• <tr id="yyy80"></tr>
  • <sup id="yyy80"></sup>
  • <tfoot id="yyy80"><noscript id="yyy80"></noscript></tfoot>
  • 99热精品在线国产_美女午夜性视频免费_国产精品国产高清国产av_av欧美777_自拍偷自拍亚洲精品老妇_亚洲熟女精品中文字幕_www日本黄色视频网_国产精品野战在线观看 ?

    Modeling and analysis of car-following behavior considering backward-looking effect?

    2021-03-19 03:23:48DongfangMa馬東方YueyiHan韓月一FengzhongQu瞿逢重andShengJin金盛
    Chinese Physics B 2021年3期

    Dongfang Ma(馬東方), Yueyi Han(韓月一), Fengzhong Qu(瞿逢重), and Sheng Jin(金盛)

    1Ocean College,Institute of Marine Information Science and Technology,Zhejiang University,Hangzhou 310058,China

    2College of Civil Engineering and Architecture,Institute of Intelligent Transportation System,Zhejiang University,Hangzhou 310058,China

    Keywords: traffic flow,car-following model,visual angle,backward-looking effect

    1. Introduction

    Car-following models have always been an important research topic in the field of microscopic traffic flow and statistical physics. It has significant application in complex systems,traffic simulation,and intelligent driving.Numerous carfollowing models have been developed in the past decades,and they have important theoretical significance for a profound understanding of complex random characteristics,critical phase transitions,and stability of traffic flow.

    Since the concept of car-following model was proposed in the early 1950s,hundreds of important car-following models have been proposed.[1-8]In recent years, many different car-following models[9-18]are put forward under the modern traffic situation and lots of scholars devote themselves in investigating the behavior of car following.[19]To enrich the traffic flow theory, some scholars proposed extended traffic flow models.[20]However, most models only consider the driver’s perception of the leading vehicle’s motion status information and completely ignore the stimulation of rear vehicle’s motion.In an actual driving environment, the driver will also observe the movement of rear vehicle through the rearview mirror and use it as a basis for adjusting the speed of own vehicle to prevent the rear vehicle from rear-end collision due to its own emergency braking. Especially when traffic is congested or vehicles frequently change lanes,the rear-view effect is more obvious. At the same time, when the distance between two vehicles in the front and rear is small,a severe deceleration of the front vehicle may cause the rear vehicle to brake urgently or even cause a rear-end collision, which may cause congestion and security risks. Therefore, to ensure their own safety and smooth traffic,the drivers will comprehensively judge the state of leading and following vehicles to adjust their driving behavior when driving.The rear-view effect helps the driver to maintain the distance between front and rear vehicles as consistent as possible which effectively stabilizes the traffic flow.

    In recent years,the effect of rear car’s movement on carfollowing behavior has been gradually considered, and relevant models have been established. Sun et al.[21]proposed a car-following model by considering the backward-looking effect and speed difference and showed that the consideration of above factors simultaneously can enhance the stability of traffic flow and effectively suppress the traffic jam. Ge et al.[22]proposed a feedback control car-following model that considers the headway of front and rear vehicles and conducted a stability analysis of the model from a cybernetic perspective, confirming the important role of feedback control signals in suppressing and alleviating traffic congestion. Chen et al.[23]proposed a car-following model that considers driver’s sensory memory and the backward-looking effect, exhibiting good traffic flow stabilization performance. Based on the GIPPS safety distance model, Yang et al.[24]further considered the backward-looking effect and developed a new model.They found that the consideration of backward-looking effect has three types of effects on traffic flow:stabilization,destabilization, and nonphysical phenomena. These models provide important reference value for understanding the effect of rear car on the car-following behavior and have important significance for understanding the factors influencing car-following behavior in a complex environment.

    However,all the above-mentioned models assume that the drivers can accurately perceive the speed, distance, and other variables of the rear car, contrary to the driver’s perception characteristics.Related studies in the field of psychology[25,26]show that the driver cannot accurately perceive the state parameters such as speed and distance; therefore, the following car model based on accurate parameters cannot describe the real driving behavior well. In an actual driving environment,the driver mainly relies on visual information to perceive the vehicle’s movement and then changes his or her own acceleration. In recent years, studies have also been conducted on the visual angle car-following model.[27-30]Jin et al.[28]introduced visual angle as a stimulus for driver to perceive the leading vehicle and built a relevant car-following model. The simulation results show that the model can better describe the characteristics of driver’s asymmetry in acceleration and deceleration and speed fluctuation. Recently, Ma et al.[29]proposed that the solid angle composed of the width and height of the vehicle is a factor influencing the car-following behavior and established the new model using the solid angle as the influencing factor. However,all the car-following models based on visual information focus on the influence of leading vehicle. It is still theoretically significant how to reconstruct the car-following behavior under the combined influence of front and rear vehicles and define more reliable stability conditions of traffic flow parameters.

    Based on the above analysis,this study assumes that during the car-following behavior,the driver judges the movement of leading car and following car depending on visual information and establishes the calculation method of visual angle and its rate of change of front and rear vehicles. On this basis,a fusion mechanism is proposed for the visual information of front and rear vehicles,and then a car-following model is developed under the influence of backward-looking effect. Further analysis of model characteristics and advantages such as stability analysis, numerical simulation, and model comparison provides a basis for building a more realistic and effective microscopic traffic flow model.Using visual angle as influencing factor is a better indicator of human drivers’ awareness.Especially for the rear car, the drivers are harder to perceive the headway and the velocity. The only way to judge the rear car’s behavior is observing the change of the rear car’s size in the driving mirror. Therefore, adopting visual angle in the model can more truly describe the backward-looking effect.Although the theoretical analysis and numerical simulation are similar to those in previous works, there are differences in some aspects such as the asymmetry of the stability and the size of the stability region. So that the results in our paper are more consistent with the actual situation. The backwardlooking effect formed by visual angle can better describe the influence of the oversize vehicles. In reality,the oversize vehicles such as trucks and buses occupy certain proportion which the traditional models cannot reflect the influence of.

