A traffic capacity analysis method and system based on driver trust probability

Abstract

The present invention discloses a capacity analysis method and system based on driver trust probability, comprising the following steps: obtaining vehicle types, the vehicle types including manually driven vehicles and intelligent vehicles, and setting basic vehicle parameters and a following model based on the vehicle types; setting the trust type of the driver of the manually driven vehicle in the intelligent vehicle, and configuring the following model parameters of the manually driven vehicle based on the trust type; obtaining the number of vehicles in a queue at a signal-controlled intersection, the penetration rate of intelligent vehicles in the queue, and setting the queuing method of the vehicles in the queue; setting the trust probability and distrust probability of the drivers in the queue in the intelligent vehicles, and performing simulation based on the penetration rate of intelligent vehicles, the trust probability and distrust probability of the drivers, the basic vehicle parameters, the following model, and the following model parameters to obtain the expected capacity and the capacity improvement rate.

Description

Technical Field

[0001] The present invention belongs to the technical field of mixed traffic, and in particular relates to a method and system for analyzing traffic capacity based on driver trust probability. Background Art

[0002] With the rapid development of my country's economic and technological level, the mixed traffic of vehicles with different levels of intelligence will become a reality, and the penetration rate of intelligent vehicles will continue to rise, but this change will not be completed in a short period of time, and the traffic state of intelligent vehicles and manually driven vehicles will be mixed for a long time. Due to the different types of trust drivers have in the technical level of intelligent vehicles, they will take different behavioral measures to follow the intelligent vehicles. For example, trusting drivers will follow the intelligent vehicles in front with a smaller distance, and untrusting drivers will follow the intelligent vehicles in front with a larger distance, which will cause errors between the actual traffic capacity and the calculated traffic capacity, affecting the actual traffic control measures.

[0003] Especially at signal-controlled intersections, since intersections have always been the bottleneck of urban traffic, when intelligent vehicles and manually driven vehicles mix, it is very easy to affect the traffic capacity at the intersection, causing road congestion and other problems. Therefore, it is necessary to design an analysis method that can analyze the traffic capacity of signal-controlled intersections based on the different probabilities of trust in intelligent vehicles by drivers in the queue, so as to more accurately design control measures and road channelization solutions for signal-controlled intersections. Summary of the invention

[0004] In view of the above problems, the present invention designs a capacity analysis method and system based on driver trust probability.

[0005] In a first aspect, the present invention designs a traffic capacity analysis method based on driver trust probability, comprising the following steps:

[0006] Obtaining a vehicle type, which includes a manually driven vehicle and an intelligent vehicle, and setting basic vehicle parameters and a following model based on the vehicle type, wherein the basic vehicle parameters include acceleration, maximum vehicle speed, start reaction time, and maximum deceleration;

[0007] Set the trust type of the driver of the manually driven vehicle towards the intelligent vehicle, and configure the parameters of the following model of the manually driven vehicle based on the trust type;

[0008] Obtain the number of vehicles in the queue at the signal-controlled intersection, the penetration rate of smart vehicles in the queue, and set the queueing mode of the vehicles in the queue;

[0009] The trust probability and distrust probability of drivers in the queue towards the intelligent vehicles are set, and simulation is performed based on the penetration rate of intelligent vehicles, the trust probability and distrust probability of drivers, the basic parameters of vehicles, the car-following model, and the car-following model parameters to obtain the expected capacity and capacity improvement rate.

[0010] Compared with the prior art, the beneficial effects of the present invention are as follows: the trust probability is the probability that the driver of the manually driven vehicle in the queue is a trusting driver, the distrust probability is the probability that the driver of the manually driven vehicle in the queue is a distrusting driver, and the indifference probability is the probability that the driver of the manually driven vehicle in the queue is an indifference driver. The trust probability, distrust probability and indifference probability of the drivers of the manually driven vehicles in the queue are set at a specific penetration rate, and after simulation and calculation, the traffic capacity expectation and traffic capacity improvement rate of the signal-controlled intersection corresponding to different trust probabilities at a specific penetration rate can be obtained.

[0011] The present invention incorporates the driver's trust type and trust probability in intelligent vehicles into the calculation system of the traffic capacity of signal-controlled intersections. In the future mixed traffic environment, it is beneficial to improve the analysis and calculation accuracy of the traffic capacity at signal-controlled intersections, provide assistance for signal timing and road channelization at signal-controlled intersections, and provide a theoretical basis for future traffic flow guidance, improve the traffic capacity at intersections, avoid traffic jams, and cause traffic accidents.

[0012] Preferably, the trust types include trust type, indifference type and distrust type.

[0013] A trust level is set for a trusting driver, and a distrust level is set for a distrusting driver; and parameters of a car-following model of a manually driven vehicle are configured based on the trust level and the distrust level.

[0014] The beneficial effects of this preferred solution are as follows: trusting drivers have a trusting attitude towards the technical level of smart vehicles, and tend to maintain a smaller headway or minimum vehicle distance during the process of following smart vehicles; distrusting drivers have a distrusting attitude towards the technical level of smart vehicles, and tend to maintain a larger headway or minimum vehicle distance during the process of following smart vehicles to ensure that there is enough distance to ensure their own safety; indifferent drivers regard smart vehicles as regular vehicles, and tend not to maintain a smaller headway or minimum vehicle distance, nor seek a larger safety distance. Therefore, the following model parameters of the indifferent manually driven vehicle will not change due to the type of vehicle in front;

[0015] The quantitative index of the degree of trust of the trusting driver in the technical level of the intelligent vehicle is called the degree of trust. Under different degrees of trust, the trusting driver will adopt different headway time or minimum vehicle distance when following the intelligent vehicle in front. When the trust level is low, although the driver has a trusting attitude towards the intelligent vehicle, he will not change his driving state and still maintain the same headway time or minimum vehicle distance as the indifferent driver. When the trust level is high, the driver will maintain a smaller headway time or minimum vehicle distance with the intelligent vehicle in front. When the trust level rises to a certain critical point, the driver will maintain the minimum headway time or minimum vehicle distance with the intelligent vehicle in front within a safe range, and will not change with the increase of the trust level again. Similarly, the quantitative index of the degree of trust of the distrusting driver in the technical level of the intelligent vehicle is called the degree of distrust. Under different degrees of distrust, the distrusting driver will adopt different headway time or minimum vehicle distance when following the intelligent vehicle in front.

