A signal light detection method and tool based on RSU-like tool

Abstract

The present invention relates to the field of vehicle-road collaboration technology, and more specifically, to a signal light detection method and tool based on a RSU-like tool, the tool comprising a signal light learning machine and an industrial computer, and the method is as follows: a signal light strategy model is loaded on the signal light learning machine, and during the learning process, signal light status information is obtained from traffic lights in real time, and the signal light learning machine uses the signal light status information as a training sample to self-study the signal light strategy model; after obtaining the signal light status information, the signal light learning machine determines the signal light strategy of the intersection according to the trained signal light strategy model, and uploads the signal light strategy to the Internet of Vehicles application service platform in a standard format through the industrial computer. The present invention does not require improvement of existing traffic lights, and the information is parsed through the Internet of Vehicles application service platform, and the restored information is pushed to the vehicle-mounted terminal within the coverage area, so as to facilitate the driver and passengers to make application decisions.

Description

Technical Field

[0001] The present invention relates to the field of vehicle-road collaboration technology, and in particular to a signal light detection method and tool based on a RSU-like tool. Background Art

[0002] In the vehicle-road cooperative system, it is necessary to obtain the status of traffic lights. The Internet of Vehicles service platform needs to parse the received status of traffic lights, restore the traffic light information, and push the status of traffic lights to the smart terminals within the coverage area. However, the traffic lights at existing intersections are different in models. To realize the signal collection of traffic lights at intersections, more intelligent equipment needs to be added. The newly added intelligent equipment needs to adjust the existing traffic lights. Upgrading or modifying the existing traffic lights will inevitably increase the workload of secondary development and the subsequent maintenance costs. Summary of the invention

[0003] The purpose of the present invention is to provide a signal light detection method and tool based on RSU-like tools, which installs a signal light learning machine and an industrial computer device at the intersection, connects the existing traffic signal machine through the signal light learning machine, performs signal collection and calibration self-learning to obtain the status and status duration information of the signal light, interconnects the signal light learning machine through the industrial computer, formats the data information according to the standard format, and uploads the information to the Internet of Vehicles application service platform. The Internet of Vehicles application service platform parses the information, restores the signal light information, and pushes the information to the vehicle-mounted terminal within the coverage area, so as to facilitate the driver and passengers to make application decisions, thereby solving the problems pointed out in the background technology.

[0004] The embodiment of the present invention is implemented by the following technical solution: a signal light detection method based on a RSU-like tool is applied to a RSU-like communication tool, wherein the RSU-like communication tool includes a signal light learning machine and an industrial computer, and the method is as follows:

[0005] Traffic light strategy self-learning steps:

[0006] The signal light status information is obtained from the traffic light in real time, and the signal light learning machine uses the signal light status information as a training sample to perform self-learning of the signal light strategy model loaded on the signal light learning machine;

[0007] Traffic light strategy detection steps:

[0008] After obtaining the traffic light status information, the traffic light learning machine determines the traffic light strategy at the intersection based on the trained traffic light strategy model, and uploads the traffic light strategy in a standard format to the Internet of Vehicles application service platform through an industrial computer.

[0009] According to a preferred embodiment, the signal light strategy includes at least a state of the signal light, a state duration, and a phase cycle of the signal light.

[0010] According to a preferred embodiment, the method also includes: when the signal light phase cycle changes, the traffic light learning machine records the phase cycle change time node while calibrating the signal light phase cycle; if the phase cycle changes at the same time node within the set time, this time node is used as the baseline of the signal light phase cycle, and the corresponding traffic light strategy is fed back to the Internet of Vehicles application service platform at the corresponding time node.

[0011] According to a preferred embodiment, the following method is used to obtain the traffic light status information from the traffic light in real time: the voltage and current information of the traffic light is obtained in real time through a voltage and current detection device connected to the traffic light control circuit, and the voltage and current information of the traffic light is sent to a logic operator; the logic operator calculates the status information of the current traffic light based on the received voltage and current information of the traffic light, and sends the status information of the traffic light to the control main board through the I / O board; the control main board packages the status information of all traffic lights at the intersection and sends it to the traffic light learning machine.

[0012] According to a preferred embodiment, the method further comprises: the traffic light learning machine uses the data packet composed of all the traffic light status information at the intersection as a training sample to perform self-learning of the traffic light strategy model.

[0013] According to a preferred implementation, the standard format meets the interface specification requirements between the RSU and the central subsystem.

