A smart city management and control system based on digital twin technology

Abstract

The present invention discloses a smart city management and control system based on digital twin technology, relates to the field of digital twin technology, and improves the response efficiency of unmanned vehicle management. The present invention generates historical road image data based on historical road video data, extracts background pixels from the historical road image data, and generates a real-time driving image model centered on the unmanned vehicle based on the real-time driving image data, maps each real-time driving image model into the city road image model, and then generates the best vehicle driving route for each unmanned vehicle, and then obtains the real-time traffic light status of each city road through the Internet, and then generates driving instructions according to the real-time traffic light status of each city road and the best driving route.

Description

Technical Field

[0001] The present invention relates to the field of digital twin technology, and in particular to a smart city management and control system based on digital twin technology. Background Art

[0002] Digital twin technology creates virtual copies of real-world objects or systems, allowing real-time data to be synchronized to virtual models for real-time monitoring, simulation, and optimization. In smart city unmanned vehicle traffic management, digital twin technology can comprehensively model the entire unmanned vehicle traffic system, including road networks, traffic signals, vehicle flow, pedestrian flow, etc.

[0003] Existing smart city traffic management technologies have the following defects:

[0004] Data accuracy and completeness: The accuracy of digital twins is highly dependent on the quality of data. If there are errors in sensor data or data is missing, the accuracy and reliability of the virtual model will be affected. In addition, the heterogeneity and inconsistency of data sources may also increase the difficulty of data fusion.

[0005] Slow response speed: In practical applications, delays in data transmission and processing may affect the real-time and accuracy of unmanned vehicle management, especially in large-scale urban traffic management.

[0006] Therefore, how to ensure the accuracy of the construction of the virtual city road model while improving the response efficiency of unmanned vehicle management is a difficulty of the existing technology. For this purpose, a smart city management and control system based on digital twin technology is provided. Summary of the invention

[0007] In order to solve the above technical problems, the purpose of the present invention is to provide a smart city management and control system based on digital twin technology.

[0008] In order to achieve the above object, the present invention provides the following technical solutions:

[0009] A smart city management and control system based on digital twin technology includes a cloud computing platform, wherein the cloud computing platform is communicatively connected to a city road image model building module, a vehicle path generation module, and a driving supervision module;

[0010] The urban road image model building module is used to connect the roadside units on both sides of the urban road, and then generate a partial road background image model according to the historical road video data collected by each roadside unit, and extract background pixels from the historical road image data;

[0011] The vehicle path generation module is provided with an edge vehicle unit and a driving decision generation unit;

[0012] An edge vehicle unit is installed on each unmanned vehicle, and an identity number is set for each edge vehicle unit. The edge vehicle unit is used to collect real-time driving image data of the unmanned vehicle through a camera, and generate a real-time driving image model centered on the unmanned vehicle according to the real-time driving image data;

[0013] The driving decision generating unit is used to generate an optimal vehicle driving route for each unmanned vehicle by mapping each real-time driving image model into an urban road image model;

[0014] The driving supervision module is used to obtain the real-time traffic light status of each city road through the Internet, and then generate driving instructions according to the real-time traffic light status of each city road and the best driving route.

[0015] Furthermore, the urban road image model building module installs a number of roadside units on both sides of the urban road, and divides the urban road into n road areas according to the distribution of the urban road, respectively sets a signal base station for each road area, and sets a number for each signal base station, where n is a natural number greater than 0;

[0016] The roadside unit is equipped with a high-speed camera, a GPS positioning device and a wireless communication device. Each roadside unit is communicated with the signal base station corresponding to the road area where it is located. At the same time, each roadside unit obtains the location information of its location and several copies of historical road video data through the GPS positioning device and the high-speed camera.

