Prediction of Vacant Parking Spaces in the City based on LSTM

With the continuous progress of modern society, more and more families have their own private cars for the convenience of travel and other needs, making it easier and faster to go anywhere. However, with the increase in the number of private cars, parking spaces in parking lots cannot keep up with the growing number of cars. It is increasingly difficult to find parking spaces everywhere when driving, and even if they are found, they are often full. The problems of chaotic management in parking lots have become more serious, greatly affecting people's pace of life. This study aims to predict the vehicle access situation of a parking lot in the future by constructing a prediction model. Based on the collected information of nearly 9,000 vehicles entering and leaving a parking lot, this paper uses the LSTM model to analyze the data by comparing the entry and exit times of different vehicles and whether the vehicles leave the parking lot, and obtain a highly fitted parameter model. The results show that the distribution state of the fitted value of the model is quite close to the real value. The model predicts the data of different time periods in the next 6 days, and obtains relatively accurate prediction results.