Multi Communities Parking Space Scheduling System Based On Alibaba Cloud

In a city the number of fixed-point parking spaces is in a dynamic process of changing with time and crowd behavior.A single fixed mode parking space allocation method can not meet the parking needs of different groups.In this paper,under the background of the rapid development of cloud computing platform,cloud computing and traditional parking operations are combined to carry out research work.Around the communities parking problem,this paper has done the following work:1.Apply Alibaba Cloud Auto Scaling service to parking space flow calculation,introduce the multi scene learning model of LSTM—AE to predict the parking space according to the parking space information and the prediction of parking flow.so as to provide the basis for parking space allocation.2.Introduce a kind of intelligent parking scheduling scheme based on user’s expectation linear optimal allocation strategy(LOAS),and establishes a multi-agent coordinated parking management model under the cloud.To solve the sudden and preemptive problems in the process of parking.In this paper,we use Auto Scaling service to optimize and improve the parking space prediction and parking scheduling.Solve the problems of parking space status update,peak valley statistics and processing of classification and storage in the parking space scheduling system.Make sure the system well adapt to the uncertain interference which caused by unexpected events in the parking process.In the aspect of path planning,the classical ant colony algorithm is studied and introduced.Through the research and improvement of the classical ant colony algorithm,the more scientific path optimization is realized in the path planning algorithm.