Research On Computational Assembly Workflow Model And Cost Optimization Based On Novel Membrane Materials

The research and development of new MOFs membrane materials is of great value in modern material science.With the introduction of the material genome concept,scientific computing has become a hot spot in the field of material research and development.Manual assembly of MOFs membrane materials has problems such as high experimental cost,large time consumption,complexity and poor reproducibility.Therefore,this paper uses computer technology to mass-produce MOFs membrane material structures,which can improve efficiency,reduce costs,reduce experimental trial and error processes,and improve the accuracy and feasibility of structural optimization.The main contents of this article include:(1)In this paper,in response to the large demand for new membrane materials,MOFs membrane materials are designed with the help of scientific calculations,and new membrane material calculation and assembly algorithms and software are developed.Based on the principle of topology-oriented material assembly,this paper uses Network X to build a topological network model;proposes a screening algorithm to match building blocks for topological networks;designs a scaling algorithm to adjust the network model by scaling factors;assembles building blocks,and the algorithm can generate new MOFs membrane materials in batches Cell description file.This work has realized the computational assembly simulation of MOFs membrane materials,and based on the simulation results,it provides an important guiding significance for the design of material experiments.(2)In this paper,considering the cost efficiency of materials computing assembly applications,an improved workflow scheduling model is proposed to achieve the performance cost optimization goal of the NMSC algorithm.This paper proposes two de-circulation methods to improve the workflow into two different DAG models,define performance cost constraints,and use scheduling algorithms to optimize the cost of DAG models.In this paper,cyclic decomposition technology and performance modeling algorithm are used to perform cyclic digestion operation on computing assembly workflow,generate DAG＿1 and DAG＿2 workflow models,and design model performance comparison experiments.The results show that DAG＿1 is better than DAG＿2.This paper defines the optimization objective of minimizing the cost of meeting deadline constraints,and uses HEFT algorithm,IC-PCP algorithm and Pro Lis algorithm to conduct DAG scheduling experiments.The results show that the Pro Lis algorithm is superior to the IC-PCP algorithm,and the model reduces the cost of the algorithm while maintaining high performance to a certain extent.(3)In this paper,a new membrane material computing assembly system based on microservice architecture is designed and implemented,which is applied in actual material experiments.First of all,this paper analyzes the function and performance requirements,designs the system architecture in detail,and uses the microservice transition technology to realize the core microservice function of the system’s material calculation assembly.Secondly,this paper uses the gRPC protocol to establish communication between microservice modules.The system uses Docker to deploy microservice modules and K8 s for container management.System testing demonstrates proper functionality and good performance.