Hybrid traffic flow simulation system and method for port intelligent driving

Abstract

The invention relates to the technical field of port simulation, and particularly discloses a mixed traffic flow simulation system and method for port intelligent driving, and the system comprises a data twin display module, a simulation operation module, a scene rendering module, an AI fusion base, and an instruction distribution module. The simulation method comprises the following steps: step 1, preparation before simulation; step 2, scene construction and virtual port establishment; step 3, importing data; and step 4, task scheduling and distribution. The method has the beneficial effects that analogue simulation is carried out based on historical traffic flow data, a vehicle model and a port model, the mixed traffic flow integrating the unmanned vehicle, the outer container truck vehicle and the inner container truck vehicle is established, the disturbance event is introduced to interfere the mixed traffic flow, so that the simulation operation data of the unmanned vehicle under the disturbance event is obtained, and the disturbance event is simulated. Through analysis and optimization of the data, prediction and deduction data can be provided for operation scheduling of the unmanned vehicle, and the scheduling efficiency of the unmanned vehicle can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of port simulation, and in particular to a hybrid traffic flow simulation system and method for port intelligent driving. Background Art

[0002] In recent years, road mixed traffic flow simulation based on multi-agent systems (MAS) has become a research hotspot. This type of simulation system can simulate the interactive behaviors between different types of traffic participants (such as motor vehicles, non-motor vehicles, pedestrians, etc.), and reproduce complex mixed traffic conditions by simulating the decision-making process of each individual. For example, the use of multi-agents to simulate different driving strategies, reaction times, and path selection behaviors improves the authenticity and accuracy of the simulation model. Advanced traffic flow simulation software, such as VISSIM and SUMO, is continuously updated and developed, providing a powerful graphical interface and a rich model library, supporting user-defined models, making the simulation of mixed traffic flow more intuitive and efficient. These software are widely used in urban traffic planning, traffic design evaluation, autonomous driving technology testing and other fields, becoming a bridge between theoretical research and practical application. With the popularization of electric vehicles, mixed traffic flow simulation has also begun to consider the charging behavior of electric vehicles and its impact on traffic flow, as well as the potential improvement of electric vehicle fleets on urban energy consumption and air quality. In addition, simulation models are also used to evaluate the contribution of different traffic policies to emission reduction targets, such as the policy effect of encouraging non-motorized transportation.

[0003] In the port scenario, the particularity of the port scenario should be taken into consideration. For example, there are no traffic lights, the roads are narrow, and there are many trailers. Considering factors such as unrestricted traffic control, with the construction of the port area's autonomous driving demonstration area, the intelligent driving vehicle simulation technology for the port has also begun to develop gradually. However, the existing vehicle simulation model pays too much attention to the kinematic model and mechanical model of the vehicle itself, and the efficiency and operational issues that are concerned about the port scenario cannot be reflected. The detour of interference events is too random. The use of machine learning deduction capabilities cannot be combined with the existing actual operating conditions to obtain interference events that are more in line with the timeline. The simulation results are not analyzed in combination with the actual situation of the port area, and the subsequent operation plan construction cannot be supported. Summary of the invention

[0004] The present invention proposes a hybrid traffic flow simulation system and method for port intelligent driving in order to solve the above problems.

[0005] The technical solution of the present invention is achieved in this way:

[0006] A hybrid traffic flow simulation system and method for port intelligent driving, wherein the simulation system includes a data twin display module, a simulation operation module, a scene rendering module, an AI fusion base, and an instruction distribution module;

[0007] Digital twin display module: Based on the web UI and UE5 engine, a three-dimensional scene is created to display the simulation process in a data-driven way in the port modeling scene. The operation of vehicles in the port area and the distribution of the overall traffic flow can be seen. The simulation process can be replayed and accelerated, so that port personnel can intuitively see the simulation results and analyze the mixed traffic flow in the simulation.

