Intelligent traffic scheduling optimization method based on cloud-side cooperative computing

Abstract

The invention relates to an intelligent traffic scheduling optimization method based on cloud-side cooperative computing, and belongs to the field of computer systems based on a specific computing model, and the method comprises the steps: monitoring each part of previous data corresponding to each previous day through a side end monitoring device above the foremost end of a set road section; a far-end cloud analysis device is used for intelligently predicting the total number of left-turning passing vehicles, the total number of straight-going vehicles and the total number of passing electric bicycles at the foremost end of a set road section in a fixed time interval of the day by adopting an intelligent traffic prediction model; and dynamic optimization of an intelligent traffic scheduling strategy in a fixed time interval of the day is realized based on an intelligent prediction result. According to the invention, various basic data screened in a targeted manner can be introduced, and the intelligent traffic prediction model with a customized structure is adopted to predict the traffic data of various vehicles in a future time interval, so that a corresponding traffic scheduling optimization strategy is formulated, and the utilization rate of limited lane resources is improved.

Description

Technical Field

[0001] The present invention relates to the field of computer systems based on specific computing models, and in particular to a smart traffic scheduling optimization method based on cloud-edge collaborative computing. Background Art

[0002] Cloud-edge collaborative computing is a computing model that combines cloud computing and edge computing, aiming to optimize data processing and transmission through collaborative work to meet the needs of different scenarios. Cloud-edge collaborative computing realizes a distributed computing model by sinking computing power from the centralized cloud to edge nodes close to the data source, thereby optimizing data processing capabilities. Cloud-edge collaborative computing is the fusion of cloud computing and edge computing, aiming to achieve efficient data processing and transmission. Cloud computing usually refers to storing data and applications on servers in remote data centers and providing services through the network, while edge computing pushes data processing to edge devices closer to the data source, such as sensors, embedded systems, etc. This collaborative model not only improves the performance and reliability of the system, but also meets complex and changing application requirements. For example, cloud-edge collaborative computing can be used to complete the targeted optimization of urban smart traffic scheduling.

[0003] For example, the Chinese invention patent publication CN118711373A proposes a smart traffic dispatch optimization system based on cloud-edge collaborative computing, which includes an edge monitoring module and a cloud analysis module. The system connects multiple edge processing devices through the cloud management center network, and each edge processing device is provided with a camera device and an edge monitoring module. After acquiring real-time traffic images, the edge monitoring module generates a feature data group through CNN algorithm model analysis, and transmits it to the cloud management center through the network. The cloud management center is provided with a cloud analysis module, which records the receiving time point and calculates the standard signal transmission time, clarifies the measurement standard of the data transmission quality of each edge processing device, judges the degree of collaboration between the cloud and the edge, and efficiently manages the cloud-edge data interaction with high accuracy. The collaborative analysis unit calculates and generates a traffic flow index, and generates a corresponding management report after comparing the traffic flow index threshold. The cloud-edge collaborative dispatch optimization effect is good.

[0004] For example, the Chinese invention patent publication CN118522141 A proposes a method and system for intelligent traffic monitoring, identification and control. The solution includes setting up sensors for information collection, dividing regions, forming remote areas and non-remote areas; continuously monitoring the remote areas to form a basic data group; online data collection for non-remote areas to form a driving estimate data group; estimating remote areas in combination with the basic data group to form the form estimate data for remote areas, and updating the basic data group when the vehicle passes through; correcting other driving times that need to pass through the corresponding remote areas; completing the traffic scheduling plan based on the final vehicle driving time, so that the client using the corresponding software can display the route with the optimal time. The solution performs online correction and analysis through the vehicle-mounted GPS and the pre-entered vehicle driving environment information, thereby completing the intelligent traffic scheduling for remote and non-remote areas.

[0005] It can be seen from this that the various technical solutions involving urban traffic scheduling in the above-mentioned prior art either only focus on the speed and efficiency of cloud-edge collaborative computing itself, or belong to the optimization of traffic routes without applying cloud-edge collaborative computing. They are unable to predict the traffic conditions of each "cell" of urban traffic, that is, each traffic section in the future, and naturally cannot optimize the corresponding intelligent traffic scheduling of each traffic section based on the prediction results. For example, due to the relatively limited number of lanes in each traffic section and the different congestion states of various types of vehicles in different time intervals of the same traffic section, it is difficult to accurately predict the congestion states of various types of vehicles in the future time interval of each traffic section. Under the premise that it is difficult to accurately predict the congestion states of various types of vehicles in the future time interval of each traffic section, it is impossible to flexibly adjust the traffic directions of each lane in the future time interval of the traffic section to meet the different traffic needs of different types of vehicles in the future time interval, resulting in the rigid and fixed traffic scheduling of each traffic section in the prior art, and it is difficult to make full use of limited lane resources. Summary of the invention

