Urban tunnel diversion area traffic capacity calculation method in intelligent network connection environment
By conducting simulation analysis and model construction of urban tunnel diversion zones in an intelligent connected environment, the problem of difficulty in calculating traffic capacity in the existing technology is solved, and refined planning of tunnel diversion zones and traffic efficiency improvement is achieved.
Patent Information
- Application Number
- CN202410935817.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-12
- Publication Date
- 2024-11-15
AI Technical Summary
It is difficult for the prior art to accurately calculate the traffic capacity of urban tunnel diversion zones in an intelligent connected environment, especially because the existing model is not suitable for this special complex scenario, resulting in relatively scarce research results.
采用一种智能网联环境下城市隧道分流区通行能力计算方法,通过仿真分析、参数敏感性分析、影响因素分析和模型构建,结合SUMO仿真模型和灰色关联度分析法,确定关键影响因素并构建多元回归模型,验证通行能力模型。
The refinement analysis and planning of the traffic capacity of urban tunnel diversion areas has been achieved, ensuring safe and efficient traffic operation, and improving the traffic capacity and traffic operation efficiency of the tunnel diversion areas.
Smart Images

Figure CN118965954A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of tunnel capacity calculation, and in particular to a method for calculating the capacity of a diversion area of an urban tunnel in an intelligent network environment. Background Art
[0002] At present, the research on traffic capacity mainly focuses on exploring the various factors that affect traffic capacity and building calculation models. At the same time, in the intelligent connected environment, the current research results mainly focus on the improvement of road traffic capacity by intelligent connected vehicles.
[0003] Existing research mainly focuses on analyzing and studying the capacity of common diversion areas in highways and expressways. Most of the research on capacity in the intelligent network environment is based on the basic sections of single-lane and multi-lane highways. A small number of studies are based on signal intersections, merging areas, and diversion areas. There is a lack of research results on the capacity of mixed traffic flow in special and complex scenarios such as urban tunnel diversion areas. There are three main existing model construction methods: estimation based on measured data, theoretical derivation analysis, and simulation modeling analysis to accurately calculate the capacity of diversion areas. However, the existing capacity model is not suitable for urban tunnel diversion areas, especially in the intelligent network environment. For this scenario, it is urgent to build a mixed traffic capacity model. Summary of the invention
[0004] In view of the above-mentioned deficiencies in the prior art, the technical problem to be solved by this patent application is how to provide a method for calculating the capacity of urban tunnel diversion areas under an intelligent connected environment, which can help explore the traffic flow operation status and capacity evolution law of urban tunnel diversion areas under an intelligent connected environment.
[0005] In order to solve the above technical problems, the present invention adopts the following technical solutions:
[0006] A method for calculating the capacity of a diversion area in an urban tunnel under an intelligent network environment analyzes the road characteristics and traffic features of the diversion area in an urban tunnel, collects road data such as the number of upstream and downstream lanes, the number of ramp lanes, and the connection form in the actual tunnel diversion area, and realizes the analysis and modeling of the capacity of the diversion area in an urban tunnel under an intelligent network environment through the following steps. The specific steps are: S1: simulation analysis of traffic flow in the diversion area of the tunnel, mainly including simulation scene construction, parameter sensitivity analysis and parameter calibration; S2: analysis of factors affecting the capacity; S3: construction of a capacity model; S4: verification of the capacity model.
[0007] As an optimization, when setting up the simulation scene in step S1, the actual tunnel is selected as the target model construction object, a road network file is created based on the static road data collected on site, and a SUMO simulation model of the urban tunnel diversion area is constructed.
[0008] As an optimization, the parameter sensitivity analysis in step S1 includes the following steps: A1: Determine the input flow of the upstream of the main line when the tunnel diversion area reaches its capacity; A2: Sensitivity test plan; Select the SUMO default Krauss car-following model and LC2013 lane-changing model as the sensitivity analysis objects, and select the following model parameters such as Mingap and sigma and the lane-changing model parameters such as lcStrategic and lcCooperative according to the relevant descriptions of each parameter in the above two types of vehicle behavior models in the SUMO official document, and formulate an analysis plan based on the default values of each parameter and the actual situation; A3: Conduct sensitivity analysis on the relevant parameters of the expected speed, the following model and the lane-changing model, and select the traffic volume in the diversion area and the average vehicle speed as evaluation indicators;
[0009] According to the comparative analysis of the test results in the above steps, the expected speed, Mingap (minimum vehicle spacing), sigma (driver defect), tau (expected headway), lcStrategic (strategic lane change), lcCooperative (cooperative lane change), lcSpeedGain (speed gain lane change), and lcAssertive (decisive lane change) are selected as the key influencing parameters of the diversion area simulation model for model calibration.
[0010] As an optimization, the parameter calibration of the simulation model in step S1 is based on the established static simulation model of the urban tunnel diversion area, and the key influencing parameters of the simulation models of manually driven vehicles and intelligent connected vehicles are calibrated; the following model of manually driven vehicles uses the SUMO default Krauss model, and the lane changing model of manually driven vehicles uses the SUMO default LC2013 model; the following model of intelligent connected vehicles uses the CACC model proposed by the PATH Laboratory in Berkeley, California, and the lane changing model of intelligent connected vehicles uses the LC2013 model.
