Traffic signal control method, device and equipment based on large language model

Abstract

The invention discloses a traffic signal control method, device and equipment based on a large language model, and belongs to the technical field of traffic control. Information data of information media are obtained, first road information of a target area road is collected, and the first road information comprises image data collected by a traffic camera and vehicle sensing data collected by a vehicle-mounted sensor; determining traffic fusion data according to the information data, the first road information and a preset large language model; predicting virtual traffic flow data of a target area road according to the traffic fusion data; road control data are generated according to the virtual traffic flow data, the road control data are data for controlling traffic signal equipment of the target area, and the situation of reducing road congestion is achieved.

Description

Technical Field

[0001] This invention relates to the field of traffic control, and in particular to a traffic signal control method, apparatus, and equipment based on a large language model. Background Technology

[0002] With the continuous increase in the number of vehicles in cities, the efficiency of road traffic affects the overall operational efficiency of the city. Optimizing traffic signal control is a lower-cost option than expanding roads and does not produce negative impacts. Currently, traffic signal control relies on traffic flow prediction. However, current traffic flow prediction is often based on the traffic flow of neighboring roads. Therefore, the accuracy of the prediction is limited, and during peak hours, the lack of timely updates to traffic control data leads to increased road congestion.

[0003] The above content is only used to help understand the technical solution of the present invention and does not represent an admission that the above content is prior art. Summary of the Invention

[0004] The main objective of this invention is to provide a traffic signal control method, apparatus, and device based on a large language model, aiming to reduce road congestion. To achieve the above objective, this invention provides a traffic signal control method based on a large language model, which includes the following steps:

[0005] Acquire information data from information media and collect first road information of the target area, the first road information including: image data collected by traffic cameras and vehicle perception data collected by vehicle sensors;

[0006] Traffic fusion data is determined based on the aforementioned information data, the first road information, and the preset large language model;

[0007] Based on the traffic fusion data, predict virtual traffic flow data for roads in the target area;

[0008] Road control data is generated based on the virtual traffic flow data, and the road control data is data used to control traffic signal equipment in the target area.

[0009] Optionally, the step of determining traffic fusion data based on the information data, the first road information, and the preset large language model includes:

[0010] The event features of the information data are extracted based on the preset large language model. The event features include: event location features, event marker features, and event time features.

[0011] Based on the event characteristics and the first road information, the second road information is obtained;

[0012] The traffic fusion data is generated based on the second road information and map data.

[0013] Optionally, the step of obtaining the second road information based on the event characteristics and the first road information includes:

[0014] The target fusion range of the first road information is determined based on the preset influence range corresponding to the event location features, the event time features, and the event marker features;

[0015] Based on the event marker features, generate road data coefficients within the target fusion range;

[0016] The second road information is generated based on the road data coefficients and the first road information.

[0017] Optionally, if the number of event features is more than one, before the step of determining the target fusion range of the first road information data fusion based on the event location feature, the event time feature, and the preset influence range corresponding to the event marker feature, the method further includes:

[0018] Based on any two event characteristics, the corresponding event interaction factors are determined, resulting in multiple event interaction factors;

[0019] The preset influence range is adjusted based on multiple event interaction factors.

[0020] Optionally, the step of adjusting the preset influence range based on the multiple event interaction factors includes:

[0021] Based on the event characteristics, related event interaction factors are determined among multiple event interaction factors, including: time influence factors and scope influence factors;

[0022] The preset influence range is adjusted based on the time influence factor and the range influence factor to obtain the updated preset influence range.

[0023] Optionally, the step of predicting virtual traffic flow data for roads in the target area based on the traffic fusion data includes:

[0024] Normalize the traffic fusion data;

[0025] Input the traffic fusion data into the traffic flow prediction model and obtain the output results of the traffic flow prediction model;

[0026] The virtual traffic flow data is determined based on the output results.

[0027] Optionally, the step of generating road control data based on the virtual traffic flow data includes:

[0028] The duration of traffic lights at each intersection in different states is adjusted based on the virtual traffic flow data.