[0006] In order to solve the technical problems in the prior art, the present invention provides a smart traffic scheduling optimization method based on cloud-edge collaborative computing, which can introduce the congestion status of various types of vehicles on a set section at the same time in the past days, whether left turns are allowed in the front and rear sections of the set section, and the section length of the set section, the width of a single lane and the number of surrounding sections as basic data, and use a smart traffic prediction model with targeted structural design to intelligently predict the congestion status of various types of vehicles on the set section within a fixed time period on the day, and dynamically revise the number of left-turn lanes and the electric bicycle borrowing mode of the set section based on the intelligent prediction results, so as to make full use of the relatively limited lane resources of the set section and improve the travel speed and efficiency of various vehicles in each future time period of the set section.

[0007] According to the present invention, a smart traffic scheduling optimization method based on cloud-edge collaborative computing is provided, and the method comprises:

[0008] Using an edge monitoring device disposed above the front end of a set road section to monitor each set of past data corresponding to each past day, the single set of past data corresponding to each past day includes the total number of left-turning vehicles, the total number of straight-going vehicles, and the total number of electric bicycles passing at the front end of the set road section within a fixed time interval on that day, wherein the set road section includes a set number of lanes;

[0009] Capturing various section parameters of the set section, wherein the various section parameters of the set section are the section length of the set section, the width of a single lane, the number of surrounding sections, the left turn enable mark of the front section, and the left turn enable mark of the rear section;

[0010] Performing multiple training operations on the BP neural network to obtain the BP neural network after the multiple training operations are performed, and outputting the BP neural network after the multiple training operations as the intelligent traffic prediction model, wherein the number of training operations performed by the BP neural network is proportional to the time length of the fixed time interval;

[0011] A cloud analysis device disposed at a remote end is used to use a smart traffic prediction model to intelligently predict the total number of left-turning vehicles, the total number of straight-going vehicles, and the total number of electric bicycles passing at the front end of the set road section within the fixed time interval on the day based on the set number, the time length of the fixed time interval, the respective copies of past data corresponding to the past days, and the various road section parameters of the set road section;

[0012] The sum of the total number of left-turning vehicles at the front end of the set road section within the fixed time interval on the day and the total number of straight-going vehicles at the front end of the set road section within the fixed time interval on the day is taken as the predicted total number of vehicles, and the percentage of the total number of left-turning vehicles at the front end of the set road section within the fixed time interval on the day to the predicted total number of vehicles is calculated as a reference left-turn percentage, and when the total number of left-turning vehicles at the front end of the set road section within the fixed time interval on the day exceeds a set number threshold, a left-turn lane is configured for the set road section within the fixed time interval on the day;

[0013] When the ratio of the total number of electric bicycles passing at the front end of a set road section within a fixed time interval on that day to the total number of straight-going vehicles at the front end of the set road section within a fixed time interval on that day exceeds a preset ratio limit, the electric bicycles passing at the front end of the set road section within the fixed time interval on that day are allowed to use the leftmost straight-going lane of the set road section;

[0014] When the total number of left-turn vehicles at the front end of a set road section within a fixed time interval on the day exceeds a set number threshold, configuring a left-turn lane for the set road section within the fixed time interval on the day includes: configuring a matching number of left-turn lanes for the set road section within the fixed time interval on the day based on a reference left-turn percentage;

[0015] Wherein, configuring a matching number of left-turn lanes for a set road section within a fixed time interval of the day based on the reference left-turn percentage includes: the number of left-turn lanes configured for the set road section within the fixed time interval of the day is monotonically positively correlated with the reference left-turn percentage;

[0016] Wherein, in the set road section, the arrangement order of the left turn lane, the straight lane and the electric bicycle lane is from left to right;

[0017] The cloud analysis device is connected to each edge monitoring device corresponding to each road section via a wireless network.

