To explore the size effect of asphalt mixture, asphalt mastic asphalt mixture was taken as the research object, and the uniaxial compressive strength was used as the evaluation index. The applicability of the three size effect theories of Bazant, Carpinteri, and Weibull were evaluated, and the corresponding size effect law parameters were calculated. The critical dimensions and critical strength of the asphalt mixture were determined, and the effects of coarse aggregate properties, maximum nominal particle size, and loading rate on the size effect of asphalt mixture were analyzed. Results show that the asphalt mixture has an obvious size effect phenomenon, and the uniaxial compressive stress-strain curves of asphalt samples with different sizes have the characteristics of compaction stage, elastic stage, and transition from plastic deformation to failure stage. The uniaxial compressive strength of the asphalt mixture is negatively correlated with the sample size, and as the sample size increases, the strength reduction decreases gradually. With the increase in the elastic modulus of the coarse aggregate, the critical strength value increases continuously, and the critical dimension decreases. As the maximum nominal particle size increases, the critical dimension and critical strength increase gradually. With the increase in the loading rate, the critical strength continues to increase, and the critical dimension gradually decreases. The Bazant and Carpinteri size effect laws are more suitable for the analysis of the size effect of asphalt mixture, while the Weibull size effect law has a large error. When SMA is used as the pavement structural layer, the ratio S of the structural layer thickness to the maximum nominal particle size calculated by the Bazant size effect law is 2.6 to 3.5, and the S range calculated by the Carpinteri size effect law is 2.6 to 4.0.

To quantitatively analyze and study the water stability of crushed gravel asphalt mixture by using surface energy theory, the surface energy parameters of crushed gravel samples at 20℃ are tested by using the gas sorption method, and the surface energy parameters of four anti-stripping agent asphalts with different dosages (0%, 0.2%, 0.4%, 0.6%) at 20℃ are tested by using the Wilhelmy plate method. Then, the binding energy of crushed gravel and anti-stripping agent asphalts with different dosages and the surface energy evaluation index of water stability are calculated. On this basis, quantitative analysis and sequencing of water stability of crushed gravel asphalt mixture were carried out. The microscopic mechanism of the anti-stripping agent's improvement on adhesion between crushed gravel and asphalt was analyzed from the angle of surface energy. Marshall test (immersion Marshall test, vacuum saturation Marshall test, freeze-thaw splitting test) with anti-stripping agent contents of 0%, 0.2%, 0.4%, and 0.6% was conducted through strict control of crushed gravel material, gradation, and oil-stone ratio. Results show that (1) the surface energy parameters of crushed gravel are dominated by polar alkali component; (2) the addition of anti-stripping agent will reduce the surface energy of asphalt, thus increasing the wettability of asphalt to aggregate; (3) the improvement of the adhesion between crushed gravel and asphalt by adding anti-stripping agent is mainly reflected in two aspects:reducing the cohesive energy of asphalt itself to increase its wettability to aggregate and increasing the surface energy acid component of asphalt to increase its adhesion binding energy to aggregate; and (4) the order of microscopic surface energy evaluation index of water stability of crushed gravel asphalt mixture is the same as that of macroscopic performance test index, thus demonstrating the accuracy of the surface energy system for quantitative analysis of the water stability of the crushed gravel asphalt mixture.

To solve the problem of the low accuracy of the traditional qualitative method for highway asphalt pavement performance, a prediction model based on improved firefly algorithm (IFA)-support vector machine (SVM) is established by combining SVM theory and IFA. First, firefly field search is introduced into the prediction model to overcome the random movement of fireflies with the increase in the number of iterations in the optimization process. Second, in the subsequent optimization process, the dynamic adjustment algorithm is used to search the step size to balance the global search ability, which accelerates the optimization selection of the performance parameters of the SVM model. Finally, the example is verified and compared with the standard FA-SVM prediction method to verify the validity of the IFA-SVM model and the feasibility of prediction accuracy. The result shows the following:(1) The maximum relative error is 2.5435% and the minimum is 0.8206% when the standard FA-SVM is used to predict the pavement performance of the Baiyin section of the G6 expressway. The maximum relative error is 1.0858% and the minimum is 0.3654%, and their root mean square error is smaller than that of the standard FA-SVM method. (2) The IFA-SVM model has a faster convergence rate and a higher accuracy than the standard FA-SVM when predicting the performance of asphalt pavement on highways. The prediction result is not only closer to the measured value but also provides effective support for the maintenance decision of asphalt pavement on highways.