[0017] Furthermore, the generation process of the partial road background image model includes:

[0018] Obtain the pixel change direction and pixel difference N of the pixels at the same position of two adjacent historical road image data, and then obtain the pixel change speed v, establish a two-dimensional coordinate system, and convert the pixel change speed into a pixel change speed vector and map it to the two-dimensional coordinate axis, and obtain the angle θ between the pixel change speed vector and the x-axis according to the pixel change direction;

[0019] Get the original vectors of the pixel change speed vector along the x-axis and y-axis, and record them as u x and u y , and then establish the pixel speed iteration formula in two directions, obtain the pixel change speed of each pixel in the x-axis and y-axis directions corresponding to each historical video data, and accumulate and average the pixel change speeds of the pixels at the same position in the historical road video data corresponding to the same period but different collection times to obtain the characteristic pixel change speed of the corresponding pixel in the x-axis and y-axis directions;

[0020] Set the pixel change speed threshold. If the characteristic pixel change speed in any direction of the pixel is greater than or equal to the pixel change speed threshold, the corresponding pixel is marked as a dynamic pixel. If the characteristic pixel change speed in any direction of the pixel is less than the pixel change speed threshold, the corresponding pixel is marked as a background pixel.

[0021] A corresponding road background image is generated according to the historical road video data of each roadside unit and the corresponding background pixels, and then a corresponding partial road background image model is established according to the road background image of each roadside unit.

[0022] Furthermore, the process of establishing the urban road image model includes:

[0023] A plurality of connection matching points are set at the edge positions of each partial road background image model, and the connection matching points at the edge positions of the partial road background image models generated by two opposing roadside units are matched with each other. According to the matching results of the connection matching points, the partial road background image models generated by the two opposing roadside units are spliced ​​at the positions where the matching connection matching points are located, so as to obtain the section road background image model where the corresponding roadside unit is located;

[0024] According to the location information of each roadside unit, the road background image models of each section are spliced ​​in sequence to obtain the urban road image model.

[0025] Furthermore, the process of establishing the real-time driving image model includes:

[0026] When the unmanned vehicle is traveling on a city road, the edge vehicle unit communicates with the vehicle path generation module through the roadside unit, and then the vehicle path generation module sets an identity number for the edge vehicle unit;

[0027] At the same time, each edge vehicle unit retrieves the real-time driving image data of its location captured by the camera on the unmanned vehicle, and then uses a process similar to the urban road image model building module to build an urban road image model, annotate dynamic pixels from the collected real-time driving image data, perform Gaussian blur processing on each dynamic pixel in the real-time driving image data, and divide a number of feature components from the real-time driving image data after the Gaussian blur processing, classify each dynamic pixel into each feature area through the Gaussian mixture model, and then build a corresponding dynamic target image model according to the dynamic elements contained in each feature component;

[0028] Taking the unmanned vehicle where the camera is located as the center, each dynamic target image model is spliced ​​together to obtain the corresponding real-time driving image model.

[0029] Furthermore, the process of mapping the real-time driving image model to the urban road image model includes:

[0030] The driving decision generation unit matches the real-time driving image models uploaded by each roadside unit, and then marks the unmanned vehicle model and the non-unmanned vehicle model in each real-time driving image model, marks the current driving speed and driving direction of each unmanned vehicle model and the non-unmanned vehicle model, and marks the corresponding identity number of the unmanned vehicle model;

[0031] Mapping each real-time driving image model into the urban road image model according to the location of the roadside unit used when uploading each real-time driving image model;

[0032] Each unmanned vehicle uploads its driving destination through the edge vehicle unit, and then the driving decision generation unit marks the driving destination of each unmanned vehicle on the urban road image model. At the same time, according to the current location of the corresponding unmanned vehicle model in the urban road image model, a number of prepared vehicle driving routes are generated.

[0033] Furthermore, the process of generating the optimal vehicle driving route includes:

[0034] A unit driving speed is set for each unmanned vehicle model, and several groups of vehicle driving combinations are generated according to the prepared vehicle driving routes of each unmanned vehicle and the best vehicle driving route currently being driven. Each group of vehicle driving combinations is respectively input into the urban road image model for driving simulation, and the best vehicle driving combination is selected from the vehicle driving combinations according to the driving simulation, and the prepared vehicle driving route in the best vehicle driving combination is recorded as the best vehicle driving route.

[0035] Furthermore, the generation process of the driving instruction includes:

[0036] Obtain the city road image model and all real-time driving image models from the vehicle path generation module, and obtain the real-time traffic light status of each city road through the Internet, and then map the real-time traffic light status of each city road into the city road image model;

[0037] Then, the current driving speed and driving direction of the vehicles around the unmanned vehicle are determined based on the real-time driving image model and the best driving route, and driving instructions are set for the unmanned vehicle.

