Dynamic Simulation And Soft Measurement For Two Phase Anaerobic Digestion Of Organic Wastes

Anaerobic digestion process is a typical green and environmental protectionproject, and this circular economy technology as a major agricultural organic wasteenergy utilization has been widely used. Two-phase anaerobic fermentation processwould separate acid-producing phase and methane-producing phase in the process offermentation, which has higher efficiency and better treatment effect.Anaerobic digestion process is a series process of complicated biochemicalreactions, and this thesis studies the dynamic model of anaerobic digestion process.Such a model can not be used in the application of practical production because of toomany operating parameters which are difficult to obtain; mechanism model requires acertain simplification and assumption, and there are some errors between actualworking conditions. The present research of the microbial fermentation process,which use support vector machine, pattern recognition, neural network intelligentalgorithm for fermentation process modeling, good results have been achieved. Whilethe research of two phase anaerobic digestion is staying in the experimental stage, thework in the establishment of the model is little.This study has designed and set up two phase anaerobic digestion system testbench and software system which under Delphi7.0development platform: includingproduce acid system, gas system, on-line monitoring system, data storage,temperature control system, power control system and computer interface. Using theself-built dynamic simulated test, we arrangement the experiment based on theoptimization design of experiment and using the orthogonal table; Data acquisitionpriority depend on the online collection, the rest depend on the tie offline artificialanalysis, and the results have been unified.In order to achieve optimization controlling to the anaerobic fermentation ofstraw and pig dung and methane-producing of production process, this research usetwo phase anaerobic digestion process system dynamic simulation tests have beencarried out. We have studied the method of support vector machine (SVM) applyingto anaerobic digestion system parameter prediction, selecting the optimal parametercombination of LIBSVM regression and established the model. Hydrolysis acidification phase,use the pH, C/N (ratio of carbon and nitrogen) and TS (total solidconcentration) as the input variables, COD concentration as the goal;Methane-producing phase with the T (temperature), COD concentration of theentering water and HRT (hydraulic retention time) as the input variables, gas rate asthe target volume, respectively, the corresponding mathematical model is established.The experimental results show that the model in the effect of prediction in systemtarget quantity has certain applicability.