    2. Model

    In addition to adjusting the speed and acceleration by observing the leader’s movement,the driver will also observe the change in the movement of rear vehicle through the rearview mirror. Especially when the rear vehicle is a large vehicle,the driver is more willing to increase the distance from the rear car to improve safety.

    Therefore, this study proposes a car-following model framework that considers both front and rear vehicle’s visual information. In this framework,the vehicle size can be introduced into the new model. During the car-following process,the driver’s perception of visual angle information of front and rear vehicles is shown in Fig.1.

    Fig.1. Driver’s sensitivity of forward and backward visual information.

    The formula for calculating the visual angle of leading vehicle (n+1)-th vehicle) to the target vehicle (n-th vehicle)can be given as follows:

    where θnF(t) means the visual angle formed by the n-th car and the car in front of it,wn+1is the width of leading vehicle and ln+1is the length of leading vehicle. Δxn(t)is the distance between n-th car and its leading car at time t. The calculation formula is as follows:

    Similarly, the formula for calculating the visual angle of following vehicle(the(n?1)-th vehicle)relative to the target vehicle can be expressed as follows:

    where θnB(t) means the visual angle formed by the n-th car and the car behind it and wn?1is the width of the following vehicle.

    The driver optimizes the speed according to the visual angle of front and rear vehicles;therefore,the optimized velocity function can be expressed as follows:

    Equation (4) shows the optimal velocity function determined by the forward visual information, while equation (5)shows the optimal velocity function determined by the backward visual information. V1F, V2F, C1F, C2F, V1B, V2B, C1B,and C2Bare the parameters of optimal velocity function to be calibrated. The values are V1F=V1B=6.75 m/s,V2F=V2B=7.91 m/s,C1F=C1B=0.13 m?1,C2F=C2B=1.57.

    Equations(4)and(5)show that the expressions of VBand VFare almost the same. The main difference is the difference in overall value direction because the influence of backwardlooking and forward-looking effects on car-following behavior is opposite.For example,when the front car is close to the current car,the forward visual angle increases,and the current car will respond to deceleration. When the rear car is close to the current car,the rear visual angle will also increase,but the current car will respond to acceleration.The reason why the same optimal velocity function as the function in previous work is adopted is to contrast the new model with the FVD model directly.It is perfectly acceptable to assume the optimal velocity function of other forms rather than the form of headway, but then the parameters must be recalibrated.

    Combining the influence of optimized velocity of front and rear vehicles and the rate of change of visual angle, the car-following model under the influence of backward-looking effect can be expressed as follows:

    where p is the parameter of influencing factors of backwardlooking effect. When p=0, the driver is not influenced by backward-looking effect;α is the sensitivity coefficient,which is the reciprocal of reaction time τ;λ is the response intensity coefficient to the stimulation of visual angle change rate. The value of λ is larger than α, which need to be explained as follows. Firstly,the value of λ is derived according to the calibration of the actual driving data(from NGSIM)and the general range is 40-60, which is not set casually. Secondly, the value of λ is unable to decide whether the changing rate of visual angle(as shown in Eq.(7))can be neglected. Particularly,in the case of the relatively small headway, the changing rate of the forward visual angle can get large, which affected the acceleration deeply. Thirdly, the changing rate of visual angle is used as the parameter influencing acceleration and has a significant influence on the stability analysis result.

    Equation(6)describes the calculation method of acceleration of target vehicle with the front and rear vehicle visual angle information as the stimulus factor. The optimized velocity in the model is obtained from the weighted average of optimized velocities decided by the forward and backward visual angles,i.e.,the driver can obtain the optimal velocity based on the forward and backward visual angle information at the same time and adjust his or her own speed. Essentially, this model indicates that the driver comprehensively considers the visual angle information of front and rear vehicles and tries to maintain the same visual angle forwards and backwards to maintain the stability of headway. When p=0, the backward-looking effect completely disappears, and the model is simplified to only consider the leading car.[28]

    3. Linear stability analysis

    The stability of car-following model is one of its important characteristics. The propagation law of density waves in a traffic flow can be obtained by analyzing the stability of carfollowing model, which has great significance for analyzing the formation mechanism of traffic congestion and proposing related improvement measures.