To accurately obtain the waveform and spectral characteristics of saline soil subgrade, on the basis of the provincial highway widening project of Ruoqiang-Weili county in Sinkiang, the field detection of coarse-grained weak sulfuric acid saline soil subgrade was carried out by using the type LTD-2100 ground-penetrating radar (GPR) to obtain the original measured files under different water content (ω) levels. IDSP analysis program was used to process the measured file and obtain the initial characteristics of waveform and frequency spectrum. To verify the initial characteristics, a rectangular model box was created in the laboratory with the field packing (the size was 1×0.8×0.8 m), and the water content of the model packing and temperature were manually adjusted. The model property of the waveform and frequency spectrum was obtained by detecting the model. A comparison of the initial property with the model property shows that under a positive temperature, the increase in ω will enhance the electromagnetic reflection, and when 8% < ω < 27%, the reflection amplitude reaches the maximum, the maximum normalized amplitude is close to 1.0, the spectrum energy is dispersed, and the main frequency is not critical while it is lower than 200 MHz. When ω>32%, the electromagnetic loss is significantly intensified as the line mapping shows typical snowflake-like characteristics, and the point mapping shows typical linear characteristics, the spectrum energy is concentrated, the main frequency distribution range is 20-65 MHz, and the low frequency characteristics are evident. The results can provide a relevant reference for GPR detection and interpretation of coarse-grained weak sulfuric acid saline soil subgrade.

To study the influence factors of low-temperature performance of semi-flexible pavement materials, a freezing test was conducted, and the cracking temperature was used as the evaluation index to study the influence of asphalt type (90 matrix asphalt, SBS-modified asphalt, TPS high-viscosity modified asphalt), cement-based grout material types (ordinary grout, rubber powder grout, emulsified asphalt grout), and skeleton void ratio (20%, 24%, 28%, 32%) on the low-temperature performance of semi-flexible pavement materials. Results indicate that the freezing test can effectively distinguish the low-temperature crack resistance improvement benefits of semi-flexible pavement materials with different asphalt, cement-based grouting materials, and skeleton void ratios. Cracking temperature can effectively evaluate the low-temperature performance of semi-flexible pavement materials. The substrate void ratio is in the range of 20%-32%, and the cracking temperature changes with the increase in the substrate void ratio, decreasing initially and then rising. When the void ratio is around 24%, the best low-temperature performance of semi-flexible materials is achieved. The improvement effect of the elasticity of grout on the low-temperature crack resistance of semi-flexible pavement materials is obvious; the improvement effect of emulsified asphalt grout on the low-temperature performance of semi-flexible pavement materials is better than that of rubber powder grout. After the modified asphalt is used in the substrate, the cracking temperature is reduced by 31.4%-38.3%, which is much greater than the effect of grouting. At the same time, the elasticity of the base asphalt is more effective than the viscosity increase in improving the low-temperature crack resistance of semi-flexible materials.

To study the effect law of asphalt mixture composition design and static/dynamic test methods on resilient modulus, and consummate the design parameters of asphalt mixture, the laboratory tests on static/dynamic compression resilient moduli and strength are conducted by the uniaxial unconfined method and top measuring method by MTS-810. The result shows that (1) there is no direct relationship between strength and resilient modulus of asphalt mixture; (2) both the aggregate gradation and asphalt quality can influence the value of resilient modulus, and the influence from asphalt quality is more significant. By studying the deformation law of the material under static and dynamic loading throughout the test, and combining with the Kelvin mechanic model and viscoelastic theory of asphalt mixture, the limitations of the static and dynamic moduli are demonstrated. It shows that the viscoelastic property is the important factor affecting the values of static and dynamic moduli of the mixture, and resulting in smaller values of moduli. In the end, the linear relationship between static and dynamic moduli is researched, and the conversation equation of static and dynamic moduli for different nominal maximum aggregate sized mixture is obtained, which is more accurate compared to the existing formula.

