The Platform Construction Of Combine Harvester Cleaning Flow Field CAE Post-processing And Cleaning Device Optimization

Combine harvester is a typical product of agricultural mechanization and modernization.As an important part of combine harvester,the cleaning effect of cleaning device directly affects the operation quality of the whole machine.It is necessary to study the design process of cleaning device and select the optimal design scheme.With the continuous development of computer technology,CAE technology is used more and more in the design and optimization process of the combine harvester cleaning device.However,there are some problems in the process of CAE postprocessing,such as difficult experience inheritance,long time-consuming and so on.The subsequent cleaning device optimization design process also involves complex and diverse technologies.Therefore,this paper constructed a platform of combine harvester cleaning flow field CAE post-processing and cleaning device optimization in order to realize procedural and intelligent CAE post-processing process of cleaning flow field and multi-objective optimization process of cleaning device.The specific research contents are as follows.(1)Bench test and CFD simulation of airflow velocity measurement in cleaning chamberCombined with the test requirements of air velocity measurement in cleaning room,the existing cleaning device test bench was modified.The experimental scheme of airflow velocity measurement and flow field simulation in cleaning room were designed.The distribution method was used to measure the air velocity.Fluent software was used to simulate the flow field of the cleaning chamber,and the simulation results of the cleaning flow field were exported to ASCII format file,which provided test data for the platform,and also provided knowledge reserve for the reasoning process of the CAE post-processing.(2)Knowledge base construction and reasoning mechanism research of cleaning flow field CAE post-processingBased on the theory of knowledge base system,the overall structure of cleaning flow field CAE post-processing knowledge base system was studied.The knowledge was widely acquired through multiple channels and divided into case knowledge and rule knowledge.Frame and producing representation were used to represent the knowledge respectively.The knowledge base of cleaning flow field CAE postprocessing was constructed.The forward reasoning strategy was used to infer the cleaning flow field CAE post-processing process.The reasoning process of comparing the cleaning flow field simulation data with the air velocity measurement experimental data was studied.In the same way,the reasoning process of air flow analysis above the screen surface of cleaning device was studied.(3)Research on key technology of multi-objective optimization for cleaning deviceCombined with the CFD-DEM coupling simulation results of grain cleaning,the PSO-SVR proxy model of impurity rate and loss rate was established based on the cleaning simulation results.SPEA2 algorithm was used to optimize the cleaning device,and the Pareto optimal solution set was obtained and a group of optimal parameters were selected.The bench test of grain cleaning was carried out by using the optimal parameter combination,which verified the feasibility of cleaning device optimization design based on PSO-SVR agent model and SPEA2 multi-objective optimization algorithm.(4)Realization of combine harvester cleaning flow field CAE post-processing and cleaning device optimization platformBased on B/S architecture,My SQL database technology,Java language,springboot and mybatis plus back-end design framework,and H5(HTML/CSS/JS)front-end technology,a platform of combine harvester cleaning flow field CAE postprocessing and cleaning device optimization was constructed,including knowledge base management system,cleaning flow field CAE post-processing module and multiobjective optimization module.In these three parts,the opening and sharing of CAE post-processing knowledge of cleaning flow field on Web platform,intelligent screening of simulation data and intelligent comparison and analysis with test data,construction of PSO-SVR agent model of cleaning impurity rate and loss rate,and multi-objective optimization design of cleaning device using SPEA2 algorithm were realized.