Marialisa Nigro

An Adaptive Bi-Level Gradient Procedure for the Estimation of Dynamic Traffic Demand

This paper presents an in-depth analysis of the bi-level gradient approximation approach for dynamic traffic demand adjustment and the development of new adaptive approaches. Initially, a comparison between the simultaneous perturbation stochastic approximation (SPSA), asymmetric design (AD), polynomial interpolation (PI) method, which was first proposed by authors in 2010-2011, and its second-order development is presented; then, a sensitivity analysis of the parameters of the SPSA AD-PI is reported; finally, some new advances of the estimation method based on an adaptive approach are proposed and evaluated on a real test network.

Investigating the Efficiency of a Gradient Approximation Approach for the Solution of Dynamic Demand Estimation Problems

In this chapter a gradient approximation approach has been applied to solve the dynamic origin-destination (O-D) simultaneous adjustment problem. It is formulated as an optimization problem aiming at minimizing the error between actual and estimated link observations (speed and volumes) and the distance between estimated and a priori O-D demand flows in a dynamic context. Different novelties, both in the problem formulation and in the solution procedure, have been introduced with the aim of improving the solution and reducing the computational times.

Two-Step Approach for Correction of Seed Matrix in Dynamic Demand Estimation

In this work, deterministic and stochastic optimization methods are tested for solving the dynamic demand estimation problem. All the adopted methods demonstrate difficulty in reproducing the correct traffic regime, especially if the seed matrix is not sufficiently close to the real one. Therefore, a new and intuitive procedure to specify an opportune starting seed matrix is proposed: it is a two-step procedure based on the concept of dividing the problem into small problems, with a focus on specific origin–destination (O-D) pairs in different steps. Specifically, the first step focuses on the optimization of a subset of O-D variables (the ones that generate the higher flows or the ones that generate bottlenecks on the network). In the second step the optimization works on all the O-D pairs, with the matrix derived from the first step as starting matrix. In this way it is possible to use a performance optimization method for every step; this technique improves the performance of the method and the quality of the result with respect to the classical one-step approach. The procedure was tested on the real-world network of Antwerp, Belgium, and demonstrated its efficacy in combination with different optimization methods.

A bi-level gradient approximation method for dynamic traffic demand estimation: Sensitivity analysis and adaptive approach

The paper presents an in-depth analysis of the bi-level gradient approximation approach for dynamic traffic demand estimation. Initially, a sensitivity analysis of the parameters of Simultaneous Perturbation Stochastic Approximation method (SPSA) with Asymmetric Design (AD) and Polynomial Interpolation (PI), firstly proposed by authors in 2011, is presented. Then, an adaptive method based on the second order SPSA AD-PI approach, is explored; finally, some new advances of the estimation method are proposed in order to keep under control traffic phenomena during the estimation.

Walkability Indicators for Pedestrian-Friendly Design

This paper presents an analysis of walking indicators related to the structure of the road network, differentiated by measures of the quality, connectivity, and proximity of the road network. These measures were computed for various zones of the city of Rome and for the historical center of Lucca and Venice in Italy. The aims of the study were (a) to understand whether some measures were more suitable than others for describing the walkability (i.e., accessibility to walking) of an area, (b) to define the best single measure or the optimal combination of measures to describe the walkability of an area, (c) to define some benchmark values for the analyzed walkability measures, and (d) to obtain valuable guidelines to define a pedestrian-oriented road network. The results showed the importance of variables such as the number of nodes and the size of the blocks. Moreover, the results demonstrated that although single measures of connectivity were not self-explanatory for describing walkability, the combination of various measures could be more effective. Finally, the study derived the benchmark values of 3 to 6 nodes per hectare for the density of nodes, 0.5 to 0.9 hectare per block for the size of blocks, and 800 m for the maximum walking distance for a pedestrian-friendly, transit-oriented development.

A new method to recover the correct land use and public transport interaction

In the last years the increasing use of private vehicles in urban areas creates negative impacts on the society particularly for the congestion, implying an increase of travel times, of air and noise pollution, of accidents, and the excessive production of greenhouse gases and land consumption. Public transport could represent a more efficient mode of travel with respect to the car, playing an important role to provide a more sustainable transport system. However, infrastructural actions operated on the public transport system are usually of long term, with respect to actions operated on land use, thus creating a considerable temporal gap between the land use and the transport system development. Starting from these remarks, the present study proposes a new method to overcome this temporal discrepancy, using the residual capacity of the mass transit system (existing or its short term development) as a variable to indicate the location and the magnitude of new residential and activities developments. The method has been applied to the city of Rome (Italy), suggesting how the Local Authority could guide the development of the urban area in a sustainable way for the next years.

A dynamic mesoscopic emission model for signalized intersections

The paper develops a mesoscopic emission model for signalized intersections that takes into account the dynamic variability of traffic conditions. It starts from an analytical model based on Akcelic theory and it permits to distinguish between vehicles in queue and vehicles entering/exiting the queue (deceleration and acceleration phases) using data derived from a Dynamic Traffic Assignment (DTA).

The value of real-time traffic information in urban freight distribution

ABSTRACT Routes optimization in urban freight distribution is usually an off-line process based on the knowledge of historical conditions on the network. Real-time data provided by Intelligent Transportation Systems (ITSs) enable online re-optimization on the basis of actual traffic conditions.  This paper evaluates the added value generated by re-optimizing the off-line solution with real-time information. The study is carried out for a practical application to the freight distribution of perishable goods in the city of Rome (Italy). The off-line problem is formulated as a vehicle routing problem with soft time windows while in the online problem it is also allowed to skip some customers or to re-sequence the deliveries. Both versions are solved with different algorithms and with different data sets. Results can be used to evaluate the potential return on investment on the acquisition of different kinds of traffic data. At the same time, results can be of interest for information providers, to fix the price of off-line and online information and/or to estimate the associated potential market share.

Evaluation of the impact of e-mobility scenarios in large urban areas

In this paper we present a comparison, based on pollutant emissions and fuel consumptions, of two alternative hypotheses for private cars renewal in a large urban area. One of the two scenarios includes a partial replacement of the internal combustion engine vehicles with the electric ones. Following a data-driven approach, we calculate well-to-wheel emissions and consumption using real traffic patterns from Floating Car Data. Results show that a 10% penetration of electric vehicles can induce a significant reduction in the primary energy consumption and in the climate-change and pollutant emissions.

Exploiting floating car data for time-dependent Origin–Destination matrices estimation

ABSTRACT The study evaluates the added value generated by estimating dynamic demand matrices by information gathered from Floating Car Data (FCD). Firstly, adopting a large dataset of FCD collected in Rome, Italy, during May 2010, all the monitored trips on a specific district of the city (Eur district) have been collected and analysed in terms of (i) spatial and temporal distribution; (ii) actual route choices and travel times. The data analysis showed that demand data from FCD are usually not suitable to retrieve directly demand matrices, due to a strong dependence of this information from the penetration rate of the monitoring device. Instead, origin–destination travel times and route choice probabilities from FCD are a much more reliable and powerful information with respect to FCD origin–destination flows, since they represent the traffic conditions and behaviors that vehicles experiment along the path. Thus, several synthetic experiments have been conducted adopting both travel times and route choice probabilities as additional information, with respect to standard link measurements, in the dynamic demand estimation problem. Results demonstrated the strength and robustness associated to these network based data, while link measurements alone are not able to define the real traffic pattern. Adopting both the information of origin–destination travel times and route choice probabilities during the demand estimation process, the spatial and temporal reliability of the estimated demand matrices consistently increases.