[0029] Furthermore, to achieve the above objectives, the present invention also provides a traffic signal control device based on a large language model, the traffic signal control device based on a large language model comprising:

[0030] The acquisition module is used to acquire information data from information media and to acquire first road information of the target area, which includes: image data acquired by traffic cameras and vehicle perception data acquired by vehicle sensors.

[0031] The fusion module is used to determine traffic fusion data based on the information data, the first road information, and the preset large language model;

[0032] The prediction module is used to predict virtual traffic flow data for roads in the target area based on the traffic fusion data.

[0033] The control module is used to generate road control data based on the virtual traffic flow data, wherein the road control data is data for controlling traffic signal equipment in the target area.

[0034] Furthermore, to achieve the above objectives, the present invention also provides a traffic signal control device based on a large language model, the traffic signal control device based on a large language model comprising: a memory, a processor, and a traffic signal control program based on a large language model stored in the memory and executable on the processor, the traffic signal control program based on a large language model being configured to implement the steps of the traffic signal control method based on a large language model as described above.

[0035] In addition, to achieve the above objectives, the present invention also provides a storage medium storing a traffic signal control program based on a large language model, wherein when the traffic signal control program based on the large language model is executed by a processor, the traffic signal control program based on the large language model implements the steps of the traffic signal control method based on the large language model described above.

[0036] This invention proposes a traffic signal control method based on a large language model. This method acquires information data from media and collects first road information for a target area, including image data from traffic cameras and vehicle perception data from onboard sensors. Based on the information data, the first road information, and a preset large language model, traffic fusion data is determined. Compared to current methods that estimate traffic flow based on neighboring roads, this method combines the actual situation in the target area to determine various types of traffic flow data, effectively expanding the data dimensions. Furthermore, based on the traffic fusion data, virtual traffic flow data for the target area is predicted, improving the accuracy of the virtual traffic flow data. Road control data is generated based on the virtual traffic flow data, which controls the traffic signal equipment in the target area, thereby effectively reducing road congestion. Attached Figure Description

[0037] Figure 1 This is a schematic diagram of the structure of a traffic signal control device based on a large language model, which is part of the hardware operating environment of the embodiment of the present invention.

[0038] Figure 2 This is a flowchart illustrating the first embodiment of the traffic signal control method based on a large language model according to the present invention.

[0039] Figure 3 This is a flowchart illustrating the second embodiment of the traffic signal control method based on a large language model according to the present invention.

[0040] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0041] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0042] Reference Figure 1 , Figure 1 This is a schematic diagram of the structure of a traffic signal control device based on a large language model, which is part of the hardware operating environment of the embodiment of the present invention.

[0043] like Figure 1As shown, the traffic signal control device based on a large language model may include: a processor 1001, such as a central processing unit (CPU), a communication bus 1002, an interactive device 1003, a network interface 1004, and a memory 1005. The communication bus 1002 is used to enable communication between these components. The interactive device 1003 may include a display screen or an input unit such as a keyboard. Optionally, the interactive device 1003 may also be connected to the communication bus via standard wired or wireless interfaces. The network interface 1004 may optionally include standard wired or wireless interfaces (such as a Wi-Fi interface). The memory 1005 may be a high-speed random access memory (RAM) or a stable non-volatile memory (NVM), such as a disk drive. Optionally, the memory 1005 may also be a storage device independent of the aforementioned processor 1001.

[0044] Those skilled in the art will understand that Figure 1 The structure shown does not constitute a limitation on traffic signal control devices based on large language models and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0045] like Figure 1 As shown, the memory 1005, which serves as a storage medium, may include an operating system, a data storage module, a network communication module, a user interface module, and a traffic signal control program based on a large language model.

[0046] exist Figure 1 In the traffic signal control device based on a large language model shown, the network interface 1004 is mainly used for data communication with other devices; the interactive device 1003 is mainly used for data interaction with the user; the processor 1001 and memory 1005 in the traffic signal control device based on a large language model of the present invention can be set in the traffic signal control device based on a large language model. The traffic signal control device based on a large language model calls the traffic signal control program based on a large language model stored in the memory 1005 through the processor 1001 and executes the traffic signal control method based on a large language model provided in the embodiment of the present invention.