[0018] Compared with the prior art, the present invention has at least the following five main inventive concepts:

[0019] Inventive concept A: Based on the total number of left-turning vehicles, the total number of straight-going vehicles, and the total number of electric bicycles passing through the set road section at fixed time intervals on the same day in the past days and various road section parameters of the set road section, the total number of left-turning vehicles, the total number of straight-going vehicles, and the total number of electric bicycles passing through the set road section at a fixed time interval on the same day are intelligently predicted, thereby completing the intelligent prediction of the total number of left-turning vehicles, the total number of straight-going vehicles, and the total number of electric bicycles passing through the future time interval based on the past simultaneous traffic conditions of the set road section and various road section parameters inherent to the set road section, providing valuable reference data for the optimization of intelligent traffic scheduling of the set road section in the future time interval;

[0020] Inventive concept B: In addition to the total number of left-turning vehicles, the total number of straight-going vehicles, and the total number of electric bicycles passing through the set road section at fixed time intervals on the past days and various road section parameters of the set road section, the basic data used for intelligent prediction also include the number of lanes of the set road section and the time length of the fixed time interval. At the same time, the various road section parameters of the set road section are the section length of the set road section, the width of a single lane, the number of surrounding road sections, the left-turn enabling mark of the road section ahead, and the left-turn enabling mark of the road section behind, thereby ensuring the sufficiency and comprehensiveness of the basic data and laying a foundation for the reliability and stability of the intelligent prediction results;

[0021] Inventive concept C: structural customization of a smart traffic prediction model for performing intelligent prediction, specifically, performing multiple training operations on a BP neural network to obtain a BP neural network after performing multiple training operations, and outputting the BP neural network after performing multiple training operations as a smart traffic prediction model, wherein the number of training operations performed by the BP neural network is proportional to the length of a fixed time interval, thereby establishing smart traffic prediction models of different structures for time intervals of different lengths, further ensuring the reliability and stability of the smart prediction results;

[0022] Inventive concept D: In each training operation performed on the BP neural network, the total number of left-turning vehicles, the total number of straight-going vehicles, and the total number of electric bicycles passing at the front end of a set road section within a fixed time interval on a certain historical day are used as output content of the BP neural network, and the set number, the length of the fixed time interval, the past data corresponding to each of the past days corresponding to the certain historical day, and the various road section parameters of the set road section are used as input content of the BP neural network to complete this training operation, thereby ensuring the training effect of each training operation of the BP neural network;

[0023] Inventive concept E: Based on the prediction results, a corresponding optimization strategy for intelligent traffic scheduling is established to ensure the speed and efficiency of various vehicles on the set road section within the fixed time period of the day, and reduce the congestion probability of the set road section. The specific optimization strategy is to calculate the percentage of the total number of left-turning vehicles at the front end of the set road section within the fixed time period of the day to the total number of passing vehicles as a reference left-turn percentage. When the predicted total number of left-turning vehicles exceeds the set number threshold, a left-turn lane is configured for the set road section within the fixed time period of the day, and a matching number of left-turn lanes are configured for the set road section within the fixed time period of the day based on the reference left-turn percentage. At the same time, when the proportion of the predicted total number of passing electric bicycles to the predicted total number of straight-moving vehicles exceeds the preset proportion limit, the passing electric bicycles at the front end of the set road section within the fixed time period of the day are allowed to use the leftmost straight-moving lane of the set road section. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The embodiments of the present invention will be described below with reference to the accompanying drawings, wherein:

[0025] Figure 1 This is a technical flow chart of the intelligent traffic scheduling optimization method based on cloud-edge collaborative computing according to the present invention.

[0026] Figure 2 This is a flowchart of the steps of a smart traffic scheduling optimization method based on cloud-edge collaborative computing according to Example 1 of the present invention.

[0027] Figure 3 This is a flowchart of the steps of a smart traffic scheduling optimization method based on cloud-edge collaborative computing according to Example 2 of the present invention.

[0028] Figure 4 This is a flowchart of the steps of a smart traffic scheduling optimization method based on cloud-edge collaborative computing according to Example 3 of the present invention.

[0029] Figure 5 This is a flowchart of the steps of a smart traffic scheduling optimization method based on cloud-edge collaborative computing according to Example 4 of the present invention.

[0030] Figure 6 A flowchart of a smart traffic scheduling optimization method based on cloud-edge collaborative computing according to Example 5 of the present invention DETAILED DESCRIPTION

[0031] like Figure 1 As shown, a technical flow chart of the intelligent traffic scheduling optimization method based on cloud-edge collaborative computing according to the present invention is given.

[0032] exist Figure 1 In the example, the city where smart traffic dispatch optimization is performed includes various road sections, and each road section is equipped with an edge monitoring device, for example, Figure 1 Two edge monitoring devices are given, serving two road sections respectively. At the same time, a cloud analysis device is set up on the management side. The cloud analysis device is connected to each edge monitoring device on each road section through a wireless network, thereby building the hardware resources of the smart traffic scheduling optimization mechanism based on cloud-edge collaborative computing.

