A parameter-adaptive method for estimating headway distribution on highways

Abstract

The present invention belongs to the field of intelligent transportation and discloses a parameter-adaptive method for estimating headway time distribution on highways, comprising: S1. Based on radar-visual fusion trajectory data of a target highway section, extracting headway time data for vehicles passing through different lanes of the target section; S2. Fitting the headway time data of the target section using a three-parameter Burr distribution model, and calculating the parameters of the three-parameter Burr distribution model using the maximum likelihood estimation method; S3. Establishing a parameter-adaptive three-parameter Burr distribution model for headway time distribution. The present invention can construct a parameter-adaptive headway time distribution model that is flexible and applicable to different time periods and road sections of highways. This model can accurately characterize the distribution characteristics of headway time distribution under different highway conditions using a small number of traffic flow parameters.

Description

Technical Field

[0001] The invention belongs to the field of intelligent traffic information technology, and in particular relates to a parameter-adaptive highway headway distribution estimation method. Background Art

[0002] Headway distribution can characterize the overall operational status of vehicles on a road section. Studying its distribution patterns helps relevant management departments assess road safety and implement traffic control. However, the actual distribution of headway is susceptible to various factors, such as traffic flow conditions, and exhibits temporal and spatial variations. Establishing a realistic headway distribution model that does not require extensive data calculations and can characterize different traffic conditions on highways can provide traffic management departments with accurate and reliable traffic information, enabling them to implement refined operational control.

[0003] However, most existing headway distribution models are based on traditional parametric models, relying on fixed parameters to determine the shape of the selected distribution model. These models are overly idealistic and fail to account for the complex and changing traffic conditions of real roads. Furthermore, these methods place significant demands on the field information and equipment coverage of headway data collection. They first need to calculate the headway of a single vehicle passing through the cross-section, then aggregate the headway of multiple vehicles across each lane to fit the distribution model. This requires a high data collection scale and a cumbersome computational process.

[0004] Patent CN110070733B provides a headway modeling method that collects the headway from the first vehicle to the kth vehicle passing the stop line at a signalized intersection, fits the collected data with different probability distribution functions, and selects a suitable probability distribution function for headway modeling. However, the headway distribution on highways varies with traffic volume and road section type. Constructing a distribution model based on fixed parameters may not be able to accurately and effectively describe the distribution characteristics of the actual headway on highways. Moreover, the process of summarizing all headway times and then fitting the distribution model has high requirements for data scale and collection, and the calculation process is cumbersome. In the above method, this factor is less considered. Summary of the Invention

[0005] In view of this, the present invention aims to provide a parameter-adaptive highway headway distribution estimation method. The present invention aims to solve the problems of existing methods such as poor accuracy, strong dependency, high data collection requirements and cumbersome calculation process.

[0006] The present invention provides a parameter-adaptive highway headway distribution estimation method, comprising the following steps:

[0007] S1. Based on the radar-visual fusion trajectory data of the target section of the highway, extract the headway data of vehicles passing through different lanes of the target section;

[0008] S2. Use the three-parameter Burr distribution model to fit the headway data of the target section and use the maximum likelihood estimation method to calculate the parameters of the three-parameter Burr distribution model;

[0009] S3. Establish a three-parameter burr distribution model for headway with parameter adaptability.

[0010] Furthermore, in step S1, the headway data of vehicles in different lanes are calculated using the following expression:

[0011] T i =t i -t i-1

[0012] Where, t i is the moment when the current vehicle passes through the section, t i-1 It is the moment when the preceding vehicle in the same lane passes the section.

[0013] Furthermore, step S2 includes the following sub-steps:

[0014] S2.1 uses the three-parameter burr distribution model to fit the headway data of the acquisition section;

[0015] The probability density function of the three-parameter burr distribution model is expressed as follows:

[0016]

[0017] Where x≥0, a≥0, c≥0, k≥0, the three parameters are used to control the shape, location and scale of the distribution; c and k are shape parameters, and a is the scale parameter;

[0018] S2.2 Use the maximum likelihood estimation method to calculate the parameters of the three-parameter burr distribution model.

[0019] Furthermore, step S3 includes the following sub-steps:

[0020] S31. Extract traffic flow characteristic variables with a 30-minute detection period based on radar-visual fusion trajectory data;

[0021] The traffic flow characteristic variables include traffic volume, average speed, occupancy rate, large vehicle penetration rate and average vehicle length;

[0022] S32. Use the random forest model to evaluate the impact of the above five traffic flow characteristic variables on the three-parameter burr distribution;

[0023] S33. Calculate the importance index of each traffic flow characteristic variable;

[0024] S34. Select the three characteristic parameters that have the most significant impact on the model to construct a parameter-adaptive distribution model;

[0025] S35. An adaptive relationship between the parameters of the three-parameter burr distribution model of vehicle headway and the selected traffic flow characteristic variables is established through the MLP model, so that the parameters of the three-parameter burr distribution model can be automatically adjusted according to changes in traffic flow f, occupancy rate O, and large vehicle penetration rate P to fit the actual vehicle headway distribution.

[0026] Beneficial effects:

[0027] Taking into account the polymorphic heterogeneity of highway headway under different traffic environments, the present invention aims to construct a parameter-adaptive headway distribution model with low data dependence that can be flexibly applied to different time periods and road sections of highways. This model can accurately depict the distribution characteristics of headway under different highway conditions using a small number of traffic flow parameters.

