Green wave optimization method based on bim-gis-mlp under cellular automaton

Abstract

The application discloses a green wave optimization method based on a cellular automaton under BIM-GIS-MLP, relates to the field of highway digitization, and comprises the following steps: segmenting a road according to a BIM and a GIS platform; based on IOT, the end point information of the multi-section journey of each vehicle is counted and fed back to corresponding traffic light calibration information of subsequent vehicles, and the subsequent vehicles access information through IOT; the BIM and the GIS platform analyze the average speed range through the multi-section continuous journey and calculate the total duration of the green wave section; and the optimization problem of the total duration of the green wave section is calculated through linear programming, and the maximum value of the objective function is solved. The application accesses the Internet of Vehicles and the Internet of Things data network, realizes data fusion intercommunication, accesses an MLP model to analyze individuals, then obtains the maximum value of the objective function about the green wave section time through linear programming, improves traffic fluency, reduces energy loss and exhaust emission, and improves the urban environmental quality.

Description

TECHNICAL FIELD

[0001] The application relates to the field of highway digitization, in particular to a green wave optimization method based on a cellular automaton under BIM-GIS-MLP. BACKGROUND

[0002] In the process of highway digitization, the measurement of the green wave section is affected by traffic flow fluctuations. Traffic flow is constantly changing. If the fluctuations are not considered during measurement, the measurement results may not match the actual situation. In addition, vehicle speed is affected by various factors such as road conditions and weather conditions, which can cause changes in vehicle speed. At the same time, unreasonable signal timing is also an important factor. If the signal timing scheme is not adjusted in a timely manner according to the actual traffic situation, the green wave effect may be poor, thereby affecting the accuracy of the measurement.

[0003] Nowadays, the signal timing does not conform to the normal operation rules of the road, and the inaccuracy of the timing of the traffic light leads to measurement deviation of the green wave section, causing vehicles to exit the green wave section. SUMMARY

[0004] The application provides a green wave optimization method based on a cellular automaton under BIM-GIS-MLP, which solves the problems of the prior art.

[0005] In a first aspect, the application provides a green wave optimization method based on a cellular automaton under BIM-GIS-MLP, comprising:

[0006] segmenting the road according to the BIM and GIS platforms;

[0007] based on IOT, the end point information of the multi-section journey of each vehicle is counted and fed back to the corresponding traffic light calibration information of the subsequent vehicle, and the subsequent vehicle accesses the information through IOT;

[0008] the BIM and GIS platforms analyze the average speed range through the multi-section continuous journey and calculate the total duration of the green wave section;

[0009] the optimal problem of the total duration of the green wave section is calculated through linear programming, and the maximum value of the objective function is solved.

[0010] Further, the road segmentation according to the BIM and GIS platforms comprises:

[0011] Determine whether the current vehicle can pass through the subsequent journey of the red light, when it can pass, merge the multiple red light roads that can pass through into a journey, and count the green wave road section duration, when it cannot pass, set the current journey of a red light road as a journey, and do not count the green wave road section duration, and divide the physical road section boundary according to the bridge, tunnel, and intersection BIM component semantic information.

[0012] Further, the current vehicle accesses the red light calibration information of the previous vehicle, judges whether it can still pass through the subsequent journey of the red light at the current location, when it can pass, the auxiliary driving information is not modified, when it cannot pass, the required vehicle speed range for passing through the merged journey is recalculated and a prompt is given to the driver.

[0013] Further, the recalculation of the required vehicle speed range for passing through the merged journey and the prompt to the driver include that the required vehicle speed range includes multiple sub-sections of different vehicle speed ranges, and it is judged whether each sub-section of different vehicle speed range is overspeed, when it is overspeed, the merged journey is divided into two or even multiple journeys at the overspeed section, when it is not overspeed, each sub-section of different vehicle speed range is merged into a journey.

[0014] Further, the road segmentation according to the BIM and GIS platform includes:

[0015] The road is segmented according to the red light setting, it is judged whether to merge each adjacent section according to the current vehicle speed and traffic flow data, when the vehicle speed and traffic flow data meet the merging requirements, the adjacent sections are merged into a journey, and the driver is given a corresponding vehicle speed prompt, when the vehicle speed and traffic flow data do not meet the merging requirements, the journey that does not meet the merging requirements is separately segmented, the adjacent sections that meet the merging requirements are merged into a journey, the IoT traffic density heat map is fused in real time, and the automatic merging is triggered when the adjacent section density difference is less than a threshold value.