[0033] Specifically, in cities with optimized smart traffic dispatch, corresponding edge monitoring equipment is installed above the front end of each road section, and Figure 1 In the embodiment, each edge monitoring device includes a camera mechanism, a mobile base station and a traffic light device, and may also include a counting mechanism (not shown) and a timing mechanism (not shown);

[0034] like Figure 1 As shown, the specific technical process of the present invention is as follows, wherein technical processes one and two can be selected to be executed at the edge monitoring device, and technical processes three, four, and five can be selected to be executed at the cloud analysis device:

[0035] Technical process 1: For the set road sections in the city where smart traffic dispatch optimization is implemented, the total number of left-turning vehicles, the total number of straight-going vehicles, and the total number of electric bicycles passing in a fixed time interval in the past days are collected, and at the same time, various road section parameters of the set road section are obtained, as well as the number of lanes of the set road section and the length of the fixed time interval;

[0036] Specifically, the various section parameters of the set section are the section length of the set section, the width of a single lane, the number of surrounding sections, the left turn enable mark of the front section, and the left turn enable mark of the rear section;

[0037] In this way, through the full and comprehensive selection of multiple basic data, the foundation is laid for the reliability and stability of subsequent intelligent prediction results;

[0038] Technical process 2: Customize smart traffic prediction models with different structures for time intervals of different lengths to further ensure the reliability and stability of smart prediction results;

[0039] Specifically, the structural customization of the intelligent traffic prediction model is reflected in the following aspects:

[0040] First, perform multiple training operations on the BP neural network to obtain the BP neural network after the multiple training operations are performed, and use the BP neural network after the multiple training operations as the intelligent traffic prediction model;

[0041] Second, the number of training operations performed by the BP neural network is proportional to the length of the fixed time interval;

[0042] Third, in each training operation performed on the BP neural network, the total number of left-turning vehicles, the total number of straight-going vehicles, and the total number of electric bicycles passing at the front end of a set road section within a fixed time interval on a certain historical day are used as output content of the BP neural network, and the set number, the length of the fixed time interval, the past data corresponding to each of the past days corresponding to the certain historical day, and the various road section parameters of the set road section are used as input content of the BP neural network to complete this training operation, thereby ensuring the training effect of each training operation of the BP neural network;

[0043] Technical process three: Adopt the customized intelligent traffic prediction model of technical process one, based on the sufficient and comprehensive basic data selected in technical process one, to intelligently predict the total number of left-turning vehicles, the total number of straight-going vehicles, and the total number of electric bicycles passing through the set road section within a fixed time interval on the day;

[0044] Technical process 4: Based on the total number of left-turning vehicles, straight-going vehicles and electric bicycles on a set road section within a fixed time interval on that day obtained by intelligent prediction in technical process 3, a dynamic optimization plan is provided for the intelligent traffic dispatch of each lane within a fixed time interval on the set road section on that day;

[0045] Specifically, the percentage of the total number of left-turn passing vehicles at the front end of a set road section within a fixed time interval on that day to the total number of passing vehicles is calculated as a reference left-turn percentage, and when the predicted total number of left-turn passing vehicles exceeds a set number threshold, a left-turn lane is configured for the set road section within the fixed time interval on that day, and a matching number of left-turn lanes are configured for the set road section within the fixed time interval on that day based on the reference left-turn percentage;

[0046] Specifically, when the ratio of the predicted total number of electric bicycles passing through to the predicted total number of through vehicles exceeds a preset ratio limit, the electric bicycles passing through at the front end of a set road section within a fixed time interval on that day are allowed to use the leftmost through lane of the set road section;

[0047] In this way, the traffic needs of different vehicles can be met as much as possible under limited traffic resources.

[0048] The key points of the present invention are: full and comprehensive selection of multiple basic data including the total number of left-turning vehicles, the total number of straight-going vehicles and the total number of electric bicycles passing through the set road section in the same time period in the past days, various road section parameters of the set road section, the number of lanes of the set road section and the time length of the fixed time interval, structural customization of the intelligent traffic prediction model, construction of hardware resources for the intelligent traffic scheduling optimization mechanism based on cloud-edge collaborative computing, and dynamic adjustment of the intelligent traffic scheduling optimization strategy based on intelligent prediction results.

[0049] Below, the intelligent traffic scheduling optimization method based on cloud-edge collaborative computing of the present invention will be specifically described in the form of an embodiment.