[0016] Preferably, the following model parameters include expected headway; the relationship between the driver's trust in the smart vehicle and the following model parameters is:

[0017]

[0018] Where t(x) is the expected headway time of following the intelligent vehicle ahead when the trust level of the trusting driver is x, t 漠 is the expected headway distance when the indifferent driver follows the smart vehicle, a max is the maximum rate of change of the expected headway, -0.32≤a max ≤0, σ1 is the model parameter;

[0019] The probability density function of the trust level x in the value range of [0%, 100%] is:

[0020]

[0021] Wherein, f(x) represents the probability density of a trusting driver at a trust level of x, σ2 is the standard deviation of the trust level distribution, and α is the correction coefficient. The calculation formula of α is:

[0022]

[0023] The beneficial effect of this preferred solution is: calculating the relationship between the headway time when the driver chooses to follow and the degree of trust is conducive to more accurately configuring the following model parameters for the trust-based manually driven vehicle, and preparing for subsequent simulation experiments.

[0024] Preferably, the relationship between the driver's distrust of the smart vehicle and the parameters of the car-following model is:

[0025]

[0026] Where t(x) is the expected headway time of the following vehicle selected by the distrustful driver when the distrust level of the intelligent vehicle ahead is x, t 漠 is the expected headway distance when the indifferent driver follows the smart vehicle in front, a max is the expected maximum rate of change of headway, 0≤a max ≤2.3, σ3 is the model parameter;

[0027] The probability density function of the distrust degree x in the value range of [0%, 100%] is:

[0028]

[0029] Where f(x) represents the probability density of a distrustful driver at a distrust level of x, σ4 is the standard deviation of the distrust level distribution, and α is the correction coefficient, which is used to set the integral result of the distrust level probability density to 1.

[0030] The calculation formula of α is:

[0031] The beneficial effect of this preferred solution is: calculating the relationship between the headway time when the driver chooses to follow and the degree of distrust is conducive to more accurately configuring the following model parameters for the distrustful manually driven vehicle, and preparing for subsequent simulation experiments.

[0032] Preferably, the formula for obtaining the number N of vehicles in the queue at the signal-controlled intersection is: N=λ*(Cg)=λ*(CGY-r+L),

[0033] Among them, λ is the vehicle arrival rate at the signal-controlled intersection, g is the effective green light time, C is the cycle duration of the signal-controlled intersection cycle timing plan, G is the actual green light display time of the phase, Y is the actual yellow light clearing time, r is the signal light full red time, and L is the total loss time of the phase.

[0034] The beneficial effect of this preferred solution is that the number N of signal-controlled intersection queues is obtained through the above formula, which is conducive to determining the number of intelligent vehicles and manually driven vehicles when specifying the penetration rate of intelligent vehicles, and then determining the vehicle queuing method.

[0035] Preferably, the calculation formula of λ is:

[0036]

[0037] Among them, G sis the green light time of the upstream intersection through phase of the signalized intersection, G l is the green light time of the left turn phase at the upstream intersection, λ r is the arrival rate of right-turn vehicles at the upstream intersection, λ s is the arrival rate of through vehicles at the upstream intersection, λ l is the arrival rate of left-turn vehicles at the upstream intersection;

[0038] The lambda r The calculation formula is:

[0039] Among them, the q r is the flow rate of right-turn traffic at the upstream intersection;

[0040] The lambda s The calculation formula is:

[0041] Among them, the q s is the flow rate of through traffic at the upstream intersection, and S s is the saturation flow rate of through traffic at the upstream intersection, The green light start time for the upstream intersection through phase. The green light end time for the straight phase at the upstream intersection;

[0042] The lambda l The calculation formula is:

[0043] Among them, the q l is the flow rate of left-turn traffic at the upstream intersection, and S l is the saturation flow rate of left-turn traffic at the upstream intersection, The green light start time for the left turn phase at the upstream intersection. The green light end time for the left turn phase at the upstream intersection.

[0044] The beneficial effects of this preferred solution are as follows: λ is related to the leaving traffic flow of the upstream intersection, which can be regarded as related to the combined traffic flow formed by the straight, left and right turns of the upstream intersection. Since the right-turning traffic flow of the upstream intersection is not controlled by the signal light, the right-turning traffic flow leaves the intersection at a constant flow rate, which is a constant arrival rate;

[0045] The process of traffic flow controlled by signals leaving the upstream intersection is divided into two stages: the beginning of the green light time and the end of the green light time. In the beginning of the green light time, vehicles will leave the upstream intersection at a saturated flow rate, and in the end of the green light time, vehicles will leave the upstream intersection at the entrance arrival rate. By calculating the arrival rates of straight, left-turn, and right-turn traffic at the upstream intersection, the combined arrival rate of the desired intersection can be calculated, which is conducive to calculating the number of queues at the signal-controlled intersection.

[0046] Preferably, based on the number N of vehicles in the queue at the signal-controlled intersection and the penetration rate P of smart vehicles in the queue, the queuing mode of the vehicles is set to a hybrid mode:

[0047] If the penetration rate of intelligent vehicles in the queue is P<0.5, then each of the intelligent vehicles will be followed by a manually driven vehicle;

[0048] If the penetration rate of intelligent vehicles in the queue is P = 0.5, then each of the intelligent vehicles is followed by a human-driven vehicle;

[0049] If the penetration rate P of intelligent vehicles in the queue is greater than 0.5, each of the manually driven vehicles will follow the intelligent vehicle.

[0050] The beneficial effects of this preferred solution are: selecting a mixed queuing method, when the penetration rate of intelligent vehicles is <0.5, is conducive to ensuring that each intelligent vehicle is followed by at least one manually-driven vehicle, and the intelligent vehicles do not follow each other; when the penetration rate of intelligent vehicles is ≥0.5, it is conducive to ensuring that each manually-driven vehicle is followed by an intelligent vehicle, and the manually-driven vehicles do not follow each other; therefore, no matter what the penetration rate of intelligent vehicles is, the mixed queuing method can maximize the effect of the driver's trust probability and distrust probability on the fleet's traffic capacity.

