Research On The Driver Shift Quality Based On Fuel Economy

Automobile fuel economy is one of the most important indexes of vehicle performance, especially with the rapid development of automobile industry and the shortage of oil resources, the significance of reducing vehicle fuel consumption is particularly important. Driver plays an important role in the driver - vehicle - road system, as the driver is not only a road traffic system information processing and decision-makers, but also regulates and controls it. Therefore, academicians have devoted much into the field of driving behavior characteristics study.The driver's behavior of shifting process is an important branch. Accelerate shifting is a common shifting behavior. The operation of the driver in the process of accelerating shifting can be used to determine the quality of shifting and also the operation proficiency of the driver. Because of this, the research of process of shifting driving behavior particular the shifting quality in the field of driver - vehicle - road system has a significant and far-reaching significance.The process of change and vehicles running trend in accelerated shifting is a certain complexity, and the operation of the driver behavior in the process of the driver accelerating shifting is difficult to quantify by the authority's standard, with fuzzy and uncertainty. Therefore, it's difficult to estimate and modeling particularity, impersonality, accurately on driver's shifting behavior characteristics by the traditional study methods.Through the analysis the theory and research methods of driving behavior and vehicle fuel economy test, this paper illustrates the basic theory of driver shifting quality based on fuel economy and gives its evaluation indexes. In order to researching the relationship between driving shifting behavior and fuel consumption characteristics, we do a lot of vehicle road test, and build the BP neural network evaluation model of driver shifting quality based fuel economy. Finally, compared with the theory has been for driver shifting quality evaluation, it proves that the BP neural network evaluation model is able to evaluate the quality of the driver shift and the fuel economy better. This will not only benefit the road transport industry to save transportation costs, but also help to regulate the operation of driver behavior and the classification of them.Through the analysis and illustrate the fuel economy theory, the theoretical study of driver behavior as well as the BP neural network theory, research was carried out by doing as follows:(1) In this paper, we analysis the status and trends of driver behavior and fuel economy at home and abroad and find it's necessary for researching the driver shifting behavior. According to the existing equipment and basic theory we determine the overall outlook of this paper. (2) And then this paper describes the fuel economy theory of shift quality and the basic methods of fuel economy test and determines the evaluation index of driver shifting quality based on fuel economy.(3) Based on previous experience, test method in the regular driving behavior characteristic study and fuel economy test methods, vehicle road test was designed and car driving test is conducted in a free up to 80 kilometers per hour the typical test cases. And we record the acts of different driver's shift operation.(4) By selecting the appropriate membership function, the network structure and algorithm,we establish the BP neural network evaluation model of driver shifting quality based fuel economy. The training through the BP neural network conclusion that the relationship between driver shifting quality and fuel consumption. The theory to evaluate the quality of the shifting quality has been on the evaluation as a tested driver's subjective evaluation, the establishment of the driver shifting of BP neural network model of the evaluation as a tested driver's objective evaluation. Comparative with these test results, it gets better test results.The research in this paper will prompt the development of optimization study on performance of driver in the traffic system. As the complicacy of the driver behavior and the variety of the vehicle, the BP neural network evaluation model of driver shifting quality based fuel economy found in the paper need to be perfected. For example, abound the training sample data-base and the vehicle model data-base, and make up the interface plat-form during the different modules.