[0038] Compared with the prior art, the present invention has the following beneficial effects:

[0039] The present invention generates historical road image data based on historical road video data, extracts background pixels from the historical road image data, and generates a real-time driving image model centered on an unmanned vehicle based on real-time driving image data, maps each real-time driving image model into an urban road image model, and then generates an optimal vehicle driving route for each unmanned vehicle, and then obtains the real-time traffic light conditions of each urban road through the Internet, and then generates driving instructions based on the real-time traffic light conditions of each urban road and the optimal driving route, thereby ensuring the accuracy of the construction of the virtual urban road model while improving the response efficiency of the unmanned vehicle management. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention.

[0041] Figure 1 It is a schematic diagram of the present invention. DETAILED DESCRIPTION

[0042] To make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be described in detail below. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other implementation methods obtained by ordinary technicians in this field without creative work belong to the scope of protection of the present invention.

[0043] like Figure 1 As shown, a smart city management and control system based on digital twin technology includes a cloud computing platform, and the cloud computing platform is communicatively connected with an urban road image model building module, a vehicle path generation module, and a driving supervision module;

[0044] The urban road image model building module is used to connect the roadside units on both sides of the urban road, and then generate a partial road background image model according to the historical road video data collected by each roadside unit, and extract background pixels from the historical road image data, and then build the urban road image model. The specific process includes:

[0045] The urban road image model building module installs several pairs of roadside units on both sides of the urban road, and divides the urban road into n road areas according to the distribution of urban roads. A signal base station is set for each road area, and each signal base station is numbered s1, s2, s3, ..., s n , where n is a natural number greater than 0;

[0046] The roadside unit is provided with a high-speed camera, a GPS positioning device and a wireless communication device. Each roadside unit is connected to a signal base station corresponding to the road area where it is located. At the same time, each roadside unit obtains the location information of its location and several copies of historical road video data through the GPS positioning device and the high-speed camera, and sends the location information and historical road video data to the signal base station;

[0047] It should be noted that the distances between roadside units on the same side of the urban road are equal, and the diameter of the shooting range of the high-speed camera of each roadside unit is greater than the distance between adjacent roadside units;

[0048] Each signal base station sends the location information and historical road video data generated by each roadside unit to the urban road image model building module.

[0049] Furthermore, the urban road image model building module builds an urban road image model according to the location information generated by each roadside unit and the historical road video data, and the specific process includes:

[0050] Each historical road video data is divided into m parts of historical road image data by frame, and the pixel value of each pixel in the historical road image data is obtained, and the pixel value of each pixel is subtracted from the pixel value of its surrounding pixels to obtain the pixel gradient value T of each pixel in different directions -x , T x , T -y , T y , where -x and -y represent the pixel gradient value between the pixel and the pixel to its left and below;

[0051] The corresponding historical road image data are arranged in chronological order, and a pixel change speed initially set to 0 is set for each pixel in the first historical road image data. It should be noted that the pixel change speed is a vector value;

[0052] During the acquisition process of historical road video data, the influence of non-human environmental factors such as sunlight is not considered within a certain acquisition time period. When there is an object position in the scene, the light and shadow of the same acquisition position will change in adjacent time periods, which will cause the pixel value of the pixel at the corresponding position in the corresponding historical road image data to change, and the change process has a direction;

[0053] Obtain the pixel change direction and pixel difference N of the pixels at the same position of two adjacent historical road image data, and then obtain the pixel change speed v, where the scalar value of v is equal to N / t, establish a two-dimensional coordinate system, and convert the pixel change speed into a pixel change speed vector and map it to the two-dimensional coordinate axis, and obtain the angle θ between the pixel change speed vector and the x-axis according to the pixel change direction, where t is the time difference between the two adjacent historical road image data;

[0054] Then, the original vectors of the pixel change speed vector along the x-axis and y-axis are obtained and recorded as u x and u y ,in , .