[0047] This invention provides a traffic signal control method based on a large language model, referring to... Figure 2 , Figure 2 This is a flowchart illustrating the first embodiment of a traffic signal control method based on a large language model according to the present invention.

[0048] In this embodiment, the traffic signal control method based on a large language model includes:

[0049] Step S1: Obtain information data from information media and collect first road information of the target area. The first road information includes: image data collected by traffic cameras and vehicle perception data collected by vehicle sensors.

[0050] In this embodiment, the information data can be obtained in various ways. The information media can be media that publish traffic-related content. Optionally, text and audio information from multiple channels, such as official traffic department announcements, news media, social media platforms, real-time traffic broadcasts, and navigation app user reports, can be obtained through crawling. Preferably, the crawling process needs to filter information data for the local area. Commonly, for areas with venues or exhibition halls, activity information for these venues and exhibition halls needs to be obtained. The first road information includes road information such as national highways, provincial highways, and rural roads in the target area. Further, for large communities or schools, it can also include internal roads within the community or school. Generally, traffic cameras are installed at traffic intersections, allowing the acquisition of traffic flow data from multiple roads through image data. Further, vehicle perception data collected by vehicle sensors can be obtained through the vehicle internet. Optionally, the first road information can also include road update information for more detailed map data of roads in the target area.

[0051] Step S2: Determine traffic fusion data based on the information data, the first road information, and the preset large language model;

[0052] Specifically, the information data is converted into text format to obtain text information. The preset large language model receives the text information, performs semantic analysis, and extracts structured information such as key events, locations, times, and impact levels. Preferably, the preset large language model can be of multiple types. Different types of preset large language models are used for processing according to different data types. For image data, a visual language model can be used to extract traffic-related data from the image; alternatively, a preset image extraction algorithm can also be used to extract traffic-related data.

[0053] Step S3: Predict virtual traffic flow data for roads in the target area based on the traffic fusion data;

[0054] In this embodiment, the virtual traffic flow data refers to the traffic flow data of various roads within one or more time periods after the current moment. The traffic flow data can be expressed as the number of vehicles, vehicle waiting times, or other specific indicators describing traffic flow. In this embodiment, a machine learning method can be used to specifically input the traffic fusion data into a pre-trained traffic recognition model, and output the virtual traffic flow data based on the machine learning model.

[0055] Step S4: Generate road control data based on the virtual traffic flow data. The road control data is data used to control traffic signal equipment in the target area.

[0056] The virtual traffic flow data mentioned here reflects the traffic flow of various roads over a future period of time. Road control data is generated based on the virtual traffic flow data, thereby enabling the adjustment of traffic signals at various intersections.

[0057] In this embodiment, information data from information media is acquired, and first road information of the target area is collected. The first road information includes image data collected by traffic cameras and vehicle perception data collected by vehicle sensors. Traffic fusion data is determined based on the information data, the first road information, and a preset large language model. Compared with the current method of estimating traffic flow based on neighboring roads, this method can combine the actual situation in the target area to determine various types of traffic flow data, thereby effectively expanding the dimensions of the data. Based on the traffic fusion data, virtual traffic flow data of the target area is predicted, thereby improving the accuracy of the virtual traffic flow data. Road control data is generated based on the virtual traffic flow data. The road control data is the data for controlling the traffic signal equipment in the target area, thereby effectively reducing the problem of road congestion.

[0058] Furthermore, based on the first embodiment, a second embodiment of the traffic signal control method based on a large language model of the present invention is proposed. In this embodiment, reference is made to... Figure 3 The step of determining traffic fusion data based on the information data, the first road information, and the preset large language model includes:

[0059] Step S21: Extract event features from the information data according to the preset large language model. The event features include: event location features, event marker features, and event time features.

[0060] Specifically, the event location feature is the actual coordinates of the event's occurrence, requiring precise location on a map. The event time feature is preferably the time the event occurred, which can be before, during, or after the current time. The time marker is determined using a preset large language model and a preset flag database. Specifically, the information data is input into the preset large language model, and the matching degree between the information data and the flags in the preset flag database is indicated. Thus, the event marker feature of the event is determined according to the preset large language model.

