Research On Multi-Objective Optimization Of Compound Muffler

Automobile noise is one of the main noise sources in the environment,which seriously affects people’s physical and mental health and life.With the increasingly stringent environmental noise regulations,exhaust noise control has received extensive attention from enterprises and society.Exhaust noise is the main way for engine noise to radiate to the environment,and it is the simplest and most effective way to control exhaust noise through a muffler.At present,the research and development,design and optimization of mufflers are mainly based on experience design and trial tuning,which is difficult to meet the requirements of high efficiency and low cost.This paper applies a multi-objective optimization algorithm to propose a multi-objective optimization design method for the acoustic performance,aerodynamic performance,and vibration performance of a composite muffler based on a proxy model,which has important engineering significance for guiding the optimization design of the muffler.This article firstly analyzes the acoustic performance and aerodynamic performance of a composite muffler that is preliminarily designed to match a certain type of engine.A one-dimensional acoustic simulation model of the engine and exhaust system was established using GT-power,and the acoustic performance of the initial plan of the muffler was analyzed.Then the aerodynamic performance of the initial plan was analyzed through STAR-CCM+,and the simulation analysis results Compare with the muffler design and development index to determine the optimization direction.Secondly,analyze the influence of the structural parameters of the basic muffler unit on the acoustic performance and aerodynamic performance,find out the structural parameters that have the greatest impact on the acoustic performance and aerodynamic performance as the optimization variables,and collect them through the optimal Latin hypercube sampling method.In the test samples,three proxy models of response surface,radial basis neural network and Kriging were constructed,and the accuracy of the model was verified.The Kriging surrogate model with the highest accuracy is selected and combined with the non-inferior ranking genetic algorithm(NSGA-Ⅱ)with elite retention strategy for multi-objective optimization of the tailpipe noise and back pressure of the initial plan.The optimized muffler was analyzed for aerodynamic noise and wall radiation noise,and then the tail pipe noise and back pressure of the optimized scheme were tested and verified.The one-dimensional noise,aerodynamic noise,and radiation noise calculated by the joint simulation were superimposed and combined with the test.The comparison error of the values is within 2%.The error between the exhaust back pressure test value and the simulated value at rated speed is 0.86%.After optimization,the noise and back pressure are both within the target limit.Finally,the vibration performance research and multi-objective optimization of the compound muffler after the optimization of noise and back pressure are carried out.A finite element model of the exhaust system including the compound muffler is established and its reliability is verified by experiments.The dynamic force and vibration isolation rate of the lifting lugs under the excitation of the engine are analyzed for the entire exhaust system that meets the requirements.The calculation shows that the dynamic force of the lifting lug 3 is 11.03 N at 28.7Hz,and the dynamic force of the lifting lug 4 is 20.1Hz.It is12.9N,which exceeds the 10 N target limit requirement.The vibration isolation rate of lifting lug 5 and lifting lug 6 is lower than 20 d B at 25 Hz and 52 Hz respectively.Aiming at the vibration problem,the dynamic stiffness of the lifting lug was selected to construct a proxy model and the vibration performance of the exhaust system was optimized through a multi-objective optimization algorithm.After multi-objective optimization,the dynamic force and vibration isolation rate of each lifting lug meet the target limit requirements.The constrained modal analysis of the optimized exhaust system shows that the entire system does not have a modal in the engine idle speed range,so no idle resonance will occur,thus achieving the optimization goal.