[0008] Simulation operation module: Based on traffic flow simulation technology, the unmanned truck model generated by the AI ​​fusion base and the interference models of various other traffic participants are imported, and interference events generated based on neural network technology are used to simulate mixed traffic flow conditions in the port area;

[0009] Scene rendering module: Based on scene modeling technology, the port scene model is generated according to the port's structural distribution and CAD drawings, and the lanes, yards, dock surfaces and other areas that affect traffic flow in the port area are segmented and modeled to provide a basic operation scene model for simulation operation;

[0010] AI Fusion Base: Based on deep learning and neural network technology, by introducing historical data of the port area, combined with mathematical modeling capabilities and traffic flow simulation technology, the vehicle kinematic model and vehicle data model in the time dimension, as well as the interference model of various interference events, are obtained as the basis for subsequent simulation operations;

[0011] Task scheduling module: Based on the port area's transportation plan and the dispatching task scheduling data of the TOS port dispatching system, the vehicle's operation tasks are generated according to the daily operation plan and provided to the instruction distribution module as a data support basis;

[0012] Instruction distribution module: obtains the job instruction data of each node by parsing the task data, issues job instructions to each model, and instructs each simulated model to install instructions to perform simulation operations;

[0013] The simulation method comprises the following steps:

[0014] Step 1: Preparation before simulation: start the simulation system, load the initialization settings, and configure the integration framework so that the AI ​​fusion base can receive and process data streams from different models;

[0015] Step 2: Build the scene, establish the virtual port, import the map data and build the port model;

[0016] Step 3: Data import: import vehicle information and historical traffic information into the AI ​​fusion base to generate a kinematic model of the vehicle;

[0017] Step 4: Task scheduling and distribution: import the transportation plan into the AI ​​fusion base, generate operation instructions based on the transportation plan, and send the operation instructions to the unmanned truck model;

[0018] Step 5: Traffic flow simulation: the unmanned truck model moves in the port model according to the operation instructions to form traffic flow. According to the historical traffic information, the external truck model and the internal truck model are imported into the port model to form a mixed traffic flow. The driving status of the unmanned truck model is observed to obtain the simulation data of each model.

[0019] Step 6: Generate disturbance events. When the unmanned truck model completes the operation instruction, import the disturbance events into the port model to perform operation disturbance, affect the traffic flow, observe the driving status of the unmanned truck model, and obtain simulation data of each model.

[0020] Step 7: Complete the operation instructions. The unmanned vehicle model completes the loading, transportation and unloading operation simulation according to different operation instructions, and updates the vehicle status and task status;

[0021] Step 8: Generate a simulation report. When the preset simulation time is reached or other termination conditions are met, the simulation is terminated. All data generated during the simulation process are collected and sorted. The simulation results are analyzed in depth using statistical analysis, visualization technology, etc., the effectiveness and accuracy of the simulation model are evaluated, and valuable insights and suggestions are extracted to generate a simulation report.

[0022] Step 9: Optimize and adjust the simulation model. During the simulation process, data such as traffic flow, vehicle status, and task completion are collected in real time. Based on these data, the simulation model is dynamically adjusted and optimized.

[0023] Furthermore, the generation of the work instructions in step 4 is based on the transportation plan and the current traffic conditions, and the optimized work tasks are assigned to the vehicles through an optimization algorithm, and the optimization algorithm takes minimizing the total transportation time or cost as the optimization goal.

[0024] By adopting the above technical scheme, the beneficial effects of the present invention are as follows: the present invention performs simulation based on historical traffic data, vehicle models, and port models, establishes a mixed traffic flow that combines unmanned vehicles, external container trucks, and internal container trucks, and interferes with the mixed traffic flow by introducing disturbance events, thereby obtaining simulated operation data of unmanned vehicles under disturbance events, and by analyzing and optimizing the data, it can provide predictive deduction data for the operation scheduling of unmanned vehicles, thereby facilitating improving the scheduling efficiency of unmanned vehicles. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0026] Figure 1 It is a simulation system framework diagram of the present invention;

[0027] Figure 2 It is a flow chart of the simulation method of the present invention. DETAILED DESCRIPTION

[0028] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the 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 the present invention.

[0029] like Figure 1-Figure 2 As shown, a hybrid traffic flow simulation system and method for port intelligent driving, the simulation system includes a data twin display module, a simulation operation module, a scene rendering module, an AI fusion base, and an instruction distribution module;

[0030] Digital twin display module: Based on the web UI and UE5 engine, a three-dimensional scene is created to display the simulation process in a data-driven way in the port modeling scene. The operation of vehicles in the port area and the distribution of the overall traffic flow can be seen. The simulation process can be replayed and accelerated, so that port personnel can intuitively see the simulation results and analyze the mixed traffic flow in the simulation.