    Suppose that the initial operating state of vehicle is stable,i.e., the headway is h (h=L/N, L is the total length of road and N is the total number of vehicles), and all the cars maintain the same velocity,the optimized velocity(1?p)VF(θ0F)+pVB(θ0B). Then,the initial position of each vehicle is

    At this time, a small disturbance yn(t) is applied in the traffic flow so that the vehicle position changes from the steady state,written as

    And the headway can be expressed as follows:

    By substituting Eq. (10) into Eqs. (1) and (3), the front and rear visual angles can be calculated as follows:

    Because yn(t)is very small,its quadratic term is ignored.The rate of change of visual angle can be obtained by the derivation of θnF(t)and θnB(t):

    By substituting the above two sets of expressions into Eq.(6)and linearizing the optimized velocity function by Taylor expansion

    Thus,

    The small disturbance yn(t) can be expressed with Fourier mode as follows:

    The first and second derivatives can be calculated and substituted into Eq.(14)

    z can be expanded as follows:

    By substituting z into Eq.(19),we have

    Let z2=0,and the critical condition of linear stability of the model is

    When z2>0, the traffic flow will gradually return to a stable state under the influence of small disturbance, so the condition that makes the model in a linear stable state is

    Figure 2 shows the stability critical curve of the model under different conditions, presenting the relationship between sensitivity coefficient and headway. The four solid lines in Fig.2(a)from the top to the bottom are the critical curves obtained by linear analysis with the conditions λ =20, 40, 60,80,and the values of p are all 0.2. In Fig.2(b),the four solid lines from the top to the bottom are the critical curves obtained by linear analysis with the conditions p=0.2,0.1,0.05,0,and all the λ values are 40. The whole plot is divided into upper and lower regions by the curve. The stable region is above the curve,and its characteristic is that the traffic flow can run stably without obstruction. The unstable area is under the curve,and its characteristic is that a stop-and-go phenomenon will appear in the traffic stream with the gradual expansion of small disturbance.

    Each curve in Fig.2 has a vertex,which is a critical point.Multiple curves indicate that the critical point is different under different parameter conditions. For the critical sensitivity coefficient αc, when α >αc, i.e., the driver’s sensitivity is high,and the sensitivity coefficient is greater than the critical value. Thus, the traffic flow can be smoothly maintained regardless of the headway. If α <αc, the state of traffic flow mainly depends on the headway. Taking the critical headway as a reference,increase or decrease of the headway will reduce the requirement for the sensitivity coefficient, and the traffic flow will more easily enter a stable state. When the headway is near the critical value hc,the sensitivity coefficient must be increased to αcto ensure the stability of traffic flow.

    Figure 2(a)shows that with the increase of λ,the movement direction of critical point is generally downward and slightly to the right, indicating that the critical headway is slightly increased,the critical sensitivity coefficient is reduced to a large extent, and the stable area in the figure becomes larger, i.e., the driver’s attention is more concentrated. When the driver is more sensitive to changes in visual angle, it becomes easier for the traffic flow to reach a stable state. As the value of p decreases in Fig.2(b),the critical point moves to the upper left, the headway decreases, and the critical sensitivity coefficient increases, indicating that the degree of consideration of rear-view effect is gradually reduced. This will make it more difficult for traffic flow to run in a stable state at a high density.

    Fig.2. Linear critical curves of different parameters: different λ (a) and different p(b).

    Figure 3 compares the linear stability of critical curves of different models. Except for the FVD model,[8]all other models consider the influence of backward-looking effect.The stability curves of BLVD model[21]and BL&OVC model[31]almost coincide when the critical headway is consistent, indicating that these models did not substantially improve the stability conditions. When the headway is small, the stability region formed by the stability curve of model proposed in this study is relatively large,indicating that the driver is more concerned about the behavior of rear car in a congested state and uses its behavior as a basis to adjust the speed to avoid collisions.

    Fig.3. Critical curves of different models.

    4. Nonlinear stability analysis

    Nonlinear analysis is widely used in the field of nonlinear systems. In order to analyze the nonlinear characteristic of the new model describing the co-effect of forward and backward visual angle, we research on its gradually changing behavior for long waves of time and space variables around the critical point(αc,hc)using the mKdV equation.

    To do a relatively simple calculation,we use the VF[xn(t)]and VB[xn(t)] to take the place of VF[θn(t)] and VB[θn(t)] respectively and equation(6)can be converted to the following equation:

    then we can gain the expression of the(n+1)-th car further

    Subtract Eq.(24)from Eq.(25)and we can derive

    The condition α =αcis satisfied near the critical point and we define the slow variables X and T as follows using the space variable n and the time variable t in the range 0 <ε ≤1.

    Thereinto, b is a constant coefficient undetermined and we assume that the headway between two cars is

    Expanding Eq. (26) in Taylor’s series and taking to the fifth term of ε,we can get the following equation:

    The expressions of parameters mi(i=1,2,3,4,5,6,7)are

    where

    Here P and Q are constants which are expressed as P=?wn+1/(hc?ln+1)2and Q=?3wn+1/(hc?ln+1)4.

    Thereinto,the expression of gi(i=1,2,3,4,5)can be written as follows by calculation:

    As we can see in the following equation, the variables are supposed to be converted to another form and the purpose of the transformation is to derive the modified KdV equation which is with infinitesimal of higher orders.

    Then we deduce the following regularized equation

    which includes the correction term O(ε). What we do next is ignoring the term O(ε)to get further the mKdV equation with the kink-antikink soliton solution

    Then we assume and reconsider the O(ε)correction term. As a result, the propagation speed c can be composed of the parameters gi

    Therefore, the kink-antikink soliton solution of the headway is expressed in the following form

    the amplitude of the solution is equal to

    The expressions Δx=hc+A and Δx=hc?A represent the headways of two opposite phases, the low-density free flow phase and the high-density congested phase. The two mentioned phases can be combined into one coexisting phase which can be indicated by the kink-antikink soliton solution.Consequently,we achieve our goal of depicting the traffic flow state around the critical point(αc,hc)adopting the solution of the mKdV equation.

