A Forecasting Model for Parking Guidance System

This paper researches prediction problems of Parking Occupancy Rate (POR) for city Parking Guidance System (PGS). First, giving the POR model based on BP Neural Network. Second, discuss the structure of BP Neural Network. At last, make a compare for different BP methods. Experimental results show that the POR predicting BP networkpsilas topology structure is 5-5-1, the generalization capacity is improved highly by combining self-adaptable studying rate and momentum.