[0050] Example 1

[0051] Figure 2 This is a flowchart of the steps of a smart traffic scheduling optimization method based on cloud-edge collaborative computing according to Example 1 of the present invention.

[0052] like Figure 2 As shown, the intelligent traffic scheduling optimization method based on cloud-edge collaborative computing includes the following specific steps:

[0053] Step S21: using an edge monitoring device disposed above the front end of a set road section to monitor the past data corresponding to each day in the past, the single past data corresponding to each day in the past includes the total number of left-turning vehicles, the total number of straight-going vehicles, and the total number of electric bicycles passing at the front end of the set road section within a fixed time interval on that day, wherein the set road section includes a set number of lanes;

[0054] For example, an edge monitoring device arranged above the front end of a set road section is used to monitor each set of past data corresponding to each of the past days. The single set of past data corresponding to each of the past days includes the total number of left-turning vehicles, the total number of straight-going vehicles and the total number of electric bicycles passing at the front end of the set road section within a fixed time interval on that day. The set road section includes a set number of lanes, including: intelligent prediction of the traffic data of various types of vehicles of the set road section within the fixed time interval from 9 a.m. to 9:30 a.m. on that day, and the each set of past data corresponding to each of the past days is the traffic data of various types of vehicles within the fixed time interval from 9 a.m. to 9:30 a.m. in the past 8 days. Since the fixed time interval from 9 a.m. to 9:30 a.m. on that day has not yet arrived, it belongs to the future time interval and needs to be intelligently predicted.

[0055] Specifically, the city where the intelligent traffic dispatch optimization is performed includes multiple road sections, and the set road section is one of them, which has an end and a front end of the road section. All types of vehicles pass from the end of the set road section to the front end of the set road section. At the front end of the set road section, some vehicles have a left turn demand, some vehicles have a straight-going demand, and there is also a passing area demand for electric bicycles;

[0056] Step S22: capturing various section parameters of the set section, wherein the various section parameters of the set section are the section length of the set section, the width of a single lane, the number of surrounding sections, the left turn enable mark of the front section, and the left turn enable mark of the rear section;

[0057] For example, capturing various section parameters of the set section, wherein the various section parameters of the set section are the section length of the set section, the width of a single lane, the number of surrounding sections, the left turn enable flag of the front section, and the left turn enable flag of the rear section, including: when the distance from the end of the set section to the front end of the set section is 800 meters, the section length of the set section is 800 meters;

[0058] Step S23: performing multiple training operations on the BP neural network to obtain the BP neural network after the multiple training operations are performed, and outputting the BP neural network after the multiple training operations as the intelligent traffic prediction model, wherein the number of training operations performed by the BP neural network is proportional to the time length of the fixed time interval;

[0059] Specifically, multiple training operations are performed on the BP neural network to obtain the BP neural network after the multiple training operations are performed, and the BP neural network after the multiple training operations are performed is output as the intelligent traffic prediction model, and the number of training operations performed by the BP neural network is proportional to the time length of the fixed time interval, including: the time length of the fixed time interval is 15 minutes, the number of training operations performed by the BP neural network is 150, the time length of the fixed time interval is 30 minutes, that is, half an hour, the number of training operations performed by the BP neural network is 300, the time length of the fixed time interval is 60 minutes, that is, 1 hour, the number of training operations performed by the BP neural network is 600, and so on;

[0060] Step S24: using a remote cloud analysis device to use a smart traffic prediction model to intelligently predict the total number of left-turning vehicles, the total number of straight-going vehicles, and the total number of electric bicycles passing at the front end of the set road section within the fixed time interval on the day based on the set number, the length of the fixed time interval, the respective past data corresponding to the past days, and the various road section parameters of the set road section;

[0061] For example, a cloud analysis device disposed at a remote end is used to adopt a smart traffic prediction model to intelligently predict the total number of left-turning vehicles, the total number of straight-going vehicles, and the total number of electric bicycles passing at the front end of a set road section within a fixed time interval on that day based on the set number, the time length of the fixed time interval, the respective copies of past data corresponding to the past days, and the various road section parameters of the set road section, including: the set number is the number of lanes of the set road section, for example, 10 lanes;

[0062] Step S25: taking the sum of the total number of left-turning vehicles at the front end of the set road section within the fixed time interval on the day and the total number of straight-going vehicles at the front end of the set road section within the fixed time interval on the day as the predicted total number of vehicles, calculating the percentage of the total number of left-turning vehicles at the front end of the set road section within the fixed time interval on the day to the predicted total number of vehicles as a reference left-turn percentage, and when the total number of left-turning vehicles at the front end of the set road section within the fixed time interval on the day exceeds a set number threshold, configuring a left-turn lane for the set road section within the fixed time interval on the day;

