| Cooking oil fumes(COFs),which are hazardous to human health,has become the main source of ambient air pollution.The separation and purification of particulate pollutants from COFs has become a common concern of researchers and technology developers.In this study,the characteristics and shortcomings of mainstream purification technologies,such as mechanical separation,filtration,washing absorption and electrostatic deposition are discussed in detail.In view of the lack of front-end separation technology of COFs purification system,a rotary separated range hood integrating separation and COFs exhaust is proposed.Then,based on computational fluid dynamics,multi-objective inverse optimization design is implemented for its core components by coupling Kriging model,fuzzy control and genetic algorithm.The main results are as follows:Firstly,the rotating impeller combined with the stationary cascade technology(rotating impeller module,RIM)is used to centrifuge and intercept grease particles by the centrifugal force.This technology sloves the problem of efficient separation and COFs exhaust at the front end of COFs exhaust system.Based on the flow process in RIM,the coupling theoretical model between geometry and aerodynamic parameters is derived and established.This model reveals the coupling and restriction laws between geometry and aerodynamic performance of RIM.According to this model,the feasible design space of RIM is explored theoretically.Moreover,in view of the complex structure of RIM,the simplified methods of“pseudo lumped parameters”and“escape coefficient”are proposed.Based on this method,the theoretical model of the coupling of RIM geometry and grade separation efficiency is derived and established.The validation results indicate that the average relative error between the calculated results of the theoretical model and the experimental results is 1.7-5.4%,and the coefficient of determination is greater than 0.95.The above models collectively constitute the coupled theoretical model for coupling RIM geometry,aerodynamics and separation parameters.It provides theoretical support for design and optimization of the rotary separated range hood from the physical mechanism level.Secondly,a parallel sampling algorithm(KPS-MEIGF)based on Kriging model is proposed.The parallel sampling criterion(Pa S)is constructed by using correlation function and EIGF(expected improvement global fit)criterion to obtain multiple updating points in one iteration.Compared with EIGF sampling method,the running speed of KPS-MEIGF is increased by 2-10 times.To realize the dynamic balance of local and global exploitation in sampling process,cross validation and threshold strategy are introduced through algorithm design.This study further proposes an adaptive parallel sampling algorithm(KAPS-MEIGF)based on Kriging model.Numerical experiments show that,especially for high-dimensional and complex global fitting problems,KAPS-MEIGF has higher sampling efficiency than sequence sampling algorithm on the basis of ensuring global fitting accuracy.When the number of samples obtained is the same,the running time of KAPS-MEIGF is reduced by66-80%compared with MMSE sequence sampling approach.Then,a multi-objective inverse optimization design theory and method of RIM based on adaptive parallel sampling is developed.By introducing fuzzy control technology,the crossover and mutation probability are adaptively adjusted according to population and individual information.Combined with gradient mutation algorithm,an optimization algorithm(FCGM)coupling fuzzy control and gradient mutation is proposed.High dimensional and nonlinear mathematical examples are used to verify that FCGM has high efficiency and good robustness.By combing KAPS-MEIGF and the theory-based and the surrogate-model-based geometry/aerodynamic and separation performance coupling data mining methods with FCGM,an adaptive parallel sampling multi-objective inverse optimization theory method is established.Based on this method,the data mining and inverse optimization design of RIM are implemented.Finally,the experimental test is carried out based on the optimized RIM.The results show that when the exhaust airflow rate and impeller input power are 0.140m3/s and 125.5 W respectively,the separation efficiency of PM2.5 is as high as 40.8%.Moreover,the RIM-SC also shows 100%separation efficiency of grease particles larger than 5μm at the impeller input power range of 67.8-127.1 W and the exhaust airflow rate range of 0.140-0.185 m3/s.By simple quantitative estimation,it found that,when the rotary separation range hood is used in the COFs exhaust system,the exhaust airflow rate load of the end fan and the total power consumption of the system can be reduced by 44.5%and 16.7%,respectively.That is,the rotary separated range hood has great application potential. |