[0016] Further, for the two adjacent merged journeys, access the red light correction data of the previous vehicle to judge whether the vehicle speed can pass through the red light of the next journey when it is in the calculated value range, when the vehicle speed is not overspeed, continue to judge whether the traffic flow in the next journey meets the current vehicle without waiting for the red light to enter the next journey, when the vehicle speed is overspeed, exit the current judgment condition, when the traffic flow meets, merge the two journeys into one journey, judge the subsequent journey in a loop, and calculate the green wave road section duration according to the auxiliary driving vehicle speed prompt, when the traffic flow does not meet, keep the original journey segmentation.

[0017] Further, for each vehicle, the vehicle speed, navigation route of the vehicle is counted, the traffic flow prediction is trained by the multi-layer perceptron in the server, the traffic flow of the time period of the subsequent driving to different trip locations is predicted, and the speed range of the intermittent multi-section continuous trip is given.

[0018] Further, the multi-layer perceptron trains the traffic flow prediction, specifically including the following steps:

[0019] Access data collection information, including current vehicle speed, historical traffic flow and timestamp data;

[0020] Clean and normalize the data, including: clean by IoT data cleaning specification: eliminate abnormal sudden stop records with acceleration > 3m / s², road sensor: compensate for detection errors in rainy and foggy weather;

[0021] Build a multi-layer perceptron model, select multiple data type features in the input layer according to nodes and time windows, use activation functions for non-linear transformation in the hidden layer, and the output layer is the speed range corresponding to the number of time windows;

[0022] Based on historical data, train the MLP model, and encode the GIS road network topology including road segment ID and upstream and downstream connection relationship as MLP input features, jointly train with IoT speed time series data, iteratively perform forward propagation and back propagation, update model parameters, and until the loss function converges;

[0023] Based on the current speed data, use the trained MLP model to predict the speed range of the future multi-section trip;

[0024] For multiple trips without waiting for traffic lights, continuously set the required range of vehicle speed;

[0025] Output the speed range result of the multi-section trip of the current vehicle.

[0026] Further, in the process of training the MLP model, the historical traffic flow data is screened out, the traffic flow data and timestamp data of the corresponding trip of the target vehicle to the end information road segment are selected, the MLP model is trained for the target vehicle, the multi-section continuous trip is set on the multi-core processor for independent training, and the trained MLP model is output after the results are summarized.

[0027] Further, the optimization problem of calculating the total length of green wave road segments by linear programming and solving the maximum value of the objective function includes:

[0028] For a single vehicle, the MLP model is used for auxiliary analysis of the speed range of the future green wave road trip, and the mutual correlation between the speed ranges of multiple vehicles is analyzed, including the following steps:

[0029] Setting the cellular automaton model to solve the mutual correlation influence, determining the model parameters of the cellular automaton model, including road length, number of lanes, vehicle density, maximum speed, random deceleration probability;

[0030] According to the traffic observation data, the vehicle distribution on the road is initialized and the data information is called to give each vehicle an initial speed;

[0031] Setting the update rule of the cellular automaton model, including acceleration rule, deceleration rule, random deceleration rule, lane changing rule, also including extracting lane number, slope curvature from BIM model and writing into cellular automaton rule library, and calling auxiliary green wave section to prompt vehicle speed data of each vehicle;

[0032] Dynamic graph generation: real-time update data blood relationship based on Apache Kafka.

[0033] # Cellular speed update rule (affected by BIM road slope)

[0034] def update_speed(cell):

[0035] max_speed = BIM.get_max_speed(cell.road_segment_id) # Get the design speed limit from BIM

[0036] if cell.speed<max_speed:

[0037] cell.speed += 1 * (1 - BIM.get_gradient(cell.position) / 10) # The greater the slope, the slower the acceleration

[0038] Divide the road according to the GIS coordinate grid, each cell carries the BIM lane attribute and the real-time traffic state of IoT, update the vehicle state of all vehicles at each time stamp including position and speed synchronously, get the latest simulation traffic flow dynamic evolution data;

[0039] Get the vehicle flow trend of the number of vehicles and vehicle speed evolution with time;

[0040] Load multiple vehicle speed data and vehicle data into the vehicle flow trend, get the prediction of the future trip vehicle flow data.

