Hybrid traffic fusion simulation system based on digital base

Abstract

The invention discloses a mixed traffic fusion simulation system based on a digital base, and relates to the field of mixed traffic analysis, and the system comprises a mixed traffic model and a digital base fusion simulation frame. The digital base fusion simulation framework provides a base for the mixed traffic model; the mixed traffic model comprises an information layer and a model iteration layer; the digital base fusion simulation framework comprises a data access layer, a business logic layer and a user interaction interface layer, an IDM model is optimized, an IDM-AEB model is provided, a mixed traffic flow model capable of being optimized in real time through an artificial intelligence algorithm is constructed, the overall framework of the digital base fusion simulation system is elaborated in detail, and the system has the advantages of being high in practicability and easy to popularize. A mixed traffic flow simulation standardization process based on a digital base is established, an efficient and reliable platform is provided for traffic state analysis of the mixed traffic flow, and closed-loop simulation of human-vehicle-road-environment simulation is achieved.

Description

Technical Field

[0001] The invention relates to the field of mixed traffic analysis, in particular to a mixed traffic fusion simulation system based on a digital base. Background Art

[0002] Mixed traffic flows are complex and changeable, and their impact on traffic flow and safety is unclear. There is no consensus or standard procedures on how to comprehensively test and evaluate them. The development of autonomous driving and vehicle-road collaboration technologies indicates that mixed traffic with intelligent connected vehicles (CAVs) and human-driven vehicles (HVs) will be common in the future. This mixed driving scenario is complex and changeable, and its impact on traffic efficiency and safety is unclear. There is no consensus or standard procedures on how to comprehensively test and evaluate it.

[0003] CAVs and HVs have different behavioral characteristics. Scholars have used a variety of systems to model vehicles. In the HV field, the GHR model assumes that the acceleration of the vehicle is positively correlated with the speed difference, and the response of the following vehicle is related to its acceleration or deceleration behavior. The safe distance Gipps model assumes that the driver controls the speed so that the vehicle can stop safely in the event of a sudden brake of the vehicle in front, and can describe the behavior of vehicles maintaining a safe distance in the traffic flow. The optimal velocity model (OVM) can describe many properties of real traffic flow, such as traffic flow instability, traffic congestion, etc. With the development of CAVs, ACC and CACC models have emerged. A cellular automaton model is used to determine the driving behavior of ACC vehicles and HV vehicles, which can reveal the characteristics of mixed traffic flow. An agent-based microscopic traffic simulation model is designed to study the impact of CACC vehicles on traffic flow characteristics. However, the new traffic flow characteristics that appear on the road when CAVs and HVs are mixed still need to be further studied.

[0004] Therefore, there is an urgent need for a hybrid traffic fusion simulation system based on a digital base to solve the above problems. Summary of the invention

[0005] To solve the above problems, this application provides a hybrid traffic fusion simulation system based on a digital base, explores the feasibility of applying the digital base to mixed traffic analysis, provides an efficient and reliable comprehensive simulation platform for traffic analysis of mixed traffic flows, and realizes a closed-loop simulation of people-vehicle-road-environment.

[0006] In the present application, a hybrid traffic fusion simulation system based on a digital base is provided, including a hybrid traffic model and a digital base fusion simulation framework;

[0007] The digital base fusion simulation framework provides a basis for the hybrid traffic model;

[0008] The hybrid traffic model includes an information layer and a model iteration layer, wherein the information layer provides facility information for the model iteration layer;

[0009] The digital base fusion simulation framework includes a data access layer, a business logic layer and a user interaction interface layer.

[0010] Preferably, the information layer is used to obtain infrastructure information, and the infrastructure information includes static infrastructure information and dynamic infrastructure information;

[0011] The model iteration layer includes a vehicle traffic flow simulation module, a model parameter and facility information storage and learning module, and a fitting iteration module;

[0012] The vehicle traffic flow simulation module includes a human-driven vehicle traffic flow simulation unit and an autonomous driving vehicle traffic flow simulation unit;

[0013] The human-driven vehicle traffic flow simulation unit includes the improved human-driven following sub-model IDM-AEB;

[0014] The model parameter and facility information storage and learning module includes a weight network GNN neural network, a LSTM neural network and a CNN neural network, which are used to train the parameters of the vehicle traffic flow simulation module;

[0015] The fitting iteration module includes a decision network, a DQN learning network, a DDPG learning network, and an A3C learning network, which are used to update the vehicle traffic flow simulation module and parameters.