[0011] As an optimization, in an intelligent connected environment, the research scope of the diversion area will change with the permeability. Therefore, the diversion area of urban tunnels is redefined in an intelligent connected environment. When the permeability is less than 0.5, the research scope of the diversion area is from 550m upstream of the diversion point to 50m downstream of the diversion point, including the deceleration lane and the two outermost lanes of the main line. When the permeability is greater than or equal to 0.5, the research scope of the diversion area is from 750m upstream of the diversion point to 50m downstream of the diversion point, including the deceleration lane and the two outermost lanes of the main line. For the diversion area of an urban tunnel with two lanes of the main line, the main line refers to all lanes.
[0012] As an optimization, when analyzing the factors affecting the capacity in step S2, by analyzing the road characteristics and traffic characteristics of the urban tunnel diversion area and the mixed traffic flow characteristics, six factors affecting the capacity are selected for study: permeability, deceleration lane length, diversion ratio, number of main line lanes, number of ramp lanes, and lane width.
[0013] As an optimization, the main steps for analyzing the influencing factors of the capacity in step S2 include B1: in-depth analysis of the impact of a single factor on the capacity of the diversion area under different permeability conditions; B2: using the grey correlation analysis method to analyze the degree of influence of each factor on the capacity; B3: calculating the correlation between the reference sequence (parent sequence) and the characteristic sequence (comparison series, subsequence), sorting based on the calculation results of the correlation, and identifying the main factors that have a greater impact on the target value; B4: selecting 3 main factors, and using two-factor combination analysis and multi-factor coupling analysis to analyze the impact on the capacity of the diversion area and draw conclusions.
[0014] As an optimization, in step S3, when constructing the capacity model, the permeability is 0.5 as the boundary, and based on the multivariate regression fitting method, the capacity model is constructed under different permeabilities. The capacity model is determined based on the small residual value and good goodness of fit of the capacity hypothesis model.
[0015] In summary, the method for calculating the capacity of urban tunnel diversion areas under the intelligent connected environment determines the key factors affecting the capacity of urban tunnel diversion areas, which is conducive to the rational planning and design of tunnel diversion areas, ensuring the safe and efficient operation of tunnel traffic, and enabling refined management and control, improving the capacity of tunnel diversion areas, and improving traffic operation efficiency. BRIEF DESCRIPTION OF THE DRAWINGS
[0016] Figure 1 This is a flow chart of a method for calculating the traffic capacity of a diversion area in an urban tunnel under an intelligent network environment described in the present invention.
[0017] Figure 2 It is the test result of tunnel diversion area capacity under default parameters.
[0018] Figure 3 These are some results of sensitivity testing.
[0019] Figure 4 Following mode for intelligent connected vehicles
[0020] Figure 5 Distribution of vehicle lane-changing positions under different penetration rates
[0021] Figure 6 This is the research scope of urban tunnel diversion area in intelligent connected environment.
[0022] Figure 7This is the flow chart of grey relational analysis.
[0023] Figure 8 This is a correlation analysis diagram.
[0024] Fig. 9 It is the experimental result of permeability and split ratio combination.
[0025] Fig.10 Permeability and deceleration lane combination test results
[0026] Fig.11 Experimental results of the combination of deceleration lane length and diversion ratio
[0027] Fig.12 Capacity surface diagram under the coupling of multiple factors DETAILED DESCRIPTION
[0028] The present invention is further described in detail below in conjunction with the accompanying drawings. In the description of the present invention, it should be understood that the directions or positional relationships indicated by directional words such as "upper, lower" and "top, bottom" are usually based on the directions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description. Unless otherwise stated, these directional words do not indicate or imply that the devices or components referred to must have a specific direction or be constructed and operated in a specific direction, and therefore cannot be understood as limiting the scope of protection of the present invention; the directional words "inside and outside" refer to the inside and outside relative to the contours of each component itself.
[0029] like Figure 1-6 As shown, a method for calculating the capacity of urban tunnel diversion areas under an intelligent network environment is used to analyze the road characteristics and traffic characteristics of urban tunnel diversion areas, collect road data such as the number of upstream and downstream lanes, the number of ramp lanes, and the connection form of the actual tunnel diversion area, and realize the analysis and modeling of the capacity of urban tunnel diversion areas under an intelligent network environment through the following steps. The specific steps are: S1: traffic flow simulation analysis of tunnel diversion areas, mainly including simulation scene construction, parameter sensitivity analysis and parameter calibration; S2: analysis of factors affecting capacity; S3: construction of capacity model; S4: verification of capacity model.
[0030] In this embodiment, when building the simulation scene in step S1, an actual tunnel is selected as the target model building object, a road network file is created based on static road data collected on site, and a SUMO simulation model of the urban tunnel diversion area is constructed.
