Application Research On The Overall Removal Method For Biomass Gasification Tar And Pollutants

Biomass gasification technology is one of the important means to achieve efficientutilization of bioenergy. However, tar and gas pollutants restrict the popularization andapplication of gasification gas. At present, both at home and abroad for biomassgasification gas purification research mainly focus on the field of purification process、catalyst properties and test methods. Researches from the angle of optimal control onbiomass tar and pollutants removal process are still insufficient.For that reason, this paper summarizes and analyses characteristics、factors、removalmethod and process of the gasification of biomass tar and pollutant removal and thendesigns the method for removing the gasification of biomass tar and pollutants. With thismethod, testing scheme、 modeling and optimization of operating conditions arediscussed.With reference to the gasification of biomass tar and pollutants removal process, thispaper presents a new grey relational analysis (GRA) with least squares support vectormachine (LS-SVM) modeling method. The modeling method based on the analysis oftest data of GRA extracts the strong related factors of gasification tar and pollutantremoval process as the training sample model to avoid the interference of redundantinformation in the sample data. Based on the test process of literature [23] and [67],biomass gasification process in furnace、tar removal process of catalytic cracking out offurnace and ammonia removal process LS-SVM models are respectively established.Results show that LS-SVM model based on GRA has better fitting and generalizationabilities than LS-SVM model alone and the fitting errors and extrapolation errors are lessthan10%which can meet the engineering requirements of forecasting.Multiobjective optimization functions for biomass gasification tar removal processin furnace、tar removal process of catalytic cracking out of furnace and ammonia removalprocess are established according to the three LS-SVM models. And then with theparallel genetic algorithms for optimizing, Pareto optimal solution sets of the gasificationbiomass gasification tar removal process in furnace、tar removal process of catalyticcracking out of furnace and ammonia removal process are obtained. Optimization resultsshow that under the Pareto optimal conditions, overall performance is better than the testvalue of tar and pollutant removal process. The tar content and calorific value of gas canbe effectively controlled and catalytic decomposition of ammonia conversion rate canreach more than98.5%.