[0016] Preferably, the specific contents of improving the human driving following sub-model IDM-AEB are:

[0017] The acceleration equation of the vehicle in the improved human driving following sub-model is as follows:

[0018]

[0019] Where: a is the maximum acceleration of the vehicle, v is the current speed of the vehicle, and v 0 is the desired speed of the vehicle, δ is the acceleration index, Δv is the speed difference between the vehicle and the front vehicle, s is the current distance between the vehicle and the front vehicle, and s * (v,Δv) is the expected following distance.

[0020] The expression of the expected vehicle distance of the IDM-AEB model is:

[0021]

[0022] Among them, s * (v,Δv) is the expected following distance, v is the current speed of the vehicle, and v 0is the desired speed of the ego vehicle, Δv is the speed difference between the ego vehicle and the front vehicle, s is the current distance between the ego vehicle and the front vehicle, TTC is the conflict collision time, a is the maximum acceleration of the ego vehicle, b is the comfortable deceleration, s 0 The minimum vehicle distance.

[0023] Preferably, the data access layer supports real-time input of multi-configuration data, including simulated AC data input and 3D engine simulated driving data input;

[0024] Simulation communication data input combines SUMO, VISSIM, AirSim and Unity3D software to provide traffic simulation data flow;

[0025] The 3D engine simulated driving data input combines Unreal, Unity3D and SiLab simulation rendering engines to provide simulated driving data sources.

[0026] Preferably, the business logic layer includes a simulated driving simulation fusion terminal and a multi-language interactive client;

[0027] The simulated driving simulation fusion terminal is integrated with a mixed traffic model.

[0028] Preferably, the user interaction interface layer includes a simulated traffic flow and 3D simulated driving data interaction platform;

[0029] The simulated traffic flow and 3D simulated driving data interaction platform includes a simulated traffic flow data control platform and a 3D simulated driving data control platform;

[0030] The user interaction interface layer is associated with HTML client, C++ client, Python client and other clients.

[0031] Preferably, the hybrid traffic fusion simulation system based on the digital base uses SUMO as the simulation engine for micro traffic simulation and selects CARLA as the simulation engine for vehicles and traffic scenes.

[0032] Preferably, the data access layer selects TCP / IP protocol to implement data exchange of fusion simulation.

[0033] In summary, compared with traditional technologies, the hybrid traffic fusion simulation system based on the digital base of the present invention optimizes the intelligent driver model (IDM) to form an IDM-AEB model, and theoretically constructs a hybrid traffic flow model. In order to establish a standardized process for micro-traffic simulation and driving fusion simulation in hybrid traffic scenarios, a fusion simulation system framework based on the digital base is established. The feasibility of applying the digital base to hybrid traffic analysis is explored, providing an efficient and reliable comprehensive simulation platform for traffic analysis of hybrid traffic flows, and realizing a closed-loop simulation of people-vehicle-road-environment.

[0034] In addition, this study solved the defect that the IDM model is not suitable for actual road conditions and proposed the IDM-AEB model. Then, the study successfully constructed and verified the traffic flow model under mixed traffic scenarios, and conducted a detailed analysis of the traffic flow model under different environments and penetration rates by integrating different vehicle following models.

[0035] This study established a fusion simulation system framework to achieve high-fidelity simulation and traffic analysis of mixed driving scenarios in a simulation environment. It solved the defects that exist when using microscopic traffic simulation and driving simulation separately, and can provide an effective platform support and experimental basis for subsequent simulation experiments.

[0036] This paper proves the feasibility of applying digital base to mixed traffic analysis and proposes a standardized process for integrating micro-traffic simulation and driving simulation in mixed traffic scenarios, which can improve the efficiency of mixed traffic simulation. It is of great significance for the construction of digital base in the field of transportation in the intelligent era.

[0037] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 It is a framework diagram of a hybrid traffic fusion simulation system based on a digital base of the present invention;

[0039] Figure 2 It is a module diagram of the mixed traffic model of the present invention;

[0040] Figure 3 It is a comparison chart of statistical indicators of the simulation results of the present invention;

[0041] Figure 4 It is a speed-flow relationship diagram in the mixed driving scenario of the present invention;

[0042] Figure 5 This is a flow-density relationship diagram under the mixed driving scenario of the present invention;

[0043] Figure 6 This is a density-speed relationship diagram for the mixed driving scenario of the present invention. DETAILED DESCRIPTION

[0044] The technical solution of the present invention is further described below by means of the accompanying drawings and embodiments. It should be noted that unless otherwise specifically stated, the relative arrangement of components and steps, numerical expressions and numerical values ​​described in these embodiments do not limit the scope of the present application.

