Sociological and Behavior Models for Connected and Autonomous Vehicles
Sociological and Behavior Models for Connected and Autonomous Vehicles
Sociological and Behavior Models for Connected and Autonomous Vehicles
Sociological and Behavior Models for Connected and Autonomous Vehicles
Organization:
Arizona State University
Organization:
Arizona State University
Organization:
Arizona State University
Duration:
Jun 2021 – Sep 2021
Duration:
Jun 2021 – Sep 2021
Duration:
Jun 2021 – Sep 2021
Overview
Investigated a structure that, through integrating computational graph with three discrete choice models (multinomial logit model, nested logit model, and integrated choice and latent variable model), is capable of dealing with complex patterns and large-scale travel datasets with statistics-oriented features. Conducted a self-designed experiment whose performance assessment function is calibrated using automatic differentiation
Integrated long short time memory (LSTM) model with a linear regression model for demand prediction to interpret significant factors of the travel demand and capture complex patterns in large datasets and improve forecasting performance
Conducted a case study using weather and holiday information, NYC taxi dataset, and Limousine Commission dataset • Reviewed the car-following models and traffic flow fundamental diagrams; explored undiscovered model features as well as physical meanings of model parameters and mathematical forms
Investigated connections between the microscopic car-following model and the macroscopic fundamental diagram; proposed a new potential model that could combine human-driven vehicles and connected autonomous vehicle
Overview
Investigated a structure that, through integrating computational graph with three discrete choice models (multinomial logit model, nested logit model, and integrated choice and latent variable model), is capable of dealing with complex patterns and large-scale travel datasets with statistics-oriented features. Conducted a self-designed experiment whose performance assessment function is calibrated using automatic differentiation
Integrated long short time memory (LSTM) model with a linear regression model for demand prediction to interpret significant factors of the travel demand and capture complex patterns in large datasets and improve forecasting performance
Conducted a case study using weather and holiday information, NYC taxi dataset, and Limousine Commission dataset • Reviewed the car-following models and traffic flow fundamental diagrams; explored undiscovered model features as well as physical meanings of model parameters and mathematical forms
Investigated connections between the microscopic car-following model and the macroscopic fundamental diagram; proposed a new potential model that could combine human-driven vehicles and connected autonomous vehicle
Overview
Investigated a structure that, through integrating computational graph with three discrete choice models (multinomial logit model, nested logit model, and integrated choice and latent variable model), is capable of dealing with complex patterns and large-scale travel datasets with statistics-oriented features. Conducted a self-designed experiment whose performance assessment function is calibrated using automatic differentiation
Integrated long short time memory (LSTM) model with a linear regression model for demand prediction to interpret significant factors of the travel demand and capture complex patterns in large datasets and improve forecasting performance
Conducted a case study using weather and holiday information, NYC taxi dataset, and Limousine Commission dataset • Reviewed the car-following models and traffic flow fundamental diagrams; explored undiscovered model features as well as physical meanings of model parameters and mathematical forms
Investigated connections between the microscopic car-following model and the macroscopic fundamental diagram; proposed a new potential model that could combine human-driven vehicles and connected autonomous vehicle
Overview
Investigated a structure that, through integrating computational graph with three discrete choice models (multinomial logit model, nested logit model, and integrated choice and latent variable model), is capable of dealing with complex patterns and large-scale travel datasets with statistics-oriented features. Conducted a self-designed experiment whose performance assessment function is calibrated using automatic differentiation
Integrated long short time memory (LSTM) model with a linear regression model for demand prediction to interpret significant factors of the travel demand and capture complex patterns in large datasets and improve forecasting performance
Conducted a case study using weather and holiday information, NYC taxi dataset, and Limousine Commission dataset • Reviewed the car-following models and traffic flow fundamental diagrams; explored undiscovered model features as well as physical meanings of model parameters and mathematical forms
Investigated connections between the microscopic car-following model and the macroscopic fundamental diagram; proposed a new potential model that could combine human-driven vehicles and connected autonomous vehicle
