Application Of Improved GM(1,N)-weighted Markov Chain Model In Traffic Noise Prediction

Grey Markov chain prediction model has important value in theory and application.On the basis of the existing research,this paper starts from the idea of "model research,improvement and case verification",and deeply studies the model in order to enrich and perfect the model,and to improve the application practicability.The main research work is as follows:(1)According to the three axioms of buffer operator theory,a kind of enhanced and weakened logarithmic buffer operator is constructed to weaken the impact disturbance.The example shows that the precision of the constructed weakening buffer operator after continuous action is higher than that of the previous one.It can reduce the impact of shock disturbance to a certain extent,indicating its effectiveness and applicability.(2)By the matrix perturbation theory,it is proved that the perturbation bound of the sample size and the solution is proportional under the condition of the same index perturbation.In addition,based on the idea of simpson formula in numerical integration,the background value of traditional GM(1,N)model is improved.An example shows that the precision of the traditional GM(1,N)model is higher than that of the traditional GM(1,N)model.(3)Finally,the improved GM(1,N)model is used to study the traffic noise.Because the abnormal value of the simulation results will lead to the large error,the weighted Markov chain model is used to modify the abnormal value of the simulated value.The accuracy of the system is improved greatly,and the steady-state analysis of traffic noise is carried out.The conclusion is in line with the actual situation,which shows that it has a good practicability.