Then, the lateral placement of different types of vehicles on the entire cross-section of roads of different widths is studied. The driving behavior of different types of vehicles is analyzed using the friction circle concept (g-g diagram) based on the dynamic parameters obtained from the field data.
Introduction
Problem Statement
Motivation
Research Objective
The study of the lateral placement of different types of vehicles across the cross-section of roads of different types (roads of different widths). Study of side clearance behavior of non-median vehicles based on median geometry.
Scope of the Research Work
Study the impact of dynamic parameters (operating speed, lateral acceleration and longitudinal acceleration/deceleration) on lateral maneuvering of various vehicles on Indian roads. Evaluation of the importance of the microscopic parameters studied from the field traffic using a commercially available microsimulation framework, VISSIM.
Outline of the Thesis
The influence of any external features such as crossing or merging traffic, pedestrians and other road obstructions, grade, horizontal or vertical bends, pavement conditions, environmental conditions or land use conditions are not considered on these microscopic parameters except for traffic, width roads. and the type of median, which influence the driver's behavior in the traffic flow. Traffic data should be collected from different sites relevant to this area in order to study each microscopic parameter in detail and establish relationships between them.
Introduction
One-Dimensional Microscopic Models
- GHR Model
- Linear Model
- Safe-Distance Model
- Optimal Velocity Model
- Intelligent Driver Model
- Psychophysical Model (Action-Point Model)
- Fuzzy logic Model
- Cellular Automata (CA) model
- Summary
To improve OVM and consider ITS applications, two velocity difference model (TVDM) was developed by (Ge, et al., 2008), and the formulation of the model is;. The validation of the model was performed on a Dutch highway for different percentages of cars and trucks (Treiber et al., 2000b).
Two-Dimensional Microscopic Models
- Modified GHR Approach
- Modified Safe-Distance Approach
- Modified Optimal Velocity Approach
- Modified Cellular Automata Approach
- Other Modelling Approaches Developed for 2D Traffic
- Summary
However, the model does not take into account the lateral separation of the immediate leader from the following vehicle. Most of these models did not represent the driving behavior of the mixed traffic flow.
Studies on Lane Changing Models
Lane Selection Model
Gap Acceptance Model
The important issues to consider when applying traditional gap acceptance models in mixed traffic conditions are;. Gap acceptance cannot be treated as a separate step in modeling lateral movement behavior for drivers in mixed traffic, as they tend to perceive both speed increase and space gap simultaneously before making a lane change decision.
Parameters Used in Existing Microscopic Models
Parameters used in Existing Modified 1D Models (CF/Lane Changing)
The parameters mentioned in Table 2.3 describe the behavior of vehicle movement and vehicle lane changes in the traffic flow. Also, lane changing maneuvers of vehicles have been included by researchers in basic car-following models to describe the lateral behavior of vehicles.
Parameters Used in 2D Models Developed by different Authors
To model such a traffic condition, a detailed micro-level knowledge of the existing traffic flow and the behavior of vehicles during maneuvering is therefore necessary. Mathew et al., (2013), used very fine strips to identify the lateral movement of the vehicles.
Parameters required to Model Indian Mixed Traffic
In order to study the driving behavior in mixed traffic without a lane and to incorporate the above mentioned traffic behaviour, the following maneuvering/behaviour are important to study from field traffic. Therefore, longitudinal acceleration/deceleration, lateral acceleration and the impact of road characteristics are important parameters to study such traffic flow, and their accurate measurement and analysis at the micro level is necessary to understand the driver's behavior.
Studies on Vehicle Manoeuvre
Studies on Lateral Acceleration
The main source of information on which the driver decides his maneuvers is the lateral acceleration. The lateral acceleration as a measurement of lateral comfort is directly related to the horizontal alignment and lateral stability of a vehicle.
Studies on Longitudinal Acceleration
In some studies, acceleration/deceleration rates are modeled for different curved road segments (Shao et al., 2011; Altamira et al. used acceleration/deceleration data of mountain roads to analyze driver behavior.
Methodologies Used for Collecting Lateral and Longitudinal Acceleration Data
The lateral behavior of different vehicles is studied by analyzing the lateral acceleration of different types of vehicles on such roads. The ratio of lateral and longitudinal acceleration is studied for different types of vehicles.
Studies on Vehicles’ Lateral Placement
- Effect of lateral placement due to edge or centre line of the road
- Effect of Lateral Placement due to Traffic and Stream Parameters
- Methodologies Used for collecting Lateral position data of vehicles
- Summary
Impact of traffic and watercourse parameters on lateral positioning of vehicles 2.7.1 Effect of lateral positioning due to roadside or center line. Armor (1985) observed that the curb line had no effect on the lateral location of the vehicles from an experimental study.