[0063] For example, when the total number of left-turn passing vehicles at the front end of a set road section in a fixed time interval on the day exceeds a set number threshold, configuring a left-turn lane for the set road section in the fixed time interval on the day includes: the set number threshold may be 500, and when the total number of left-turn passing vehicles at the front end of the set road section in the fixed time interval on the day exceeds 500, configuring a left-turn lane for the set road section in the fixed time interval on the day;

[0064] Step S26: when the ratio of the total number of electric bicycles passing at the front end of the set road section within the fixed time interval on the day to the total number of straight-going vehicles at the front end of the set road section within the fixed time interval on the day exceeds a preset ratio limit, the electric bicycles passing at the front end of the set road section within the fixed time interval on the day are allowed to use the leftmost straight-going lane of the set road section;

[0065] For example, when the ratio of the total number of electric bicycles passing at the front end of a set road section within a fixed time interval on a day to the total number of straight-going vehicles at the front end of the set road section within a fixed time interval on a day exceeds a preset ratio limit, the electric bicycles passing at the front end of the set road section within the fixed time interval on a day are allowed to borrow the leftmost straight-going lane of the set road section, including: Here, the electric bicycles passing are only borrowing the lane, and the leftmost straight-going lane of the set road section is a driving area where the electric bicycles can borrow the lane, which needs to be clearly stated by the motor vehicle driver;

[0066] When the total number of left-turn vehicles at the front end of a set road section within a fixed time interval on the day exceeds a set number threshold, configuring a left-turn lane for the set road section within the fixed time interval on the day includes: configuring a matching number of left-turn lanes for the set road section within the fixed time interval on the day based on a reference left-turn percentage;

[0067] Wherein, configuring a matching number of left-turn lanes for a set road section within a fixed time interval of the day based on the reference left-turn percentage includes: the number of left-turn lanes configured for the set road section within the fixed time interval of the day is monotonically positively correlated with the reference left-turn percentage;

[0068] Specifically, the number of left-turn lanes configured for a set section within a fixed time interval of the day and the reference left-turn percentage are monotonically positively associated, including: when the number of lanes on the set section is 10, when the value of the reference left-turn percentage is 20%, the number of left-turn lanes configured for the selected set section within the fixed time interval of the day is 2, when the value of the reference left-turn percentage is 35%, the number of left-turn lanes configured for the selected set section within the fixed time interval of the day is 3, and when the value of the reference left-turn percentage is 50%, the number of left-turn lanes configured for the selected set section within the fixed time interval of the day is 4;

[0069] In this way, considering that there are left-turn lanes in the sections ahead and behind the set section, vehicles that will have left-turn traffic requirements in the future can choose to turn left. Through the design of the above numerical correspondence, the traffic requirements of the straight lanes of the set section can be met as much as possible under the premise of ensuring the matching of the traffic requirements of left-turn vehicles, thereby ensuring the traffic efficiency of the set section.

[0070] Wherein, in the set road section, the arrangement order of the left turn lane, the straight lane and the electric bicycle lane is from left to right;

[0071] The cloud analysis device is connected to each edge monitoring device corresponding to each road section through a wireless network;

[0072] Wherein, the edge monitoring device arranged at the front end of the set road section is used to monitor the past data corresponding to each day in the past, and the single past data corresponding to each day in the past includes the total number of left-turning vehicles, the total number of straight-going vehicles and the total number of electric bicycles passing at the front end of the set road section within a fixed time interval on that day, and the set road section includes a set number of lanes, including: the smaller the value of the set number, the fewer the days of the selected past days;

[0073] For example, the smaller the value of the set number is, the fewer the number of days in the past selected are, including: when the set number is 10, the number of days in the past selected is 8 days, when the set number is 8, the number of days in the past selected is 7 days, when the set number is 6, the number of days in the past selected is 6 days, and so on;

[0074] The capturing of various road section parameters of the set road section includes the road section length of the set road section, the width of a single lane, the number of surrounding road sections, the left turn enable flag of the road section ahead, and the left turn enable flag of the road section behind, including: among a set number of lanes in the set road section, the width of each lane is the same, and the length of each lane is the same, and the length of each lane is equal to the road section length of the set road section;

