Font Size: a A A

Study On Effect Factors And Prediction Method Of Diesel Engine Combustion Noise

Posted on:2023-07-18Degree:MasterType:Thesis
Country:ChinaCandidate:Y DongFull Text:PDF
GTID:2532307154968929Subject:Power Engineering and Engineering Thermophysics
Abstract/Summary:PDF Full Text Request
With the rise of international vehicle ownership,increasingly stringent noise emission laws,and consumers’ higher comfort requirements,the improvement on the engine NVH performance is becoming more and more important.Because of the complex structure and various transmission paths of internal combustion engine,the frequency components and noise sources of internal combustion engine are very complicated,so it is hard to predict mechanical noise or combustion noise.Therefore,based on signal processing technology,a quantitative analysis method of combustion noise was proposed to analyze the effect of load,speed and injection parameters on diesel engine combustion noise.At the same time,the combustion excitation was quantified by combustion characteristic parameters,and the nonlinear prediction model of combustion noise was established,which is of great significance in the prediction of engine combustion noise.Based on the load,speed and fuel injection parameters which are related to the cylinder pressure,the influence of these on the combustion noise is analyzed.And the relationship between the combustion noise and the cylinder pressure,cylinder pressure rise rate,the high frequency oscillation of pressure and heat release rate were summarized.The results showed that the contribution of combustion noise was greater at medium speed,while mechanical noise was dominant at high speed.The effect of load on combustion excitation was higher than that of rotational speed.Advancing main injection timing and rising fuel rail pressure would worsen the combustion noise.To a certain extent,pilot injection strategies can reduce combustion noise.Pilot injection timing had no obvious effect on combustion noise and small effects on combustion excitation.In particular,increase of the pilot injection quantity was beneficial to reducing combustion noise,but smaller pilot injection quantity had no apparent effect.Under the assumption of linear transitive relation between combustion excitation and radiation noise.Transfer function prediction model between cylinder pressure and combustion noise was set up.The transfer function spectrum of exhaust side,intake side and top side were calculated under different working conditions.The transfer function consistency in three directions under the experimental operating conditions was found.Finally the combustion noise transfer function of diesel engine was determined.Cylinder pressure and transfer function were used to predict combustion noise,but the deviation of prediction results was large,which shows the limitation of linear transfer hypothesis,and also indicated the method of transfer function prediction was more suitable for qualitative analysis of combustion noise.By extracting peak pressure,the maximum cylinder pressure rise rate,pressure second derivative,high-frequency oscillation peak,high frequency oscillation energy and peak heat release rate to quantify the combustion excitation.Based on the results of single factor correlation analysis between combustion characteristic parameters and combustion noise,peak heat release rate,peak pressure and the maximum cylinder pressure rise rate had a good relationship with the combustion noise and were selected as inputs of the prediction model.Then the nonlinear SVR prediction model of combustion noise is established,and the parameter optimization algorithm of PSO and GA were also introduced to further improve the model.Each prediction model was trained by training sets and the results were evaluated from perspective of multiple indicators.Finally using the test set data to evaluate the performance of all the models.It was found that GA-SVR prediction model can better reflect the nonlinear relationship between combustion characteristic parameters and combustion noise,it has better prediction accuracy of combustion noise.
Keywords/Search Tags:Diesel Engine, Combustion noise, Transfer function, Prediction model of combustion noise, Support vector regression
PDF Full Text Request
Related items