[0051] Preferably, when the penetration rate of smart vehicles P<0.5, the calculation formula of the expected traffic capacity E(ω) is:

[0052]

[0053] Where E(ω) is the expected capacity of the signal-controlled intersection under a specific intelligent vehicle penetration rate and driver trust probability, N is the number of vehicles in the queue at the intersection, P is the penetration rate of intelligent vehicles in the queue, i is the number of manually driven vehicles that are trustworthy and follow intelligent vehicles at the same time, j is the number of manually driven vehicles that are untrustworthy and follow intelligent vehicles at the same time, P1, P2, and P3 are the driver's trust probability, untrust probability, and indifference probability, respectively, and P1+P2+P3=1, P4=1-P3, ω(i,j) is the capacity of the signal-controlled intersection corresponding to any set of parameters i and j under a specific penetration rate P;

[0054] Based on the calculation formula of expected capacity E(ω) and the driver's trust probability and distrust probability, the simulation system is run to output the expected capacity at the signalized intersection when the permeability P<0.5;

[0055] When the penetration rate of intelligent vehicles P ≥ 0.5, the calculation formula of the expected traffic capacity E(ω) is:

[0056]

[0057] Where E(ω) is the expected capacity of the signal-controlled intersection under a specific intelligent vehicle penetration rate and driver trust probability, N is the number of vehicles in the queue at the intersection, P is the penetration rate of intelligent vehicles in the queue, i is the number of manually driven vehicles that are trustworthy and follow intelligent vehicles at the same time, j is the number of manually driven vehicles that are untrustworthy and follow intelligent vehicles at the same time, and ω(i,j) is the capacity of the signal-controlled intersection corresponding to any set of parameters i and j under a specific penetration rate P;

[0058] Based on the calculation formula of the expected capacity E(ω) and the driver's trust probability and distrust probability, the simulation system is run to output the expected capacity at the signal-controlled intersection when the permeability P≥0.5.

[0059] The beneficial effects of this preferred solution are: when the penetration rate of smart vehicles is less than 0.5, the mixed queuing method has multiple queuing forms, and it cannot be guaranteed that both trusted or untrusted manually driven vehicles can follow the smart vehicles. Therefore, under specific penetration rates and trust probabilities, there will be corresponding multiple capacities. Therefore, it is necessary to calculate the expected capacity to obtain the capacity under specific penetration rates and trust probabilities, thereby calculating the corresponding capacity improvement rate.

[0060] Similarly, when the penetration rate of smart vehicles is greater than 0.5, the mixed queuing mode has multiple queuing forms, so under a specific penetration rate and trust probability, there will also be multiple corresponding capacities. Therefore, it is necessary to calculate the expected capacity to obtain the capacity of the signal-controlled intersection under a specific penetration rate and trust probability.

[0061] Preferably, at a specific permeability, the functional relationship between the output capacity improvement rate y and the trust probability x1 and the distrust probability x2 is: y=cx1+dx2+k;

[0062] Obtain the coefficient c of the trust probability x1 under different penetration rates, and output the functional relationship between the coefficient c and the penetration rate of smart vehicles;

[0063] According to the functional relationship between coefficient c and the penetration rate of intelligent vehicles in the queue, based on the penetration rate of intelligent vehicles in the queue at the actual signal-controlled intersection, the coefficient c that represents the influence of the driver's trust probability on the traffic capacity is obtained.

[0064] The beneficial effects of this preferred solution are as follows: after obtaining the capacity improvement rate of the signal-controlled intersection under different trust probabilities, a linear regression is performed on the operation results to obtain the functional relationship between the capacity improvement rate y and the trust probability x1 and the distrust probability x2, wherein the coefficient of the trust probability x1 is the degree of influence of the trust probability on the capacity; at different penetration rates, the simulation results are output to obtain the coefficient of the trust probability x1 under different penetration rates, and then the functional relationship between the coefficient and the penetration rate can be calculated, so that the degree of influence of the trust probability of the drivers of manually driven vehicles in the queue on the capacity at different penetration rates at the signal-controlled intersection can be calculated.

[0065] In a second aspect, the present invention further provides a traffic capacity analysis system based on driver trust probability, comprising:

[0066] The parameter configuration module is configured to obtain the vehicle type, which includes manually driven vehicles and intelligent vehicles; based on the vehicle type, set the basic vehicle parameters and the following model; the basic vehicle parameters include acceleration, maximum vehicle speed, start reaction time, and maximum deceleration; set the trust type of the driver of the manually driven vehicle in the intelligent vehicle, which includes trust type, indifference type, and distrust type; based on the trust type, configure the following model parameters of the manually driven vehicle; preferably, the parameter configuration module includes:

[0067] A car-following model parameter configuration unit configured to set a trust level for a trusting driver, set a distrust level for a distrusting driver, and configure car-following model parameters for a manually driven vehicle based on the trust level and the distrust level;

[0068] A queuing mode setting module is configured to obtain the number of vehicles in a queue at a signal-controlled intersection, the penetration rate of intelligent vehicles in the queue, and set a queuing mode for the vehicles in the queue;

[0069] A simulation module is configured to set the trust probability and distrust probability of drivers in the queue, and simulate based on the penetration rate of intelligent vehicles, the trust probability and distrust probability of drivers, basic vehicle parameters, a car-following model, and car-following model parameters to obtain the expected capacity and the capacity improvement rate;

[0070] The function relationship output module is configured to output the function relationship y=cx1+dx2+k between the capacity improvement rate y and the trust probability x1 and the distrust probability x2 under a specific penetration rate; obtain the coefficient c of the trust probability x1 under different penetration rates, output the function relationship between the coefficient c and the penetration rate of smart vehicles, and according to the function relationship between the coefficient c and the penetration rate of smart vehicles in the queue, based on the actual penetration rate of smart vehicles in the queue at the signal-controlled intersection, derive the coefficient c that represents the degree of influence of the driver's trust probability on the capacity.

[0071] Compared with the prior art, the beneficial effects of the present invention are as follows: through the simulation experiment conducted by the system, it is helpful to calculate the influence of the penetration rate of intelligent vehicles in the mixed queue and the trust probability of drivers of manually driven vehicles on the traffic capacity at the signal-controlled intersection. In the future mixed traffic environment, it is helpful to improve the analysis and calculation accuracy of the traffic capacity at the signal-controlled intersection, provide assistance for the signal timing and road channelization at the signal-controlled intersection, and provide a theoretical basis for the future traffic flow guidance, so as to improve the traffic capacity at the intersection and avoid traffic jams and traffic accidents. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] Figure 1 Flow chart of the capacity analysis method according to an embodiment of the present invention;

[0073] Figure 2 This is a schematic diagram of an upstream intersection according to an embodiment of the present invention;

[0074] Figure 3 The vehicle queuing mode under a specific penetration rate of an embodiment of the present invention;

[0075] Figure 4 This is the vehicle queuing mode when the penetration rate is 0.4 in the embodiment of the present invention;

[0076] Figure 5 This is the vehicle queuing mode when the penetration rate is 0.6 in the embodiment of the present invention; DETAILED DESCRIPTION

[0077] In order to make the objectives, technical solutions and advantages of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the embodiments of the present invention.