[0045] The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the present application, its application, or uses.

[0046] Technologies, systems, and devices known to ordinary technicians in the relevant art may not be discussed in detail, but where appropriate, the technologies, systems, and devices should be considered part of the specification.

[0047] In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not limiting. Therefore, other examples of the exemplary embodiments may have different values.

[0048] Unless otherwise defined, technical or scientific terms used in the present invention shall have the common meanings understood by one having ordinary skills in the field to which the present invention belongs.

[0049] like Figure 1 As shown, the present invention provides a hybrid traffic fusion simulation system based on a digital base, including a hybrid traffic model and a digital base fusion simulation framework;

[0050] The digital base fusion simulation framework provides a basis for the hybrid traffic model;

[0051] The hybrid traffic model includes an information layer and a model iteration layer, wherein the information layer provides facility information for the model iteration layer;

[0052] Furthermore, the information layer is used to obtain infrastructure information, and the infrastructure information includes static infrastructure information and dynamic infrastructure information;

[0053] The model iteration layer includes a vehicle traffic flow simulation module, a model parameter and facility information storage and learning module, and a fitting iteration module;

[0054] The vehicle traffic flow simulation module includes a human-driven vehicle traffic flow simulation unit and an autonomous driving vehicle traffic flow simulation unit;

[0055] The human-driven vehicle traffic flow simulation unit includes the improved human-driven following sub-model IDM-AEB;

[0056] The model parameter and facility information storage and learning module includes a weight network GNN neural network, a LSTM neural network and a CNN neural network, which are used to train the parameters of the vehicle traffic flow simulation module;

[0057] The fitting iteration module includes a decision network, a DQN learning network, a DDPG learning network, and an A3C learning network, which are used to update the vehicle traffic flow simulation module and parameters.

[0058] Furthermore, the specific contents of improving the human driving following sub-model IDM-AEB are as follows:

[0059] The acceleration equation of the vehicle in the improved human driving following sub-model is as follows:

[0060]

[0061] Where: a is the maximum acceleration of the vehicle, v is the current speed of the vehicle, and v 0 is the desired speed of the vehicle, δ is the acceleration index, Δv is the speed difference between the vehicle and the front vehicle, s is the current distance between the vehicle and the front vehicle, and s * (v,Δv) is the expected following distance.

[0062] The expression of the expected vehicle distance of the IDM-AEB model is:

[0063]

[0064] Among them, s * (v,Δv) is the expected following distance, v is the current speed of the vehicle, and v 0 is the desired speed of the ego vehicle, Δv is the speed difference between the ego vehicle and the front vehicle, s is the current distance between the ego vehicle and the front vehicle, TTC is the conflict collision time, a is the maximum acceleration of the ego vehicle, b is the comfortable deceleration, s 0 The minimum vehicle distance.

[0065] The digital base fusion simulation framework includes a data access layer, a business logic layer and a user interaction interface layer.

[0066] Furthermore, the data access layer supports real-time input of multi-configuration data, including simulated AC data input and 3D engine simulated driving data input;

[0067] Simulation communication data input combines SUMO, VISSIM, AirSim and Unity3D software to provide traffic simulation data flow;

[0068] The 3D engine simulated driving data input combines Unreal, Unity3D and SiLab simulation rendering engines to provide simulated driving data sources.

[0069] Furthermore, the business logic layer includes a simulated driving simulation fusion terminal and a multi-language interactive client;

[0070] The simulated driving simulation fusion terminal is integrated with a mixed traffic model.

[0071] Furthermore, the user interaction interface layer includes a simulated traffic flow and 3D simulated driving data interaction platform;

[0072] The simulated traffic flow and 3D simulated driving data interaction platform includes a simulated traffic flow data control platform and a 3D simulated driving data control platform;

[0073] The user interaction interface layer is associated with HTML client, C++ client, Python client and other clients.

[0074] Furthermore, the hybrid traffic fusion simulation system based on the digital base uses SUMO as the simulation engine for micro traffic simulation and selects CARLA as the simulation engine for vehicles and traffic scenarios.

[0075] Furthermore, the data access layer selects TCP / IP protocol to implement data exchange of fusion simulation.