[0031] Simulation operation module: Based on traffic flow simulation technology, the unmanned truck model generated by the AI ​​fusion base and the interference models of various other traffic participants are imported, and interference events generated based on neural network technology are used to simulate mixed traffic flow conditions in the port area;

[0032] Scene rendering module: Based on scene modeling technology, the port scene model is generated according to the port's structural distribution and CAD drawings, and the lanes, yards, dock surfaces and other areas that affect traffic flow in the port area are segmented and modeled to provide a basic operation scene model for simulation operation;

[0033] AI Fusion Base: Based on deep learning and neural network technology, by introducing historical data of the port area, combined with mathematical modeling capabilities and traffic flow simulation technology, the vehicle kinematic model and vehicle data model in the time dimension, as well as the interference model of various interference events, are obtained as the basis for subsequent simulation operations;

[0034] Task scheduling module: Based on the port area's transportation plan and the dispatching task scheduling data of the TOS port dispatching system, the vehicle's operation tasks are generated according to the daily operation plan and provided to the instruction distribution module as a data support basis;

[0035] Instruction distribution module: obtains the operation instruction data of each node by parsing the task data, issues operation instructions to each model, including the destination, container number to be transferred, etc., and instructs each simulated model to install the instructions to perform simulation operations;

[0036] The simulation method comprises the following steps:

[0037] Step 1: Preparation before simulation: Start the simulation system, load the initialization settings, including single simulation time, time step, simulation end conditions, etc., configure the integration framework, and define the data exchange protocol and interface standards so that the AI ​​fusion base can receive and process data streams from different models;

[0038] Step 2: Build the scene and virtual port, import map data and use the map generation algorithm to build the road network and port layout, including road length, width, number of lanes, intersection location, traffic light information, port shape length, yard shape, location capacity, throughput, etc., to build the port model;

[0039] Step 3: Data import, import vehicle information (vehicle type, size, initial position, speed, acceleration, braking parameters, initial power, energy consumption, etc.), historical traffic information to the AI ​​fusion base, generate the vehicle's kinematic model, including unmanned truck vehicle model, external truck vehicle model, and internal truck vehicle model. Each vehicle has a unique ID. By importing vehicle information, a vehicle behavior prediction model can be formed to simulate the movement process of the predicted vehicle at different speeds. Importing historical vehicle data can analyze traffic flow, average speed, etc. These data can be used to verify the accuracy of the model, and can also be used to add vehicle data to the traffic flow to form mixed traffic;

[0040] Step 4: Task scheduling and distribution: import the transportation plan into the AI ​​fusion base, generate operation instructions based on the transportation plan, and send the operation instructions to the unmanned truck model;

[0041] Step 5: Traffic flow simulation: the unmanned truck model moves in the port model according to the operation instructions to form traffic flow. According to the historical traffic information, the external truck model and the internal truck model are imported into the port model to form a mixed traffic flow. The driving status of the unmanned truck model is observed to obtain the simulation data of each model.

[0042] Step 5.1: At each time step, update the positions and velocities of all vehicles;

[0043] Step 5.2: Use a traffic flow model to simulate the interaction between vehicles, including following, lane changing, crossing, etc. The traffic flow model used includes any one of the LWR model, macro model, and micro model;

[0044] Step 5.3: Calculate indicators in traffic flow, such as the number of vehicles passing a certain end surface per unit time (flow), the number of vehicles per unit length of road (density), the average speed of vehicles on the road (speed), etc.;

[0045] Step 6: Generate disturbance events. When the unmanned truck model completes the operation instruction, the disturbance event is imported into the port model to perform operation disturbance, affect the traffic flow, observe the driving status of the unmanned truck model, and obtain the simulation data of each model. The generation probability of the disturbance event is generated according to the probability distribution (Poisson distribution is used for the accident rate). For example, if a traffic accident causes lane congestion or closure, the traffic wave theory such as the traffic wave propagation speed formula under the Greenhill speed-density linear relationship is used to simulate the behavior of the vehicle;

[0046] Step 7: Complete the operation instructions. The unmanned vehicle model completes the loading, transportation and unloading operation simulation according to different operation instructions. This includes dispatching the vehicle to the designated loading / unloading area, simulating the waiting state required for the loading / unloading process, updating the vehicle status and task status, such as removing the completed task from the task list, updating the vehicle's cargo status, etc., so as to simulate the port operation process within the set time;

[0047] Step 8: Generate a simulation report. When the preset simulation time is reached or other termination conditions are met, the simulation ends. Collect and organize all data generated during the simulation process, including traffic flow data, vehicle trajectories, task completion status, etc. Use statistical analysis, visualization technology and other means to conduct in-depth analysis of the simulation results, evaluate the effectiveness and accuracy of the simulation model, extract valuable insights and suggestions, generate a simulation report, summarize the simulation process, results and analysis conclusions, and provide reference and support for decision makers.