[0031] In this embodiment, the parameter sensitivity analysis in step S1 includes the following steps: A1: determining the input flow of the main line upstream when the tunnel diversion area reaches its traffic capacity; A2: sensitivity test plan; selecting the SUMO default Krauss car-following model and LC2013 lane-changing model as sensitivity analysis objects, according to the relevant description of each parameter in the above two types of vehicle behavior models in the SUMO official document, selecting Mingap, sigma and other car-following model parameters and lcStrategic, lcCooperative and other lane-changing model parameters, and formulating an analysis plan based on the default values of each parameter and the actual situation; A3: performing sensitivity analysis on the relevant parameters of the expected speed, car-following model and lane-changing model respectively, and selecting the diversion area traffic volume and the average vehicle speed as evaluation indicators;
[0032] Specifically, in step A1, the speed is set to 60 km / h for the main line and 40 km / h for the ramp, and the rest of the parameters are set to default values. The capacity test results show that the simulation model is inferred to have the set parameters, and when the upstream input flow is 3700 pcu / h, the diversion area will reach its capacity. Therefore, in the subsequent sensitivity analysis of the simulation model, the input flow upstream of the main line is fixed to this value.
[0033] Specifically, the values of the sensitivity test scheme in step A2 are shown in Table 1 below:
[0034]
[0035]
[0036] Table 1 Sensitivity test scheme
[0037] Specifically, sensitivity analysis was performed on the parameters of the expected speed, the following model, and the lane change model, and the traffic volume in the diversion area and the average vehicle speed were selected as evaluation indicators. Based on the sensitivity analysis of 13 model parameters, the expected vehicle speed and expected headway test results are shown in the attached figure. Figure 3 shown.
[0038] Specifically, in step A3, based on the sensitivity analysis of 13 model parameters, the sensitivity of each parameter in the simulation model was evaluated from two dimensions: the downstream flow rate of the diversion area and the average vehicle speed in the diversion area. The results are shown in Table 2
[0039]
[0040]
[0041] Table 2 Comparison of sensitivity test results
[0042] According to the comparative analysis of the test results in the above steps, the expected speed, Mingap (minimum vehicle spacing), sigma (driver defect), tau (expected headway), lcStrategic (strategic lane change), lcCooperative (cooperative lane change), lcSpeedGain (speed gain lane change), and lcAssertive (decisive lane change) are selected as the key influencing parameters of the diversion area simulation model for model calibration.
[0043] In this embodiment, the parameter calibration of the simulation model in step S1 is based on the established static simulation model of the urban tunnel diversion area, and the key influencing parameters of the simulation models of manually driven vehicles and intelligent connected vehicles are calibrated; the following model of the manually driven vehicle uses the SUMO default Krauss model, and the lane changing model of the manually driven vehicle uses the SUMO default LC2013 model; the following model of the intelligent connected vehicle uses the CACC model proposed by the PATH Laboratory in Berkeley, California, and the lane changing model of the intelligent connected vehicle uses the LC2013 model.
[0044] Specifically, the following model of manually driven vehicles uses the default Krauss model of SUMO. This model is a safety distance model. The model assumes that the driver always maintains a safe headway between the driver and the vehicle in front. In the event of emergency braking of the vehicle in front, the driver still has enough time to react and slow down to stop. The safe speed calculation formula of the Krauss model is as follows:
[0045]
[0046] Where: V safe is the safe speed; V a (t) is the speed of the preceding vehicle at time t; V f (t) is the speed of the rear vehicle at time t; g(t) is the distance between the front and rear vehicles at time t; λ is the driver's reaction time; b is the maximum deceleration of the vehicle.
[0047] The lane-changing model of manually driven vehicles uses the default LC2013 model of SUMO, which can effectively simulate the lane-changing behavior of drivers in actual driving. Based on the results of the sensitivity analysis of the simulation model, the key influencing parameters such as the expected speed and expected headway are first calibrated using measured data. In addition, the key parameters such as Mingap, sigma, lcStrategic, and lcCooperative of manually driven vehicles need to be calibrated. Based on the analysis of the traffic operation characteristics of the diversion area of the urban tunnel and the description of each parameter in the official documents of SUMO, the remaining key parameters are repeatedly tested, and the key parameters are calibrated by comparing the speed, flow, and lane-changing distribution of the diversion area with the actual operation. The calibration results of the key parameters of manually driven vehicles are shown in Table 3 below:
[0048] parameter Parameter meaning default value Determine the parameters Mingap Minimum distance between vehicles 2.5 2.50 Sigma Driver Defects 0.5 0.30 tau Expected headway 1.0 1.89 lcStrategic Strategic Lane Change 1.0 0.50 lcCooperative Cooperative Lane Change 1.0 0.60 lcSpeedGain Speed Gain Lane Change 1.0 3.00 lcAssertive Decisive lane change 1.0 3.00
[0049] Table 3 Calibration results of key parameters of manually driven vehicles
[0050] Specifically, the intelligent connected vehicle following mode can be divided into two types: intelligent connected vehicles follow intelligent connected vehicles and manually driven vehicles, as shown in the attached figure. Figure 4 As shown in the figure, when the leading vehicle is an intelligent connected vehicle, the vehicle can be operated and controlled collaboratively through inter-vehicle communication, and its headway is recorded as tau cc ; When the leading vehicle is a manually driven vehicle, inter-vehicle communication cannot be achieved, and some functions of the intelligent connected vehicle will be degraded. The headway time is recorded as tau ch .