    In Fig.4, the black curves are the linear stability critical curves and the red curves with the same relative line style are the nonlinear stability critical curves. Obviously,the addition of nonlinear stability critical curves under the same conditions divides the whole picture into three regions, namely, the stable region(the common region above the two curves under the same conditions), the metastable region (the region between the two curves) and the unstable region (the common region below the two curves). Since the driver’s behavior is asymmetrical in the process of acceleration and deceleration, the nonlinear stability curves show the characteristics of asymmetry which is similar to the linear stability curves.

    Fig.4. Linear critical curves of different parameters: different λ (a) and different p(b).

    5. Numerical simulation

    To further illustrate that the introduction of rear-view effect can more accurately describe the real driving behavior and its spatial and temporal evolution, the model characteristics were analyzed in depth by numerical simulation. In the simulation,all vehicles have the same size: The length is l=5 m,and the width is w=1.8 m. A periodic boundary condition is set as the length of ring road L=1500,and the total number of vehicles N =100. The initial state is stable, i.e., the vehicles are evenly distributed,and the running velocity is the same. In the initial state, the initial position of the first car is 1, indicating that a small disturbance is applied to the first car. The initial state of the vehicles can be expressed as follows:

    The running state parameters of the vehicles can be expressed as follows:

    This section aims to verify the effect of different factors by setting different experimental conditions to prove that the proposed model can further enhance the stability of traffic flow by considering the forward and backward vision effects.

    Figure 5 shows the position distribution of each car under different values of p for a period after the disturbance is applied. The period is selected as t=2000 s-2200 s; p is 0,0.1,0.2, and 0.3; λ is 5, and other parameters remain the same.The aim is to directly analyze the influence of rate of change of visual angle in the model. In the three pictures shown in Figs. 5(a)-5(c), the uneven distribution of vehicles can be clearly observed,showing that the vehicles on the ring road are sparsely or densely distributed on different road sections,i.e.,the stop-and-go phenomena. Beyond that, with the increase of p,the driver’s visual perception of rear car becomes strong,and the degree of sparse or dense lines of the vehicle gradually slows down. The width and number of stripes that can describe the traffic congestion decrease,i.e.,the region of congestion becomes smaller,and the stop-and-go phenomenon is alleviated. When p=0.3, i.e., the stability condition is satisfied, the disturbance of traffic flow disappears after 2000 s,the headway between vehicles remains equal and unchanged,and the traffic flow enters a stable state. This shows that it can effectively enhance the stability of traffic flow for the driver to adjust the observation ratio of forward and backward information to a certain extent.

    Because the model proposed in this study can describe the effect of different car sizes on the driver’s rear-view effect, it was considered to add large cars and set different proportions in the simulation process. The size of a large car was set to 5-m long and 2.5-m wide. In the simulation,the proportion of large vehicles was set to 0.2, 0.1, 0.05, and 0.02; λ is 10; p is 0.2; other parameters were kept the same. Figure 6 shows the change trend of vehicle headway under different proportions of large vehicles after a disturbance is applied. The time points of the two figures are t=300 s and t=2000 s. Figure 6 shows that with the increase in the proportion of large vehicles, the amplitude of up and down fluctuations of headway decreases, and the stop-and-go phenomenon gradually slows down. This indicates that the stability of traffic flow can be improved if there is a certain proportion of large vehicles on the road, whether the large vehicle is observed as the leading car or the rear car because large cars can increase the driver’s attention and sensitivity to changes in their behavior.

    Fig.5. Spatiotemporal vehicle trajectory diagrams for different values of p: (a) p=0,(b) p=0.1,(c) p=0.2,and(d) p=0.3.

    Fig.6. Stop-and-go waves of different truck ratios: (a) t = 300 s, (b)t=2000 s.

    Because there are certain differences in the driving habits of drivers,there will be different decisions on whether to consider the backward-looking effect. Figure 7 shows the spatial and temporal evolution of a headway within a period after the disturbance is applied under different vehicle proportions considering the backward-looking effect. In the simulation, the proportion of vehicles considering the backward-looking effect was set to 0, 0.5, 0.8, and 0.95; the p value is 0.25; the λ value is 5; other parameters were also consistent. All the curves in the three graphs of Figs.7(a)-7(c)fluctuate significantly,indicating that the traffic flow is unstable. In Fig.7(a),all vehicles are only influenced by the front visual angle. The headway greatly fluctuates in the traffic flow. The vehicle is in a stop-and-go state, and the traffic flow is congested. In Fig.7(b),the number of vehicles that only considers the front visual angle and the number of the vehicles that considers both the front and rear visual information make up half of the total number of vehicles. There is still a large fluctuation in the headway. The distribution of headway is discontinuous and varies between two sets of similar data.Such a ratio makes the situation of traffic flow more complex.In Fig.7(c),the ratio of vehicles considering front and rear visual information simultaneously is 80%. The distance between the heads still fluctuates,but the amplitude is significantly reduced compared to the previous two cases. An uneven distribution phenomenon still exists, but it is obviously alleviated. The traffic flow tends to be more stable. In Fig.7(d),95%of the vehicles consider both the front and rear visual information. The head-to-head distance remains unchanged. The small fluctuations are caused by a small number of vehicles that only observe the vehicle in front. The traffic flow is basically stable. Based on the above analysis, the influence of driver’s perception of the rear-view effect on car-following behavior can promote the stability of the traffic flow.This also provides a basis for building an intelligent vehicle control model with the traffic flow stabilization.

    Fig.7. Spatiotemporal evolution of distance headway considering different proportions of visual information forward and backward: (a)rate=0,(b)rate=0.5,(c)rate=0.8,(d)rate=0.95.

