Research On The Intelligent Distribution System Cloud Computing Architecture Platform

With the development of modern industrial production,the scale of intelligent storage is more and more large.Intelligent warehousing and distribution is a very complicated process.The traditional Intelligent distribution system solutions can not flexibly meet the expanding requirements of production scale.Therefore,how to design a solution to meet the needs of such large-scale distribution tasks has become particularly needed.The traditional non-cloud distribution system has a good performance when the distribution scale is small.But with the storage scale becoming larger and larger,the problems of poor system scalability,insufficient capacity of the algorithm strategy,low efficiency and so on begin to emerge.This thesis focuses on the large-scale warehousing scenario to design a smart warehousing and distribution system which is based on cloud platform.And on the basis of third-party cloud resources,this thesis implements the system prototype.At the same time,this thesis selects the key issue of distribution task---the path planning of distribution vehicle.The simulation experiments are carried out to verify the engineering feasibility and algorithm validity of the intelligent distribution system which is designed on the basis of cloud platform and deep reinforcement learning in large-scale storage scenarios.Specifically,this thesis carries on the research analysis according to the following three logical levels:Firstly,the author studies the traditional intelligent delivery technology,including the key issues and technical points of the smart delivery vehicle and the intelligent warehouse delivery system,and probes into the shortcomings and causes of the problems.Then the author studies the core of cloud computing and reinforcement learning and focuses on the technology of the application of intelligent distribution system.Secondly,according to the advanced cloud computing architecture platform,the demand of large-scale distribution system and the traditional distribution system technology,the author designs an intelligent distribution system with multi-delivery vehicle which is based on cloud platform.It is worth mentioning that,this program focuses on advanced architecture design.This program can not only meet the needs of intelligent distribution,but also be as part of the entire smart factory cloud module.It seamlessly docks other cloud services and breaks the situation of data barrier in traditional architecture.Thirdly,with the development of the scale of distribution system,the path planning task of the delivery vehicle becomes very difficult.The computational complexity which is based on the deterministic programming algorithm and the heuristic intelligent search algorithm is greater.We abandon such traditional algorithm programs and model intelligent delivery tasks on the basis of the powerful computing power of cloud platform,the respective advantages of deep learning and enhanced learning,and the depth-intensive algorithms.The final simulation experiment shows that our algorithmx model does very well on large-scale intelligent distribution tasks.