Research And Application Of Intelligent Dispatching Platform Based On Big Data Of Shared Bicycle

As the concept of sharing is deeply rooted in the hearts of the people,the sharing economy has developed rapidly in my country,which not only improves people’s quality of life,but also improves the overall utilization of urban resources.As the pacesetter of the sharing economy,shared bicycles have now become an important means of transportation for residents.At present,the resource allocation of shared bicycles in my country’s cities is relatively stable,but there are still some operational problems,such as traffic congestion and waste of public resources caused by the accumulation and parking of shared bicycles.These problems cannot be solved by bicycle-sharing companies alone,and require the government to carry out global scheduling to solve them.However,due to the different economic strength and traffic conditions of each city,how to intelligently dispatch the shared bicycle resources of different enterprises in different regions requires big data support.The government integrates the operation data of all shared bicycle enterprises to start from the whole,and conduct overall and reasonable scheduling,so as to improve the scientific management and efficient use of shared bicycle resources,and promote the sustainable and healthy development of the shared bicycle industry.For urban shared bicycle resource scheduling,it is necessary to first predict the dispatching site and dispatch amount.Therefore,this thesis analyzes the seven main characteristics that affect the demand for shared bicycles from the three dimensions of time,climate and brand,and uses the Pearson correlation coefficient to improve the random forest algorithm.In the process of random feature selection,the improved random forest algorithm is finally compared with the original random forest algorithm and BP neural network,which proves that the improved random forest algorithm has superiority in predicting the demand for shared bicycles.Then,this thesis analyzes the problems and reasons of multi-enterprise scheduling,and establishes a shared bicycle multi-enterprise scheduling model based on government control to solve the global scheduling problem.Since the scheduling path solving problem belongs to NP-Hard problem,this thesis adopts the hybrid ant colony algorithm to solve the problem of global scheduling.model solution.The hybrid ant colony algorithm utilizes the genetic algorithm to generate the initial pheromone distribution,thereby enhancing the global search and iterative ability of the ant colony algorithm.After that,the hybrid ant colony algorithm to solve the model is compared with the improved ant colony algorithm of other scholars.It is found that the hybrid ant colony algorithm in this thesis has the lowest cost and high solution efficiency.Finally,according to the research content,this thesis designs and implements a shared bicycle intelligent dispatching platform based on the use of government control departments,which is used to practically solve the problem of unbalanced utilization of shared bicycle resources in the entire city due to the lack of data communication between enterprises.The use of this platform by the government for overall control can not only ensure the interests of enterprises,but also better manage the urban shared bicycle resources,improve the green travel level of urban residents,and help the government to promote the allocation and management of shared resources,thereby promoting the construction of beautiful and smart cities.