[0078] Example 1

[0079] This embodiment provides a traffic capacity analysis method based on driver trust probability, such as Figure 1 As shown, the following steps are included:

[0080] Step 1: Obtain the vehicle type, which includes manually driven vehicles and intelligent vehicles. The vehicle type can be obtained through field surveys or retrieved from a third-party database. The specific acquisition method is not limited in this embodiment. Based on the vehicle type, set the basic vehicle parameters and the following model; the basic vehicle parameters include acceleration, maximum vehicle speed, start-up reaction time, maximum deceleration, etc. In this embodiment, the manually driven vehicle adopts the IDM following model, and the intelligent vehicle includes a smart car or a networked car, and adopts the ACC following model.

[0081] Step 2: Obtain the trust type of the driver of the manually driven vehicle, and configure the parameters of the following model of the manually driven vehicle based on the trust type; preferably, in this embodiment, the trust type includes trust type, distrust type and indifference type, and the trust level is set for the trust type driver and the distrust level is set for the distrust type driver; configure the parameters of the following model of the manually driven vehicle based on the trust level and the distrust level. The following model parameters include the minimum vehicle-to-vehicle distance, acceleration index, expected headway, etc. The relationship between the trust level and the following model parameters is:

[0082]

[0083] Where t(x) is the expected headway time of following the intelligent vehicle ahead when the trust level of the trusting driver is x, t 漠 is the expected headway distance when the indifferent driver follows the smart vehicle, a max is the maximum rate of change of the expected headway, -0.32≤a max ≤0, σ1 is the model parameter, and its value range is (0,0.16],;

[0084] The probability density function of the trust level x in the value range of [0%, 100%] is:

[0085]

[0086] Among them, f(x) represents the probability density of a trusting driver at a trust level of x, σ2 is the standard deviation of the trust level distribution, and α is the correction coefficient, which is used to set the integral result of the trust level probability density to 1.

[0087] The calculation formula of α is:

[0088] Similarly, the relationship between the degree of distrust and the parameters of the car-following model is:

[0089]

[0090] Where t(x) is the expected headway time of the following vehicle selected by the distrustful driver when the distrust level of the intelligent vehicle ahead is x, t 漠 is the expected headway distance when the indifferent driver follows the smart vehicle in front, a max is the expected maximum rate of change of headway, 0≤a max ≤2.3, σ3 is the model parameter;

[0091] The probability density function of the distrust degree x in the value range of [0%, 100%] is:

[0092]

[0093] Where f(x) represents the probability density of a distrustful driver at a distrust level of x, σ4 is the standard deviation of the distrust level distribution, and α is the correction coefficient, which is used to set the integral result of the distrust level probability density to 1.

[0094] The calculation formula of α is:

[0095] Step 3: Obtain the number N of vehicles in the queue at the signal-controlled intersection, the penetration rate P of smart vehicles in the queue, and set the queuing mode of the vehicles; preferably, the formula for obtaining the number N of the queue at the signal-controlled intersection is:

[0096] N = λ*(Cg) = λ*(CGY-r+L),

[0097] Among them, λ is the vehicle arrival rate at the signal-controlled intersection, g is the effective green light time, C is the cycle duration of the signal-controlled intersection cycle timing plan, G is the actual green light display time of the phase, Y is the actual yellow light clearing time, r is the signal light full red time, and L is the total loss time of the phase.

[0098] like Figure 2 As shown, λ is related to the traffic flow leaving the upstream intersection, and can be regarded as being related to the combined traffic flow formed by the upstream intersection going straight, turning left, and turning right. The calculation formula of λ is:

[0099]

[0100] Among them, G s is the green light time of the upstream intersection through phase of the signalized intersection, G l is the green light time of the left turn phase at the upstream intersection, λ r is the arrival rate of right-turn vehicles at the upstream intersection, λ s is the arrival rate of through vehicles at the upstream intersection, λ l is the arrival rate of left-turn vehicles at the upstream intersection;

[0101] The lambda r The calculation formula is:

[0102] Among them, the q r is the flow rate of right-turn traffic at the upstream intersection;

[0103] The lambda s The calculation formula is:

[0104] Among them, the q s is the flow rate of through traffic at the upstream intersection, and S s is the saturation flow rate of through traffic at the upstream intersection, The green light start time for the upstream intersection through phase. The green light end time for the straight phase at the upstream intersection;

[0105] The lambda l The calculation formula is:

[0106] Among them, the q l is the flow rate of left-turn traffic at the upstream intersection, and S l is the saturation flow rate of left-turn traffic at the upstream intersection, The green light start time for the left turn phase at the upstream intersection. The saturation flow rate described in this embodiment refers to the number of vehicles per hour obtained by equivalently converting the traffic volume in a time period of less than 1 hour (such as 5 minutes, 15 minutes). When the traffic light turns green, the traffic flow originally waiting behind the stop line begins to move forward, and the vehicles cross the stop line in a line, and the flow rate quickly increases from 0 to a stable value, that is, the saturation flow rate.

[0107] Based on the number of vehicles N in the queue at the signal-controlled intersection and the penetration rate P of smart vehicles in the queue, the vehicle queuing mode is set to a hybrid type:

[0108] If the penetration rate of smart vehicles is less than 0.5, each smart vehicle is followed by a human-driven vehicle; if the penetration rate of smart vehicles in the queue is P<0.5, each smart vehicle is followed by a human-driven vehicle; if the penetration rate of smart vehicles in the queue is P=0.5, each smart vehicle is followed by a human-driven vehicle; if the penetration rate of smart vehicles in the queue is P>0.5, each human-driven vehicle is followed by a smart vehicle. This is conducive to ensuring that no matter what the penetration rate of smart vehicles is, the mixed queuing method can maximize the effect of the driver's trust probability and distrust probability on the fleet's traffic capacity.

[0109] Step 4: Set the trust probability, distrust probability, and indifference probability of the drivers in the queue towards the intelligent vehicle, where trust probability + distrust probability + indifference probability = 1.

[0110] Based on the penetration rate of intelligent vehicles, the probability of trust and distrust of drivers, basic vehicle parameters, the car-following model, and the parameters of the car-following model, simulation is performed to obtain the traffic capacity and traffic capacity improvement rate.

[0111] When the penetration rate of intelligent vehicles is P<0.5, the mixed queuing mode has multiple queuing forms, and in one queuing form, it cannot be guaranteed that every trusted or untrusted manually driven vehicle can follow the intelligent vehicle. Therefore, under a specific penetration rate and trust probability, there will be multiple corresponding capacities. Therefore, it is necessary to calculate the expected capacity. The calculation formula of the expected capacity E(ω) is:

[0112]

[0113] Where E(ω) is the expected capacity of the signal-controlled intersection under a specific intelligent vehicle penetration rate and driver trust probability, N is the number of vehicles in the queue at the intersection, P is the penetration rate of intelligent vehicles in the queue, i is the number of manually driven vehicles that are trustworthy and follow intelligent vehicles at the same time, j is the number of manually driven vehicles that are untrustworthy and follow intelligent vehicles at the same time, P1, P2, and P3 are the driver's trust probability, untrust probability, and indifference probability, respectively, and P1+P2+P3=1, P4=1-P3, ω(i,j) is the capacity of the signal-controlled intersection corresponding to any set of parameters i and j under a specific penetration rate P;

[0114] Based on the calculation formula of expected capacity E(ω) and the driver's trust probability and distrust probability, the simulation system can be run to output the expected capacity at the signalized intersection when the permeability P<0.5.

