Research And Design Of Cloud Portal System For Intelligent Stereo Garage

With the rapid growth of car ownership,“parking difficulty” has affected the development of urban traffic planning and become a common problem.Developing a stereo garage with large storage space and small footprint can effectively help solve this problem.However,at present,functions such as remote parking reservation,remote management of the garage can not be realized because most of the stereo garages are not connected to the Internet,which affects the intelligent level of the stereo garage.And some garages have low access efficiency due to the disordered use of access strategies.In view of this situation,with the support of the provincial key research and development plan project,this paper studies several algorithms such as the prediction algorithm of the number of free parking spaces,garage recommendation algorithm,and implements a cloud portal system for intelligent stereo garage.Firstly,after analyzing the existing problems of the stereo garage,according to the development needs of the system,the overall architecture and functional module structure of the system are planned,the key technical schemes such as garage recommendation strategy,prediction method for the number of free parking spaces are analyzed,and the system frame based on cloud platform with high scalability,low cost and high availability is designed.Secondly,based on the prediction algorithm of the number of free parking spaces,this paper studies the garage recommendation algorithm,which provides algorithm support for the realization of system functions.On the issue of prediction algorithm,the LSTM network is improved based on Ensemble Learning,and the prediction model is established.Compared with the prediction model based on LSTM / BP network / RF,the superiority of the improved model is verified.Based on the prediction algorithm,the attribute value of the predicted number of free parking spaces is obtained,and the garage recommendation problem is further researched.The candidate garage selection model based on geohash and the garage selection model based on multi-attribute decision are established.Thirdly,this paper studies the access strategy of lift-sliding stereo garage.Based on the principle of the shortest storage time and the lowest center of gravity,the parking space allocation strategy is formulated.On the basis of determining the strategy of parking space allocation,this paper further researches the strategy of garage access based on queuing theory.Each access strategy is modeled and simulated in different periods of time.Through comparative analysis,the access strategy of the garage is determined to further reduce the operating cost and improve the service performance of the garage.Finally,the cloud portal system for intelligent stereo garage is designed and implemented.The system client includes Android App,IOS App,We Chat applet and browser client.And it realizes garage recommendation,parking space reservation,historical order management and other functions.After joint debugging and testing,the results show that the system can meet the relevant requirements.The cloud portal system for intelligent stereo garage developed in this project improves the intelligence and network level of the stereo garage,and helps to improve the service efficiency of the stereo garage and the traffic conditions of the city.