Analysis And Prediction Of Family Influencing Factors On The Number Of Household E-bikes

In recent years,e-bikes have become popular as a means of transportation that is easy to operate,economical,convenient and flexible,and have become one of the main modes of transportation for people to travel short distance.More and more families have begun to own one or more electric bicycles.While e-bikes bring convenience to people’s daily travel,they not only affect the urban traffic structure,but also cause many traffic management and road safety problems.In order to formulate more reasonable e-bikes management measures,it is important to study the influencing factors of household e-bike ownership and accurately predict its ownership.First of all,based on reading and sorting out relevant domestic and foreign documents,in view of the deficiencies of existing research,combined with the actual conditions of cities in China,it is proposed to explore the influencing factors of the number of e-bikes from the perspective of the family.Based on the 2017 resident trip sample survey data in Zhoukou City,Henan Province,the survey data is collected by households,and the information obtained is divided into four categories: demographic factors,income and expenditure factors,housing factors and vehicle factors,and analyzes the change trend of the difference of various family factors on the ownership of e-bikes.Secondly,after screening through variables,the family’s total population,monthly income,housing area,the number of private cars and bicycles are used as independent variables to establish a multinomial logistic regression model.According to the parameter calibration results of the model,the influence mechanism between various family factors and the number of e-bikes is quantitatively analyzed.The results show that the probability of a family owning e-bikes is positively correlated with the family’s total population,monthly income and housing area,and negatively correlated with the number of private cars and bicycles.Among them,the positive correlation degree of housing area is the smallest,and the negative correlation degree of private cars is greater than that of bicycles.Next,in view of the numerous influencing factors and complex relationships among household e-bikes,a prediction model of household e-bikes is constructed based on BP neural network algorithm.After many attempts,it is determined that the number of nodes in the hidden layer of the neural network is 30,Tanh function is selected as the excitation function,and Purelin function is selected as the output layer,and the prediction effect is good.In order to improve the convergence speed and prediction accuracy,the particle swarm optimization algorithm is used to optimize the initial weights and thresholds of the BP neural network,and the network training process is improved.The results show that the prediction accuracy of the optimized model has been greatly improved,and it can be applied to predict the number of household e-bikes under different family characteristics.Finally,according to the above-mentioned research content,combined with the family characteristics of different number of e-bike holdings,relevant suggestions and measures are put forward from the aspects of e-bike sales management,transportation planning and travel concept.