    6. Conclusion

    This study fully considers the characteristics of driver’s corresponding car-following behavior by visually acquiring the front and rear vehicle information during driving, introduces a concept of visual angle to describe the visual stimuli of front and rear vehicles,and proposes a car-following model under the influence of rear-view.Through linear stability analysis, a critical stability condition of the model was obtained.The stable area of traffic flow can be expanded by increasing the driver’s sensitivity to the rate of change of visual angle and increasing the ratio of considering the backward-looking effect. The result of nonlinear stability analysis is consistent with that of linear stability analysis. Using numerical simulation to further analyze the characteristics of the model, the results show that different parameter settings, different sizes of large vehicles,and the proportion of backward-looking effects have a significant effect on the spatiotemporal evolution of traffic flow. Therefore, consideration of the visual characteristics of human driver in the model can be closer to the real situation. The driver observes the traffic conditions forwards and backwards with a reasonable allocation ratio and then makes driving decisions,conducive to the stability of traffic flow. This study also provides a basis for studying vehicle control rules in an intelligent networked environment.

    色综合站精品国产| 女生性感内裤真人,穿戴方法视频| 欧美日韩中文字幕国产精品一区二区三区| av超薄肉色丝袜交足视频| 欧美乱妇无乱码| av片东京热男人的天堂| 精品国内亚洲2022精品成人| 身体一侧抽搐| 精品乱码久久久久久99久播| 午夜福利视频1000在线观看| 蜜桃久久精品国产亚洲av| 国产熟女xx| 少妇粗大呻吟视频| 18禁国产床啪视频网站| 亚洲黑人精品在线| 欧美另类亚洲清纯唯美| 国产亚洲精品av在线| 亚洲成人久久性| 久久99热这里只有精品18| 国产黄色小视频在线观看| 亚洲人与动物交配视频| 午夜精品一区二区三区免费看| av免费在线观看网站| 国产三级在线视频| 少妇裸体淫交视频免费看高清 | xxx96com| 成人特级黄色片久久久久久久| 国产不卡一卡二| 一个人免费在线观看的高清视频| 日韩有码中文字幕| 国产av麻豆久久久久久久| 亚洲成人免费电影在线观看| 他把我摸到了高潮在线观看| 免费看十八禁软件| 成人18禁高潮啪啪吃奶动态图| 不卡av一区二区三区| 国产精品久久久久久亚洲av鲁大| 久久久久久国产a免费观看| 亚洲欧美日韩高清专用| 久久热在线av| 老汉色∧v一级毛片| 国产欧美日韩一区二区三| 亚洲欧美一区二区三区黑人| 99久久综合精品五月天人人| 不卡av一区二区三区| 国产野战对白在线观看| 制服丝袜大香蕉在线| 在线观看免费午夜福利视频| 精品久久久久久久人妻蜜臀av| 亚洲精品中文字幕在线视频| 中文字幕人妻丝袜一区二区| 久久久国产成人免费| 99热这里只有是精品50| 亚洲成人久久性| 波多野结衣高清作品| 十八禁人妻一区二区| 国产片内射在线| 色老头精品视频在线观看| 日韩欧美精品v在线| 听说在线观看完整版免费高清| 精品熟女少妇八av免费久了| 中文字幕av在线有码专区| 悠悠久久av| 亚洲精品色激情综合| 