[0014] The present invention also provides an RSU-like communication tool based on an industrial computer, which is applied to the method described above, and the tool comprises:

[0015] A traffic light learning machine is used to load the traffic light strategy model. During the learning process, the traffic light status information is obtained from the traffic light in real time. The traffic light learning machine uses the traffic light status information as a training sample to perform self-learning of the traffic light strategy model.

[0016] After obtaining the signal light status information, the signal light learning machine determines the signal light strategy of the intersection according to the trained signal light strategy model;

[0017] The industrial computer is used to receive the traffic light strategy sent by the traffic light learning machine and upload the traffic light strategy to the Internet of Vehicles application service platform in a standard format.

[0018] The technical solution of the embodiment of the present invention has at least the following advantages and beneficial effects: the present invention provides a signal light detection method and tool based on a RSU-like tool, which installs a signal light learning machine and an industrial computer device at the intersection, connects the existing traffic signal machine through the signal light learning machine, performs signal collection and calibration self-learning to obtain the status and status duration information of the signal light, interconnects the signal light learning machine through the industrial computer, formats the data information in a standard format, and uploads the information to the Internet of Vehicles application service platform, which parses the information, restores the signal light information, and pushes the information to the vehicle-mounted terminals within the coverage area, so as to facilitate the drivers and passengers to make application decisions. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 A schematic diagram of a flow chart of a signal light detection method based on a RSU-like tool provided in Embodiment 1 of the present invention;

[0020] Figure 2 A schematic diagram of the structure of an RSU-like communication tool based on an industrial computer provided in Example 1 of the present invention. DETAILED DESCRIPTION

[0021] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0022] Example 1

[0023] refer to Figure 1 As shown, Figure 1 A flowchart of a signal light detection method based on an RSU-like tool provided in an embodiment of the present invention.

[0024] A signal light detection method based on a RSU-like tool provided in an embodiment of the present invention is applied to a RSU-like communication tool, as shown in FIG. Figure 2 As shown, the RSU-like communication tool includes a signal light learning machine and an industrial computer; the present invention aims to install a signal light learning machine and an industrial computer in the traffic signal box of each intersection without upgrading or improving the existing traffic signal machine at the intersection, and the signal light learning machine is connected to the existing traffic signal machine and interconnected with the signal light learning machine through the industrial computer. The specific method is as follows:

[0025] Traffic light learning machine self-learning steps:

[0026] The signal light learning machine is loaded with a signal light strategy model. During the learning process, the signal light status information is obtained from the traffic light in real time. The signal light learning machine uses the signal light status information as a training sample to self-learn the signal light strategy model. It should be noted that the signal light strategy refers to the strategy set for the state, state duration and phase cycle of the signal light according to the different traffic flows during the morning peak, normal period and evening peak. Specifically in the embodiment of the present invention, the signal light status is read by hardware detection to realize the millisecond-level real-time acquisition of the signal light status, as follows:

[0027] The phase board transmits the collected signal light voltage and current information to the logic operator through the signal light control circuit through the internal voltage and current detection device; the logic operator calculates the current status information of the signal light based on the received signal light voltage and current information, and sends the signal light status information to the control main board through the I / O board; it should be noted that the present embodiment adopts the strategy of single light collection, first collects the single light status, and further collects the signal light voltage and current information of the 36 lines at the current intersection, forms a data packet, and calculates the status of all lights through the logic operator; further, the control main board packages the status information of all signal lights at the intersection and sends it to the signal light learning machine, and the signal light learning machine uses the data packet composed of the packaged status information of all signal lights at the intersection as a training sample for self-learning of the signal light strategy model.

[0028] Traffic light learning machine detection steps:

[0029] After obtaining the signal light status information, the traffic light learning machine determines the traffic light strategy of the intersection according to the trained signal light strategy model, and uploads the traffic light strategy to the Internet of Vehicles application service platform in a standard format through an industrial computer; the standard format meets the interface specification requirements between the RSU and the central subsystem; after uploading the information to the Internet of Vehicles application service platform, the Internet of Vehicles application service platform parses the information, restores the traffic light information, and pushes the information to the vehicle-mounted terminals within the coverage area, so as to facilitate the drivers and passengers to make application decisions.

[0030] It should be noted that, in an embodiment of the present invention, the traffic light learning machine is embedded with a traffic light cycle self-learning AI algorithm, which is as follows: the traffic light learning machine uses the time axis and baseline offset method to identify and learn the traffic light strategy. When the traffic light phase cycle changes, the traffic light learning machine calibrates the traffic light phase cycle and records the time node of the phase cycle change. If the phase cycle changes at the same time node within the set time, this time node is used as the baseline of the traffic light phase cycle, and the corresponding traffic light strategy is fed back to the Internet of Vehicles application service platform at the corresponding time node, so as to avoid the adaptation period delay generated by the traffic light learning machine when the traffic light phase cycle changes, and realize the real-time detection of the traffic light strategy.