[0076] To verify the simulation effect of the IDM-AEB model, this paper selects the HighD dataset as the natural driving dataset and compares the IDM model with the IDM-AEB model proposed in this paper. For the model setting, the 30% quantile of the speed in its statistical significance is used as the maximum speed limit, that is, it is set to 160km / h. The headway time (THW) parameter in the IDM is modified to the conflict collision time (TTC) parameter in the IDM-AEB model, and the other initialization parameters remain the same.

[0077] Through simulation experiments, the statistical indicators of the simulation results are shown in Table 1. The statistical results of the HighD dataset, IDM model and IDM-AEB model are shown in Table 1. Figure 3 , Figure 3 a is the velocity distribution histogram of the three. Figure 3 Figure b is the acceleration distribution histogram of the three models. It can be seen that the speed data of the IDM model has a higher variance and standard deviation, indicating that the speed data is more volatile, while the average vehicle speed of the IDM-AEB model is closer to the average vehicle speed in the HighD data set. The speed distribution results have a good fit except for the large difference between 80-100km / h. The variance and standard deviation of the acceleration data of the three models are not much different, indicating that the stability of the acceleration data is relatively high in the IDM and IDM-AEB models, and the performance is good.

[0078] Table 1

[0079]

[0080]

[0081] In order to verify the effect of the hybrid traffic fusion simulation system based on the digital base, a mixed traffic scenario was created, and the IDM-AEB model and CACC model were loaded for simulation analysis, and the experimental results of the basic diagram of traffic flow in the hybrid driving scenario under the fusion simulation environment were obtained.

[0082] The impact of different autonomous driving penetration rates on traffic speed and flow in mixed driving scenarios Figure 4 As shown, Figure 4 Figure a is the speed-flow relationship diagram in a single lane scenario. Figure 4 Figure b is the speed-flow relationship diagram in a two-lane scenario. Figure 4 Figure c is a speed-flow relationship diagram in a three-lane scenario. When the flow is small, the vehicle speed is close to 120 km / h, indicating that most vehicles are in the free flow stage and are not affected or constrained by the operation of other vehicles. When the traffic volume increases, the speed tends to decrease due to the mutual constraints between vehicles. The higher the penetration rate, the more gradual the trend of vehicle speed decreasing with the increase in traffic volume, indicating that on a single lane, the higher the penetration rate of autonomous driving, the higher the speed under high traffic conditions. As the number of lanes increases, such as Figure 4 (b)(c), vehicles can change lanes. Therefore, when the penetration rate of autonomous driving is low, the vehicle speed decreases more significantly with the increase in traffic. However, when the penetration rate of autonomous driving is high, lane selection is more reasonable and the vehicle speed is relatively less affected.

[0083] Figure 5 The impact of different autonomous driving penetration rates on traffic density and flow in mixed driving scenarios is demonstrated. Figure 5 Figure a is a flow-density relationship diagram in a single lane scenario. Figure 5 Figure b is the traffic-density relationship diagram in a two-lane scenario. Figure 5 Figure c is a flow-density relationship diagram for a three-lane scenario. In the case of low penetration, the relationship between traffic density and flow shows a steeper increase, indicating that a small increase in traffic density will lead to a significant increase in flow. As penetration increases, the curve flattens, indicating that the addition of autonomous vehicles can effectively improve the lane capacity, that is, it can handle higher traffic at the same traffic density. In two-lane and three-lane environments, such as Figure 5 (b)(c), vehicles have more maneuvering space, so at low penetration, the increase in density can bring higher flow improvement. This may be because vehicles can avoid diverting traffic by changing lanes and reducing congestion.

[0084] Figure 6 It depicts the relationship between speed and density at different autonomous driving penetration rates. Figure 6 Figure a is the density-speed relationship diagram in a single lane scenario. Figure 6 Figure b is the density-speed relationship diagram in a two-lane scenario. Figure 6Figure c is a density-speed relationship diagram for a three-lane scenario. At low penetration rates, vehicle speed drops rapidly as traffic density increases, reflecting that in an environment dominated by manual driving, a small increase in traffic density will significantly affect vehicle speed. In the case of high penetration rates, the downward trend in vehicle speed becomes more gradual. This means that autonomous vehicles are able to maintain higher speeds at higher traffic densities, which may be due to the fact that the more precise and consistent driving behavior of autonomous vehicles reduces fluctuations and instabilities in traffic flow. This shows that the application of autonomous driving technology can help improve the smoothness of high-density traffic flow.

