Research On Key Technologies Of Intelligent Compaction System For The Asphalt Pavement

Rolling is a key process in the compaction of asphalt mixtures and the realization of road performance,affecting the strength,stability and fatigue resistance of asphalt pavements directly.Therefore,it is necessary to pay attention to and strengthen compaction quality control in the construction of asphalt pavement.At present,the compaction quality management of asphalt pavement is mainly based on post-detection methods yet,which is difficult to acquire the compaction status in time and achieve the process control.Therefore,Intelligent Compaction,with ability of continuously monitoring and real-time feedback of compaction status,has attracted attention gradually.This dissertation focuses on the rolling process of asphalt mixture,and conducts research on the key technologies of Intelligent Compaction System with "Roller-Material" coupling system model and analysis,vibration feedback signal processing,compaction state perception and intelligent quality evaluation,etc.First,the compaction mechanism of asphalt mixtures was described,the action principles of three resistances in the compaction process were analyzed,and a law that the compaction effect is better under resonance was summarized;A one-dimensional rheological model was selected to analyze the rheological behavior of the asphalt mixture during the rolling process,and a nonlinear model of the "Roller-Material" coupling system of vibration compaction was established.Then,model analysis was performed under linear,nonlinear and general conditions.The results show that while the vibration parameters are determined,there is a linear relationship between the resistance of the asphalt mixture to the compaction machine and the inertial force of the vibrating wheel;By measuring change information of the feedback response of vibrating wheel,the changes of the asphalt mixture structure can be analyzed.Then,the compaction status is obtained,which provides a theoretical basis for subsequent intelligent compaction monitoring.Second,through the gyratory compaction experiments,it was concluded that there is a good logarithmic relationship between rolling counts and the compaction degree;A new outdoor vibration compaction test method was proposed and field tests were carried out to obtain vibration feedback signals during the rolling process.A vehicle-mounted detector was designed based on the dual-processor architecture,and a low-cost coordinated positioning solution was proposed to meet the needs of rolling detection and positioning;A remote monitoring system for pavement construction was developed to realize the real-time non-destructive monitoring of construction parameters continuously.Third,the vibration feedback signal collected from the experiments was processed and analyzed.A finite impulse response digital band-pass filter was designed based on the Hamming window,which suppressed high-frequency noise components effectively while ensuring linear phase characteristics of the original signal;The polynomial least square method was used to eliminate the trend terms,and the five-point three-times averaging method was used as smoothing tool,removing the zero drift and burrs in the signal and smoothing the vibration signal waveform.The Ensemble Empirical Mode Decomposition(EEMD)method was selected to decompose the vibration feedback signal for the nonlinearity and non-stationary.Based on Hilbert-Huang Transform(HHT),the effective IMF components were obtained,and then performed time-frequency analysis.The researches show that EMD method regards the IMF components as "basis functions" to reconstruct the signal which can improve the signal quality,reduce the errors caused by spectrum leakage and fence effect,and also has advantages of strong adaptability,as well as good Signal-to-Noise Ratio.Fourth,according to the Parseval's law of energy conservation,a compaction state perception method based on energy distribution was proposed,as well as a new index Vibration Compaction Value(VCVe).Based on the results of vibration signal processing,Compaction Meter Value(CMV),Continuous Compaction Value(CCV),and VCVe index values were calculated.Researches show that,with the increase of rolling pass,CMV,CCV and VCVe values show a gradual increase trend,improving the stability and consistency of harmonic analysis indexes while describing of compaction state;The calculation of VCVe,CMV and CCV index are independent of each other,and can be used for compaction monitoring independently or jointly;Compared with conventional coring detection methods,CMV,CCV and VCVe index can describe the compaction state change of asphalt mixture.Although cannot be used as the acceptance criteria directly,it can be used for compaction state perception and quality process control.Finally,multi-source monitoring data was integrated,and intelligent compaction quality evaluation was carried out based on Support Vector Machine(SVM)and Hidden Markov Model(HMM).The characteristics of training samples were selected,the data was preprocessed,and the sample data was identified based on the Real-time Kinematic(RTK)GPS calibration system;The Fuzzy C-means Method was used to calculate the membership degree of sample data,which suppresses the influence of noise and isolated points;Based on Radial Basis Function kernel function,A Fuzzy Support Vector Machine classifier was designed to classify the compaction state effectively,with the accuracy of 72.6%;Using RTKGPS positioning data to calculate the hidden compaction state sequence,taking SVM state classification result as the observation sequence,the parameters of HMM were calculated based on Maximum Likelihood Estimation algorithm;In the end,the transition probability matrix and the observation probability matrix were obtained;Based on the HMM decoding algorithm,the hidden state sequence of the rolling construction process was calculate,and the accuracy rate was 78.3%;Compared with the FSVM compaction state classification,the accuracy of SVM-HMM was improved greatly,and the Quality Evaluation of the entire rolling process was realized.The research has basically realized primary intelligent compaction technology such as continuous and non-destructive perception and intelligent quality evaluation during the rolling process of asphalt pavement.Nevertheless,the feedback control of intelligent compaction system has not been explored in depth yet;In the future,new technologies such as artificial intelligence,adaptive feed control theory can be combined to achieve the advanced intelligent compaction technology,promoting the development of intelligent construction for the transportation infrastructure.