In order to achieve a quantitative evaluation of adhesion from the microscopic point of view, based on the variation law of the contact angle in the process of asphalt spreading at the aggregate surface, an evaluation index that can quantitatively characterize the adhesive properties of asphalt and aggregate is presented. The image information of the contact angle is collected based on wetting theory and sessile drop method at different temperatures. The technical parameters are extracted from the fitting equation of the contact angle, and the quantitative analysis of the adhesive performance is realized. The result shows that (1) In the process of the asphalt spreading at the aggregate surface, the equation for the relationship between the contact angle and time is in accordance with power function y=Axa, the constant term A in the power function represents the size of the initial contact angle, and the initial state of contact angle is related to the viscosity state of the asphalt at the same temperature, which represents of the asphalt. There is a linear relationship between the constant term A and the asphalt viscosity η at log-log plot. (2) The index a of the power function represents the diffusion rate of asphalt at the aggregate surface and reflects the wetting and the adhesive properties between the asphalt and the aggregate. (3) The diffusion rate of 3 graded matrix asphalts at the surface of diabase experimented in this research has a higher temperature sensitivity, and the wetting and the adhesive properties between the asphalt and the diabase is better than those of limestone.

The hysteresis curve of dynamic resilient modulus (DRM) is the load and deformation curve of asphalt mixture under dynamic loading, it can reflect the evolution rule of internal stress-train state with load and temperature. Based on viscoelastic theory, 3 morphological parameters could be determined:slope k of major axis, eccentricity e of ellipse, and area S of HC, which characterizes DRM, viscosity and energy dissipation capacity of asphalt mixture respectively. Thus the DRM test under various temperatures and loads are carried out. The result shows that there is a significant difference in the morphological parameters of SBS and asphalt mixture No.70 with the increase of temperature, and a inflection point exists at 15~20℃ in both hysteresis curves, the values of k and S change regularly but e keeps unchanged as load level increases. It indicates that (1) asphalt quality has more influence on the property than grade, and the viscosity is unrelated to load size; (2) between 15~20℃, the mechanical properties of asphalt mixture vary greatly, and their sensitivity to temperature are higher; (3) different values of morphological parameters of HC reflect the performance difference of asphalt mixture, and 3 morphological parameters could be used to evaluate the basic mechanical performance of asphalt mixture include viscosity, fatigue durability, etc. Finally, based on the changing rule of test curves, the stress dependence model of resilience modulus which reflects the response behaviors of asphalt mixture under various stress level conditions is established, and the prediction results are close to the real state. The variable load cumulative dissipation energy prediction equation is derived based on the Miner linear damage criterion, and the fatigue equation based on the theory of energy dissipation is established. The morphological parameters S obtained by the dynamic modulus test can be used to predict the fatigue performance of asphalt mixture.

In order to objectively characterize the nonlinear viscoelastic behavior of asphalt mixture and its dependence on temperature and load, the two-point dynamic flexural tensile modulus test of asphalt mixture at different strain levels, temperatures and loading frequencies is carried out by using French trapezoidal beam tester. By introducing the method of nonlinear operator, the dynamic modulus dependency modeling method for asphalt mixture based on temperature and strain parameters is put forward, the expression of dynamic modulus principal surface of asphalt mixture is established, and the reliability of the dynamic modulus dependent model is evaluated by correlation coefficient test and variance analysis method. The result shows that (1) for the two-point flexural tensile test of trapezoidal beam based on asphalt mixture of full-scale test track, the mean values of dynamic modulus variation coefficient at 0℃ and 45℃ are about 2.5% and 6.2% respectively, the mean values of dynamic modulus variation coefficient at 10 test temperatures is about 4%, the variation level of the test method is low, the error is small, which has a very good parallelism and reliability; (2) for the established dynamic modulus dependent model expression based on temperature and strain parameters, the decision coefficient R² can reach more than 0.99, the P value of F test is 0, the fitting effect of the model on the test result is good, and there is a significant correlation of dynamic modulus with temperature and strain level; (3) the model can be used to describe the dynamic flexural tensile modulus of asphalt mixture, and the description model of the dynamic modulus can be transformed from principal curve to principal surface. In pavement structure analysis, this model can reflect the nonlinear response phenomenon in actual pavement more objectively, and can avoid the problems such as distortion of the calculation result caused by unreasonable modulus value in structural analysis effectively.

