F2010A043
Development of Ecological Driving System Using Model Based Prediction of Traffics and Information of Road Shapes
This paper presents a novel development of ecological driving system using model predictive control by addressing driving styles. Anticipation of traffic situations and information of road shapes such as slopes and curvatures are used for optimizing the vehicle control actions. The proposed system can be realized in near future through the advancement of new sensing and information technologies in intelligent transportation systems. Ecological driving is a way of driving intended to enable energy conservation, cut air pollution by reducing emissions of carbon dioxide, and contribute to the effort to reduce global warming. Besides many physical factors, driving styles have a great influence on vehicle emissions and energy consumption. Proper driving styles may improve the travel economy or driving efficiency further. It is obvious that a skilled driver can drive more economically than an average or new driver. Generally, fuel economy is maximized when acceleration and braking are minimized. Therefore, a fuel-efficient or ecological strategy is to anticipate what is happening ahead, and drive in such a way so as to minimize acceleration and braking, cruising at the optimal speed.
The existing ecological driving approaches provide very general suggestions, for example, not to accelerate aggressively when the system detects abrupt acceleration. None of them gives the driver proper advice specific to the current situation and also anticipated situations from the current situation and based on rigorous reasoning in the form of quantitative values. Preliminary development of ecological driving system revealed significant improvement in fuel economy on a flat and straight road on a single lane. Vehicle control strategies due to variations in road conditions, such as flat-road, road with up-down slopes and curvatures, have not been studied so far. In this paper, the model predictive control approach of ecological driving system is further enhanced by incorporating the road shapes.
The proposed system senses status of the host vehicle, its preceding vehicle, the traffic signal ahead, and the road slopes and curvatures, which are used to constitute the model of the vehicle-traffic non-linear system. At any instant, using the model, the system generates a course of actions required in the predicted horizon for long run optimum economy. The performance index includes terms related to both fuel consumptions and safety factors. Once a sequence of optimum actions is generated, only the instant action is executed to drive the vehicle, and the whole process is repeated at each sampling interval throughout the driving.
The proposed ecological driving system is tested on AIMSUN, a microscopic traffic simulator, through an extension developed by application program interface. The performance of ecological vehicle is compared with that of Gipps model based vehicles in AIMSUN. The Fuel saving characteristics are illustrated graphically and analyzed. The ecological driving system generates appropriate control inputs and avoids unnecessary braking and acceleration in various driving situations to minimize the fuel consumption per unit distance travelled. Comparative results reveal the significant fuel savings in long run travel as a result of proper anticipation of traffics and road shapes ahead.
This abstract is supplemented by a PDF, which can be viewed here.
Session: Intelligent Transportation Systems