[0051] Specifically, the intelligent connected vehicle following model uses the CACC model proposed by the PATH Laboratory in Berkeley, California. This model is widely used in academic research on intelligent connected vehicles and autonomous vehicles. The CACC model is divided into four modes: speed control mode, spacing control mode, spacing approach control mode and collision avoidance control mode. The basic expression of the CACC following model is as follows:
[0052]
[0053] Where: x i is the displacement of the following vehicle; v i is the speed of the following vehicle; e is the error between the actual distance and the expected distance; is the differential of the error between the actual distance and the expected distance; i-1 represents the front vehicle; v kprev is the speed of the vehicle at the previous moment; k p is the inter-vehicle distance error control coefficient, k p >0;k d is the control coefficient of the differential term of the inter-vehicle distance error, k d >0;
[0054] When an intelligent connected vehicle follows a manually driven vehicle, its following model uses the ACC model proposed by the PATH laboratory to simulate the following behavior of the degraded intelligent connected vehicle.
[0055] The lane-changing model of the intelligent connected vehicle adopts the LC2013 model, and its expected speed is consistent with that of the manually driven vehicle. The CACC following model has no sigma parameter. According to the results of the model parameter sensitivity analysis, the six key parameters of Mingap, tau, lcStrategic, lcCooperative, lcSpeedGain, and lcAssertive are repeatedly tested to complete the calibration of the intelligent connected vehicle parameters. The calibration results of the intelligent connected vehicle simulation parameters are shown in Table 4 below:
[0056] parameter Parameter meaning default value CACC Vehicles ACC Vehicles Mingap Minimum safety distance 2.5 2.0 2.0 tau Expected headway 1.0 0.6 1.1 lcStrategic The desire to perform strategic lane changes 1.0 2.5 2.5 lcCooperative Willingness to make cooperative lane changes 1.0 1.0 1.0 lcSpeedGain Eager to make lane changes to gain speed 1.0 1.0 1.0 lcAssertive Willing to accept lower front-to-rear clearance on target lane 1.0 2.0 2.0
[0057] Table 4. Calibration results of intelligent connected vehicle simulation parameters
[0058] In this embodiment, under the intelligent network environment, the research scope of the diversion area will change with the permeability, so the diversion area of the urban tunnel is redefined under the intelligent network environment; when the permeability is less than 0.5, the research scope of the diversion area is 550m upstream of the diversion point to 50m downstream of the diversion point, including the deceleration lane and the two outermost lanes of the main line; when the permeability is greater than or equal to 0.5, the research scope of the diversion area is 750m upstream of the diversion point to 50m downstream of the diversion point, including the deceleration lane and the two outermost lanes of the main line. For the diversion area of the urban tunnel with two lanes of the main line, the main line is all lanes.
[0059] Specifically, the diversion area is determined as an area extending 750m upstream and 50m downstream from the diversion point. Experimental tests have found that when the penetration rate is greater than 50%, a large number of intelligent connected vehicles change lanes in the range of 700-750m upstream of the diversion point. At the same time, the distribution of CAV lane changes at different penetration rates during peak hours is statistically analyzed, and the results are shown in the attached figure. Figure 5 shown.
[0060] Specifically, the reference line of CAV lane-changing position is drawn based on the decile value, that is, the auxiliary line with a cumulative frequency of 0.1 in the figure. Using this reference line, combined with the horizontal axis, the starting position of 90% of CAV lane changes can be obtained, and then the overall lane-changing behavior characteristics and trends of CAV can be studied. Analysis shows that the lane-changing distribution has significant differences when the permeability is less than 0.5 and greater than or equal to 0.5, and the permeability of 0.5 is the critical point where the lane-changing characteristics change.
[0061] In this embodiment, when analyzing the influencing factors of traffic capacity in step S2, by analyzing the road characteristics and traffic characteristics of the urban tunnel diversion area and the mixed traffic flow characteristics, six traffic capacity influencing factors are selected for study: permeability, deceleration lane length, diversion ratio, number of main line lanes, number of ramp lanes, and lane width.
[0062] In this embodiment, the main steps of analyzing the factors affecting the traffic capacity in step S2 include:
[0063] B1: In-depth analysis of the impact of a single factor on the capacity of the diversion area under different permeability conditions;
[0064] B2: Use grey correlation analysis to analyze the influence of various factors on traffic capacity;
[0065] B3: Calculate the correlation between the reference sequence (parent sequence) and the characteristic sequence (comparison sequence, subsequence), sort them based on the calculation results of the correlation, and identify the main factors that have a greater impact on the target value;
[0066] B4: Select three main factors, use two-factor combination analysis and multi-factor coupling analysis to analyze their impact on the diversion area capacity and draw conclusions.
[0067] Specifically, the actual calculation process includes.
