Research On Threshold Setting Of ACNS Based On Pavement Roughness Recognition

China's rapid development of road transport industry in improving travel quality of citizens also brought about the serious impact of frequent traffic accidents and casualties.The Automatic Crash Notification System(ACNS)can transmit the collision information to the rescue center in a timely manner after the collision of vehicles,so as to reduce the casualty rate,which is increasingly becoming an important post-accident safety technology for vehicles.The new type of ACNS needs to provide stable and reliable accident information by different road conditions,avoiding the interference hazard by mistakenly triggering to the driver and passengers.Therefore,this paper studied the identification of pavement roughness and the threshold setting in the new ACNS.Firstly,the effect of pavement roughness on ACNS performance was studied.Based on the theory of pavement roughness,this paper analyzed and compared its different evaluation indexes and detection methods,and discussed the rationality of the International Roughness Index(IRI)as the evaluation index of this paper on the study of pavement roughness,and wrote the calculation program according to the calculation model of IRI,and deduced the calculation model of acceleration power spectral density and IRI value based on the theory of power spectral density,further explored the correlation between the mean square root of acceleration power spectral density and IRI value.Secondly,the CarSim pavement and vehicle modeling was studied.Based on the harmonic superposition method,the multi-level pavement elevation data were generated,and the validity of the generated road elevation was verified by comparing the power spectral density of standard pavement,and the modular modeling process of the CarSim vehicle was described.In order to verify the reliability of the data generated by the vehicle model,a real vehicle test system was built to collect the road surface elevation with precision level,and with NI acquisition card,triaxial acceleration sensor,notebook installed with Labview software and others,and the CarSim simulation model and the reliability of simulation data were verified by comparison simulation and measured data.Thirdly,the correlation between body acceleration and IRI was studied.The regression analysis of simulation data in CarSim was conducted by SPSS software,and the correlation model between the mean square root and the IRI value of the body acceleration power spectral density was established,and the body acceleration and pavement roughness detection system was built,and the reliability of the correlation model was verified,and the velocity sensitivity of the correlation model was analyzed by CarSim simulation model.Finally,the setting problem of ACNS threshold was studied.The influence of threshold value on ACNS anti-interference and accident information accuracy was analyzed.The model of setting different threshold values according to different IRI values and different speeds was established,and finally the ACNS trigger decision algorithm based on threshold dynamic matching was proposed.