Research And Development Of Key Technologies For Programmable Controller Integrated Development Environment With Cloud-Edge Collaboration

Programmable logic controller integrated development environment(platform software)is the basic industrial software,which realizes the design,compilation and debugging of control program.For a long time,the platform software adopts the centralized development mode of single architecture,which has the problems of high coupling degree of the whole architecture,difficult to expand functions,limited computing resources and so on.With the development of intelligent manufacturing,higher requirements are put forward for the high integration and rapid expansion capability of platform software intelligence technology.In this thesis,the design idea of platform software microservice architecture based on cloud-edge collaboration is proposed.The massive computing resources and dynamic scaling mechanism provided by cloud computing are used to alleviate the limitation of computing resources of platform software under single architecture.Meanwhile,the advantages of loose coupling of modules in microservice architecture are used to effectively solve the problems of high coupling degree and difficult function expansion of platform software modules under single architecture.On this basis,the theory and technology of graph theory,graph database,knowledge graph,model-driven and so on are used to realize the collaborative and efficient operation of platform software cloud edge under microservice architecture.The main content of this thesis is divided into the following parts:(1)Aiming at the lack of objective basis for the existing platform software single architecture module splitting method and the cloud-edge collaborative microservice architecture scheme is not used in the platform software,this thesis proposes a single module splitting algorithm driven by data flow graph and a platform software microservice architecture scheme based on service governance.In that method,firstly,the data dependency relationship between the module is taken as an objective splitting,and a microservice module set is obtained through the construction of a zero-layer data flow graph,the subdivision of a first-layer data flow graph and the module aggregation based on rules in sequence,so that the interference of human factor on a result is reduced,and meanwhile,the splitting efficiency is considered;Then,according to the real-time difference requirements of the working scenarios of the split service modules.The architecture scheme is built from two main aspects of microservice deployment and communication based on the service governance concepts such as service registration and discovery,service fuse,etc..(2)Aiming at the problems of high model coupling,poor maintainability and poor computing performance after the platform software monomer model is transferred to the microservice architecture,this thesis proposes a microservice model reconstruction method based on graph theory,and introduces the graph database as the optimization scheme of model representation and the knowledge graph theory to realize the interaction between different microservice models.Firstly,according to graph theory,the complex model under the original monomer architecture is reconstructed to reduce the coupling between models.Secondly,the model’s maintainability and response speed are improved by using the visualization and high performance of the graph database.Finally,a microservice model interaction method based on knowledge graph theory is designed.Each microservice model is abstracted into a knowledge graph node and a connection is established in the method.By using the attributes and relationships in the knowledge graph,the cross-service model interaction under the premise of resource isolation is realized by using an event-driven mechanism.(3)Aiming at the problem that the original single model-driven architecture does not match the platform software under microservices,this thesis proposes a modeldriven architecture based on cloud-edge collaboration with operation separation.According to the real-time requirements of the work scenarios for each microservice,based on the design scheme of the new model-driven architecture,two driving modes,synchronous and asynchronous,are proposed respectively to achieve the consistency of the cloud-edge model in terms of structure and data.At the same time,a model incremental matching algorithm based on VF2 is designed.Combined with the structural characteristics of the reconstructed model,the algorithm is optimized based on heuristic search strategy,which ensures the efficient operation of the platform software under the cloud-edge collaborative microservice architecture.(4)Experiments are carried out on the above architecture and optimization methods,including module coupling evaluation,single point function testing,collaborative function testing and performance testing.The method of code static analysis is used to evaluate the coupling degree of modules,which verifies the effectiveness of the microservice architecture in reducing the coupling degree of platform software modules;For single point function verification,black box testing is used to verify that each microservice can run successfully and provide corresponding services;The collaboration function is tested by an actual project,and the collaboration between the cloud and the edge end and between different microservices is verified.The experimental data show that the proposed model optimization representation scheme and the model incremental matching optimization method have significantly reduced response time compared with the original method.The performance of the new architecture is tested by stress test,which verifies that the new architecture can fully use cloud computing resources and has the ability to provide concurrent services for multiple users.The overall experimental results show that the proposed cloud-edge collaboration platform software microservice architecture scheme has good availability and reliability,and the proposed optimization method has superiority in computing performance compared with the original method,which provides a new idea and method for the development of platform software,and provides a feasible technical solution for industrial applications.