Concluding Remarks
But due to laneless discipline and mixed traffic condition, lateral movement of vehicles is high in India. No studies have been found regarding the lateral position behavior of vehicles in disciplined traffic without (or weak) lanes.
Introduction
In this chapter, the method of data collection to achieve the research objectives is described in three subsections. a) Investigation of vehicle maneuvering using vehicle dynamic parameters. To achieve a higher accuracy in lateral gap data, the lateral gap maintenance behavior of different vehicles is measured from the median using a sensor-based assembly with a video camera.
Study of Vehicular Manoeuvre Using Dynamic Parameters
Methodology for Data Collection and Extraction
However, some of the authors used the vehicle-fixed GPS-based instrument to collect vehicle trajectory data or vehicle dynamic parameters from highly accurate vehicle position data. The accuracy of the VBOX data recording is tested by the VBOX automotive manufacturer using the Doppler Shift method in the GPS carrier signal.
Details of Sites Selected for Data Collection
The details of the data collection locations (data collection route details) and the different vehicle types used for data collection are shown in Table 3.3. The length of the entire road section used for data collection across the country was approximately 255 km for motorized 3Ws, SUVs and sedan cars, 215 km for H-back cars and 175 km for motorized 2Ws.
Study of Vehicles Lateral Placement on Entire Road Cross-section
Methodology for Data Collection and Extraction
Most of the studies used the image-based technologies or video recording technique to measure the lateral placement. The video is collected in such a way that the four corners of the vehicle and both the edge of the road boundary can be marked.
Details of Sites Selected for Data Collection
The software gives real coordinates and calculates parameters such as vehicle size, inter vehicle gap, center line separation between the vehicles, spacing between the road edge and the vehicle, the instantaneous speed of the vehicle along with the user classified vehicles.
Lateral Gap Maintaining Behaviour of Vehicles from Road Edge (Median)
Methodology for Data Collection and Extraction
Based on the travel time of the transmitted and reflected wave, the lateral distance of the vehicle from the sensor is calculated. Thus, microseconds divided by 58 (=29X2) gives the distance of the object from the sensor in cm.
Details of Sites Selected for Data Collection
Concluding Remarks
Introduction
Study of Longitudinal Acceleration/Deceleration (A/D) Behaviour of Vehicles
- Variation of Longitudinal A/D Behaviour with Different Road Width
- Variation of Longitudinal A/D for Different Type of Vehicles
- Variation of Longitudinal A/D for Different Cities
- Study of Longitudinal A/D for Combined Data from all Cities
The data without significant difference between different cities are combined to see the general behavior of different vehicle types. 𝐴 𝑜𝑟 𝐷 = 𝑚(𝑣) + 𝑛……….4-4 where A or D=acceleration or deceleration; v = vehicle speed; m and n=regression coefficients Table 4.7 details the coefficients of the A/D models and the statistical evaluation results (ie the r-squared value and Root Mean Squared Error values) for the models describing the A/D behavior respectively.
Study of lateral Acceleration
- Variation of Lateral Acceleration Behaviour with Different Road Width
- Variation of Lateral Acceleration for Different Type of Vehicles
- Variation of Lateral Acceleration for Different Cities
- Study of Lateral Acceleration with Combined Data of Different Cities
Therefore, the lateral acceleration of 2-lane and 3-lane roads is investigated separately for each type of vehicle for further analysis. Therefore, the lateral acceleration of all five types of vehicles on both 4-lane and 6-lane roads (2-lane and 3-lane roads) are significantly different at the 5% significance level.
Concluding remarks
These A/D values and lateral acceleration can be used in models to predict operating speed and also to set thresholds in the case of simulation models. The lateral acceleration model can be used to estimate the lateral weaving of vehicles in simulated trajectories for such traffic.
Introduction
In the following sections, the longitudinal and lateral acceleration for all five types of vehicles are studied based on the combined data obtained from the five different cities in India. To see the interaction between lateral and longitudinal acceleration more clearly, the relationship between the two dynamic parameters has been determined for all five vehicle types.
Interaction between the Vehicle Dynamic Parameters
The decreasing trend at higher speeds indicates that the drivers experience a greater risk with speeds over a certain limit which produces a reduction in the acceleration range (lateral and longitudinal). The average lateral maneuver for different acceleration or deceleration values of different vehicle types is presented in the following Figure 5-4.
Driving Behaviour Analysis using g-g Diagram
Representation of g-g Diagram in three-Dimensions (3D)
The g-g diagram can be seen in more detail by plotting the lateral and longitudinal acceleration in three-dimensional plots for all five vehicle types. The resulting graphs show the lateral and longitudinal acceleration for different velocity groups plotted along the z-axis (Figure 5-7).
