MDOE Application Of Optimization Methods Bag Filter Experiment

Bag filter has high efficiency and stable operation,and it has been widely used in industrial field.However,the filtration and cleaning mechanism is complex and has not yet been unified,and there are still some problems such as running resistance and short service life of filter bag.Experimental research is one of the important methods to explore the mechanism and performance of bag dust removal.In the development of preliminary scheme,design optimization and performance evaluation,it is necessary to carry out a large number of bag dust collector test,and the reasonable test design will achieve a multiplier effect.To this end,the bag filter 1:1 test platform for the development of the test system,and to study the optimization of experimental design methods.Modern design of experiments(MDOE)than the traditional one factor at a time(OFAT)method,the test times less high precision.In order to solve the problem of variable coupling in the complex condition of bag filter,the MDOE method of bag filter,(1)The test apparatus and filter jet cleaning process was simulated,and the resulting data to build response surface model to compare the Latin hypercube Sampling(LHS)of the different criteria and uniform design for jet dust filter and experimental design The further optimization of the experimental sample points is given,and the applicability of the optimized design method is analyzed.(2)The paper sets up the comprehensive test platform of the bag filter and develops the corresponding test system.The host computer uses LabVIEW to carry on the data acquisition,the processing,the storage,the lower position machine PLC to the actuator control.Taking the performance test of the blowing valve as an example,the correctness and validity of the designed MDOE design method are verified.(3)Application results show that:on the filtration test,the blowing cleaning test and the blowing valve performancethe test,number of MDOE method tests is only 46%,25%and 28%of Full factor test;The variance regression analysis showed that the correlation coefficients were 0.9966,0.993 and 0.9973,and the overall root mean square errors were0.51min?17.7608Pa and 0.0071m~3/s.The LHS based on the maximum and minimum distance criterion and uniform design experiment optimization scheme can fully reflect the variation law of filtration,cleaning and characteristics.MDOE method can be widely used in large-scale equipment prototype testing test optimization,can significantly shorten the experiment cycle,reduce experiment costs for the mechanism of the product to deepen the experiment and lay the foundation for rapid innovation.