[0061] Step S22: Obtain second road information based on the event characteristics and the first road information;

[0062] It should be noted that the first road information often only represents the situation of vehicles on the road, lacking a standardized measurement of peak traffic hours, i.e., the phenomenon of a large number of vehicles entering the road network from buildings within a certain period of time, also known as departure congestion. Similarly, it does not include the phenomenon of a large number of vehicles entering buildings from the road network within a certain period of time, i.e., arrival congestion. Here, buildings can include: parking lots, residential areas, commercial centers, etc. Optionally, the first road information can be corrected based on the event characteristics, and the first road information can be supplemented to include the number of vehicles about to enter the first road from adjacent buildings, to obtain the second road information. Optionally, the data corresponding to the event characteristics can be merged with the first road information to obtain the second road information.

[0063] Step S23: Generate the traffic fusion data based on the second road information and map data.

[0064] Specifically, the map data here is the map data of roads in the target area. Commonly, the map data stores road data up to the current time. The map data is updated using the second road information to generate the traffic fusion data.

[0065] In this embodiment, event features of the information data are extracted according to the preset large language model, and second road information is obtained based on the event features and the first road information; thus, more accurate road information in terms of dimension is obtained, and the traffic fusion data is generated based on the second road information and map data, thereby improving the accuracy of the traffic fusion data.

[0066] Furthermore, the step of obtaining the second road information based on the event characteristics and the first road information includes:

[0067] The target fusion range of the first road information is determined based on the preset influence range corresponding to the event location features, the event time features, and the event marker features;

[0068] Based on the event marker features, generate road data coefficients within the target fusion range;

[0069] The second road information is generated based on the road data coefficients and the first road information.

[0070] In this embodiment, the first road information can be data from a road within a certain time interval in the target area. Generally, traffic flow data corresponding to the current moment and before the current moment can be obtained through data preprocessing and feature extraction. The location range of the influence is determined based on the event location features and the preset influence range, and the time range of the influence is determined based on the event time features and the preset influence range. The target fusion range is determined based on the location range and the time range of the influence. The preset influence range can include a spatial radius and a temporal length. Each event marker feature corresponds to a road data coefficient. If multiple event marker features exist for a road adjacent to a building in the first road information, the markers for the buildings are marked on the road. Specifically, adjacent buildings are buildings within a preset distance from the road. It should be noted that in the process of generating the second road information, the image data and vehicle perception data are not directly multiplied by the road data coefficient. Instead, the traffic flow data extracted from the image data and vehicle perception data at the current moment and before the current moment are multiplied by the corresponding road data coefficient to obtain the second road information.

[0071] In this embodiment, the target fusion range of the first road information is determined by the preset influence range corresponding to the event location feature, the event time feature, and the event marker feature; thereby, the updated data range can be accurately determined, road data coefficients in the target fusion range are generated according to the event marker feature, and the data fusion result is determined by the road data coefficients. The second road information is generated according to the road data coefficients and the first road information, thus achieving accurate data fusion.

[0072] Furthermore, based on the first or second embodiment, a third embodiment of the traffic signal control method based on a large language model of the present invention is proposed. In this embodiment, the number of event features is more than one. Before the step of determining the target fusion range of the data fusion of the first road information according to the preset influence range corresponding to the event location feature, the event time feature, and the event marker feature, the method further includes:

[0073] Based on any two event characteristics, the corresponding event interaction factors are determined, resulting in multiple event interaction factors;

[0074] The preset influence range is adjusted based on multiple event interaction factors.

[0075] Specifically, complex nonlinear coupling effects exist between events. The mutual influence of two events can lead to road anomalies. For example, the end of a concert and construction on a main road can cause non-main roads to bear the increased traffic flow caused by these two events, resulting in congestion. In this embodiment, corresponding event interaction factors are determined based on the characteristics of any two events, resulting in multiple event interaction factors. The preset influence range is adjusted based on these multiple event interaction factors, thereby improving the accuracy of data fusion by calculating the impact of each event on road traffic flow during the data fusion process.

