Research On Online Trajectory Optimizing Strategy Of Bus Engine Operating Condition With The Driving Style Improving

Improving the driving style based on driving assistance is one of the key on road fuel saving technologies for in-use buses,which is of a positive significance in reduce operations costs for buses.This paper researches the approach to lower fuel consumption of buses by improving the driving style based on driving assistance.In regard of this topic,the author firstly studies and puts forward the fuel-saving driving style characteristic parameters;then an evaluation model of driving style is built combined the parameters with the real-word vehicle operating data and the model pays a foundation for the optimization algorithm of driving assistance;finally,the driving assistance system is developed featuring such algorithm and it has been validated.The main idea is as follows:The platform for data collection and analysis is exploited by this research.With the platform more than 100,000 km vehicle operating data was collected.Besides,8 algorithms in 3 types concerning reconstruction of vehicle state,driver’s operation and environment information have been developed.The combination of raw data and reconstruction algorithm will fully restore the condition of the driver,the vehicle and the road.Based on a mechanism analysis of the pattern of the engine track influenced by driving style,the author lists 14 fuel-saving characteristic parameters related to fuel saving,including the average accelerator pedal depth in different gears,the positive change rate of the accelerator pedal and the distribution of shift points.Considering the vehicle operating data,the fuel-saving’s impacts on sensitivity of those parameters and the pattern of operating point distribution have been though the statistics process and analysis,from which the key features of driving style affecting the fuel saving are concluded under the influence of the combination of the driver,the vehicle and the road.The C4.5 algorithm of decision tree leads to the evaluation model of driving style.The model input consists of two types,i.e.,the environment and the characteristic parameters of driving style.Upon analyzing the input and the pattern of influence of the sample number on model accuracy,the author obtains results that shows that vehicle loading has the greatest impact on the accuracy compared with other environment factors;the predicted accuracy rate approximating confusion of the model is 96% via the combination of character parameters of 6 driving styles and samples from 2335 bus stations.An optimization algorithm of driver assistance has been put forward based on that model and orientation-distance indicators of the difference characterization of driving style.This algorithm is able to plan a most fuel-efficient and simplest improvement for drivers.As the optimization algorithm of driving assistance acts as the core,a fuel-saving driving assistance system with robustness and adaptability targeted is built,including human-machine interface,real-time data collection and operation planning,dynamic model building and optimization.The prediction accuracy of the evaluation model is the trigger condition of the system to plan a new path for improving the driving style.This mechanism could secure real-time recognition of changes in the combination of the driver,the vehicle and the road.Finally,a verification is conducted concerning the tips effectiveness and fuel saving results of the driver assistance system with the driving simulation system and simulation platform of GT-SUITE.The results demonstrate that the drivers’ execution rate for the tips is over 74.58%.After adopting the system,the speed has scarcely changed,but 6.25% of fuel consumption can be reduced.