[0048] Step 9: Optimize and adjust the simulation model. During the simulation process, data such as traffic flow, vehicle status, and task completion are collected in real time. Based on these data, the simulation model is dynamically adjusted and optimized, such as optimizing task scheduling and adjusting traffic signal control strategies, to cope with changes in actual conditions. Real-time data analysis techniques (such as data mining and machine learning) are used to identify potential problems and opportunities and take corresponding measures.

[0049] In this embodiment, the generation of the work instructions in step 4 is based on the transportation plan and the current traffic conditions, and the optimized work tasks are assigned to the vehicles through an optimization algorithm, and the optimization algorithm takes minimizing the total transportation time or cost as the optimization goal.

[0050] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A hybrid traffic flow simulation system and method for port intelligent driving, characterized by: The simulation system includes a data twin display module, a simulation operation module, a scene rendering module, an AI fusion base, and an instruction distribution module; Digital twin display module: Based on the web UI and UE5 engine, a three-dimensional scene is created to display the simulation process in a data-driven way in the port modeling scene. The operation of vehicles in the port area and the distribution of the overall traffic flow can be seen. The simulation process can be replayed and accelerated, so that port personnel can intuitively see the simulation results and analyze the mixed traffic flow in the simulation. Simulation operation module: Based on traffic flow simulation technology, the unmanned truck model generated by the AI ​​fusion base and the interference models of various other traffic participants are imported, and interference events generated based on neural network technology are used to simulate mixed traffic flow conditions in the port area; Scene rendering module: Based on scene modeling technology, the port scene model is generated according to the port's structural distribution and CAD drawings, and the lanes, yards, dock surfaces and other areas that affect traffic flow in the port area are segmented and modeled to provide a basic operation scene model for simulation operation; AI Fusion Base: Based on deep learning and neural network technology, by introducing historical data of the port area, combined with mathematical modeling capabilities and traffic flow simulation technology, the vehicle kinematic model and vehicle data model in the time dimension, as well as the interference model of various interference events, are obtained as the basis for subsequent simulation operations; Task scheduling module: Based on the port area's transportation plan and the dispatching task scheduling data of the TOS port dispatching system, the vehicle's operation tasks are generated according to the daily operation plan and provided to the instruction distribution module as a data support basis; Instruction distribution module: obtains the job instruction data of each node by parsing the task data, issues job instructions to each model, and instructs each simulated model to install instructions to perform simulation operations; The simulation method comprises the following steps: Step 1: Preparation before simulation: start the simulation system, load the initialization settings, and configure the integration framework so that the AI ​​fusion base can receive and process data streams from different models; Step 2: Build the scene, establish the virtual port, import the map data and build the port model; Step 3: Data import: import vehicle information and historical traffic information into the AI ​​fusion base to generate a kinematic model of the vehicle; Step 4: Task scheduling and distribution: import the transportation plan into the AI ​​fusion base, generate operation instructions based on the transportation plan, and send the operation instructions to the unmanned truck model; Step 5: Traffic flow simulation: the unmanned truck model moves in the port model according to the operation instructions to form traffic flow. According to the historical traffic information, the external truck model and the internal truck model are imported into the port model to form a mixed traffic flow. The driving status of the unmanned truck model is observed to obtain the simulation data of each model. Step 6: Generate disturbance events. When the unmanned truck model completes the operation instruction, import the disturbance events into the port model to perform operation disturbance, affect the traffic flow, observe the driving status of the unmanned truck model, and obtain simulation data of each model. Step 7: Complete the operation instructions. The unmanned vehicle model completes the loading, transportation and unloading operation simulation according to different operation instructions, and updates the vehicle status and task status; Step 8: Generate a simulation report. When the preset simulation time is reached or other termination conditions are met, the simulation is terminated. All data generated during the simulation process are collected and sorted. The simulation results are analyzed in depth using statistical analysis, visualization technology, etc., the effectiveness and accuracy of the simulation model are evaluated, and valuable insights and suggestions are extracted to generate a simulation report. Step 9: Optimize and adjust the simulation model. During the simulation process, data such as traffic flow, vehicle status, and task completion are collected in real time. Based on these data, the simulation model is dynamically adjusted and optimized.

2. The hybrid traffic flow simulation system and method for port intelligent driving according to claim 1 is characterized by: The generation of the work instructions in step 4 is based on the transportation plan and the current traffic conditions. The optimized work tasks are assigned to the vehicles through an optimization algorithm, and the optimization algorithm takes minimizing the total transportation time or cost as the optimization goal.

Images