Design And Implementation Of Car Rental Platform Based On Microservice

With traffic congestion and environmental pollution becoming increasingly bixuserious,new energy vehicles have started to emerge in the field of time-sharing rental by integrating the concept of "green sharing".At present,there are not only EvCard,GoFun and other leading companies in this field,but also many small and medium-sized emerging time-sharing companies.How to help these companies quickly develop and iterate a rental system to meet the needs of users is an urgent issue to be solved.In order to improve the company's operational efficiency,this thesis focus on the one-stop car time-sharing leasing business.Compared with the traditional car rental platform,which uses a monolithic architecture with single function and poor scalability,this thesis adopts microservice architecture and adheres to the design concept of low coupling and high cohesion.The front-end adopts Wechat Mini Program and Web framework for development,and the back-end uses Spring Cloud framework for design,thus the system has high availability and strong scalability.At the same time,the system is deployed based on container technology,integrating development,testing and operation and maintenance,and realizing an enterprise-level online car rental platform.The main work of this thesis is as follows.(1)After a comprehensive analysis of the business needs and non-functional needs of the car rental platform,this thesis divides six sub-modules into user service,order service,payment service,coupon service,vehicle service and branch service by integrating the overlapping services of the client side and the operation management platform and combining with the Spring Cloud framework and related components.Meanwhile,the system architecture of the car rental platform and the E-R relationship of the database are given.(2)Aiming at the problems of unbalanced demand among time-sharing business outlets,which leads to users' difficulty in using cars and the loss of part of the company's profits,this thesis proposes a demand forecasting algorithm based on a combination model of Prophet and long short term memory network and solves its weight coefficients by particle swarm optimization.Through simulation experiments and comparison with single model,the results show that the improved algorithm has higher prediction accuracy and fits the historical operating data of the platform better.In the next system optimization work,it will be deployed as a service module to the platform to provide support for the operating company and maximize corporate revenue.(3)In this thesis,the service modules of the rental platform are realized and containerized deployment by using Kubernetes.Detailed steps,classes and interface methods are given.Finally,set up the test environment.Through a comprehensive business function test and non-functional test on the rental platform,the results show that the rental platform has been reached the level of online and has high availability.