[0068] (1) Determine the analysis sequence
[0069] The reference sequence X0 and the characteristic sequence Xi are expressed as the following formulas:
[0070] X0={X0(k)|k=1,2,...,n}
[0071] X i ={X i (k)|k=1,2,...,n}
[0072] (2) Dimensionless processing
[0073] There are two main methods for dimensionless processing of reference sequence and feature sequence data: initialization and averaging. The initialization method normalizes the data by dividing it by the initial value so that the processed data falls roughly around the order of 1. This method is suitable for data sequences that show a steady growth or decrease trend. The initialization process is shown in the following formula.
[0074] The reference sequence is denoted as:
[0075] Y0={Y0(k)}={X0(k) / X0(1)|k=1,2,...,n}
[0076] The feature sequence is recorded as:
[0077] Y i = {Y i (k)}={X i (k) / X i (1)|k=1,2,...,n})
[0078] The averaging process divides the reference sequence and feature sequence data by the mean. The processed data roughly falls around the order of 1, which is suitable for sequences with no obvious rising and falling trends. The averaging process is shown in the following formula.
[0079] The reference sequence is denoted as:
[0080]
[0081] Where: is the mean of the reference sequence data,
[0082] The feature sequence is recorded as:
[0083]
[0084] Where: is the mean of the characteristic sequence data,
[0085] (3) Calculation of correlation coefficient
[0086] The specific calculation formula of the correlation coefficient is as follows:
[0087]
[0088] Where: i (k) is the correlation coefficient between the corresponding elements of each reference sequence and the characteristic sequence; ρ is the resolution coefficient, 0<ρ<1, usually 0.5; min i min k |Y0(k)-Y i (k)| is the minimum difference between the two poles;
[0089] max i max k |Y0(k)-Y i (k)| is the maximum difference between the two poles; k = 1, 2, ..., n; i = 1, 2, ..., m.
[0090] (4) Calculate the correlation
[0091] For each indicator, calculate its correlation coefficient with the corresponding reference sequence, and obtain the weighted average of these correlation coefficients to measure the correlation between each feature sequence and the reference sequence, that is, the degree of correlation. The calculation formula is:
[0092]
[0093] (5) Relevance ranking
[0094] According to the gray weighted correlation value, the correlation of each influencing factor is ranked. A higher correlation indicates that the influencing factor has a greater impact on the reference sequence. The diversion area capacity is used as the reference sequence, and the permeability, deceleration lane length, diversion ratio, main line lane number, ramp lane number, and lane width are the characteristic sequences. In the characteristic sequence, the number of ramp lanes is divided into 2 groups of variables, and the rest are divided into 3 groups of variables. The SUMO simulation model is used to conduct experiments to obtain the capacity under different combinations of factors and perform statistics. The data are shown in Table 5.
[0095]
[0096] Table 5 Statistical data of grey correlation analysis
[0097] Based on the statistical data in the above table, the data is dimensionless processed by the averaging method, and the correlation coefficient corresponding to each reference sequence and the characteristic sequence is calculated using the specific calculation formula of the correlation coefficient as follows. The data are shown in Table 6 below.
[0098]
[0099] Table 6 Correlation coefficient calculation table
[0100] Based on the calculation results of the correlation coefficient in Table 6, the correlation degree is calculated and ranked using the correlation degree calculation formula. Combined with the correlation degree classification in Table 7, the analysis of the correlation degree results shows that: the diversion ratio and permeability have a strong impact on the diversion area capacity; the length of the deceleration lane and the number of ramp lanes have a strong impact on the diversion area capacity; the number of main line lanes and lane width have a moderate impact on the diversion area capacity;
[0101] Relevance value grade Impact 0~0.35 Low Weaker 0.35~0.65 medium medium 0.65~0.85 Higher Strong 0.85~1.00 high Very strong
[0102] Table 7 Correlation level reference table
[0103] Combined with the analysis of factors affecting capacity under different permeabilities and the results of grey correlation analysis, three parameters, namely diversion ratio, permeability and deceleration lane length, were selected as the main influencing factors of the capacity of the diversion area of urban tunnels under the intelligent network environment.
[0104] Specifically, the influence of permeability, diversion ratio and deceleration lane length on the diversion area capacity is analyzed from the perspectives of dual factors and multiple factors. The influence form and specific degree of the three main influencing factors on the capacity under combined conditions are explored by drawing a three-dimensional surface graph.
