Design And Key Technology Research Of Intelligent Charging System For Electric Bicycle

With the development of the Internet,electric bicycle charging pile industry is transforming towards the direction of wisdom and interconnection.There are many networked charging piles on the market,providing a variety of services through equipment networking.However,there are some problems in the existing networked charging system: when providing the proximity charging pile inquiry service,the method most of these systems token are based on the distance between the user and the charging pile,the efficiency is low in the case of a large amount of data.In the meantime,these charging systems have poor performance in the case of high concurrent access,and cannot ensure stable service during peak hours.This thesis designs and implements an intelligent charging system for electric bicycles,and conducts in-depth research on the above two problems and proposes solutions.The system includes hardware design,GPRS communication module design,cloud platform construction and application software development.The hardware part is designed with overcurrent,overload and short circuit protection functions to achieve safe charging.According to different roles,the system develops the user version of We Chat webpage system,operator version We Chat webpage system and web management system to meet the needs of charging users,operator and business administrators.The neighboring charging pile search method in user service is optimized.The thesis introduces spatial query and makes some improvements to existing spatial index methods,designs the fast search algorithm based on quadtree partitioning.Learning from the Geo Hash coding,the thesis develops B-Geo Hash coding to encode charging piles,reduces the dimensionality of the data.The algorithm is divided into two stages: pruning and filtering.In the pruning stage,according to the spatial relationship between the MBR and the target,uses the characteristics of the B-Geo Hash coding to filter out possible target sets from a large amount of data quickly,then uses the spherical surface distance to filter.After testing,the algorithm can achieve results in 1 second when dealing with 300,000 data,which is 84% shorter than the conventional method.Aiming at the problem of slow response caused by a large number of repeated database operations in high concurrency condition,a kind of JDBC-based query caching mechanism is designed.Through extending JDBC interface,the SQL Result Set is cached in the memory of the service server.When the next request is sent,filters it before connecting to database.If the corresponding Result Set exists in the cache,reads it directly,avoiding frequent repeated database operations and reducing system pressure.Research on the cache permutation strategy when the cache capacity reaches the threshold,improves the LRU algorithm,and increases the cache hit rate.At the same time,the cache consistency maintenance strategy is designed to ensure the accuracy of the cache.The JDBC caching mechanism reduces the times of connecting to database.Experiments show that this mechanism can double the response speed of the system.