Design And Implementation Of Process Assembly System Based On Visualization Technology

At present,China’s primary medical institutions,many supporting laws and regulations are not perfect,the price of drugs is so high,and there are problems such as unhealthy trends in the process of drug purchase and sale.The common people generally reflect that the contradiction of “expensive medical treatment” has become increasingly prominent,which has led the country to carry out medical reforms for primary medical institutions.This thesis designs and implements a private cloud platform for pharmaceutical procurement based on the SaaS model in the context of the county-level medical reform in a province.The main work of the thesis includes:(1)Based on the grassroots pharmaceutical procurement reform,combined with the SaaS maturity model,we chose to explore the SaaS third-level maturity model cloud multi-tenant.The difficulty of cloud multi-tenancy is to create system and multi-tenant data isolation.The problem of creating a system is solved by providing the tenant with a selection module and a custom template,which satisfies the requirement of the multi-tenant to create a personalized system according to the business.This thesis proposes a URL rewriting mechanism algorithm to realize a multi-tenant data isolation scheme of shared database instance and exclusive data table,which ensures that SaaS multi-tenant resource utilization is high and data isolation is high,which satisfies business needs.(2)According to the characteristics of cloud multi-tenancy,the thesis proposes a PaaS workflow solution in the cloud to solve the problem of unified management of cloud multi-tenant workflow.Based on the secondary development of MyFlow open source assembly,the workflow web modeling problem is solved.The writing of the workflow core activity algorithm solves the workflow process operation problem and is the core key to distinguish other workflows.This thesis proposes a workflow middleware to pull data from the platform to synchronize data,solve the problem of platform and workflow middleware data synchronization loss,and ensure platform and workflow data consistency.Based on DUBBO,the issue of workflow microservice cloud publishing is solved.(3)Based on the HESSION framework,the problem of drug catalogue synchronization between the provincial pharmaceutical procurement platform and the municipal pharmaceutical procurement platform was solved.Design the entire pharmaceutical procurement process so that drug information can be transmitted to health clinics,health centers,suppliers,and municipal supervisory units in a timely manner to ensure timely supply of hospital drugs,improve drug supply efficiency,and reduce circulation costs.The hospital purchases the drug details page cache design.This thesis proposes an asynchronousserialization solution to solve the problem of high-speed data cache and database double-write consistency design of the drug detail page in the high concurrency scenario.This thesis proposes an asynchronous update multi-level caching strategy to solve the data consistency problem of the drug detail page.This thesis proposes a solution based on zookeeper distributed lock to solve the concurrency conflict problem of cache reconstruction.This thesis proposes a dual zookeeper distributed lock solution to solve the distributed parallel cache warm-up problem.(4)Medical procurement data flow service,data burying point to track the use of the medical procurement cloud platform,and send the collected data to the log server to provide operational data support for the follow-up.The log scrolling design solves the problem of Nginx generating log files.For Flume log collection,this thesis proposes an IntegratedChannel solution,which is effectively compatible with both MemoryChannel and FileChannel,ensuring fast data processing and persistence,making data transmission more stable and efficient.It solves the problem of data collection and data loss during the use of Flumue to improve the fault tolerance of Flume data collection server.This thesis proposes that Redis deduplication and increased disaster tolerant interceptors solve the problem of log rolling data collection.This thesis proposes that Redis deduplication and daemon threads solve Flume data loss problems.This thesis proposes a design method combining Rowkey random hashing and pre-partitioning to solve the HBase storage hotspot problem and solve the JVM Full GC problem in the process of using HBase.The municipal supervisory unit personnel complete the supervision of medical procurement data flow by analyzing real-time and offline indicators,complete scientific decision-making,and better improve medical procurement services.At the end of the thesis,functional testing and performance testing of the medical procurement cloud platform based on SaaS model were carried out.The test results show that the platform has improved the level of medical procurement services and achieved the expected results.