Acceleration Limit Curve
The resulting acceleration limit (𝑎̅ = 𝜇. 𝑔) is calculated for various vehicle design speeds based on the given coefficient of friction (depicted below in Figure 5-9). In the case of Eboli et al. 2016) it is observed that the higher lateral maneuver is observed at higher speeds (40-60 km/h).
Spatial Representation of Acceleration Limit Curve
However, the urban acceleration shown in Figure 5.10 indicates that most data points are within the safety range for all types of cars (less than 1% of the data is outside the safety limit curve), with the exception of motorized 2Ws and 3Ws. The defined safety domain is helpful in identifying those points/sections of the roadway where the majority of motorists exhibit unsafe behavior (risk appetite).
Joint Distribution Modelling of Vehicle Dynamic Parameters Using Copula
- Concept of Copula
- Field Data and Relationship between Different Microscopic Variables
- Dependence of Microscopic Variables
- Modelling joint dependence structure using copulas
- Conditional Probability distribution using Copula
Given the domain of Kendall's 𝜏 and the estimated correlation values of the observed data (Table 5-4), some copulas (Gumbel, Clayton and Joe) could not account for the negative association between speed and lateral acceleration, speed and longitudinal /D). The estimated parameters (ρ) of the trivariate Gaussian copula and their log-likelihood values are presented below in Table 5-7.
Concluding Remarks
From this observation it can be concluded that at higher speed conditions the probability of higher lateral and longitudinal maneuvering decreases for all types of vehicles. The driving behavior (acceptance of safe or unsafe driving) is analyzed by considering dynamic parameters such as operating speed, lateral and longitudinal accelerations of the vehicles on straight roads of ordinary terrain.
Lateral Placement of Vehicles over Roads of Different Width
Introduction
Study of lateral gap-maintenance behavior of the vehicles from the roadside (median) according to their geometric variability, based on the data collected from the sensor-based assembly.
Study of Lateral Placement of Vehicles on Cross-section of Roads
- Stream Details and Initial Data Processing for Further Analysis
- Vehicle Type wise Variation in Lateral Placement on Different Roads
- Variation in Lateral Placement of Different Vehicle Types on Different Roads
- Operating speed prediction model using Lateral Placement of Vehicles on 4-
- Operating speed prediction model using Lateral Placement of Vehicles on 6-
- Operating speed prediction model using Lateral Placement of Vehicles on 8-
The lateral distribution of different types of vehicles on the entire road width of 4-lane divided roads for each location (location-1, 2 and 3) is obtained. Figure 6-8 shows the relationship between lateral alignment and speed of different types of vehicles on 4-lane roads.
Lateral Gap Maintaining Behaviour of Vehicles from Road Median
- Model Development for Minimum Lateral Gap maintained by Cars and Trucks
Similarly, the height impact is obtained and displayed by maintaining a fixed median width of 2.5 m (Figure 6-22b & 6-22d). At a median width of 2.5 m, the plantation height was smaller compared to the median width of 4.5 m.
Concluding Remarks
It is observed that as the height of the median decreases, cars move away from the median. In this part, the minimum lateral gap maintained by vehicles from the median is studied based on the type of median.
Importance of Microscopic Parameters in Simulating No-Lane Based Heterogeneous
- Introduction
- Calibration of VISSIM Parameters
- Validation of VISSIM with Field Data
- Study of other Microscopic Parameters Obtained from VISSIM
- Study of Lateral Speed and Lateral Acceleration
- Study of Longitudinal Acceleration/Deceleration (A/D)
- Study of Lateral Placement of Vehicles on Entire Road Cross-section
- Study of Minimum Lateral Gap maintained by Vehicles from Median
- Concluding Remarks
- Introduction
- Summary of the Research Work
- Study of Dynamic Parameters of Vehicles
- Study of Lateral Placement of Vehicles across the Road section
- Impact of Median Type on Lateral Gap of Vehicles from Median
- Importance of Microscopic Parameters in Simulating Non-Lane Based
- Contribution of this Research
- Future Scope
- Summary on the modification made in basic CF models for mixed traffic conditions23
- Different traffic stream criteria implemented by different authors to represent no-lane
- Acceleration/Deceleration ranges observed by different authors
- Details of the three types of cars available in India (Gear Heads, 2011)
- Accuracy level of different parameters measured by the VBOX Data logger
- Site Details & Vehicle Types
- Details of the selected road sections to study the lateral placement of vehicles
- Site Details with the Median type and their details are shown in the following table
- Site details for the Lateral Gap analysis
The lateral distribution of different types of vehicles over the entire road section is obtained from position data obtained from VISSIM. From Figure 7-5, it can be seen that the lateral placement of vehicles obtained from VISSIM extends to the entire transverse profile of the road.