日本一二三区视频观看| 中文字幕久久专区| 99久久久亚洲精品蜜臀av| 亚洲av美国av| 亚洲一卡2卡3卡4卡5卡精品中文| 亚洲男人的天堂狠狠| 久久久精品大字幕| 亚洲专区中文字幕在线| 淫秽高清视频在线观看| 国产成人影院久久av| 国产成人av教育| 校园春色视频在线观看| 日日摸夜夜添夜夜添小说| 老司机靠b影院| 国产成年人精品一区二区| 国产成人影院久久av| 久久久久精品国产欧美久久久| 亚洲午夜精品一区,二区,三区| 搡老岳熟女国产| 99国产综合亚洲精品| 欧美色欧美亚洲另类二区| 窝窝影院91人妻| 99久久久亚洲精品蜜臀av| 国产精品九九99| 国产精品美女特级片免费视频播放器 | 九色成人免费人妻av| 久久久国产成人免费| 看免费av毛片| 女人高潮潮喷娇喘18禁视频| 亚洲av熟女| 色av中文字幕| 亚洲国产高清在线一区二区三| 丰满人妻一区二区三区视频av | 两个人免费观看高清视频| 啪啪无遮挡十八禁网站| 亚洲av成人精品一区久久| 久久精品91蜜桃| 国产私拍福利视频在线观看| 可以在线观看的亚洲视频| 黑人欧美特级aaaaaa片| 特级一级黄色大片| 搞女人的毛片| 成人永久免费在线观看视频| 欧美高清成人免费视频www| av免费在线观看网站| 亚洲成人免费电影在线观看| 久久婷婷成人综合色麻豆| 国产三级中文精品| 亚洲国产精品999在线| 久久这里只有精品中国| 久久精品aⅴ一区二区三区四区| 国产亚洲欧美在线一区二区| 51午夜福利影视在线观看| 69av精品久久久久久| 嫁个100分男人电影在线观看| 国内精品久久久久久久电影| 久久久久国产一级毛片高清牌| 级片在线观看| 搞女人的毛片| av中文乱码字幕在线| 日韩欧美三级三区| 麻豆成人午夜福利视频| 特级一级黄色大片| 丰满人妻一区二区三区视频av | 岛国视频午夜一区免费看| 毛片女人毛片| 国产99久久九九免费精品| 亚洲 国产 在线| or卡值多少钱| 日本三级黄在线观看| 国产视频一区二区在线看| 久久久久久久久久黄片| 免费在线观看成人毛片| 少妇熟女aⅴ在线视频| 国产精品久久久久久亚洲av鲁大| 在线播放国产精品三级| 女人爽到高潮嗷嗷叫在线视频| 日本黄色视频三级网站网址| 国产成人系列免费观看| 亚洲精品国产精品久久久不卡| 日本 欧美在线| 99久久精品国产亚洲精品| 亚洲欧美日韩高清在线视频| 婷婷精品国产亚洲av| 成人av一区二区三区在线看| 国产精品一及| 亚洲一码二码三码区别大吗| 久99久视频精品免费| 狂野欧美白嫩少妇大欣赏| 夜夜夜夜夜久久久久| tocl精华| 老鸭窝网址在线观看| 欧美黄色片欧美黄色片| 一区福利在线观看| 欧美日本亚洲视频在线播放| 99久久久亚洲精品蜜臀av| 白带黄色成豆腐渣| 亚洲欧美日韩高清在线视频| 欧美日韩亚洲国产一区二区在线观看| 日韩大码丰满熟妇| 亚洲aⅴ乱码一区二区在线播放 | 丝袜人妻中文字幕| 久9热在线精品视频| 免费看日本二区| 精品高清国产在线一区| 亚洲专区国产一区二区| 2021天堂中文幕一二区在线观| 天堂√8在线中文| 夜夜爽天天搞| 日韩欧美在线二视频| 老司机在亚洲福利影院| 亚洲国产精品sss在线观看| 欧美日韩乱码在线| 国产99久久九九免费精品| 精品久久久久久,| www.www免费av| 精品久久久久久久人妻蜜臀av| 国产一区在线观看成人免费| 嫁个100分男人电影在线观看| 波多野结衣巨乳人妻| 国产一区二区在线观看日韩 | 男女床上黄色一级片免费看| 成年女人毛片免费观看观看9| 国产精品久久久久久久电影 | 天堂动漫精品| www.精华液| 我的老师免费观看完整版| 香蕉丝袜av| 观看免费一级毛片| 亚洲成人中文字幕在线播放| 亚洲人与动物交配视频| 窝窝影院91人妻| 夜夜夜夜夜久久久久| 91国产中文字幕| 亚洲天堂国产精品一区在线| 亚洲一区中文字幕在线| 久久久久久九九精品二区国产 | 精品午夜福利视频在线观看一区| 99热这里只有是精品50| 亚洲色图av天堂| 黄色女人牲交| 日韩欧美免费精品| tocl精华| 欧美中文综合在线视频| 欧美绝顶高潮抽搐喷水| 婷婷精品国产亚洲av在线| 日本精品一区二区三区蜜桃| 亚洲精华国产精华精| 亚洲欧美激情综合另类| 国产99久久九九免费精品| 999精品在线视频| 色综合亚洲欧美另类图片| 日韩av在线大香蕉| x7x7x7水蜜桃| 国产亚洲精品av在线| 亚洲成人久久性| 国产精品日韩av在线免费观看| 日韩有码中文字幕| 亚洲中文日韩欧美视频| 99热只有精品国产| 精品久久久久久成人av| av天堂在线播放| 亚洲五月天丁香| 成人欧美大片| 国产精品久久久人人做人人爽| 精品一区二区三区四区五区乱码| 国产精品免费视频内射| 少妇裸体淫交视频免费看高清 | 国产精品香港三级国产av潘金莲| 精品无人区乱码1区二区| av福利片在线观看| 叶爱在线成人免费视频播放| 伦理电影免费视频| 男女午夜视频在线观看| 精品熟女少妇八av免费久了| 亚洲欧美精品综合一区二区三区| 在线观看美女被高潮喷水网站 | 亚洲欧洲国产日韩| 又爽又黄无遮挡网站| 亚洲第一区二区三区不卡| 一级二级三级毛片免费看| av黄色大香蕉| 村上凉子中文字幕在线| 