[0055] Furthermore, the speed iteration formula of pixels in two directions is established according to the Gauss-Seidel iteration method, where the speed iteration formula is expressed as:

[0056] ;

[0057] ;

[0058] Where α is the weight parameter, k is a natural number greater than 1 and less than or equal to m-1;

[0059] Through the above speed iteration formula, the pixel change speed of each pixel in the x-axis and y-axis directions corresponding to each historical video data is obtained. After accumulating and averaging the pixel change speeds of the pixels at the same position in the historical road video data corresponding to the same period but different acquisition times, the characteristic pixel change speed h of the corresponding pixel in the x-axis and y-axis directions is obtained. x and h y ;

[0060] Set a pixel change speed threshold. If the characteristic pixel change speed in any direction of the pixel is greater than or equal to the pixel change speed threshold, the corresponding pixel is marked as a dynamic pixel.

[0061] If the characteristic pixel change speed in any direction of the pixel is less than the pixel change speed threshold, the corresponding pixel is marked as a background pixel;

[0062] Generate a corresponding road background image based on the historical road video data of each roadside unit and the corresponding background pixels, and then establish a corresponding partial road background image model based on the road background image of each roadside unit;

[0063] Since each roadside unit is installed oppositely on both sides of the urban road, there are identical parts between the partial road background image models generated by the two opposite roadside units, and then a number of connection matching points are set at the edge positions of each partial road background image model;

[0064] Match the connection matching points at the edge positions of the partial road background image models generated by two opposing roadside units. If the connection matching points from different partial road background image models are at the same position, the pair of connection matching points are judged to be matched, otherwise, they are judged to be unmatched;

[0065] According to the matching results of the connection matching points, the partial road background image models generated by the two opposing roadside units are spliced ​​according to the positions of the matching connection matching points, so as to obtain the road background image model of the section where the corresponding roadside unit is located;

[0066] According to the location information of each roadside unit, the road background image models of each section are spliced ​​in sequence to obtain the urban road image model, and the urban road image model is sent to the vehicle path generation module.

[0067] The vehicle path generation module is provided with an edge vehicle unit and a driving decision generation unit;

[0068] An edge vehicle unit is installed on each unmanned vehicle, and an identity number is set for each edge vehicle unit. The edge vehicle unit is used to collect real-time driving image data of the unmanned vehicle through a camera, and generate a real-time driving image model centered on the unmanned vehicle according to the real-time driving image data;

[0069] The driving decision generating unit is used to generate an optimal vehicle driving route for each unmanned vehicle by mapping each real-time driving image model into an urban road image model;

[0070] The specific process of the vehicle path generation module generating a driving decision includes:

[0071] When the unmanned vehicle is traveling on a city road, its built-in edge vehicle unit communicates with the vehicle path generation module through the roadside unit, and then the vehicle path generation module sets an identity number for the edge vehicle unit;

[0072] At the same time, each edge vehicle unit retrieves the real-time driving image data of its location captured by the camera on the unmanned vehicle, and then uses a process similar to the urban road image model building module to build an urban road image model, annotate dynamic pixels from the collected real-time driving image data, perform Gaussian blur processing on each dynamic pixel in the real-time driving image data, and divide a number of feature components from the real-time driving image data after the Gaussian blur processing, classify each dynamic pixel into each feature area through the Gaussian mixture model, and then build a corresponding dynamic target image model according to the dynamic elements contained in each feature component;

[0073] Taking the unmanned vehicle where the camera is located as the center, each dynamic target image model is spliced ​​together to obtain the corresponding real-time driving image model.

[0074] Furthermore, after each edge vehicle unit annotates the real-time driving image model generated by it with a corresponding identity number, the real-time driving image model is sent to the vehicle path generation module through the nearest roadside unit;

[0075] The vehicle path generation module sends all real-time driving image models to the driving decision generation unit, and then the driving decision generation unit matches the real-time driving image models uploaded by each roadside unit, and then marks the unmanned vehicle model and the non-unmanned vehicle model in each real-time driving image model, marks the current driving speed and driving direction of each unmanned vehicle model and non-unmanned vehicle model, and marks the corresponding identity number for the unmanned vehicle model;

[0076] Mapping each real-time driving image model into the urban road image model according to the location of the roadside unit used when uploading each real-time driving image model;

[0077] Each unmanned vehicle uploads its driving destination through the edge vehicle unit, and then the driving decision generation unit marks the driving destination of each unmanned vehicle on the urban road image model, and generates several prepared vehicle driving routes according to the current location of the corresponding unmanned vehicle model in the urban road image model;

[0078] A unit driving speed is set for each unmanned vehicle model, and a plurality of vehicle driving combinations are generated according to the prepared vehicle driving routes of each unmanned vehicle and the best vehicle driving route currently being driven;

[0079] Input each vehicle driving combination into the urban road image model for driving simulation. If it is determined that there is a collision of unmanned vehicle models during the driving simulation, the corresponding vehicle driving combination is eliminated.

