As is known to all, the parking demand is a complex problem with many influence factories, such as the society development and individual habit, and the parking demand forecast is also a difficulty task. In this paper, the influence factories which influencing the curb parking in the center district of big city are taken as the input dates, the parking spaces are taken as the output dates to establish the prediction model based on artificial neural network, and the parking spaces of the programming year can be forecasted using it. This paper studies several facets as follows:Firstly, the parking demand characteristics of the big city are researched, and the factors relating to some parking forecasting methods are analyzed, then, the influence factories which being wanted can be chosen;Secondly, the influence factories that chosen above are taken as the input nerve cells, the curb parking demand is taken as the output nerve cells, the number of nerve cells for hidden layer is made certain through tests, and the network is trained by the reformative BP arithmetic, the Matlab computer language is used to make of the model, the data in Appendix are taken as the training sample, and then the prediction model based on artificial neural network is established.Finally, the central district of Changsha city is taken as a case, the curb parking demands are forecasted using the trained artificial neural network prediction model, and the forecasting results are gained and analyzed.
|