[0105] (1) Two-factor analysis
[0106] Two-factor analysis was performed on the three combinations of permeability and diversion ratio, permeability and deceleration lane length, and diversion ratio and deceleration lane length. The traffic capacity under the combined influence of the two in each group was obtained through simulation, and the results were plotted as a three-dimensional surface graph, and fitted using a function to select the optimal fitting result. The results are shown in the attached figure. Figure 9-11 As shown;
[0107] After the permeability reaches 0.5, the length of the deceleration lane has almost no effect on the capacity, that is, when the permeability is greater than or equal to 0.5, the main factors affecting the capacity are only the permeability and the diversion ratio; when the permeability is less than 0.5, the main factors affecting the capacity are the diversion ratio, permeability, and the length of the deceleration lane. The diversion area capacity C has a quadratic polynomial correlation with the diversion ratio d and the deceleration lane length L; the diversion area capacity C has an exponential relationship or a quadratic polynomial correlation with the permeability p;
[0108] (2) Multi-factor coupling analysis
[0109] A multi-factor coupling analysis is conducted on the main influencing factors of traffic capacity (permeability, diversion ratio and deceleration lane length) when the permeability is less than 0.5. SUMO is used to conduct simulation experiments to extract the traffic capacity values under the combination of three main influencing factors. The permeability increases from 0 to 0.5 with an increase step of 0.1, the diversion ratio increases from 0 to 1 with an increase step of 0.2, and the deceleration lane length increases from 30m to 190m with an increase step of 20m. Based on the simulation experiment data, a surface diagram of the traffic capacity under the coupling of the three main influencing factors is drawn, as shown in Figure 2. Figure 6 As shown. In the figure, the X-axis represents the permeability p, the Y-axis represents the diversion ratio d, and different height levels (Z-axis) correspond to different deceleration lane lengths L, ranging from 30m at the bottom to 190m at the top, with every 20m as a height level, for a total of 9 levels. It reflects the influence of permeability and diversion ratio on traffic capacity under different deceleration lane length conditions. This method is used to observe the changing trend and pattern of traffic capacity under different combinations of main influencing factors;
[0110] The analysis shows that the diversion ratio d has the greatest impact on the capacity C, and it is inferred that under the coupling of multiple factors, the diversion ratio d and the capacity C still have a quadratic polynomial relationship; the permeability p has a smaller impact on the capacity C than the diversion ratio d, and it is inferred that there is an exponential relationship or a quadratic polynomial relationship between the permeability p and the capacity C; the deceleration lane length L has the least impact on the capacity C, and it is inferred that the deceleration lane length L and the capacity C still have a quadratic polynomial relationship.
[0111] (3) Analysis conclusion
[0112] Through two-factor combination analysis and multi-factor coupling analysis, it was found that the diversion ratio, deceleration lane length and capacity have a quadratic relationship, the permeability has an exponential or quadratic relationship with capacity, and it was found that the factors affecting capacity are different when the permeability is less than 0.5 and when the permeability is greater than or equal to 0.5.
[0113] In this embodiment, in step S3, when constructing the capacity model, the permeability is 0.5 as the boundary, and based on the multivariate regression fitting method, the capacity model is constructed at different permeabilities. The capacity model is determined based on the small residual value and good goodness of fit of the capacity hypothesis model as the judgment criteria.
[0114] Specifically, the capacity model assumes
[0115] (1) Permeability is less than 0.5
[0116] When the permeability p is less than 0.5, the factors affecting the capacity are the diversion ratio, permeability, and the length of the deceleration lane. Two hypothetical models are constructed when the permeability p is less than 0.5. Model 1 assumes that the capacity C and the permeability p present a quadratic polynomial relationship, as shown in the formula;
[0117] C=ap 2 +bd 2 +cL 2 +fpd+gdL+hp+id+jL+k
[0118] Model 2 assumes that the capacity C and permeability p have an exponential relationship, as shown in the formula
[0119] C=le mp +nd 2 +oL 2 +qdL+rd+sL+t
[0120] Where: C is the actual traffic capacity of the diversion area of the urban tunnel (pcu / h); d is the diversion ratio, which ranges from [0,1]; p is the permeability, which ranges from [0,0.5); L is the length of the deceleration lane (m); e is a natural constant; a, b, c, f, g, h, i, j, k, l, m, n, o, q, r, s, and t are all regression coefficients.
[0121] (2) Permeability greater than or equal to 0.5
[0122] When the permeability p is greater than or equal to 0.5, the factors affecting the capacity are the diversion ratio and the permeability, and the length of the deceleration lane has no effect on the capacity. When the permeability p is greater than or equal to 0.5, two hypothetical models are constructed. Model 1 assumes that the capacity C and the permeability p present a quadratic polynomial relationship, as shown in the formula;
[0123] C=ap 2+bd 2 +fpd+gp+hd+i
[0124] Model 2 assumes that the capacity C and permeability p have an exponential relationship, as shown in the formula
[0125] C=je kp +ld 2 +mpd+nd+o
[0126] Where: C is the actual traffic capacity of the diversion area of the urban tunnel (pcu / h); d is the diversion ratio, which ranges from [0,1]; p is the permeability, which ranges from [0.5,1]; e is a natural constant; a, b, f, g, h, i, j, k, l, m, n, and o are all regression coefficients.
[0127] Multiple nonlinear regression fitting
[0128] (1) Permeability is less than 0.5
[0129] The capacity data obtained from the multi-factor coupling experiment was substituted into the hypothetical model, and SPSS software was used to perform multivariate nonlinear regression fitting on the unknown parameters in the formulas of Model 1 and Model 2 with a permeability less than 0.5; the fitted model regression coefficients and corresponding confidence intervals are shown in Table 8 below.