[0041] Further, since the BIM and GIS platforms integrate the latest road construction information, the data loaded as a basic auxiliary platform provides a strong data foundation for road optimization driving, and the data information specification of the BIM and GIS platforms is helpful for road segmentation. Secondly, IOT records various data information of vehicles in multi-vehicle applications, and the fusion of IOT information facilitates the BIM and GIS platforms to assist vehicles to complete the macroscopic driving optimization of road vehicles, and provides a convenient traffic environment. Moreover, BIM provides centimeter-level road geometric precision (such as bend radius and longitudinal slope), supports accurate calculation of green wave speed interval, GIS realizes global spatial indexing, and makes real-time association of IoT vehicle trajectory and road network topology; IoT and cellular automaton closed-loop optimization: vehicles upload the actual passing time of signal lights through V2I, dynamically calibrate the preset timing scheme of BIM, and cellular model adjusts simulation parameters according to real-time traffic of IoT, thereby improving the feasibility of linear programming solution.

[0042] The BIM-GIS-MLP-based cellular automaton green wave optimization method provided by the application accesses the Internet of Vehicles and the Internet of Things data network, realizes data fusion and interconnection, accesses the MLP model to analyze individuals, and then obtains the maximum value of the objective function about the green wave section time through linear programming, thereby improving traffic flow, reducing energy loss and exhaust emission, and improving urban environmental quality. BRIEF DESCRIPTION OF DRAWINGS

[0043] The accompanying drawings, which are included to provide a further understanding of the embodiments of the application and are incorporated in and constitute a part of this application, illustrate embodiments of the application and together with the description serve to explain the principles of the application. In the drawings:

[0044] Figure 1 The BIM-GIS-MLP-based cellular automaton green wave optimization method flow chart provided by an exemplary embodiment of the application.

[0045] Figure 2 The training process flow chart of the MLP model in the BIM-GIS-MLP-based cellular automaton green wave optimization method provided by an exemplary embodiment of the application. DETAILED DESCRIPTION

[0046] The exemplary embodiments will be described in detail herein with reference to the drawings. Unless otherwise indicated, the same numbers in different drawings indicate the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the application. Rather, they are merely examples of devices and methods consistent with some aspects of the application as detailed in the appended claims.

[0047] First, the terms involved in the application are explained:

[0048] Green wave section, also known as green wave zone, refers to a traffic management measure that, on a specific road, through coordinated control of traffic lights, vehicles can continuously pass through multiple green lights within a certain speed range, reducing the number of stops during driving, thereby improving the traffic efficiency and smoothness of the road.

[0049] In the prior art, the inaccurate measurement of the second-reading data of the traffic lights on the driving section in the navigation map is the main reason for the vehicle to exit the green wave section, and for long-distance vehicle trips, the judgment of the green wave section of the vehicle is limited to the adjacent section and often appears for a moment and disappears, the present application considers that the highway construction project does not fully use various data, therefore, through the data network combined with the Internet of Vehicles and the Internet of Things, the auxiliary reminder is matched to help the vehicle enter the green wave section.

[0050] The present application indirectly uses third-party data for traffic data related to safety, calibrates the second-reading display of the traffic light, calculates the known remaining second-reading section in advance according to different settings of the traffic light, assists the driver to enter the green wave section for a longer time within the specified speed from the individual, and ensures that the total time of a batch of vehicles entering the green wave section is the longest, thereby improving the traffic smoothness.

[0051] The specific application scenario of the present application is the green wave section auxiliary navigation of highway engineering.

[0052] The green wave optimization method based on BIM-GIS-MLP under the cellular automaton provided by the present application aims to solve the above technical problems of the prior art.

[0053] The technical solutions of the present application and how the technical solutions of the present application solve the above technical problems will be described in detail below. The following specific embodiments can be combined with each other, and the same or similar concepts or processes can not be described again in some embodiments. The embodiments of the present application will be described below with reference to the accompanying drawings.

[0054] Embodiment 1:

[0055] In the present application, the data information of each platform is applied as follows:

[0056] IoT end layer: through the real-time collection of vehicle dynamic data (GPS position, instantaneous speed, acceleration, navigation path) and road environment data (signal light phase, queue length, lane-level traffic) by IoT devices such as vehicle-mounted OBD terminal, roadside unit (RSU), and geomagnetic sensor, the data flow is transmitted to the edge computing node for preprocessing through 5G / V2X.

