| The properties of F-T diesel synthesied from coal are the same as diesel,which has great low emission potention when the diesel engine fuelled with F-T diesel,so it is of great significance to improve engine emission performance and energy sustainability.However,there are some differences in the molecular structure and element composition between F-T coal-to-liquid and 0~#diesel.It is necesssary to optimize the injection parameters and develop of more targeted injection data,therefore,the excellent characteristics of F-T coal-to-liquid can be fully exploited.In this paper,the effects of injection parameters on combustion and emission performance are researched when the engine fuels with F-T diesel.By the means of CFD simulation and test bench experiment,the influennce of main injection parameters,which include main injection timing,pre-injection timing,injection pressure and pilot quantity onthe combustion and emission properties of common rail diesel fueling with F-T diesel were carefully studied under different operating conditions.At the same time,a machine learning numerical simulation model which can accurately express the relationship between emission and diesel engine operating parameters is established by ANN(BP-ANN?Wavelet-ANN)and SVM(Support Vector Machine).Finally,based on the simulation model,a multi-objective optimization algorithm for injection parameters was builded by coupling of machine learning and intelligent algorithm,also multi-attribute decision making process of Topsis was applied in selecting Pareto solutions,which determined the best combination of injection parameters when the emission performance was optimal.The main research contents and conclusions of this paper are as follows:(1)A multi-dimensional combustion model of F-T diesel engine was built,which coupled the simplified kinetic model of F-T diesel with engine combustion chamber model.CONVERGE multi-dimensional simulation software was adopted to analyze the influence of injection parameters(pre-injection timing,main-injection timing,injection pressure and pilot injection quantity)on combustion process when the F-T diesel engine operated at the typical condition.The results indicate that:among them,the injection pressure make the most significant effect on turbulent kinetic energy,meanwhile,the pre-injection timing have a remarkable influence on fuel evaporation rate.It is helpful to improve the turbulent kinetic energy of the mixing,promote the evaporation and mixing process of the fuel,and increase the generation of homogeneous mixture when advance the injection angle,increase the injection pressure and decrease the pre-injection pilot in a certain range.(2)Based on the test bench of common rail diesel engine,the combustion and emission characteries of F-T coal-to-liquid with different injection parameters under the test operating conditions were analyzed in detail.The results show that:the in-cylinder pressure and heat release rate decreased with retarded injection timing,the NOx,HC,and CO emissions decreased to some degree simultaneously,however,the SOOT exhaust and fuel consumption quantity increased;The fuel atomization was effectively improved when the injection pressure increased,then the air/fuel mixtute was improved,so the SOOT emissions decreased sharply,but the NOx emission increased obviously;Increasing the pre-injection pilot will lead to the decrease of main and pre-injection interval,the decrease of“cold flame”effect in the cylinder.The NOx emission decreased,however,the SOOT HC and CO emission showed an increasing trend.In order to make the diesel engine run with high efficient and low emission,the injection strategy with retarded injection timing and increased injection pressure within a certain range is appropriate,the proportion of pre-injection pilot should be controlled in the range of 10-18%.(3)On the basis of experiment data,a numerical simulation model was established by BP-ANN?Wavelet-ANN and support vector machine(SVM)which can effectively express the correlation between SOOT and NOx emission and the operating parameters of the common rail diesel engine studied,then the kennel function parameters under different working conditions were selceted by cross validation.The model was calibrated and analyzed with the test data.The results show that the prediction ability of the model established by SVM was different for various targets,which indicated that the maximun prediction error is not more than 10%,in addition,the prediction error is in the range of 0.21-3%except for very few points.The high prediction accuracy and excellent generalization ability of the numerical model are fully illustrated,the feasibility of using the SVM numerical model to study the performance of electronically controlled high pressure common rail diesel engine is demonstrated simultaneously.(4)A multi-objective numerical optimization model of SOOT and NOx for the electronically controlled high pressure common rail diesel engine was built by CMLIA for the common rail diesel engine,which can effectively combine the numerical model of SVM with the new intelligent algorithm to ensure the optimization results of SOOT and NOx under various operating conditions are effective.Finally,the Pareto optimization solutions of SOOT and NOx were determined by using multi-attribute decision making process of Topsis according to the properties of indictors for the electronic controlled high pressure common rail diesel engine were determined.At the same time the optimization results were validated and analyzed experimentally,the results show that the Pareto optimization solution as evenly distributed,the solutions had clear dominant relationship and there was no congestion problem.Which can effectively improve the trade-off relationship between SOOT and NOx emissions.The corresponding test experiments were conduted according to the optimized parameters.The results show that the Pareto optimal solutions were highly consistent with the experiment values.The goodness of fit of both solutions was greater than 0.95 under the test conditons,and the HC,CO emissions,fuel consumption rate and peak in-cylinder pressure fluctuated in a small range,which indicated the accurancy of Pareto optimal results.CMLIA optimization algorithm and Topsis were feasible for multi-objective optimization of electronic controlled high-pressure common rail diesel engine,the SOOT emissions decreased by 17.94 and 16.16%,however the NOx emissions decreased by 3.76 and 0.64%at the test conditions resrectively... |
PDF Full Text Request