| With the rapid increase of automobiles in china, more and more families use cars for their trips. And fuel as the main"food"for cars, which price has been a significant growth in mid-2008. Moreover, the discussion and implementation of fuel tax under the high fuel price, making the attention and argument for fuel reached an unprecedented peak.Fuel cost as one part of the most important vehicle travel costs, its price changes will directly affect the use of vehicles, similar to parking fees and congestion fees. Both are mainly to raise the cost of car travel. The impacts of fuel cost on car use include use frequency and vehicle mileage. Therefore, the main purpose of this thesis is searching the influence on the self-driving tour with the fuel price changes.This study based on the behavior theories according to many related research at home and abroad. Get the information about self-driving tour distance, frequency and traveler number through the questionnaire for persons which own car (cars). The questionnaire use stated preference(SP) survey method, also get (1)one whether choose self-driving tour or not in certain condition (it means the same distance and traveler number) (2) the self-driving tour frequency under the given furl price. According to the SP data, establish two models. One is the tour mode choice model to explain why people having cars choice self-driving tour or not, using the Binary Logit (BL) model. The other is the frequency choice model to find out what kinds of people choose which frequency, using the Ordered Logit model.Through BL model, we found that the main factors which effect people choice self-driving tour or not under certain conditions excluded travel distance and the traveler number, so we can consider that the distance and traveler number have little impact on their tour mode. While the impact produced by family property and personal property is more obvious. So the subsequent study is to find the factors impact the self-driving tour frequency through Ordered Logit model.After self-driving tour frequency choice model (Ordered Logit model), getting the main factors contain personal annual income, family size and number of family cars, vehicles tenure, the general self-driving tour distance, the way of vehicle use fees payment, generally self-driving tour cycle and fuel prices. Through the Ordered Logit model analysis, found the each factor on the role of frequency choice. For self-driving tour frequency select model, also adopted Multinomial Logit (MNL) model to estimate parameters. Judging from the model accuracy, the Ordered Logit model is more appropriate for the ordered dependent variable than the MNL model. And take the oil prices and personal annual income for self-driving tour frequency sensitivity analysis. It shows that with the rising oil prices, rapid growth of low-frequency of self-driving tour. And with the increase of annual income, the proportion of low-frequency decreases rapidly, the mid-frequency is relatively stable and slowly increase of high frequency. Last, give an example of the model application. |