[0076] Furthermore, the step of adjusting the preset influence range based on multiple event interaction factors includes:

[0077] Based on the event characteristics, related event interaction factors are determined among multiple event interaction factors, including: time influence factors and scope influence factors;

[0078] The preset influence range is adjusted according to the time influence factor and the range influence factor to obtain the updated preset influence range.

[0079] In this embodiment, typically, the location distance between the location features of two event features is calculated, and the time difference between the time features of the two event features is calculated. The interaction factor is negatively correlated with the location distance and negatively correlated with the time difference. Optionally, for two events that exacerbate road congestion, the interaction factor ranges from [1, 10]. When the two event features do not affect each other, the event interaction factor is 1. For example, when the location distance is equal to a first preset location distance, or when the time difference is equal to a first preset time difference, the event interaction factor is determined to be 1. When the location distance is greater than the first preset location distance, or when the time difference is greater than the first preset time difference, the event interaction factor is set to a null value. When the location distance is less than the first preset location distance, or when the time difference is less than the first preset time difference, the event interaction factor is determined to be negatively correlated with both the location distance and the time difference. Specifically, the preset influence range is multiplied by multiple related event interaction factors to obtain the adjusted preset influence range.

[0080] In this embodiment, the associated event interaction factors are determined among multiple event interaction factors based on the event characteristics, and the preset influence range is adjusted according to the time influence factor and the range influence factor to obtain an updated preset influence range, thereby improving the accuracy of the fused data.

[0081] Furthermore, based on any of the above embodiments, a fourth embodiment of the traffic signal control method based on a large language model of the present invention is proposed. In this embodiment, the step of predicting virtual traffic flow data of roads in the target area based on the traffic fusion data includes:

[0082] Normalize the traffic fusion data;

[0083] Input the traffic fusion data into the traffic flow prediction model and obtain the output results of the traffic flow prediction model;

[0084] The virtual traffic flow data is determined based on the output results.

[0085] The traffic flow prediction model here is used to predict traffic flow data. Since different types of models often have different effects in different scenarios, the specific traffic flow prediction model here can be selected by the management personnel or automatically selected by the program. Optionally, the traffic flow prediction model here can also be set with a learning module for dynamically adjusting parameters. When the sensor detects a sudden increase in traffic flow in a certain road segment, the model updates the weights within 10 seconds. The weights here can be only a portion of the weights, thereby improving the efficiency of weight adjustment and thus efficiently adjusting the prediction results.

[0086] Furthermore, the step of generating road control data based on the virtual traffic flow data includes:

[0087] The duration of traffic lights at each intersection in different states is adjusted based on the virtual traffic flow data.

[0088] For example, based on real-time predicted congestion trends, the system can extend the green light duration at upstream intersections, typically by 15%-20%, diverting vehicles to less congested routes and thus reducing average waiting time.

[0089] Furthermore, this invention also proposes a traffic signal control device based on a large language model, the traffic signal control device based on a large language model comprising:

[0090] The acquisition module is used to acquire information data from information media and to acquire first road information of the target area, which includes: image data acquired by traffic cameras and vehicle perception data acquired by vehicle sensors.

[0091] The fusion module is used to determine traffic fusion data based on the information data, the first road information, and the preset large language model;

[0092] The prediction module is used to predict virtual traffic flow data for roads in the target area based on the traffic fusion data.

[0093] The control module is used to generate road control data based on the virtual traffic flow data, wherein the road control data is data for controlling traffic signal equipment in the target area.

[0094] Furthermore, this embodiment of the invention also proposes a traffic signal control device based on a large language model. The traffic signal control device based on a large language model includes: a memory, a processor, and a traffic signal control program based on a large language model stored in the memory and executable on the processor. The traffic signal control program based on a large language model is configured to implement the steps of the traffic signal control method based on a large language model as described above.

[0095] Furthermore, embodiments of the present invention also propose a storage medium storing a traffic signal control program based on a large language model, wherein when the traffic signal control program based on the large language model is executed by a processor, it implements the steps of the traffic signal control method based on the large language model described above.

