Study On Chaotic Characteristics Of Biodiesel Combustion In Industrial Furnace

Biodiesel is an eco-friendly energy source,usually produced from animal and plant fats and waste oils through a transesterification reaction.Because of its wide range of raw materials,and renewable and low combustion emission products,it can be a new energy source to replace fossil diesel.There are few alternative applications of biodiesel industrial kilns at home and abroad.This paper investigates and predicts the chaotic characteristics of combustion flame and emission characteristics of Jatropha curcas biodiesel in industrial kilns.Jatropha biodiesel was selected as the fuel.The flame length,area,high-temperature area rate,and combustion product distribution were analyzed using a high-speed camera,a flue gas analyzer,and a high-temperature endoscope at different inlet volumes and oxygen-enriched levels.The results showed that too high and too low inlet volume affect combustion emissions.The best combustion effect of jatropha biodiesel was achieved at 21% oxygen concentration,and 7m~3/h inlet volume when the NO and NOx concentrations in the emission products reached the highest value of43 ppm and 45 ppm,respectively,and the lowest CO concentration was 121 ppm,which showed the best combustion effect;under the condition of 3m~3/h inlet volume,The higher the oxygen enrichment,the smaller the flame length and area,the higher the NO and NOx concentrations in the emission products,the lower the CO concentration,the higher the flame high-temperature area rate,and the best combustion condition is reached at 36% oxygen concentration.Currently,the NO and NOx concentrations in the emission products are 143 ppm and 144 ppm,respectively,and the CO concentration is 120 ppm,showing the best combustion effect.It was verified that the flame geometry and combustion emission characteristics could be used as an essential basis for evaluating the combustion performance of biodiesel.For the nonlinear kinetic characteristics of the biodiesel combustion system,the chaos identification of the flame ash time series was performed using the power spectrum method,the maximum Lyapunov exponent method,the phase space reconstruction method,and the 0-1 test method.The power spectra of each test group’s flame ash time series had no prominent single peak,and the phase space reconstruction showed that they all tended to converge to a specific attractive subset.The chaotic characteristics were generated.The statistics found that the larger the NO and NOx concentrations,the larger the values of Hi and LLE simultaneously.When the NO and NOx concentrations peaked,the Hi and LLE values reached the highest.The correlation coefficient between the maximum Lyapunov index and the high-temperature area rate Hi is 0.9897,indicating a correlation between flame chaos characteristics and combustion stability.Three neural network models,NAR,LSTM,and CNN,were used to predict the biodiesel combustion chaos time series under different operating conditions.The results indicated that the coefficient of determination R2 of the CNN neural network model had an average value of 0.958,root mean square error RMSE of 0.007,and average absolute error of 0.005 for the five operating conditions.Its performance index was better than the other models regarding prediction accuracy.This study is helpful in investigating the relationship between the chaotic characteristics of biodiesel combustion flame and the new characteristics of combustion emission and predicting the chaotic time series,which is a guideline for the research of combustion digitization and precision control.