[0057] BIM platform layer: load the detailed 3D model of highway engineering, including: lane geometry topology (curvature, slope, width), traffic facility attributes (signal light coordinates, detector distribution, speed limit signs), data flow → after model lightening, release to GIS spatial database, build attribute dependency graph: component ID → material batch → inspection report → maintenance record;

[0058] GIS platform layer: integrate BIM model and geographic spatial data, realize: road network topology analysis (path connectivity, road segment segmentation), dynamic data spatial matching (vehicle position mapping to lane-level precision), data flow → generate traffic digital twin with spatial attributes, dynamically select key fields based on graph node centrality (Degree Centrality);

[0059] Among them, the design end: BIM model; the construction end: IoT sensor; the maintenance end: GIS spatio-temporal data;

[0060] As shown in Figure 1 , the steps of the auxiliary optimization method for entering the green wave section of highway digitalization from a macroscopic point of view are as follows:

[0061] According to the BIM and GIS platforms, the road is segmented, i.e. S1, the road is segmented according to the traffic light settings in the urban road;

[0062] Based on IOT, the end point information of each vehicle's multi-section journey is counted, i.e. S2, the number of vehicles on each section of road is counted, and the driving end point of road vehicles is analyzed based on navigation data;

[0063] BIM and GIS platforms analyze the average speed range through multi-section continuous journey and calculate the total length of green wave section, i.e. S3-S7 as follows:

[0064] S3, divide the multi-section journey of each vehicle, and feed back the preset second reading data of the traffic light to the navigation system according to the driving record data of the previous vehicle of the same section journey;

[0065] S4, the navigation system corrects the traffic light record data of the subsequent vehicle of the same section journey;

[0066] S5, calculate the average speed range through multi-section continuous journey;

[0067] S6, analyze the traffic flow prediction data predicted by the multi-layer perceptron for each section of vehicle through multi-section journey to obtain the length of the green wave section in the intermittent multi-section journey;

[0068] S7, calculate the total green wave length by summing up the number of vehicles that do not need to wait for the traffic light to enter the next section of journey;

[0069] S8, S9, the optimization problem of the total length of the green wave section is calculated by linear programming, and the maximum value of the objective function is solved as follows:

[0070] S8, the optimization problem of the total length of the green wave section is calculated by linear programming, and the maximum value of the objective function is solved as follows:

[0071] S9, the maximum value of the objective function of the optimization problem is calculated by calling linear programming.

[0072] In the data verification link:

[0073] Load the metadata atlas, and automatically trigger BIM model reverse query or on-site sensor supplementary sampling for missing fields.

[0074] The five-dimensional rule verification example is as follows:

[0075]

[0076] Specifically, in step S6, the multi-layer perceptron data is intervened, and the step of training the traffic flow prediction is as shown in Figure 2 as follows:

[0077] K1, access data collection information, including current vehicle speed, historical traffic flow and timestamp data;

[0078] K2, clean and normalize the data, including: clean up through IoT data cleaning specification as follows: vehicle-mounted terminal data: eliminate abnormal sudden stop records with acceleration>3m / s², road sensor: compensate for detection errors in rainy and foggy weather;

[0079] K3, construct a multi-layer perceptron model, select multiple data type features in the input layer according to nodes and time windows, use activation function for nonlinear transformation in the hidden layer; the output layer is the speed range corresponding to the number of time windows;

[0080] K4, train the MLP model based on historical data, and encode the GIS road network topology including road segment ID and upstream and downstream connection relationship as MLP input features, jointly train with IoT speed time series data, iteratively perform forward propagation and back propagation, update model parameters until the loss function converges;

[0081] K5, based on the current speed data, use the trained MLP model to predict the speed range of multiple future trips;

[0082] K6, for multiple trips that do not need to wait for a red light, continuously set the required range of vehicle speed;

[0083] K7, output the speed range result of the multiple trips of the current vehicle.

[0084] For the K1-K7 steps, specific examples are as follows:

[0085] % Load data

[0086] data = load('traffic_data.mat'); % Assume data is stored in traffic_data.mat file

[0087] X = data.X; % Input features: historical speed, flow, signal timing, etc.