[0075] Among them, performing multiple training operations on the BP neural network to obtain the BP neural network after performing multiple training operations, and outputting the BP neural network after performing multiple training operations as a smart traffic prediction model, the number of training operations performed by the BP neural network is proportional to the time length of the fixed time interval, including: in each training operation performed on the BP neural network, the total number of left-turning vehicles, the total number of straight-going vehicles, and the total number of electric bicycles passing at the front end of a set road section within a fixed time interval on a certain day in history are used as output content of the BP neural network, and the set number, the time length of the fixed time interval, the respective copies of past data corresponding to the past days corresponding to the certain day in history, and the various road section parameters of the set road section are used as input content of the BP neural network to complete this training operation;

[0076] And wherein, in each training operation performed on the BP neural network, the known total number of left-turning vehicles, the total number of straight-going vehicles and the total number of electric bicycles passing at the front end of a set road section within a fixed time interval on a certain historical day are used as the output content of the BP neural network, and the set number, the time length of the fixed time interval, the various past data corresponding to the past days corresponding to the certain historical day and the various road section parameters of the set road section are used as the input content of the BP neural network, and completing this training operation includes: the various past data corresponding to the past days corresponding to the certain historical day are the various past data corresponding to the past days before the certain historical day.

[0077] Example 2

[0078] Figure 3 This is a flowchart of the steps of a smart traffic scheduling optimization method based on cloud-edge collaborative computing according to Example 2 of the present invention.

[0079] like Figure 3 As shown, Figure 2 Different from the embodiment in, after step S26, the method further includes:

[0080] Step S31: using a projection mechanism disposed above the front end of the set road section to perform a projection operation of a standard pattern of an electric bicycle on the leftmost straight lane of the set road section;

[0081] Specifically, the projection operation of performing the projection of the standard pattern of an electric bicycle on the leftmost straight lane of the set section of road using a projection mechanism disposed above the front end of the set section of road includes: the projection mechanism projects the standard pattern of an electric bicycle as a traffic sign onto the road surface of the leftmost straight lane of the set section of road, thereby providing a traffic lane reminder operation for passing electric bicycles and straight-moving vehicles.

[0082] Example 3

[0083] Figure 4 This is a flowchart of the steps of a smart traffic scheduling optimization method based on cloud-edge collaborative computing according to Example 3 of the present invention.

[0084] like Figure 4 As shown, Figure 2 Different from the embodiment in, after step S25, the method further includes:

[0085] Step S41: using a lane traffic sign disposed at the front end of a set road section to mark the respective traffic directions of each left turn lane and each straight lane of the set road section;

[0086] For example, using a lane traffic sign disposed at the front end of a set road section to mark the respective traffic directions of each left turn lane and each straight lane of the set road section includes: the lane traffic sign disposed at the front end of the set road section is an LED display screen.

[0087] Example 4

[0088] Figure 5 This is a flowchart of the steps of a smart traffic scheduling optimization method based on cloud-edge collaborative computing according to Example 4 of the present invention.

[0089] like Figure 5 As shown, Figure 2 Different from the embodiment in, after step S23, the method further includes:

[0090] Step S51: using a data storage mechanism to complete the model storage of the smart traffic prediction model by storing various model parameters of the smart traffic prediction model;

[0091] Specifically, the data storage mechanism is used to complete the model storage of the intelligent traffic prediction model by storing the various model parameters of the intelligent traffic prediction model, including: an MMC storage chip, a TF storage chip or a dynamic storage chip can be selected as the data storage mechanism, and the model storage of the intelligent traffic prediction model is completed by storing the various model parameters of the intelligent traffic prediction model.

[0092] Example 5

[0093] Figure 6 This is a flowchart of the steps of a smart traffic scheduling optimization method based on cloud-edge collaborative computing according to Example 5 of the present invention.

[0094] like Figure 6 As shown, Figure 2 Different from the embodiment in, before step S21, the method further includes:

[0095] Step S61: using a wireless communication link to establish a wireless network connection between the cloud analysis device and the edge monitoring device of the set road section;

[0096] For example, using a wireless communication link to establish a wireless network connection between the cloud analysis device and the edge monitoring device of the set road section includes: the wireless communication link is a frequency division duplex communication link or a time division duplex communication link.

[0097] Next, various method embodiments of the present invention are described in detail.

