On the solving method of the indirect source characteristics identification model for an engine exhaust system based on error estimation

Haitao LIU,Xiaomin LIAN,Linya LIU
DOI: https://doi.org/10.13465/j.cnki.jvs.2016.16.016
2016-01-01
Abstract:Linear time-invariant hypothetical model solved in the frequency domain for the determination of engine source characteristics plays a practical role in the engine exhaust system design.However,the error of existing methods for solving the model is large,the results are sensitive to input error,and the identification and solving processes are complicated.In order to improve the accuracy of engine source characteristics identification results,an error estimation method was established for the indirect identification model,namely source strength level dispersion estimation and source impedance deviation estimation.The equation from the over-determined nonlinear equations in the identification model was pairwise combined,which were directly solved by an analytic geometry method.Multiple source characteristic parameter solutions were obtained.And then the optimal solution was selected by the source impedance deviation estimation. According to analysis and calculation,the error value of the source characteristic parameters acquired by the method above,is significantly lower than the traditional four-load method.Finally,the three-dimensional coupled simulation of acoustic and flow method was used to verify the results of source characteristic identification.The sound pressure level of far-field response point predicted by simulations agrees well with experimental results,which indicates that the optimal source characteristic parameters selection method based on error estimation can achieve more accurate source identification results,and can efficiently reduce the identification error and complexity.
What problem does this paper attempt to address?