在线免费十八禁| 亚洲最大成人中文| a级毛片a级免费在线| 国产在视频线在精品| 日本五十路高清| 男人的好看免费观看在线视频| 国产午夜精品论理片| 国产成年人精品一区二区| 国产精品伦人一区二区| 亚洲av中文av极速乱| 国产成人a∨麻豆精品| 波多野结衣巨乳人妻| 久久中文看片网| 又爽又黄无遮挡网站| 成人毛片60女人毛片免费| 久久韩国三级中文字幕| 亚洲欧美日韩东京热| 亚洲熟妇中文字幕五十中出| 高清在线视频一区二区三区 | 亚洲综合色惰| 99热这里只有精品一区| 国产三级在线视频| 精品无人区乱码1区二区| 国产白丝娇喘喷水9色精品| 久久久久久伊人网av| 黄色欧美视频在线观看| 身体一侧抽搐| 欧美精品国产亚洲| 97热精品久久久久久| 欧美一级a爱片免费观看看| 悠悠久久av| 99久久精品热视频| 欧美在线一区亚洲| 亚洲美女搞黄在线观看| 国产午夜福利久久久久久| 亚洲精品色激情综合| 一边摸一边抽搐一进一小说| 成人av在线播放网站| 国产精品久久久久久久久免| 久久久久久久久中文| 国产探花极品一区二区| 国产不卡一卡二| 日本av手机在线免费观看| 午夜激情福利司机影院| 中文资源天堂在线| 亚洲自拍偷在线| 欧美日韩乱码在线| 成年免费大片在线观看| 国产精品一区二区在线观看99 | 国产真实乱freesex| 国产探花极品一区二区| 国产亚洲91精品色在线| 男人和女人高潮做爰伦理| 天美传媒精品一区二区| 国产乱人偷精品视频| 成人毛片a级毛片在线播放| 噜噜噜噜噜久久久久久91| 免费在线观看成人毛片| 18+在线观看网站| 久久欧美精品欧美久久欧美| 九九热线精品视视频播放| 国产中年淑女户外野战色| 国产精品国产高清国产av| avwww免费| 午夜老司机福利剧场| 国产亚洲av片在线观看秒播厂 | 亚洲精品国产成人久久av| 亚洲一区高清亚洲精品| 大香蕉久久网| 国产成年人精品一区二区| 国产麻豆成人av免费视频| 久久久久国产网址| 国产精品av视频在线免费观看| 女人被狂操c到高潮| 日日啪夜夜撸| 国产日本99.免费观看| 一区二区三区高清视频在线| 日本欧美国产在线视频| 国产成人精品久久久久久| 天天躁日日操中文字幕| 看十八女毛片水多多多| 嫩草影院入口| 欧美日韩综合久久久久久| 观看美女的网站| 久久亚洲国产成人精品v| 久久精品综合一区二区三区| 少妇的逼水好多| 99在线人妻在线中文字幕| 老师上课跳d突然被开到最大视频| 26uuu在线亚洲综合色| 亚洲av二区三区四区| 嫩草影院精品99| 国产av麻豆久久久久久久| 又粗又硬又长又爽又黄的视频 | 日韩欧美精品v在线| 国产日本99.免费观看| 亚洲av中文av极速乱| 国产麻豆成人av免费视频| 99久久久亚洲精品蜜臀av| 在现免费观看毛片| 一区二区三区高清视频在线| 国产黄色视频一区二区在线观看 | 搞女人的毛片| 亚洲精华国产精华液的使用体验 | 一本久久中文字幕| 免费看光身美女| 色噜噜av男人的天堂激情| 能在线免费看毛片的网站| 久久久久久久亚洲中文字幕| 男女啪啪激烈高潮av片| 91麻豆精品激情在线观看国产| 午夜精品在线福利| 亚洲成av人片在线播放无| 日韩制服骚丝袜av| 99热全是精品| 在线天堂最新版资源| 国国产精品蜜臀av免费| 男人狂女人下面高潮的视频| 夜夜看夜夜爽夜夜摸| a级毛片a级免费在线| 欧美高清性xxxxhd video| 日韩一本色道免费dvd| 亚洲av一区综合| av在线老鸭窝| 亚洲成人中文字幕在线播放| 在线免费十八禁| 网址你懂的国产日韩在线| 欧美性猛交黑人性爽| 国产av一区在线观看免费| 中文欧美无线码| 国产精品久久久久久精品电影小说 | 成年av动漫网址| 久久这里有精品视频免费| ponron亚洲| 狠狠狠狠99中文字幕| 中文字幕制服av| av.在线天堂| 亚洲国产欧美在线一区| 亚洲国产精品成人久久小说 | 九色成人免费人妻av| 亚洲欧美精品综合久久99| 欧美精品一区二区大全| 欧美在线一区亚洲| av专区在线播放| 99热网站在线观看| 欧美色欧美亚洲另类二区| 国产精品久久久久久久电影| а√天堂www在线а√下载| 亚洲丝袜综合中文字幕| 国产 一区精品| 精品人妻熟女av久视频| av天堂中文字幕网| 欧美性感艳星| 欧美潮喷喷水| 91麻豆精品激情在线观看国产| 性欧美人与动物交配| 国产乱人视频| 国内精品宾馆在线| 99久久九九国产精品国产免费| a级毛片a级免费在线| 国产精品不卡视频一区二区| 日本爱情动作片www.