[0031] Furthermore, considering that the industrial computer may be connected to a variety of different devices, the reported data must include SPAT, INFO reported and sent data, and may also require MAP, RSI, BSM, RSM and other messages. Therefore, the AI ​​algorithm of the present invention uses a configuration table to call the program sub-module. When the external docking device changes, the configuration file can be modified by remotely connecting to the industrial computer, thereby reducing the workload of secondary development and subsequent maintenance costs.

[0032] Furthermore, the embodiment of the present invention also deploys a traffic light learning machine networking software system at the intersection, and reserves an encrypted upgrade interface. When there is a need for an upgrade, the upgrade program requests the upgrade package on the server, receives the upgrade package, replaces the library file under lib, and automatically restarts the software after the upgrade is completed, thereby realizing remote upgrade of intersection equipment.

[0033] The embodiment of the present invention further provides an RSU-like communication tool based on an industrial computer, which is applied to the method described above, and the tool includes:

[0034] A traffic light learning machine is used to load the traffic light strategy model. During the learning process, the traffic light status information is obtained from the traffic light in real time. The traffic light learning machine uses the traffic light status information as a training sample to perform self-learning of the traffic light strategy model.

[0035] After obtaining the signal light status information, the signal light learning machine determines the signal light strategy of the intersection according to the trained signal light strategy model;

[0036] The industrial computer is used to receive the traffic light strategy sent by the traffic light learning machine and upload the traffic light strategy to the Internet of Vehicles application service platform in a standard format.

[0037] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A signal light detection method based on a RSU-like tool, characterized in that: Applied to RSU-like communication tools, the RSU-like communication tools include a signal light learning machine and an industrial computer, the signal light learning machine and the industrial computer are installed in a traffic signal box, the method is as follows: Traffic light strategy self-learning steps: The signal light status information is obtained from the traffic light in real time, and the signal light learning machine uses the signal light status information as a training sample to perform self-learning of the signal light strategy model loaded on the signal light learning machine; Traffic light strategy detection steps: After obtaining the signal light status information, the signal light learning machine determines the signal light strategy of the intersection according to the trained signal light strategy model, and uploads the signal light strategy in a standard format to the Internet of Vehicles application service platform through the industrial computer. The Internet of Vehicles application service platform is used to parse the signal light strategy, restore the signal light information, and push the signal light information to the vehicle terminals within the coverage area; The signal light strategy at least includes a state of the signal light, a state duration, and a phase cycle of the signal light.

2. The signal light detection method based on the RSU-like tool according to claim 1, characterized in that: The method also includes: when the signal light phase cycle changes, the signal light learning machine records the phase cycle change time node while calibrating the signal light phase cycle. If the phase cycle changes at the same time node within the set time, this time node is used as the baseline of the signal light phase cycle, and the corresponding signal light strategy is fed back to the Internet of Vehicles application service platform at the corresponding time node.

3. The signal light detection method based on the RSU-like tool according to claim 1, characterized in that: The signal light status information is obtained from the traffic light in real time in the following manner: the voltage and current information of the signal light is obtained in real time through a voltage and current detection device connected to the signal light control circuit, and the voltage and current information of the signal light is sent to a logic operator; The logic operator calculates the current status information of the traffic light based on the received traffic light voltage and current information, and sends the traffic light status information to the control main board through the I / O board; the control main board packages the status information of all traffic lights at the intersection and sends it to the traffic light learning machine.

4. The signal light detection method based on the RSU-like tool according to claim 3, characterized in that: The method also includes: the traffic light learning machine uses the data packets composed of all the traffic light status information at the intersection as training samples to perform self-learning of the traffic light strategy model.

5. The signal light detection method based on a RSU-like tool according to any one of claims 1 to 4, characterized in that: The standard format meets the interface specification requirements between the RSU and the central subsystem.

6. An RSU-like communication tool based on an industrial computer, characterized in that: Applied to the method according to any one of claims 1 to 5, the tool comprises: A traffic light learning machine is used to load the traffic light strategy model. During the learning process, the traffic light status information is obtained from the traffic light in real time. The traffic light learning machine uses the traffic light status information as a training sample to perform self-learning of the traffic light strategy model. After obtaining the signal light status information, the signal light learning machine determines the signal light strategy of the intersection according to the trained signal light strategy model; The industrial computer is used to receive the traffic light strategy sent by the traffic light learning machine and upload the traffic light strategy to the Internet of Vehicles application service platform in a standard format.

Images