PARKING SPACE SCHEDULING ALGORITHM BASED ON TIME SERIES PREDICTION

In view of the problems of parking conflicts and low parking space utilization in existing parking methods, a parking space scheduling algorithm based on time series prediction is proposed. The algorithm used the Holt-Winters model to predict the parking flow of monthly rental vehicles in the next period. It used two parking scheduling strategies, namely monthly rental service priority strategy and dynamic allocation strategy. It realized the dynamic planning of the berth types of available parking berths under the premise of ensuring the parking service quality of monthly rental vehicles. Experimental results show that under the conditions of different numbers of parking spaces and monthly rented users, this method has a higher cumulative utilization rate than the existing parking methods, and can effectively improve the utilization rate of the office building parking lot.