Study On Active Control Of Road Noise Inside Vehicles Based On Hybrid Control Strategy

With the current popularity of new energy vehicles and the continuous development of intelligent cabin technology,the demand for acoustic comfort within automotive cabins is increasing.Developing efficient,cost-effective,integrated,and highly robust active road noise control systems for the cabin has become a pressing technological challenge in the field of intelligent cabin research for automobiles.This thesis aims to systematically investigate active road noise control based on a hybrid control strategy,focusing on the following main research areas:Firstly,a multi-channel Remote Microphone Technology(RMT)model was established,and simulation was conducted to verify the significant impact of the number and placement of observation microphone arrays on virtual detection accuracy and regions.The Genetic Algorithm(GA)was combined to propose a method for selecting the optimal observation microphone array,effectively determining the optimal arrangement and quantity of observation microphones while minimizing virtual detection errors.Detailed experiments were designed to analyze the influence of primary sound source location,causality,and correlation on virtual detection performance.Secondly,a Hybrid Active Noise Control(HANC)algorithm combining RMT method was developed,and its computational complexity was analyzed and compared.Through simulation analysis and hardware-in-the-loop experiments based on the Speedgoat real-time controller in the laboratory,it was demonstrated that the algorithm can achieve excellent noise attenuation performance without the constraints of error microphone placement.Additionally,the Feedback control module can further attenuate peak noise in the noise spectrum that was not controlled by the pure feedforward control module.Next,optimization selection of reference sensors in the real car RNC(Road Noise Cancellation)system was addressed using the Weight Average Multiple Coherence(WAMCOH)method combined with the Fischer Information Matrix(FIM)principle to rank the importance of all vibration signals in real car testing.The optimal number and locations of sensors were determined,and the minimum number of sensor channels was identified using the critical characteristic analysis method.Consider the cost of real car applications,a distributed dual feedforward control algorithm was designed,a hybrid reference sensor scheme with a few microphones instead of part of the vibration sensors was proposed.Through performance simulation and real-car testing,it was proved that the scheme can ensure the noise reduction effect,and the performance attenuation was less than 0.7 d B(A).Subsequently,the real-car application of the new hybrid RNC algorithm was explored.The genetic algorithm was employed to select the optimal observation microphone array within the vehicle,demonstrating the robustness of the RMT method at different road and speed,ensuring good detection accuracy within the range of 50-350 Hz.Experimental verification revealed that the system equipped with the HANC-RMT algorithm achieved optimal control performance when the error microphones were placed on the roof and seats.However,compared to the control system with error microphones directly placed near the ears,there was still a 1 d B(A)attenuation in overall noise reduction performance.Finally,a Hybrid Broadband and Narrowband ANC(HBNANC)algorithm was proposed based on the HANC algorithm to achieve synchronous control of road noise and multiple-order engine noise.Through simulation and real-car testing,it was verified that the HBNANC control algorithm achieved greater attenuation pointedly for multiple narrowband noise components in steady-state and transient conditions,while not affecting road noise control performance.Furthermore,an innovative Hybrid Broadband and Narrowband Active Sound Quality Control(HBNASQC)algorithm was introduced.Compared to HBNANC algorithm,it can ensure good road noise attenuation performance,while enabling personalized design for peak noises corresponding to different engine orders to meet customer demands for sound quality and sound management.