Prediction Of Residents' Travel Behavior Based On GA-BP Neural Network

With the rapid development of the national economy,the problem of traffic congestion has become increasingly serious.The demand for urban resideots' travel has shown a diversified trend,and the choice of residents' travel has also increased.Travel behavior prediction is particularly important as an essential part of transportation planning.This paper proposes a prediction of residents' travel behavior based on GA-BP Neural network models and makes accurate predictions on travel times,mode and time.The main research points of this paper nainly include the following aspects:(1)Analyze the composition of the survey data of residents' travel and the characteristics of travel,and analyze the proportion of their common factors on the choice of travel ehoices from the three prediction models of travel mode,time and times,and obtain the general trend and regularity of residents*travel.(2)Combining the proportional distribution relationship between residents' travel behavior and common factors,and using the correlation analysis of SPSS software to determine the factors that are significantly related to the final stage of entering the travel behavior prediction model.(3)Taking GA-BP neural network as the core,establish the travel mode,times and time prediction model respectively,and eompare the prediction results with the actual choice of residents.The results show that the total accuracy of the above three models is 87.29%,82.59%and 81.78%,the results achieved the expected results.(4)Simulate the implementation of traffic demand management measures by changing the input data of the test set,and apply the predicted model of training to the limit line,bus priority policy and economic development,and predicting trends in residents' travel behavior.The research results verify that the model has certain Practical value.The research results of this paper are of great significance for the formulation of scientific and rational transportation planning and targeted traffic demand management policies.