[0080] If it is determined that there is no collision of the unmanned vehicle model during the driving simulation, the corresponding vehicle driving combination is retained;

[0081] Counting the estimated time required for all the best vehicle driving routes and the reserve vehicle driving routes in each reserved vehicle driving combination to complete the driving, selecting the vehicle driving combination with the shortest estimated time as the best vehicle driving combination, and recording the reserve vehicle driving route in the best vehicle driving combination as the best vehicle driving route;

[0082] Then the vehicle path generation module sends each optimal vehicle driving route in the optimal vehicle driving combination to the corresponding driving supervision module.

[0083] Furthermore, the driving supervision module is used to obtain the real-time traffic light status of each city road through the Internet, and then generate driving instructions according to the real-time traffic light status of each city road and the best driving route. The specific process includes:

[0084] The driving supervision module obtains the urban road image model and all real-time driving image models from the vehicle path generation module, and obtains the real-time traffic light status of each urban road through the Internet, and then maps the real-time traffic light status of each urban road into the urban road image model;

[0085] Then, the current driving speed and driving direction of the vehicles around the unmanned vehicle are determined based on the real-time driving image model and the best driving route, and driving instructions are set for the unmanned vehicle;

[0086] If the current driving direction of the vehicles around the unmanned vehicle is the same as that of the unmanned vehicle, and the current driving speed is greater than or equal to the current speed of the unmanned vehicle, and the current speed of the unmanned vehicle is equal to the unit driving speed, no operation will be performed;

[0087] If the current driving direction of the vehicles around the unmanned vehicle is the same as that of the unmanned vehicle, but the current driving speed is lower than the current speed of the unmanned vehicle, a deceleration instruction is generated and executed immediately;

[0088] If the current driving direction of the vehicles around the unmanned vehicle is the same as that of the unmanned vehicle, and the current driving speed is greater than the current speed of the unmanned vehicle, and the current speed of the unmanned vehicle is equal to the unit driving speed of 0, then an acceleration instruction is generated and executed immediately;

[0089] If the unmanned vehicle needs to turn based on the real-time driving image model and the optimal driving route, a turning instruction is generated and executed.