[0130]
[0131]
[0132] Table 8 Model regression coefficients and confidence intervals (p<0.5)
[0133] After standard error and confidence interval analysis, goodness of fit test and residual analysis, the hypothesis model 1 was selected as the actual capacity model under the condition that the permeability p was less than 0.5:
[0134] C=234.82p 2 +721.83d 2 -0.006L 2 -507.89pd+0.258dL+
[0135] 584.496p-2305.79d+2.22L+2772.31
[0136] Where: C is the actual traffic capacity of the diversion area of the urban tunnel (pcu / h); d is the diversion ratio, which ranges from [0,1]; p is the permeability, which ranges from [0,0.5); L is the length of the deceleration lane (m).
[0137] (2) Permeability greater than or equal to 0.5
[0138] Substitute the two-factor analysis simulation experimental data when the permeability is greater than or equal to 0.5 into the capacity hypothesis model, and use SPSS software to perform multivariate nonlinear regression fitting on the unknown parameters in the formulas of model 1 and model 2 when the permeability is greater than or equal to 0.5. The fitted model regression coefficients and corresponding confidence intervals are shown in Table 9 below.
[0139]
[0140]
[0141] Table 9 Model regression coefficients and confidence intervals (p ≥ 0.5)
[0142] After standard error and confidence interval analysis, goodness of fit test and residual analysis, the hypothesis model 1 was selected as the actual capacity model under the condition that the permeability p is greater than or equal to 0.5:
[0143] C=964.94p 2 +1090.97d 2 -1370pd+599.03p-2193.24d+2727.6
[0144] Where: C is the actual traffic capacity of the diversion area of the urban tunnel (pcu / h); d is the diversion ratio, which ranges from [0,1]; p is the permeability, which ranges from [0.5,1].
[0145] Capacity model determination
[0146] According to the analysis, the residual value of the capacity hypothesis model 1 under different permeabilities is smaller and the goodness of fit is better. Therefore, the actual capacity model of the urban tunnel diversion area under the intelligent network environment constructed by the present invention is as follows:
[0147]
[0148] Where: C is the actual traffic capacity of the diversion area of the urban tunnel (pcu / h); d is the diversion ratio, which ranges from [0,1]; p is the permeability, which ranges from [0,1]; and is the length of the deceleration lane (m).
[0149] In this embodiment, multiple groups of video data of actual scenes are collected, and the penetration rate of intelligent connected vehicles is adjusted in the simulation model to obtain the actual traffic capacity in the intelligent connected environment; the deceleration lane length, diversion ratio and penetration rate of the scene are substituted into the constructed capacity model to obtain the capacity model calculation values under different combinations; if the error between the model calculation value and the actual capacity meets the requirements, it indicates that the capacity model constructed by the present invention can more accurately reflect the actual traffic capacity of the diversion area of the urban tunnel under the intelligent connected environment.
[0150] Beneficial Effects
[0151] The present invention uses the collected 4 sets of data to obtain the actual traffic capacity of the tunnel diversion area in the experimental scenario under the intelligent network environment. At the same time, the deceleration lane length, diversion ratio and permeability of the tunnel diversion area are substituted into the traffic capacity model constructed by the present invention to obtain the traffic capacity model calculation value under different combinations. The comparison results of the traffic capacity model calculation value and the actual traffic capacity are shown in Table 10. It can be seen from the table that the error between the model calculation value and the actual traffic capacity does not exceed 5.29%, indicating that the traffic capacity model constructed by the present invention can more accurately reflect the actual traffic capacity of the urban tunnel diversion area under the intelligent network environment.
[0152]
[0153]
[0154] Table 10 Comparison between the calculated value of the capacity model and the actual capacity
[0155] Facing the future mixed traffic flow, combining theoretical analysis and simulation experiments, a calculation model for the capacity of urban tunnel diversion areas under intelligent network environment is constructed, and the key factors affecting the capacity of urban tunnel diversion areas are determined, which is conducive to the rational planning and design of tunnel diversion areas and ensures the safe and efficient operation of tunnel traffic. At the same time, it can grasp the actual capacity of typical scenarios in mixed traffic flow environments, conduct refined control of intelligent networked vehicles and manually driven vehicles, improve the capacity of tunnel diversion areas, and improve traffic operation efficiency.
[0156] Finally, it should be noted that those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims and their equivalents, the present invention is also intended to include these modifications and variations.
Claims
1. A method for calculating the traffic capacity of a diversion area in an urban tunnel under an intelligent network environment, characterized in that: The road characteristics and traffic characteristics of the urban tunnel diversion area are analyzed, and the road data such as the number of upstream and downstream lanes, the number of ramp lanes, and the connection form of the actual tunnel diversion area are collected. The following steps are performed to realize the analysis and modeling of the traffic capacity of the urban tunnel diversion area under the intelligent network environment. The specific steps are as follows: S1: Traffic flow simulation analysis in tunnel diversion area, mainly including simulation scene construction, parameter sensitivity analysis and parameter calibration; S2: Analyze the factors affecting traffic capacity; S3: Constructing the capacity model; S4: Validate the capacity model.
2. According to the method for calculating the traffic capacity of the diversion area of an urban tunnel in an intelligent network environment in claim 1, it is characterized in that: When building the simulation scene in step S1, the actual tunnel is selected as the target model building object, a road network file is created based on the static road data collected on site, and a SUMO simulation model of the urban tunnel diversion area is constructed.
