Abstract:There has been extensive research in traffic state estimation that accounts for the stochastic nature of traffic flow models. However, these studies often exhibit limitations such as an exclusive focus on motorway traffic and a reliance on a single data source. This paper departs from these methods by introducing a stochastic estimation framework that is specifically designed for urban arterial traffic. The framework has the capability to incorporate multiple data sources, which serves to improve its accuracy and robustness. The framework is composed of three components: (i) a stochastic traffic flow model, (ii) a filtering algorithm, and (iii) an algorithm for incorporating multi-source measurements. In terms of the traffic model, we introduce a new stochastic Switching Mode Model that can be applied to arterial roads that have both signalized and unsignalized intersections. This model does not consider uncertainty in the current mode of operation, which substantially reduces the computational complexity because there is only one mode at each time step. Furthermore, we propose three different filtering algorithms for multi-source traffic estimation, including the incremental stochastic Kalman Filter (SKF), the incremental stochastic Unscented Kalman Filter (SUKF), and the hybrid approach. Since the SKF can only deal with linear functions, non-linear measurement equations need to be linearized using first-order Taylor expansions. The SUKF is based on the Unscented Transform (UT), which enables it to work with a wider range of functions regardless of linearity, non-linearity, or non-differentiability. The hybrid algorithm is a combination of the SKF and the SUKF, in which linear equations are treated similarly to the SKF, and non-linear equations are handled with the UT in the same way as in the SUKF. The performances of these algorithms were similar when applied to the synthetic data of an urban arterial in Christchurch CBD, New Zealand. The hybrid algorithm, however, worked slightly better and was more stable than the other two.

A Mesoscopic Traffic Data Assimilation Framework for Vehicle Density Estimation on Urban Traffic Networks Based on Particle Filters

Adaptive Rolling Smoothing with Heterogeneous Data for Traffic State Estimation and Prediction.

Traffic State and Emission Estimation for Urban Expressways Based on Heterogeneous Data

An Extended Generalized Filter Algorithm for Urban Expressway Traffic Time Estimation Based on Heterogeneous Data

Real-time Freeway Traffic State Estimation Based on Extended Kalman Filter: Adaptive Capabilities and Real Data Testing

Urban Network-Wide Traffic Speed Estimation with Massive Ride-Sourcing GPS Traces

Turning traffic surveillance cameras into intelligent sensors for traffic density estimation

Freeway Traffic Estimation in Beijing Based on Particle Filter.

Data assimilation approach for addressing imperfections in people flow measurement techniques using particle filter

State Estimation of Unequipped Vehicles Utilizing Microscopic Traffic Model and Principle of Particle Filter

Network-wide Traffic Flow Estimation with Insufficient Volume Detection and Crowdsourcing Data

Stochastic Switching Mode Model based Filters for urban arterial traffic estimation from multi-source data

Particle Filter for Estimating Freeway Traffic State in Beijing

Data assimilation for online model calibration in discrete event simulation

Real-Time Traffic Estimation with Incomplete Information under Urban Traffic Network

Traffic Estimation Based on Particle Filtering with Stochastic State Reconstruction Using Mobile Network Data

Fusing Physics-Based and Data-Driven Models for Car-Following Modeling: A Particle Filter Approach

Data Assimilation in Air Contaminant Dispersion Using a Particle Filter and Expectation-Maximization Algorithm

A Traffic Density Estimation and Congestion Identification of Urban Freeways Based on Kalman Filter

Vehicle Path Reconstruction Using Automatic Vehicle Identification Data: an Integrated Particle Filter and Path Flow Estimator

Freeway network density estimation based on Dynamic Graph Hybrid Automata model by using Kalman filter