[0115] Similarly, when the penetration rate of smart vehicles is greater than 0.5, the mixed queuing mode has multiple queuing forms. Therefore, under a specific penetration rate and trust probability, there will also be multiple corresponding capacities. Therefore, it is necessary to calculate the expected capacity. The calculation formula of the expected capacity E(ω) is:

[0116]

[0117] Where E(ω) is the expected capacity of the signal-controlled intersection under a specific intelligent vehicle penetration rate and driver trust probability, N is the number of vehicles in the queue at the intersection, P is the penetration rate of intelligent vehicles in the queue, i is the number of manually driven vehicles that are trustworthy and follow intelligent vehicles at the same time, j is the number of manually driven vehicles that are untrustworthy and follow intelligent vehicles at the same time, and ω(i,j) is the capacity of the signal-controlled intersection corresponding to any set of parameters i and j under a specific penetration rate P;

[0118] Based on the calculation formula of the expected capacity E(ω) and the driver's trust probability and distrust probability, the simulation system can be run to output the expected capacity at the signal-controlled intersection when the permeability P≥0.5.

[0119] Step 5: Under a specific permeability, output the functional relationship between the capacity improvement rate y and the trust probability x1 and the distrust probability x2: y=cx1+dx2+k. In this embodiment, preferably, a number of capacity expectations and capacity improvement rates under a specific permeability can be obtained through simulation. The calculation method of the capacity improvement rate is: all drivers in the queue are set to be indifferent, and the basic capacity expectation at the intersection when the trust probability and distrust probability are 0 can be simulated. Based on the basic capacity expectation and the capacity expectation corresponding to the specific trust probability and distrust probability, the capacity improvement rate under the specific trust probability and distrust probability can be calculated.

[0120] Step 6: Obtain the coefficient c of the trust probability x1 under different penetration rates, and output the functional relationship between the coefficient c and the penetration rate of smart vehicles;

[0121] According to the functional relationship between coefficient c and the penetration rate of intelligent vehicles in the queue, based on the penetration rate of intelligent vehicles in the queue at the actual signal-controlled intersection, the coefficient c that represents the degree of influence of driver trust probability on traffic capacity can be obtained. Therefore, according to the functional relationship between traffic capacity improvement rate y and trust probability x1: y = cx1, the traffic capacity improvement rate y at the signal-controlled intersection can be calculated.

[0122] Example 2

[0123] This embodiment further illustrates the traffic capacity analysis method of a signal-controlled intersection through an actual simulation experiment, which includes the following steps:

[0124] Step 1: Get the vehicle type, which includes manually driven vehicles (MV) and intelligent vehicles. The intelligent vehicles include non-connected autonomous vehicles (AV) or connected autonomous vehicles (CAV). This embodiment uses MV and AV as examples. Based on the vehicle type, set the basic vehicle parameters and the following model; the basic vehicle parameters include acceleration, maximum speed, start reaction time, maximum deceleration, etc.

[0125] Setting up an Intelligent Driver Model (IDM) for a Human-Driven Vehicle (MV):

[0126]

[0127] Among them, a is the maximum acceleration, v0 is the free flow speed, s0 is the minimum stopping distance, T is the safe headway, b is the comfortable deceleration, Δv is the speed difference between adjacent vehicles, and α is the acceleration index.

[0128] The ACC model calibrated by the University of California, Berkeley's path laboratory through real vehicles is set for the intelligent vehicle (AV). This model can better reflect the following characteristics of the current commercial ACC control system applied to real vehicles:

[0129]

[0130] Where T is the expected headway, k1 and k2 are the gap control and speed control coefficients respectively, e i,k is the gap error between the actual distance between the rear vehicle i and the front vehicle i-1 at time k and the expected distance, v i-1,k-1 is the speed of the preceding vehicle at time k-1, v i,k-1 is the speed of the following vehicle at time k-1, d0 is the gap margin between the leading and following vehicles, T is the expected headway, x i-1,k-1 is the position of the preceding vehicle at time k-1, x i,k-1 is the position of the following vehicle at time k-1.

[0131] The calibration results of path laboratory for k1 and k2 are 0.23 and 0.07, the value of T is 1.1s, the minimum vehicle distance is 1.5m, and the maximum vehicle speed is 14m / s.

[0132] Step 2: Obtain the trust type of the MV driver, and configure the parameters of the following model of the manually driven vehicle based on the trust type; preferably, in this embodiment, the trust type includes trust type, distrust type and indifference type, and the trust level is set for the trust type driver, and the distrust level is set for the distrust type driver; configure the parameters of the following model of the manually driven vehicle based on the trust type, trust level and distrust level. The following model parameters include the minimum vehicle-to-vehicle distance, acceleration index, expected headway, etc. The relationship between the trust level and the following model parameters is:

[0133]

[0134] Where t(x) is the expected headway time of following the intelligent vehicle ahead when the trust level of the trusting driver is x, t 漠 is the expected headway time for the indifferent driver to follow the smart vehicle ahead. In this embodiment, the preferred value is 2s. max is the expected maximum rate of change of headway, at this time -0.32≤a max ≤0, the lower limit -0.32 is determined according to the actual MV driver’s following headway range, σ1 is a model parameter, and its value range is (0,0.16]. Its value can be obtained according to the actual trust level of the trusting driver in the smart vehicle and the corresponding headway survey. In this embodiment, the preferred value is 0.1.

[0135] The normal distribution is used to describe the distribution of the trust level of the trusting driver. The probability density function of the trust level x in the value range of [0%, 100%] is:

[0136]

[0137] Wherein, f(x) represents the probability density of a trusting driver at a trust level of x, σ2 is the standard deviation of the trust level distribution, which can be obtained based on the statistical analysis of the trust level survey results of trusting drivers, and the preferred value in this embodiment is 0.46; α is a correction coefficient, which is used to set the integral result of the trust level probability density to 1,

[0138] The calculation formula of α is:

[0139] Similarly, the relationship between the degree of distrust and the parameters of the car-following model is:

[0140]

[0141] Where t(x) is the expected headway time of the following vehicle selected by the distrustful driver when the distrust level of the intelligent vehicle ahead is x, t 漠 is the expected headway time when the indifferent driver follows the smart vehicle ahead, which is preferably 2s in this embodiment. max is the expected maximum rate of change of headway, at this time 0≤a max ≤2.3, the upper limit of 2.3 is determined according to the actual driver's range of variation of the following headway, σ3 is a model parameter, and its value can be obtained according to the actual distrust level of the distrustful driver towards the intelligent vehicle ahead and the corresponding headway survey, and the preferred value in this embodiment is 0.1.

