Parking Occupancy Prediction Method Based on Multi Factors and Stacked GRU-LSTM

• Published in: IEEE access Vol. 10; pp. 47361 - 47370
• Main Authors: Zeng, Chao, Ma, Changxi, Wang, Ke, Cui, Zihao
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 2022; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Data models; Deep learning; Driving conditions; Feature extraction; Meteorology; Motor vehicles; multi factors; Neural networks; Occupancy; Parking; parking occupancy; Prediction models; Predictive models; Resource utilization; stacked GRU-LSTM; Time series analysis; time series prediction; Traffic; Urbanization; Weather
• ISSN: 2169-3536; 2169-3536
• DOI: 10.1109/ACCESS.2022.3171330

With the development of society and the continuous advancement of urbanization, motor vehicles have increased rapidly, which exacerbates the imbalance between parking supply and demand. Therefore, it is very important to excavate knowledge from historical parking data and forecast the parking volume in different time periods so as to optimize parking resource utilization and improve traffic conditions. This paper proposes a new hybrid model that stacks gated recurrent unit (GRU) and long-short term memory (LSTM). The proposed stacked GRU-LSTM model combines LSTM's advantage in prediction accuracy and GRU's advantage in prediction efficiency, and uses multi factors, including occupancy, weather conditions and holiday, as input to predict parking availability. When compared against other predictive models such as stacked simple RNN, stacked LSTM-RNN, and stacked LSTM-Bi-LSTM, our experimental results indicate that the stacked GRU-LSTM model has better performance for parking occupancy prediction as it not only improves prediction accuracy, but also reduces prediction time.