Research On Reasonable Scale Of Expressway Service Area Based On Consumption Behavior Analysis

As an important “post station” and “window” on the expressway,the service area can meet the various enroute needs of expressway travelers,and it is the guarantee for safe,comfortable and convenient travel of the expressway,and an important window to reflect the quality and level of expressway services.However,there are still some problems in the development and construction of the service area in our country,for example,the size of service area does not adapt to the total demand of service area,and the type of service and the quality of service demand are not suitable.The size of the scale of the service area is closely related to the service quality,operational efficiency,resource utilization rate,traveler service satisfaction and loyalty of the service area.Therefore,in the new era of transformation and upgrading of service areas and diversification of consumer demand,it is of great value to carry out research on the reasonable size of expressway service areas based on consumer behavior analysis.Most of the existing research on the scale of the service area starts from the physical facilities and the pause rate of the service area,the parking area of the service area is calculated by the pause rate,and then qualitatively analyzes the relationship between the parking area and the area of the service area,the value of the parameters is mostly empirical,there is less research on the quantitative relationship between the reasonable size of the service area and its influencing factors,lack of quantitative model of the objective law of response;There are few studies on the consumption behavior of service areas,the combination of consumption behavior and the reasonable scale of service areas has not been studied,and the influence of consumption behavior on the scale of service areas has not been fully paid attention to,this paper takes the expressway service area as the object to carry out the following research:1.Defining the consumption behavior of the service area based on the relevant theory of consumer behavior,from the service area traveler role,travel mode,passengers' activity rules,consumption demand,consumption level,satisfaction,loyalty,etc.,the qualitative and quantitative methods are used to analyze the consumption characteristics of travellers under different angles of view,including the consumption behavior characteristics of different high-speed line service areas and different travellers(drivers,passengers,passenger cars,trucks),and the existing problems in the service area and the traveler's consumption preferences in the service area are summarized.2.On the basis of systemically defining the connotation of reasonable scale of the expressway service area,and referring to the analysis of the influencing factors of the existing service area scale,the traveler's consumption behavior is included as a key factor affecting the reasonable scale of the service area,and the main indicators affecting the reasonable scale of the service area are determined.The parameters affecting the reasonable scale of the service area are obtained,and the influence relationship between each parameter and the service area size is comprehensively analyzed,and the influence process of each parameter on the reasonable scale of the service area is clarified.3.Based on the deep learning theory,a reasonable scale measurement model of the service area based on MLP neural network is constructed,and the model is validated by the questionnaire data of the expressway service area of Guizhou Province.The reasonable scale value of the service area of the test sample is obtained by training,and the model prediction effect evaluation index shows that The absolute relative error,the maximum absolute relative error,and the average absolute relative error are all within 5%,and the equal coefficient is 0.9845,which is greater than 0.9,indicating that the reasonable scale prediction value has a higher degree of fit to the expected value,and the model prediction result is ideal.