[0142] The normal distribution is used to describe the distribution of the distrust degree of the distrustful driver. The probability density function of the distrust degree x in the value range of [0%, 100%] is:

[0143]

[0144] Wherein, f(x) represents the probability density of the distrustful driver at the level of distrust x, σ4 is the standard deviation of the distrust level distribution, which can be obtained based on the statistical analysis of the distrust level survey results of the distrustful MV drivers, and the preferred value in this embodiment is 0.34; α is the correction coefficient, which is used to set the integral result of the probability density of the distrust level to 1,

[0145] The calculation formula of α is:

[0146] Therefore, the parameter settings of the manual driving vehicle and the intelligent vehicle in this embodiment are shown in Table 1:

[0147]

[0148] Table 1

[0149] Step 3: Obtain the number N of vehicles in the queue at the signal-controlled intersection, the penetration rate P of smart vehicles in the queue, and set the queuing method of the vehicles; the formula for obtaining the number N of queues at the signal-controlled intersection is: N=λ*(Cg)=λ*(CGY-r+L). Preferably, in this embodiment, by investigating the actual signal-controlled intersection, the actual data such as the intersection vehicle arrival rate and the effective green light time are substituted into the above formula, and the number N of queues at the signal-controlled intersection is calculated to be 10.

[0150] Based on the number of vehicles N in the queue at the signal-controlled intersection and the penetration rate P of intelligent vehicles in the queue, the queuing mode of vehicles is set to a hybrid type: Figure 3 As shown in the figure, if the penetration rate of smart vehicles is P = 0.3, each smart vehicle is followed by at least one manually driven vehicle, and the remaining manually driven vehicles are arranged together in a way with the highest aggregation intensity; if the penetration rate of smart vehicles is P = 0.5, each smart vehicle is followed by a manually driven vehicle; if the penetration rate of smart vehicles is P = 0.8, each manually driven vehicle is followed by a smart vehicle, and the remaining smart vehicles are arranged together in a way with the highest aggregation intensity. This is conducive to ensuring that no matter what the penetration rate of smart vehicles is, the mixed queuing method can maximize the effect of the driver's trust probability and distrust probability on the fleet's traffic capacity.

[0151] Step 4: Set the trust probability and disregard probability of the drivers in the queue towards the smart car AV, where trust probability + disregard probability = 1 - distrust probability.

[0152] Based on the penetration rate of intelligent vehicles, the probability of trust and distrust of drivers, basic vehicle parameters, the car-following model, and the parameters of the car-following model, simulation is performed to obtain the traffic capacity and traffic capacity improvement rate.

[0153] In this embodiment, the penetration rates of smart vehicles P=0.4 and P=0.8 are used as examples for explanation:

[0154] (1) When the penetration rate of intelligent vehicles is P = 0.4, the queueing pattern of vehicles in the queue is as follows: Figure 3 As shown,

[0155] When the penetration rate of intelligent vehicles is P<0.5, the mixed queuing mode has multiple queuing forms, and in one queuing form, it cannot be guaranteed that every trusted or untrusted manually driven vehicle can follow the intelligent vehicle. Therefore, under a specific penetration rate and trust probability, there will be multiple corresponding capacity. Therefore, it is necessary to calculate the expected capacity. The calculation formula of the expected capacity E(ω) is:

[0156]

[0157] Based on the calculation formula of the expected capacity E(ω) and the driver's trust probability and distrust probability, the simulation system can be run to output the expected capacity and capacity improvement rate at the signalized intersection when the penetration rate of intelligent vehicles P = 0.4, as shown in Table 2:

[0158] It should be noted that, since the trust probability+the indifference probability+the distrust probability=1, although the distrust probability is not described in Table 2, it can be calculated based on the trust probability and the indifference probability.

[0159]

[0160]

[0161] Table 2

[0162] (2) When the penetration rate of intelligent vehicles is P = 0.8, the queueing mode of the vehicles in the queue is as follows: Figure 4 As shown,

[0163] Since the mixed queuing mode has multiple queuing forms, it will also correspond to multiple capacities under a specific penetration rate and trust probability. Therefore, it is necessary to calculate the expected capacity. The calculation formula of the expected capacity E(ω) is:

[0164]

[0165] Based on the calculation formula of the expected capacity E(ω) and the driver's trust probability and distrust probability, the simulation system can be run to output the expected capacity and capacity improvement rate at the signalized intersection when the penetration rate of intelligent vehicles P = 0.8, as shown in Table 3:

[0166] It should be noted that, since the trust probability+the indifference probability+the distrust probability=1, although the distrust probability is not described in Table 3, it can be calculated based on the trust probability and the indifference probability.

[0167]

[0168]

[0169] Table 3

[0170] Step 5: Under a specific penetration rate, output the functional relationship between the capacity improvement rate y and the trust probability x1 and the distrust probability x2: y=cx1+dx2+k.

[0171] Preferably, in this embodiment, when the penetration rate of smart vehicles P=0.4, a linear regression is performed on the data in Table 2, with the trust probability x1 and the distrust probability x2 as independent variables and the traffic capacity improvement rate y as the dependent variable. At a significance level of 0.05, the functional relationship is obtained as follows:

[0172] y=-1.3495*10 -6 +0.0465*x1-0.0663*x2

[0173] Determine whether the above functions have linear correlation:

[0174] (1) Perform an F test. When F>F1-α(m,nm-1), it is considered that there is a significant correlation between the dependent variable and the independent variable. In this embodiment, F=2109.6>F1-α(2,25)=3.3852, which means that the above function has a significant linear correlation.

[0175] (2) p in this experiment is much less than 0.05;

[0176] (3) The determination coefficient of the above function is r2 = 0.9941, which is very close to 1. Therefore, there is a strong linear correlation between the dependent variable and the independent variable.

[0177] When the penetration rate of intelligent vehicles P = 0.8, a linear regression is performed on the data in Table 3, with the trust probability x1 and the distrust probability x2 as independent variables and the traffic capacity improvement rate y as the dependent variable. At the significance level of 0.05, the functional relationship is:

[0178] y=-0.0016+0.0542*x1-0.0706*x2

[0179] Determine whether the above functions have linear correlation:

[0180] (1) Perform an F test. When F>F1-α(m,nm-1), it is considered that there is a significant correlation between the dependent variable and the independent variable. In this embodiment, F=887.1545>F1-α(2,25)=3.8853, which means that the above function has a significant linear correlation.

