Unoccupied Parking Space Prediction of Chaotic Time Series

Parking Guidance System is an important component of the Intelligent Transportation System. Unoccupied parking space prediction is a crucial technology of the system. So far there have been no reliable prediction methods. Therefore, to predict the future unoccupied parking space of parking lot becomes an important research subject. Based on the chaotic time series forecasting methods of historical data, the paper developed a weighted one-rank local-region method to predict unoccupied parking space. The model was verified with observed data of a Parking Guidance System in Beijing, China. The results show that the prediction model is precise and model has a great practical value.