3. The method for calculating the traffic capacity of a diversion area in an urban tunnel under an intelligent network environment according to claim 2 is characterized in that: The parameter sensitivity analysis in step S1 includes the following steps: A1: Determine the input flow upstream of the mainline when the tunnel diversion area reaches its capacity; A2: Sensitivity test plan; select the SUMO default Krauss car-following model and LC2013 lane-changing model as the sensitivity analysis objects. According to the relevant descriptions of the parameters in the above two types of vehicle behavior models in the SUMO official documents, select the following model parameters such as Mingap and sigma and the lane-changing model parameters such as lcStrategic and lcCooperative. According to the default values of each parameter and the actual situation, formulate an analysis plan; A3: Conduct sensitivity analysis on the parameters of the expected speed, the car-following model and the lane-changing model, and select the traffic volume in the diversion area and the average vehicle speed as evaluation indicators; According to the comparative analysis of the test results in the above steps, the expected speed, Mingap (minimum vehicle spacing), sigma (driver defect), tau (expected headway), lcStrategic (strategic lane change), lcCooperative (cooperative lane change), lcSpeedGain (speed gain lane change), and lcAssertive (decisive lane change) are selected as the key influencing parameters of the diversion area simulation model for model calibration.
4. The method for calculating the traffic capacity of a diversion area in an urban tunnel under an intelligent network environment according to claim 3 is characterized in that: In step S1, the parameters of the simulation model are calibrated, based on the established static simulation model of the urban tunnel diversion area, to calibrate the key influencing parameters of the simulation model of the manually driven vehicle and the intelligent networked vehicle; The following model of manually driven vehicles uses the SUMO default Krauss model, and the lane changing model of manually driven vehicles uses the SUMO default LC2013 model; the following model of intelligent connected vehicles uses the CACC model proposed by the PATH Laboratory in Berkeley, California, and the lane changing model of intelligent connected vehicles uses the LC2013 model.
5. The method for calculating the traffic capacity of a diversion area in a city tunnel under an intelligent network environment according to claim 4 is characterized in that: In the intelligent connected environment, the research scope of the diversion area will change with the permeability. Therefore, the diversion area of urban tunnels is redefined in the intelligent connected environment. When the permeability is less than 0.5, the research scope of the diversion area is 550m upstream of the diversion point to 50m downstream of the diversion point, including the deceleration lane and the two outermost lanes of the main line. When the permeability is greater than or equal to 0.5, the research scope of the diversion area is 750m upstream of the diversion point to 50m downstream of the diversion point, including the deceleration lane and the two outermost lanes of the main line. For the diversion area of an urban tunnel with two lanes of the main line, the main line refers to all lanes.
6. The method for calculating the traffic capacity of a diversion area in a city tunnel under an intelligent network environment according to claim 5 is characterized in that: When analyzing the factors affecting the traffic capacity in step S2, by analyzing the road characteristics and traffic characteristics of the diversion area of the urban tunnel and the mixed traffic flow characteristics, six factors affecting the traffic capacity are selected for study: permeability, deceleration lane length, diversion ratio, number of main line lanes, number of ramp lanes, and lane width.
7. The method for calculating the traffic capacity of a diversion area in a city tunnel under an intelligent network environment according to claim 6 is characterized in that: The main steps of analyzing the factors affecting traffic capacity in step S2 include: B1: In-depth analysis of the impact of a single factor on the capacity of the diversion area under different permeability conditions; B2: Use grey correlation analysis to analyze the influence of various factors on traffic capacity; B3: Calculate the correlation between the reference sequence (parent sequence) and the characteristic sequence (comparison sequence, subsequence), sort them based on the calculation results of the correlation, and identify the main factors that have a greater impact on the target value; B4: Select three main factors, and use two-factor combination analysis and multi-factor coupling analysis to analyze their impact on the diversion area's traffic capacity and draw conclusions.
8. The method for calculating the traffic capacity of a diversion area in an urban tunnel under an intelligent network environment according to claim 7 is characterized in that: In step S3, when constructing the capacity model, the permeability is 0.5 as the boundary, and based on the multivariate regression fitting method, the capacity model is constructed at different permeabilities. The capacity model is determined based on the small residual value and good goodness of fit of the capacity hypothesis model.
Citation Information
Patent Citations
Method and device for calculating road traffic capacity
CN113470363A
Method for analyzing traffic capacity of diversion influence area of multi-lane expressway
CN114582122A
Intelligent network connection special lane design method considering mixed flow theoretical traffic capacity
CN116665442A
Expressway tunnel portal vehicle speed control method based on vehicle-infrastructure cooperation
CN117392851A
Urban-region road network vehicle-passage flow prediction method and system based on hybrid deep learning model
US20240220686A1
Cited By
Flow distribution area traffic flow distribution method for novel mixed traffic flow
CN120071614A
Traffic accident five-post combined service traffic diversion dynamic adjustment method
CN121564955A
A method for dynamic adjustment of traffic diversion in five-post joint traffic accident response
CN121564955B