[0088] y = data.y; % Output target: future speed range

[0089] % Data preprocessing

[0090] % Normalize data

[0091] X = (X - min(X)) / (max(X) - min(X));

[0092] y = (y - min(y)) / (max(y) - min(y));

[0093] % Split dataset

[0094] rng(1); % Set random seed to ensure reproducibility

[0095] idx = randperm(size(X, 1));

[0096] trainRatio = 0.7;

[0097] validateRatio = 0.2;

[0098] testRatio = 0.1;

[0099] trainIdx = idx(1:floor(trainRatio * end));

[0100] validateIdx = idx(floor(trainRatio * end) + 1:floor((trainRatio + validateRatio) * end));

[0101] testIdx = idx(floor((trainRatio + validateRatio) * end) + 1:end);

[0102] XTrain = X(trainIdx, :);

[0103] yTrain = y(trainIdx, :);

[0104] XValidate = X(validateIdx, :);

[0105] yValidate = y(validateIdx, :);

[0106] XTest = X(testIdx, :);

[0107] yTest = y(testIdx, :);

[0108] % Building an MLP model

[0109] hiddenLayerSize = 10; % Number of neurons in the hidden layer

[0110] net = feedforwardnet(hiddenLayerSize);

[0111] % Set training parameters

[0112] net.trainFcn = 'trainlm'; % Levenberg-Marquardt optimization algorithm

[0113] net.trainParam.epochs = 1000; % Maximum number of training epochs

[0114] net.trainParam.goal = 1e-5; % Target mean squared error

[0115] net.trainParam.max_fail = 10; % Maximum number of failures

[0116] % Training model

[0117] [net, tr]= train(net, XTrain', yTrain');

[0118] % Evaluation Model

[0119] % Validation set evaluation

[0120] yValidatePred = net(XValidate');

[0121] validateMSE = mean((yValidatePred - yValidate').^ 2);

[0122] % Test set evaluation

[0123] yTestPred = net(XTest');

[0124] testMSE = mean((yTestPred - yTest').^ 2);

[0125] % Predict future speed range

[0126] % Assume we have new input data newX

[0127] newX =... % new input data, needs to be consistent with training data format

[0128] newX = (newX - min(X)) / (max(X) - min(X)); % Normalize processing

[0129] yPred = net(newX');

[0130] % Denormalize the prediction result

[0131] yPred = yPred * (max(y) - min(y)) + min(y);

[0132] % Output the prediction result

[0133] disp('Predicted speed range:');

[0134] disp(yPred);

[0135] For the above output results, the comparison chart of the prediction results and the actual values obtained by matlab is highly consistent. For multiple vehicles accessing linear programming to solve optimization problems, the total sum of green wave sections is obtained for the longest duration, which ensures low energy consumption of vehicles and reduces exhaust emissions. In this embodiment, the edge node processes IoT data response <50ms, meeting the real-time decision-making demand of green wave; and the BIM model needs to be converted to the CGCS2000 coordinate system and aligned with the GIS base map.

[0136] The modules described as separate components can or can not be physically separate, and the components shown as modules can or can not be physical modules, i.e. can be located in one place, or can be distributed over a plurality of network modules. Some or all of the modules can be selected according to actual needs to achieve the purpose of the embodiment of the present application.

[0137] It should also be noted that the terms "comprising", "comprises" or any other variant thereof are intended to cover non-exclusive inclusions, such that processes, methods, articles or apparatuses that comprise a list of elements do not only include those elements, but can also include other elements not expressly listed or inherent to such processes, methods, articles or apparatuses. Without further limitation, an element defined by the statement "comprising a" does not exclude the presence of additional identical elements in the process, method, article or apparatus that includes the element.

[0138] The above only describes the embodiments of the present application and is not intended to limit the present application. The present application can have various modifications and changes for those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application shall be included in the scope of claims of the present application.

[0139] Other embodiments of the application will be apparent to those skilled in the art from consideration of the specification and practice of the application disclosed herein. The specification and examples are to be considered exemplary only, with the true scope and spirit of the application indicated by the following claims. The claims should not be limited by the above description but should be given their full scope.

[0140] It should be understood that the present application is not limited to the precise structures described above and illustrated in the drawings and that various modifications and changes can be made without departing from the scope thereof. The scope of the present application is indicated by the appended claims.

