, due to physical limitations of the right-of-way or statutory restrictions, it may be the case that the existing overall width of the roadway (lanes plus shoulders) cannot exceed a certain value. For a given overall roadway width constraint, wider lanes mean there will be smaller space for shoulders, and vice versa. In such zero-sum situations, it is not certain if a cross-sectional configuration with wide lanes and narrow shoulders is safer compared with one with narrow lanes and wide shoulders. Clearly, for a given total roadway width (TRW), the trade-off between shoulder and lane widths should be evaluated in order to maximize the user benefits (safety) without unduly incurring excessive agency life-cycle costs of construction and preservation. The TRW allocations to the lane and shoulder, expressed as the ratio of lane width to shoulder width, will generally depend on the TRW; lane and shoulder pavement material types and costs; and traffic volume, among other factors. In order to optimize the lane and shoulder width allocations for a given TRW, this paper proposes a framework that formulates the problem as an optimization problem with the objective of minimizing the total (agency and user) life-cycle cost of the roadway. First, the relationships are established between lane/shoulder widths and the resulting user costs (crashes), and also between lane/shoulder widths and the associated agency costs of lane/shoulder construction and preservation. The sensitivity of the optimal solution to different evaluation inputs (such as the relative weights between the agency and user costs) is also analyzed. Using the developed framework, the paper presents a number of decision support charts that can be used by highway agencies to determine the optimal lane and shoulder widths under a given set of conditions including the highway functional class and the total available roadway width. The flexibility of the proposed optimization framework to accommodate risk compensation effects was also demonstrated.