In the process of drilling, the drilling tool contacts with the rock and soil directly. The response information of the drilling tool reflects the mechanical properties of the rock and soil comprehensively. A large amount of geological data hidden in the response information of the drilling tool can be used to analyze and determine the mechanical parameters and spatial distribution of the rock and soil, thus providing an important reference index for the identification of the formation interface and the classification of the surrounding rock. At present, the response information of drilling tools in the field of geotechnical engineering has not been collected, resulting in a great waste of data resources. To make full use of the relevant parameter information in tunnel blasthole drilling, the relationship between the parameters and the quality of surrounding rock is collected and established, and in combination with the geological situation of the face, the level of surrounding rock in front of the face is predicted. The tunnel of Sandu high-speed rail in Guizhou Province is used to conduct research on advance geological prediction technology based on borehole drilling of tunnel face, and the following results are obtained. The drillability of rock is directly related to drilling speed V, vibration acceleration a, and vibration frequency f. Theoretical research on short-distance geological prediction of borehole drilling in tunnel face is completed on this basis. A set of while-drilling monitoring equipment is developed and applied to pneumatic rock drill. While-drilling equipment can collect the real-time data of drilling speed V and vibration acceleration a of the drilling machine. The data is collected based on the analysis of drilling rate V and vibration acceleration a. The quantitative relationship between rock mass level and drilling information is established. Through application in supporting engineering, a standard database of rock antidrilling coefficient of different levels of surrounding rock based on the rock type of quartz sandstone is preliminarily established. This database is better than the geological report of the excavation site.

Reducing traffic accidents and relieving traffic congestion are two main topics for Chinese road traffic management. Vehicle vision-based vehicle detection and tracking technology is the primary link in road environment perception. The detection technology for forward vehicles is becoming increasingly mature, and commercial products have appeared. Detection is more complicated for oblique vehicles. This research mainly studies the detection and tracking of oblique vehicles. The main work is as follows:First, the image preprocessing method in image processing technology is mainly studied. With focus on research of lane line detection algorithm, an improved lane line detection algorithm combining Hough transform and K-means clustering is proposed to improve the accuracy. With the lane line taken as the standard, the vehicle detection area is divided and combined with the ROI area extraction to further narrow the detection range. With the aim of performing vehicle detection, a two-level optimization algorithm is proposed. First, the method of initial inspection and machine learning detection verification of feature detection is presented. The vehicle target feature and image vertical edge feature fusion are used to generate the suspected target region. An improved dual threshold segmentation algorithm is then proposed to improve the effect of shadow feature extraction. The cascading AdaBoost classifier is combined to generate the target verification area. To address the real-time problem, kernel principle component analysis using Gaussian kernel function is applied to reduce the dimensionality of image feature and improve real-time performance. Experimental results show that the identification accuracy is up to 90% for oblique vehicles when the optimization algorithm of the mixing of shadow feature and image vertical edge feature is used. In good weather, the accuracy is up to 95% when using the AdaBoost classifier combined with HOG+Haar-like feature; The detection time is reduced by 26% when kernel principal component analysis is used. The result of this project is valuable for dynamic oblique vehicle detection.

Nonparametric regression is an important method for short-term traffic flow forecasting, but the traditional nonparametric regression method needs a large storage space and slow query speed when the data are large and the dimension is high. In this paper, an improved nonparametric regression traffic flow forecasting algorithm is proposed. Subtraction fuzzy clustering method is used to cluster historical data to reduce the amount of data in the pattern database. Principal component analysis (PCA) is used to reduce the dimension of the pattern to overcome the problems of slow matching speed and interference of irrelevant dimension caused by the high dimension of the pattern. The support vector machine method is used to estimate the value of the final predicted variables by searching the patterns. The operation efficiency and prediction accuracy of the algorithm are improved. An online simulation-based test shows that the algorithm exhibits better efficiency and accuracy compared with traditional methods.

To research the quantitative evaluation of the effectiveness of intelligent vehicles from user operating conditions to test operating conditions in proving ground, a matching model and evaluation model of intelligent vehicle test condition and user condition was presented on the basis of theory of human-vehicle-road cooperative driving risk field by comprehensively considering the acceleration coefficient, risk coverage, maximum risk, and risk distribution of user conditions and test site conditions. With the vehicle following scene taken as an example and with the JT/T 1242-2019, GB/T 33577-2017, and ISO 22839 standards used as reference, the validity of the actual operating condition data of users and the operating condition data of the test site was analyzed and evaluated by using the correlation evaluation model. Results showed that the research results can be used to quantitatively evaluate the effectiveness of different intelligent vehicle test conditions.