在线观看| 国产成人freesex在线| 亚洲国产精品成人综合色| 久久久久国产网址| 一级毛片我不卡| 国产成人精品久久久久久| 亚洲欧美中文字幕日韩二区| 欧美又色又爽又黄视频| 国产精品蜜桃在线观看 | 免费看av在线观看网站| 午夜a级毛片| 久久精品国产亚洲av天美| 免费电影在线观看免费观看| 大又大粗又爽又黄少妇毛片口| 国国产精品蜜臀av免费| 色5月婷婷丁香| 我要看日韩黄色一级片| 国产精品乱码一区二三区的特点| 成人午夜高清在线视频| 国产熟女欧美一区二区| 国产精品美女特级片免费视频播放器| 亚洲欧美精品专区久久| 国产精品久久久久久久久免| 亚洲av男天堂| 美女被艹到高潮喷水动态| 97在线视频观看| 久久这里有精品视频免费| 国产视频首页在线观看| 黄色配什么色好看| 免费大片18禁| 午夜免费男女啪啪视频观看| 亚洲国产欧美人成| 日本熟妇午夜| 午夜免费激情av| 国产精品久久久久久久电影| 美女内射精品一级片tv| 国产亚洲精品久久久com| 亚洲在线自拍视频| 欧美+亚洲+日韩+国产| 春色校园在线视频观看| 内射极品少妇av片p| 变态另类丝袜制服| 久久精品国产亚洲av天美| 自拍偷自拍亚洲精品老妇| 能在线免费观看的黄片| 永久网站在线| 午夜免费激情av| 波野结衣二区三区在线| 在线观看美女被高潮喷水网站| 天堂av国产一区二区熟女人妻| 草草在线视频免费看| 日本撒尿小便嘘嘘汇集6| 日韩欧美 国产精品| 两个人视频免费观看高清| 午夜免费男女啪啪视频观看| 免费不卡的大黄色大毛片视频在线观看 | 美女脱内裤让男人舔精品视频 | 黄色配什么色好看| 一本久久精品| 91午夜精品亚洲一区二区三区| 亚洲一区高清亚洲精品| 久久精品夜夜夜夜夜久久蜜豆| 精品久久久久久久久久久久久| 一边亲一边摸免费视频| 伦精品一区二区三区| 午夜福利在线观看吧| 一区二区三区高清视频在线| 国产精品伦人一区二区| 日日撸夜夜添| 小蜜桃在线观看免费完整版高清| 丝袜美腿在线中文| 不卡一级毛片| 一级av片app| 精品一区二区免费观看| 赤兔流量卡办理| 69av精品久久久久久| 高清午夜精品一区二区三区 | 亚洲美女视频黄频| 免费观看精品视频网站| 日韩成人av中文字幕在线观看| 精品不卡国产一区二区三区| 亚洲人与动物交配视频| 日韩精品有码人妻一区| 亚洲丝袜综合中文字幕| 久久6这里有精品| 国产成人精品久久久久久| 国产视频内射| 精品一区二区三区视频在线| 日本色播在线视频| 国产日韩欧美在线精品| 亚洲av一区综合| 人体艺术视频欧美日本| 人妻系列 视频| 久久亚洲精品不卡| 看片在线看免费视频| 亚洲aⅴ乱码一区二区在线播放| 寂寞人妻少妇视频99o| 欧美丝袜亚洲另类| 菩萨蛮人人尽说江南好唐韦庄 | 岛国毛片在线播放| 1024手机看黄色片| 国产三级在线视频| 最近最新中文字幕大全电影3| 97在线视频观看| 亚洲色图av天堂| 午夜爱爱视频在线播放| 国内揄拍国产精品人妻在线| 99久久精品一区二区三区| 99国产精品一区二区蜜桃av| 国产精品一区二区性色av| 国产精品伦人一区二区| 欧美精品国产亚洲| 人妻制服诱惑在线中文字幕| 麻豆国产av国片精品| 美女大奶头视频| 国产色爽女视频免费观看| 美女xxoo啪啪120秒动态图| 在线a可以看的网站| 少妇熟女欧美另类| 日日撸夜夜添| 亚洲性久久影院| 99riav亚洲国产免费| 只有这里有精品99| 午夜久久久久精精品| 听说在线观看完整版免费高清| 一个人看的www免费观看视频| 亚洲av免费高清在线观看| 国产 一区精品| 国内精品久久久久精免费| 岛国在线免费视频观看| 久久久久久九九精品二区国产| 精品久久久久久成人av| 一级毛片电影观看 | 又粗又爽又猛毛片免费看| 成年av动漫网址| 丰满乱子伦码专区| 波野结衣二区三区在线| 日本撒尿小便嘘嘘汇集6| 99久久中文字幕三级久久日本| 久久亚洲国产成人精品v| 久久久久九九精品影院| 久久99蜜桃精品久久| 国产成人精品一,二区 | 99视频精品全部免费 在线| 国产黄色视频一区二区在线观看 | 午夜免费激情av| 日韩欧美精品v在线| 中文字幕久久专区| 亚洲国产精品成人综合色| 日韩av不卡免费在线播放| 尤物成人国产欧美一区二区三区| 一级二级三级毛片免费看| 亚洲第一区二区三区不卡| 国产精品久久久久久久久免| 麻豆国产97在线/欧美| 婷婷精品国产亚洲av| 日韩一区二区三区影片| 久久中文看片网| 伦精品一区二区三区| 99在线视频只有这里精品首页| 少妇丰满av| 赤兔流量卡办理| 高清毛片免费观看视频网站| 亚洲欧美精品专区久久| 青春草视频在线免费观看| av卡一久久| av在线观看视频网站免费| 亚洲在线自拍视频| 色播亚洲综合网| 欧美+亚洲+日韩+国产| 国产大屁股一区二区在线视频| 日韩成人av中文字幕在线观看| 一区二区三区高清视频在线| 久久精品国产鲁丝片午夜精品| 欧美性猛交╳xxx乱大交人| 寂寞人妻少妇视频99o| 精品久久久久久成人av| 舔av片在线| 国内精品美女久久久久久| 蜜桃久久精品国产亚洲av| 国产精品精品国产色婷婷| 身体一侧抽搐| 日本在线视频免费播放| 免费看美女性在线毛片视频| 亚洲一区二区三区色噜噜| 99热这里只有精品一区| 欧美性猛交╳xxx乱大交人| 身体一侧抽搐| 直男gayav资源| 一个人免费在线观看电影| 国产精品综合久久久久久久免费| 高清在线视频一区二区三区 | 国产亚洲精品久久久com| 男人舔女人下体高潮全视频| 91精品一卡2卡3卡4卡| 亚洲av免费高清在线观看| 热99在线观看视频| 欧美高清成人免费视频www| 久久午夜亚洲精品久久| 婷婷六月久久综合丁香| 欧美人与善性xxx| 最近2019中文字幕mv第一页| 免费电影在线观看免费观看| 黄色配什么色好看| 人妻系列 视频| 国产av不卡久久| 人体艺术视频欧美日本|