[0085] Table 2 provides the statistics of basic traffic elements at different penetration rates of autonomous vehicles, showing the changes in maximum speed, theoretical maximum density, and theoretical maximum flow. It can be seen that as the penetration rate of autonomous vehicles increases, although the maximum speed may be slightly affected, the theoretical maximum density and theoretical maximum flow show an upward trend. This shows that autonomous vehicles can improve the overall performance of the transportation system by improving traffic mobility and reducing the negative impact of density on speed, thereby increasing road capacity and traffic flow efficiency.

[0086] Table 2

[0087]

[0088]

[0089] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that they can still modify or replace the technical solution of the present invention with equivalents, and these modifications or equivalent replacements cannot cause the modified technical solution to deviate from the spirit and scope of the technical solution of the present invention.

Claims

1. A hybrid traffic fusion simulation system based on a digital base, characterized in that: Includes hybrid traffic model and digital base fusion simulation framework; The digital base fusion simulation framework provides a basis for the hybrid traffic model; The hybrid traffic model includes an information layer and a model iteration layer, wherein the information layer provides facility information for the model iteration layer; The digital base fusion simulation framework includes a data access layer, a business logic layer and a user interaction interface layer.

2. A hybrid traffic fusion simulation system based on a digital base according to claim 1, characterized in that: The information layer is used to obtain infrastructure information, and the infrastructure information includes static infrastructure information and dynamic infrastructure information; The model iteration layer includes a vehicle traffic flow simulation module, a model parameter and facility information storage and learning module, and a fitting iteration module; The vehicle traffic flow simulation module includes a human-driven vehicle traffic flow simulation unit and an autonomous driving vehicle traffic flow simulation unit; The human-driven vehicle traffic flow simulation unit includes an improved human-driven following sub-model IDM-AEB; The model parameter and facility information storage and learning module includes a weight network GNN neural network, a LSTM neural network and a CNN neural network, which are used to train the parameters of the vehicle traffic flow simulation module; The fitting iteration module includes a decision network, a DQN learning network, a DDPG learning network, and an A3C learning network, which are used to update the vehicle traffic flow simulation module and parameters.

3. A hybrid traffic fusion simulation system based on a digital base according to claim 2, characterized in that: The specific contents of improving the human driving following sub-model IDM-AEB are as follows: The acceleration equation of the vehicle in the improved human driving following sub-model is as follows: Where: a is the maximum acceleration of the ego vehicle, v is the current speed of the ego vehicle, v0 is the expected speed of the ego vehicle, δ is the acceleration index, Δv is the speed difference between the ego vehicle and the front vehicle, s is the current distance between the ego vehicle and the front vehicle, and s * (v,Δv) is the expected following distance. The expression of the expected vehicle distance of the IDM-AEB model is: Among them, s * (v, Δv) is the expected following distance, v is the current speed of the vehicle, v0 is the expected speed of the vehicle, Δv is the speed difference between the vehicle and the front vehicle, s is the current distance between the vehicle and the front vehicle, TTC is the conflict collision time, a is the maximum acceleration of the vehicle, b is the comfortable deceleration, and s0 is the minimum distance.

4. A hybrid traffic fusion simulation system based on a digital base according to claim 2, characterized in that: The data access layer supports real-time input of multi-configuration data, including simulated communication data input and 3D engine simulated driving data input; Simulation communication data input combines SUMO, VISSIM, AirSim and Unity3D software to provide traffic simulation data flow; The 3D engine simulated driving data input combines Unreal, Unity3D and SiLab simulation rendering engines to provide simulated driving data sources.

5. The hybrid traffic fusion simulation system based on a digital base according to claim 2 is characterized in that: The business logic layer includes the simulated driving simulation fusion terminal and the multi-language interactive client; The simulated driving simulation fusion terminal is integrated with a mixed traffic model.

6. The hybrid traffic fusion simulation system based on a digital base according to claim 1 is characterized in that: The user interaction interface layer includes a simulated traffic flow and 3D simulated driving data interaction platform; The simulated traffic flow and 3D simulated driving data interaction platform includes a simulated traffic flow data control platform and a 3D simulated driving data control platform; The user interaction interface layer is associated with HTML client, C++ client, Python client and other clients.

7. A hybrid traffic fusion simulation system based on a digital base according to claim 6, characterized in that: The hybrid traffic fusion simulation system based on the digital base uses SUMO as the simulation engine for micro traffic simulation and selects CARLA as the simulation engine for vehicles and traffic scenes.

8. The hybrid traffic fusion simulation system based on a digital base according to claim 4 is characterized in that: The data access layer selects TCP / IP protocol to realize data exchange of fusion simulation.

Images