22-01-2013, 11:03 AM
Developing a risk-sensitive active traffic management system
ABSTRACT
Our objective is to develop and validate a Risk-Sensitive Active Traffic Management (RSATM) system using micro simulation. This objective converges from two reasons. Firstly, we have been developing pro-active traffic risks identification models (RIM) using disaggregated traffic flow data, meteorological data and crash data collected from Swiss motorways. The results show that our models perform well with high percentage of recognized pre-crash situations and low percentage of misclassified normal situations. The outcome of the model development is that we have uncovered important factors having influence on crash occurrences. Regulating these factors will dominate the collision root. Secondly, Active Traffic Management (STM) systems have been developed and implemented in several countries. The main goal of these systems is to better manage traffic congestions which are more and more ubiquitous today. Applying our models and implementing ATM techniques as preventive measures when traffic risks are identified are the main idea of the RSATM system that we develop. The implementation and validation of the RSATM system is undertaken with the help of micro simulation.
Models for two crash types (rear-end and sideswipe) have been developed for two-lane per direction straight motorway road sections. The whole traffic is divided into eight scenarios representing eight different traffic conditions. Two models for each scenario are developed to identify traffic risks relating
to two crash types. The interpretation of important factors suggests two major causes of motorway crash occurrences: the speed variation on the right lane (the normal lane) and the speed difference between the right lane and the left lane (the overtaking lane). These two causes represent the inside-lane fluctuation of speed during a time period and the discrepancy of speed between two lanes. To prevent the risks, it is necessary to regulate the speed on both lanes using a speed harmonisation strategy. Dynamic speed limit regulation could be an option for the strategy with both dynamic lower and upper limits. Multiple gantries system will guarantee the smoothness of the traffic and is synchronized with the output of RIM.
The RSATM system with RIM as core part is implemented with the help of micro simulation. Simulated model is calibrated using real data as inputs such as OD matrix, vehicles properties. Thereafter, outputs of simulation model are compared with real data to tune models parameters. The calibration strategy starts with comparing lane capacity, speed distribution, lane distribution, and headway distribution. By calibration process, we uncover simulation issues when building simulation models for studying traffic safety. With the calibration results, we obtain the simulation models which represent the most the reality from traffic safety point of view. The next step is to implement RSATM system with different speed harmonisation strategies to find the best strategy. Finally we provide the evaluation of the outcome given by the RSATM system. Our preliminary results show that RSATM system really improves the traffic safety.
ABSTRACT
Our objective is to develop and validate a Risk-Sensitive Active Traffic Management (RSATM) system using micro simulation. This objective converges from two reasons. Firstly, we have been developing pro-active traffic risks identification models (RIM) using disaggregated traffic flow data, meteorological data and crash data collected from Swiss motorways. The results show that our models perform well with high percentage of recognized pre-crash situations and low percentage of misclassified normal situations. The outcome of the model development is that we have uncovered important factors having influence on crash occurrences. Regulating these factors will dominate the collision root. Secondly, Active Traffic Management (STM) systems have been developed and implemented in several countries. The main goal of these systems is to better manage traffic congestions which are more and more ubiquitous today. Applying our models and implementing ATM techniques as preventive measures when traffic risks are identified are the main idea of the RSATM system that we develop. The implementation and validation of the RSATM system is undertaken with the help of micro simulation.
Models for two crash types (rear-end and sideswipe) have been developed for two-lane per direction straight motorway road sections. The whole traffic is divided into eight scenarios representing eight different traffic conditions. Two models for each scenario are developed to identify traffic risks relating
to two crash types. The interpretation of important factors suggests two major causes of motorway crash occurrences: the speed variation on the right lane (the normal lane) and the speed difference between the right lane and the left lane (the overtaking lane). These two causes represent the inside-lane fluctuation of speed during a time period and the discrepancy of speed between two lanes. To prevent the risks, it is necessary to regulate the speed on both lanes using a speed harmonisation strategy. Dynamic speed limit regulation could be an option for the strategy with both dynamic lower and upper limits. Multiple gantries system will guarantee the smoothness of the traffic and is synchronized with the output of RIM.
The RSATM system with RIM as core part is implemented with the help of micro simulation. Simulated model is calibrated using real data as inputs such as OD matrix, vehicles properties. Thereafter, outputs of simulation model are compared with real data to tune models parameters. The calibration strategy starts with comparing lane capacity, speed distribution, lane distribution, and headway distribution. By calibration process, we uncover simulation issues when building simulation models for studying traffic safety. With the calibration results, we obtain the simulation models which represent the most the reality from traffic safety point of view. The next step is to implement RSATM system with different speed harmonisation strategies to find the best strategy. Finally we provide the evaluation of the outcome given by the RSATM system. Our preliminary results show that RSATM system really improves the traffic safety.