[0096] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.

[0097] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0098] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0099] The above are merely preferred embodiments of the present invention and do not limit the scope of the patent. Any equivalent structural or procedural transformations made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the scope of patent protection of the present invention.

Claims

1. A traffic signal control method based on a large language model, characterized in that, The traffic signal control method based on a large language model includes the following steps: Acquire information data from information media and collect first road information of the target area, the first road information including: image data collected by traffic cameras and vehicle perception data collected by vehicle sensors; Traffic fusion data is determined based on the aforementioned information data, the first road information, and the preset large language model; Based on the traffic fusion data, predict virtual traffic flow data for roads in the target area; Road control data is generated based on the virtual traffic flow data, and the road control data is data used to control traffic signal equipment in the target area.

2. The traffic signal control method based on a large language model as described in claim 1, characterized in that, The step of determining traffic fusion data based on the information data, the first road information, and the preset large language model includes: The event features of the information data are extracted based on the preset large language model. The event features include: event location features, event marker features, and event time features. Based on the event characteristics and the first road information, the second road information is obtained; The traffic fusion data is generated based on the second road information and map data.

3. The traffic signal control method based on a large language model as described in claim 2, characterized in that, The step of obtaining the second road information based on the event characteristics and the first road information includes: The target fusion range of the first road information is determined based on the preset influence range corresponding to the event location features, the event time features, and the event marker features; Based on the event marker features, generate road data coefficients within the target fusion range; The second road information is generated based on the road data coefficients and the first road information.

4. The traffic signal control method based on a large language model as described in claim 3, characterized in that, Before the step of determining the target fusion range of the first road information data fusion based on the preset influence range corresponding to the event location feature, the event time feature, and the event marker feature, where the number of event features is greater than one, the method further includes: Based on any two event characteristics, the corresponding event interaction factors are determined, resulting in multiple event interaction factors; The preset influence range is adjusted based on multiple event interaction factors.

5. The traffic signal control method based on a large language model as described in claim 4, characterized in that, The step of adjusting the preset influence range based on multiple event interaction factors includes: Based on the event characteristics, related event interaction factors are determined among multiple event interaction factors, including: time influence factors and scope influence factors; The preset influence range is adjusted according to the time influence factor and the range influence factor to obtain the updated preset influence range.

6. The traffic signal control method based on a large language model as described in claim 1, characterized in that, The step of predicting virtual traffic flow data for roads in the target area based on the traffic fusion data includes: Normalize the traffic fusion data; Input the traffic fusion data into the traffic flow prediction model and obtain the output results of the traffic flow prediction model; The virtual traffic flow data is determined based on the output results.

7. The traffic signal control method based on a large language model as described in any one of claims 1 to 6, characterized in that, The step of generating road control data based on the virtual traffic flow data includes: The duration of traffic lights at each intersection in different states is adjusted based on the virtual traffic flow data.

8. A traffic signal control device based on a large language model, characterized in that, The traffic signal control device based on a large language model includes: The acquisition module is used to acquire information data from information media and to acquire first road information of the target area, which includes: image data acquired by traffic cameras and vehicle perception data acquired by vehicle sensors. The fusion module is used to determine traffic fusion data based on the information data, the first road information, and the preset large language model; The prediction module is used to predict virtual traffic flow data for roads in the target area based on the traffic fusion data. The control module is used to generate road control data based on the virtual traffic flow data, wherein the road control data is data for controlling traffic signal equipment in the target area.

9. A traffic signal control device based on a large language model, characterized in that, The traffic signal control device based on a large language model includes: a memory, a processor, and a traffic signal control program based on a large language model stored in the memory and executable on the processor, wherein the traffic signal control program based on a large language model is configured to implement the steps of the traffic signal control method based on a large language model as described in any one of claims 1 to 7.

10. A storage medium, characterized in that, The storage medium stores a traffic signal control program based on a large language model, which, when executed by a processor, implements the steps of the traffic signal control method based on a large language model as described in any one of claims 1 to 7.

Images