Claims

1. A green wave optimization method based on a cellular automaton under BIM-GIS-MLP, characterized in that, Comprise: loading road digital information into BIM and GIS platform according to IOT sensor; segmenting road according to BIM and GIS platform, including: segmenting road according to traffic light setting, judging whether to merge adjacent segments according to current vehicle speed and traffic flow data, when vehicle speed and traffic flow data meet the merging requirements, merging adjacent segments into a segment of journey, at the same time giving the corresponding vehicle speed prompt to the driver, when vehicle speed and traffic flow data do not meet the merging requirements, separately segmenting the journey that does not meet the merging requirements, merging adjacent segments that meet the merging requirements into a segment of journey, real-time fusing IoT vehicle flow density heat map, triggering automatic merging when adjacent segment density difference < threshold value; for the two adjacent segments after merging, accessing the traffic light correction data of the previous vehicle to judge whether the vehicle speed can pass through the traffic light of the next segment of journey, when the vehicle speed is not overspeed, continue to judge whether the traffic flow in the next segment of journey meets the current vehicle without waiting for the traffic light to enter the next segment of journey, when the vehicle speed is overspeed, exit the current judgment condition, when the traffic flow meets, merge the two segments of journey into a segment of journey, cyclically judge the subsequent journey, and calculate the green wave road section duration according to the auxiliary driving speed prompt, when the traffic flow does not meet, keep the original journey segmentation; for each vehicle, statistics the vehicle speed and navigation route, train the traffic flow prediction through the multi-layer perceptron in the server, predict the traffic flow of the time period of the different journey positions in the subsequent driving, and give the vehicle speed range of the discontinuous multi-segment continuous journey; the specific steps include: accessing data collection information, including current vehicle speed, historical traffic flow and timestamp data; cleaning and normalizing the data, including: cleaning through IoT data cleaning specification: removing abnormal sudden stop records with acceleration > 3m / s², compensating for detection errors in rain and fog weather for roadside sensors; building a multi-layer perceptron model, selecting multiple data type features in the input layer according to nodes and time window, using activation function for non-linear transformation in the hidden layer; the output layer is the vehicle speed range corresponding to the number of time windows; training the MLP model based on historical data, at the same time, encoding the GIS road network topology including road segment ID and upstream and downstream connection relationship as MLP model input features, training jointly with IoT vehicle speed time series data, iteratively executing forward propagation and backward propagation, updating model parameters until the loss function converges; predicting the vehicle speed range of future multi-segment journey based on current vehicle speed data using the trained MLP model; for multiple journeys that do not need to wait for traffic light, continuously setting the required range of vehicle speed; outputting the vehicle speed range result of the current vehicle's multi-segment journey; based on IOT, statistics the end point information of each vehicle's multi-segment journey, and feedback the corresponding traffic light calibration information to the subsequent vehicle, the subsequent vehicle accesses the information through IOT; BIM and GIS platform analyze the average speed range of multi-segment continuous journey and calculate the total duration of green wave road section; solving the optimization problem of total duration of green wave road section through linear programming and finding the maximum value of objective function.

2. The BIM-GIS-MLP based under cellular automaton green wave optimization method according to claim 1, characterized in that, In the process of training the MLP model, the historical traffic data is screened out, the traffic data and timestamp data of the corresponding journey of the target vehicle to the end point information section are selected, the MLP model is trained for the target vehicle, the multiple continuous journeys are set on the multi-core processor for independent training, and the trained MLP model is output after the results are summarized.

3. The BIM-GIS-MLP based under cellular automaton green wave optimization method according to claim 2, characterized in that, The optimization problem of calculating the total length of the green wave section by linear programming and solving the maximum value of the objective function includes: While assisting in analyzing the vehicle speed range of the future green wave section journey for a single vehicle, the mutual correlation between the speed ranges of multiple vehicles is analyzed, including the following steps: Setting up a cellular automaton model to solve the mutual correlation, determining the model parameters of the cellular automaton model, including road length, number of lanes, vehicle density, maximum speed, and random deceleration probability; According to the traffic observation data, the vehicle distribution on the road is initialized and generated, and the data information is called to give each vehicle an initial speed; Setting the update rules of the cellular automaton model, including acceleration rules, deceleration rules, random deceleration rules, and lane changing rules, also including extracting the number of lanes and slope curvature from the BIM model and writing them into the cellular automaton rule library, and calling the auxiliary green wave section to prompt the vehicle speed data of each vehicle; The road is divided into GIS coordinate grids, each cell carries BIM lane attributes and IoT real-time traffic flow status, and the vehicle state of all vehicles at each timestamp, including position and speed, is updated synchronously to obtain the latest simulation traffic flow dynamic evolution data; Obtaining the traffic flow change trend of the number of vehicles and vehicle speed over time; Loading multiple vehicle speed data and vehicle data into the traffic flow change trend to obtain the prediction of the future traffic flow data.

Images