[0181] (2) p in this experiment is much less than 0.05;

[0182] (3) The determination coefficient of the above function is r2 = 0.9933, which is very close to 1. Therefore, there is a strong linear correlation between the dependent variable and the independent variable.

[0183] Step 6: From step 5, we can see that there is a highly linear relationship between the driver's trust probability and the traffic capacity improvement rate, and at different penetration rates, the driver's trust probability has different influences on the traffic capacity. Through multiple experiments, the coefficient of the trust probability x1 at different penetration rates (i.e., the influence of the trust probability on the traffic capacity) is obtained, as shown in Table 4:

[0184] Smart vehicle penetration rate The influence of trust probability on traffic capacity (coefficient) 1 0 0.9 0.018511886 0.8 0.028019718 0.7 0.043173374 0.6 0.054223605 0.5 0.065985836 0.4 0.046534001 0.3 0.031253443 0.2 0.01857399 0.1 0.00951663 0 0

[0185] Table 4

[0186] Taking the penetration rate P as the independent variable and the influence degree of trust probability on traffic capacity c as the dependent variable, at the significance level of 0.05, the functional relationship between the coefficient and the penetration rate of intelligent vehicles when the penetration rate P ≥ 0.5 is output:

[0187]

[0188] Determine whether the above functions have linear correlation:

[0189] (1) Perform an F test. When F>F1-α(m,nm-1), it is considered that there is a significant correlation between the dependent variable and the independent variable. In this embodiment, F=567.00>F1-α(1,4)=7.7086, which means that the above function has a significant linear correlation.

[0190] (2) p in this experiment is much less than 0.05;

[0191] (3) The determination coefficient of the above function is r2 = 0.9930, which is very close to 1. Therefore, there is a strong linear correlation between the dependent variable and the independent variable.

[0192] In this embodiment, preferably, the permeability P is used as the independent variable, the influence degree of trust probability on traffic capacity c is used as the dependent variable, and at a significance level of 0.05, the functional relationship between the coefficient and the permeability of smart vehicles when the permeability <0.5 is output:

[0193]

[0194] Determine whether the above functions have linear correlation:

[0195] (1) Perform an F test. When F>F1-α(m,nm-1), it is considered that there is a significant correlation between the dependent variable and the independent variable. In this embodiment, F=170.6971>F1-α(1,4)=7.7086, which means that the above function has a significant linear correlation.

[0196] (2) p in this experiment is much less than 0.05;

[0197] (3) The determination coefficient of the above function is r2 = 0.9771, which is very close to 1, so the linear correlation between the dependent variable and the independent variable is very strong.

[0198] In summary, the present embodiment calculates that the influence of the driver's trust probability of the intelligent vehicle on the traffic capacity at the signalized intersection is:

[0199]

[0200] According to the functional relationship between the degree of influence of trust probability on traffic capacity c and the penetration rate P of intelligent vehicles, the penetration rate P of intelligent vehicles in the actual signal-controlled intersection queue can be used to obtain the coefficient c that characterizes the degree of influence of driver trust probability on traffic capacity. According to the functional relationship between the traffic capacity improvement rate y and the trust probability x1: y=cx1, the traffic capacity improvement rate y at the actual intersection can be obtained.

[0201] This embodiment incorporates the driver's trust probability in smart vehicles into the calculation system of the traffic capacity of signalized intersections, which is conducive to more accurate analysis and calculation of the traffic capacity of signalized intersections in future mixed traffic environments, and provides assistance for signal timing and road channelization at signalized intersections.

[0202] The basic principles, main features and advantages of the present invention are demonstrated above. Those skilled in the art should understand that the embodiments are only specific explanations of the present invention, and the present invention is not limited to the embodiments. Without departing from the scope of the present invention, the technicians may also make various changes and improvements to the present invention, and these changes and improvements are all within the scope required by the claims of this application.

Claims

1. A traffic capacity analysis method based on driver trust probability, characterized in that: The following steps are involved: Obtaining a vehicle type, which includes a manually driven vehicle and an intelligent vehicle, and setting basic vehicle parameters and a following model based on the vehicle type, wherein the basic vehicle parameters include acceleration, maximum vehicle speed, start reaction time, and maximum deceleration; Setting the trust type of the driver of the manually driven vehicle in the intelligent vehicle, the trust type includes trust type, indifference type and distrust type, setting the trust level for the trust type driver and the distrust level for the distrust type driver based on the trust type, and configuring the parameters of the following model of the manually driven vehicle based on the trust level and the distrust level; Obtain the number of vehicles in the queue at the signal-controlled intersection, the penetration rate of smart vehicles in the queue, and set the queueing mode of the vehicles in the queue; Set the probability of trust and distrust of the drivers in the queue towards the smart vehicles, and perform simulation based on the penetration rate of smart vehicles, the probability of trust and distrust of the drivers, basic vehicle parameters, the car-following model, and the car-following model parameters to obtain the expected capacity and capacity improvement rate; At a specific penetration rate, the functional relationship between the output capacity improvement rate y and the trust probability x1 and the distrust probability x2 is: y = cx1 + dx2 + k; Obtain the coefficient c of the trust probability x1 under different penetration rates, and output the functional relationship between the coefficient c and the penetration rate of smart vehicles; According to the functional relationship between coefficient c and the penetration rate of intelligent vehicles in the queue, based on the penetration rate of intelligent vehicles in the queue at the actual signal-controlled intersection, the coefficient c that represents the influence of the driver's trust probability on the traffic capacity is obtained.