[0090] The above embodiments are only used to illustrate the technical method 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 the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A smart city management and control system based on digital twin technology, including a cloud computing platform, characterized in that: The cloud computing platform is communicatively connected to a city road image model building module, a vehicle path generation module, and a driving supervision module; The urban road image model building module is used to generate a partial road background image model from the historical road video data collected by the roadside units on both sides of the urban road, and to extract background pixels from the historical road image data; Several pairs of roadside units are installed on both sides of the urban road, and the urban road is divided into n road areas according to the distribution of urban roads. A signal base station is set for each road area, and each signal base station is numbered, where n is a natural number greater than 0; The roadside unit is provided with a high-speed camera, a GPS positioning device and a wireless communication device. Each roadside unit is connected to a signal base station corresponding to the road area where it is located. At the same time, each roadside unit obtains the location information of its location and several copies of historical road video data through the GPS positioning device and the high-speed camera; The generation process of the partial road background image model includes: Obtain the pixel change direction and pixel difference of pixels at the same position of two adjacent historical road image data, and then obtain the pixel change speed, establish a two-dimensional coordinate system, convert the pixel change speed into a pixel change speed vector and map it on the two-dimensional coordinate axis, and obtain the angle between the pixel change speed vector and the x-axis according to the pixel change direction; The original vectors of the pixel change speed vector along the x-axis and the y-axis are obtained, and then the pixel speed iteration formula in two directions is established to obtain the pixel change speed in the x-axis and y-axis directions corresponding to each pixel of each historical video data. The pixel change speeds of the pixels at the same position in the historical road video data corresponding to the same period but different acquisition times are accumulated and averaged to obtain the characteristic pixel change speed of the corresponding pixel in the x-axis and y-axis directions; Set the pixel change speed threshold. If the characteristic pixel change speed in any direction of the pixel is greater than or equal to the pixel change speed threshold, the corresponding pixel is marked as a dynamic pixel. If the characteristic pixel change speed in any direction of the pixel is less than the pixel change speed threshold, the corresponding pixel is marked as a background pixel. Generate a corresponding road background image based on the historical road video data of each roadside unit and the corresponding background pixels, and then establish a corresponding partial road background image model based on the road background image of each roadside unit; A plurality of connection matching points are set at the edge positions of each partial road background image model, and the connection matching points at the edge positions of the partial road background image models generated by two opposing roadside units are matched with each other. According to the matching results of the connection matching points, the partial road background image models generated by the two opposing roadside units are spliced ​​at the positions where the matching connection matching points are located, so as to obtain the section road background image model where the corresponding roadside unit is located; According to the location information of each roadside unit, the road background image models of each section are sequentially spliced ​​to obtain the urban road image model; The vehicle path generation module is provided with an edge vehicle unit and a driving decision generation unit; An edge vehicle unit is installed on each unmanned vehicle, and an identity number is set for each edge vehicle unit. The edge vehicle unit is used to collect real-time driving image data of the unmanned vehicle through a camera, and generate a real-time driving image model centered on the unmanned vehicle according to the real-time driving image data; The process of establishing the real-time driving image model includes: When the unmanned vehicle is traveling on a city road, the edge vehicle unit communicates with the vehicle path generation module through the roadside unit, and then the vehicle path generation module sets an identity number for the edge unit; At the same time, each edge vehicle unit retrieves the real-time driving image data of its location captured by the camera on the unmanned vehicle, and then uses a process similar to the urban road image model building module to build an urban road image model, annotate dynamic pixels from the collected real-time driving image data, perform Gaussian blur processing on each dynamic pixel in the real-time driving image data, and divide a number of feature components from the real-time driving image data after the Gaussian blur processing, classify each dynamic pixel into each feature area through the Gaussian mixture model, and then build a corresponding dynamic target image model according to the dynamic elements contained in each feature component; Taking the unmanned vehicle where the camera is located as the center, each dynamic target image model is spliced ​​to obtain the corresponding real-time driving image model; The driving decision generating unit is used to generate an optimal vehicle driving route for each unmanned vehicle by mapping each real-time driving image model into an urban road image model; The process of mapping the real-time driving image model to the urban road image model includes: The driving decision generation unit matches the real-time driving image models uploaded by each roadside unit, and then marks the unmanned vehicle model and the non-unmanned vehicle model in each real-time driving image model, marks the current driving speed and driving direction of each unmanned vehicle model and the non-unmanned vehicle model, and marks the corresponding identity number of the unmanned vehicle model; Mapping each real-time driving image model into the urban road image model according to the location of the roadside unit used when uploading each real-time driving image model; Each unmanned vehicle uploads its driving destination through the edge vehicle unit, and then the driving decision generation unit marks the driving destination of each unmanned vehicle on the urban road image model, and generates several prepared vehicle driving routes according to the current location of the corresponding unmanned vehicle model in the urban road image model; The driving supervision module is used to obtain the real-time traffic light conditions of various city roads through the Internet, and then generate driving instructions according to the real-time traffic light conditions of various city roads and the best driving route; The generation process of the optimal vehicle driving route includes: A unit driving speed is set for each unmanned vehicle model, and a plurality of vehicle driving combinations are generated according to the prepared vehicle driving routes of each unmanned vehicle and the best vehicle driving route currently being driven. Each vehicle driving combination is input into the urban road image model for driving simulation, and the best vehicle driving combination is selected from the vehicle driving combinations according to the driving simulation, and the prepared vehicle driving route in the best vehicle driving combination is recorded as the best vehicle driving route; The generation process of the driving instruction includes: Obtain the city road image model and all real-time driving image models from the vehicle path generation module, and obtain the real-time traffic light status of each city road through the Internet, and then map the real-time traffic light status of each city road into the city road image model; Then, the current driving speed and direction of the vehicles around the unmanned vehicle are determined based on the real-time driving image model and the best driving route, and driving instructions are set and executed for the unmanned vehicle.

Images