[0028] Other advantages, objects, and features of the present invention will be described in part in the following description and, in part, will be apparent to those skilled in the art upon examination of the following description or may be learned from practice of the present invention. The objects and other advantages of the present invention may be realized and obtained through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 This is a schematic diagram of headway time of the present invention;

[0030] Figure 2 This is the parameter fitting result diagram of the three-parameter burr distribution;

[0031] Figure 3 This is a flow chart of the headway distribution estimation of the present invention. DETAILED DESCRIPTION

[0032] To make the technical solutions, advantages, and purposes of the present invention more clear, the technical solutions of the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the described embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0033] like Figure 1 and Figure 3As shown, the present invention provides a parameter-adaptive highway headway distribution estimation method, comprising the following steps:

[0034] S1. Based on the radar-visual fusion trajectory data of the target section of the highway, extract the headway data of vehicles passing through different lanes of the target section;

[0035] The calculation expression of the headway data of vehicles in different lanes is as follows:

[0036] T i =t i -t i-1

[0037] Where, t i is the moment when the current vehicle passes through the section, t i-1 It is the moment when the preceding vehicle in the same lane passes the section.

[0038] S2. Use the three-parameter Burr distribution model to fit the headway data of the target section and use the maximum likelihood estimation method to calculate the parameters of the three-parameter Burr distribution model;

[0039] The probability density function of the three-parameter burr distribution model is expressed as follows:

[0040]

[0041] Where x≥0, a≥0, c≥0, k≥0, the three parameters are used to control the shape, location and scale of the distribution; c and k are shape parameters, and a is the scale parameter;

[0042] S2.2 Use the maximum likelihood estimation method to calculate the parameters of the three-parameter burr distribution model. The fitting results are as follows: Figure 2 As shown;

[0043] S3. Dynamically adjust the parameters of the three-parameter burr distribution model based on changes in a small number of key traffic flow characteristic parameters, thereby establishing a three-parameter burr distribution model with parameter adaptability;

[0044] S31. Extract traffic flow characteristic variables with a 30-minute detection period based on radar-visual fusion trajectory data;

[0045] Traffic flow characteristic variables include traffic volume, average speed, occupancy rate, large vehicle penetration rate, and average vehicle length;

[0046] S32. Use the random forest model to evaluate the impact of the above five traffic flow characteristic variables on the three-parameter burr distribution;

[0047] S33. Calculate the importance index of each traffic flow characteristic variable;

[0048] S34. Select the three characteristic parameters that have the most significant impact on the model to construct a parameter-adaptive distribution model;

[0049] S35. An adaptive relationship between the parameters of the three-parameter burr distribution model of vehicle headway and the selected traffic flow characteristic variables is established through the MLP model, so that the parameters of the three-parameter burr distribution model can be automatically adjusted according to changes in traffic flow f, occupancy rate O, and large vehicle penetration rate P to fit the actual vehicle headway distribution.

[0050] It is hereby stated that the above embodiments are only used to illustrate the technical solutions of the present invention and are not limiting. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the purpose and scope of the technical solutions, which should all be included in the scope of the claims of the present invention.

Claims

1. A parameter-adaptive highway headway distribution estimation method, characterized in that: The following steps are involved: S1. Based on the radar-visual fusion trajectory data of the target section of the highway, extract the headway data of vehicles passing through different lanes of the target section; S2. Use the three-parameter Burr distribution model to fit the headway data of the target section and use the maximum likelihood estimation method to calculate the parameters of the three-parameter Burr distribution model; S2.1 uses the three-parameter burr distribution model to fit the headway data of the acquisition section; The probability density function of the three-parameter burr distribution model is expressed as follows: Where x≥0, a≥0, c≥0, k≥0, the three parameters are used to control the shape, location and scale of the distribution; c and k are shape parameters, and a is the scale parameter; S2.2 Use maximum likelihood estimation to calculate the parameters of the three-parameter burr distribution model; S3. Establish a three-parameter burr distribution model for headway with parameter adaptability; S31. Extract traffic flow characteristic variables with a 30-minute detection period based on radar-visual fusion trajectory data; The traffic flow characteristic variables include traffic volume, average speed, occupancy rate, large vehicle penetration rate and average vehicle length; S32. Use the random forest model to evaluate the impact of the above five traffic flow characteristic variables on the three-parameter burr distribution; S33. Calculate the importance index of each traffic flow characteristic variable; S34. Select the three characteristic parameters that have the most significant impact on the model to construct a parameter-adaptive distribution model; S35. An adaptive relationship between the parameters of the three-parameter burr distribution model of vehicle headway and the selected traffic flow characteristic variables is established through the MLP model, so that the parameters of the three-parameter burr distribution model can be automatically adjusted according to changes in traffic flow f, occupancy rate O, and large vehicle penetration rate P to fit the actual vehicle headway distribution.

2. The parameter-adaptive highway headway distribution estimation method according to claim 1, characterized in that: In step S1, the headway data of vehicles in different lanes are calculated using the following expression: T i =t i -t i-1 Where, t i is the moment when the current vehicle passes through the section, t i-1 It is the moment when the preceding vehicle in the same lane passes the section.

Images