2. The method for analyzing traffic capacity based on driver trust probability according to claim 1, characterized in that: The following model parameters include expected headway; The relationship between the driver's trust in the smart vehicle and the parameters of the car-following model is: Where t(x) is the expected headway time of following the intelligent vehicle ahead when the trust level of the trusting driver is x, t 漠 is the expected headway distance when the indifferent driver follows the smart vehicle, a max is the maximum rate of change of the expected headway, -0.32≤a max ≤0, σ1 is the model parameter; The probability density function of the trust level x in the value range of [0%, 100%] is: Wherein, f(x) represents the probability density of a trusting driver at a trust level of x, σ2 is the standard deviation of the trust level distribution, and α is the correction coefficient. The calculation formula of α is: The relationship between the driver's distrust of the intelligent vehicle and the parameters of the car-following model is: Where t(x) is the expected headway time of the following vehicle selected by the distrustful driver when the distrust level of the intelligent vehicle ahead is x, t 漠 is the expected headway distance when the indifferent driver follows the smart vehicle in front, a max is the expected maximum rate of change of headway, 0≤a max ≤2.3, σ3 is the model parameter; The probability density function of the distrust degree x in the value range of [0%, 100%] is: Where f(x) represents the probability density of a distrustful driver at a distrust level of x, σ4 is the standard deviation of the distrust level distribution, and α is the correction coefficient, which is used to set the integral result of the distrust level probability density to 1. The calculation formula of α is:

3. The method for analyzing traffic capacity based on driver trust probability according to claim 1, characterized in that: The formula for obtaining the number N of vehicles in the queue at the signal-controlled intersection is: N = λ*(Cg) = λ*(CGY-r+L), Among them, λ is the vehicle arrival rate at the signal-controlled intersection, g is the effective green light time, C is the cycle duration of the signal-controlled intersection cycle timing plan, G is the actual green light display time of the phase, Y is the actual yellow light clearing time, r is the signal light full red time, and L is the total loss time of the phase.

4. The method for analyzing traffic capacity based on driver trust probability according to claim 3 is characterized in that: The calculation formula of the vehicle arrival rate λ at the signal-controlled intersection is: Among them, G s is the green light time of the upstream intersection through phase of the signalized intersection, G l is the green light time of the left turn phase at the upstream intersection, λ r is the arrival rate of right-turn vehicles at the upstream intersection, λ s is the arrival rate of through vehicles at the upstream intersection, λ l is the arrival rate of left-turning vehicles at the upstream intersection; r The calculation formula is: Among them, the q r is the flow rate of right-turn traffic at the upstream intersection; The lambda s The calculation formula is: Among them, the q s is the flow rate of through traffic at the upstream intersection, and S s is the saturation flow rate of through traffic at the upstream intersection, The green light start time for the upstream intersection through phase. The green light end time for the straight phase at the upstream intersection; The lambda l The calculation formula is: Among them, the q l is the flow rate of left-turn traffic at the upstream intersection, and S l is the saturation flow rate of left-turn traffic at the upstream intersection, The green light start time for the left turn phase at the upstream intersection. The green light end time for the left turn phase at the upstream intersection.

5. The method for analyzing traffic capacity based on driver trust probability according to claim 1, characterized in that: Based on the number of vehicles N in the queue at the signal-controlled intersection and the penetration rate P of smart vehicles in the queue, the vehicle queuing mode is set to a hybrid type: If the penetration rate of intelligent vehicles in the queue is P<0.5, then each of the intelligent vehicles is followed by a manually driven vehicle; if the penetration rate of intelligent vehicles in the queue is P=0.5, then each of the intelligent vehicles is followed by a manually driven vehicle; If the penetration rate P of intelligent vehicles in the queue is greater than 0.5, each of the manually driven vehicles will follow the intelligent vehicle.

6. The method for analyzing traffic capacity based on driver trust probability according to claim 1, characterized in that: When the penetration rate of intelligent vehicles P<0.5, the calculation formula of the expected capacity E(ω) is: Where E(ω) is the expected capacity of the signal-controlled intersection under a specific intelligent vehicle penetration rate and driver trust probability, N is the number of vehicles in the queue at the intersection, P is the penetration rate of intelligent vehicles in the queue, i is the number of manually driven vehicles that are trustworthy and follow intelligent vehicles at the same time, j is the number of manually driven vehicles that are untrustworthy and follow intelligent vehicles at the same time, P1, P2, and P3 are the driver's trust probability, untrust probability, and indifference probability, respectively, and P1+P2+P3=1, P4=1-P3, ω(i,j) is the capacity of the signal-controlled intersection corresponding to any set of parameters i and j under a specific penetration rate P; Based on the calculation formula of expected capacity E(ω) and the driver's trust probability and distrust probability, the simulation system is run to output the expected capacity at the signalized intersection when the permeability P<0.5; When the penetration rate of intelligent vehicles P ≥ 0.5, the calculation formula of the expected traffic capacity E(ω) is: Where E(ω) is the expected capacity of the signal-controlled intersection under a specific intelligent vehicle penetration rate and driver trust probability, N is the number of vehicles in the queue at the intersection, P is the penetration rate of intelligent vehicles in the queue, i is the number of manually driven vehicles that are trustworthy and follow intelligent vehicles at the same time, j is the number of manually driven vehicles that are untrustworthy and follow intelligent vehicles at the same time, and ω(i,j) is the capacity of the signal-controlled intersection corresponding to any set of parameters i and j under a specific penetration rate P; Based on the calculation formula of the expected capacity E(ω) and the driver's trust probability and distrust probability, the simulation system is run to output the expected capacity at the signal-controlled intersection when the permeability P≥0.

5.

7. A traffic capacity analysis system based on driver trust probability, characterized in that: include: A parameter configuration module, configured to obtain a vehicle type, the vehicle type includes a manually driven vehicle and an intelligent vehicle, and to set basic vehicle parameters and a following model based on the vehicle type, the basic vehicle parameters including acceleration, maximum vehicle speed, start reaction time, and maximum deceleration; Setting the trust type of the driver of the manually driven vehicle towards the intelligent vehicle, wherein the trust type includes trust type, indifference type and distrust type, and configuring the parameters of the following model of the manually driven vehicle based on the trust type; A queuing mode setting module is configured to obtain the number of vehicles in a queue at a signal-controlled intersection, the penetration rate of intelligent vehicles in the queue, and set a queuing mode for the vehicles in the queue; A simulation module is configured to set the trust probability and distrust probability of drivers in the queue, and simulate based on the penetration rate of intelligent vehicles, the trust probability and distrust probability of drivers, basic vehicle parameters, a car-following model, and car-following model parameters to obtain the expected capacity and the capacity improvement rate; The function relationship output module is configured to output the function relationship y=cx1+dx2+k between the capacity improvement rate y and the trust probability x1 and the distrust probability x2 under a specific penetration rate; obtain the coefficient c of the trust probability x1 under different penetration rates, output the function relationship between the coefficient c and the penetration rate of smart vehicles, and according to the function relationship between the coefficient c and the penetration rate of smart vehicles in the queue, based on the actual penetration rate of smart vehicles in the queue at the signal-controlled intersection, derive the coefficient c that represents the degree of influence of the driver's trust probability on the capacity.

8. The traffic capacity analysis system based on driver trust probability according to claim 7 is characterized in that: The parameter configuration module includes: The car-following model parameter configuration unit is configured to set a trust level for a trusting driver, set a distrust level for a distrusting driver, and configure the car-following model parameters of a manually driven vehicle based on the trust level and the distrust level.

Images