[0098] In the intelligent traffic scheduling optimization method based on cloud-edge collaborative computing according to various method embodiments of the present invention:

[0099] The edge monitoring device disposed above the front end of a set road section is used to monitor the respective sets of past data corresponding to each past day, wherein the single set of past data corresponding to each past day includes the total number of left-turning vehicles, the total number of straight-moving vehicles, and the total number of electric bicycles passing at the front end of the set road section within a fixed time interval on that day, wherein the set road section includes a set number of lanes and further includes: the edge monitoring device has a built-in camera mechanism, a counting mechanism, and a timing mechanism, which are used to monitor the total number of left-turning vehicles, the total number of straight-moving vehicles, and the total number of electric bicycles passing at the front end of the set road section within a fixed time interval on each day;

[0100] For example, the edge monitoring device has a built-in camera mechanism, a counting mechanism and a timing mechanism, and is used to monitor the total number of left-turning vehicles, the total number of straight-going vehicles and the total number of electric bicycles passing at the front end of a set road section within a fixed time interval every day. The camera mechanism has a built-in wide-angle lens and a CMOS image sensor;

[0101] Among them, the edge monitoring equipment has a built-in camera mechanism, a counting mechanism and a timing mechanism, which are used to monitor the total number of left-turning vehicles, the total number of straight-moving vehicles and the total number of electric bicycles passing at the front end of a set road section within a fixed time interval every day. The counting mechanism is respectively connected to the camera mechanism and the timing mechanism.

[0102] In the intelligent traffic scheduling optimization method based on cloud-edge collaborative computing according to various method embodiments of the present invention:

[0103] Capturing various section parameters of the set section, wherein the various section parameters of the set section are the section length of the set section, the width of a single lane, the number of surrounding sections, the left turn enable flag of the front section and the left turn enable flag of the rear section, and further comprising: the left turn enable flag of the front section of the set section is used to indicate whether the front section of the set section allows the vehicle to turn left, and the left turn enable flag of the rear section of the set section is used to indicate whether the rear section of the set section allows the vehicle to turn left;

[0104] The left-turn enabling flag of the road section ahead of the set road section is used to indicate whether the road section ahead of the set road section allows the vehicle to turn left, and the left-turn enabling flag of the road section behind the set road section is used to indicate whether the road section behind the set road section allows the vehicle to turn left, including: when the left-turn enabling flag of the road section ahead of the set road section is 0B0001, it indicates that the road section ahead of the set road section allows the vehicle to turn left, and when the left-turn enabling flag of the road section ahead of the set road section is 0B0000, it indicates that the road section ahead of the set road section prohibits the vehicle from turning left;

[0105] Among them, the left turn enable flag of the front section of the set road section is used to indicate whether the vehicle is allowed to turn left on the front section of the set road section, and the left turn enable flag of the rear section of the set road section is used to indicate whether the vehicle is allowed to turn left on the rear section of the set road section, including: when the left turn enable flag of the rear section of the set road section is 0B0011, it indicates that the vehicle is allowed to turn left on the rear section of the set road section; when the left turn enable flag of the rear section of the set road section is 0B0010, it indicates that the vehicle is prohibited from turning left on the rear section of the set road section.

[0106] And in the intelligent traffic scheduling optimization method based on cloud-edge collaborative computing according to various method embodiments of the present invention:

[0107] Performing multiple training operations on the BP neural network to obtain the BP neural network after the multiple training operations, and outputting the BP neural network after the multiple training operations as the intelligent traffic prediction model, wherein the number of training operations performed by the BP neural network is proportional to the time length of the fixed time interval, further comprising: using a content mapping formula to represent a content mapping relationship in which the number of training operations performed by the BP neural network is proportional to the time length of the fixed time interval;

[0108] For example, a numerical simulation mode may be selected to implement testing and simulation of a data processing process that uses a content mapping formula to represent a content mapping relationship in which the number of training operations performed by the BP neural network is proportional to the time length of the fixed time interval;

[0109] And wherein, multiple training operations are performed on the BP neural network to obtain the BP neural network after the multiple training operations are performed, and the BP neural network after the multiple training operations is output as a smart traffic prediction model, and the number of training operations performed by the BP neural network is proportional to the time length of the fixed time interval. It also includes: in the content mapping formula, the number of training operations performed by the BP neural network is the output data, and the time length of the fixed time interval is the input data.

[0110] In addition, the present invention may also cite the following technical contents to highlight the contribution of the present invention to the prior art:

[0111] The cloud analysis device arranged at a remote end is used to adopt a smart traffic prediction model to intelligently predict the total number of left-turning vehicles, the total number of straight-moving vehicles and the total number of electric bicycles passing at the front end of the set road section within the fixed time interval on the day based on the set number, the time length of the fixed time interval, the respective copies of past data corresponding to the past days and the various road section parameters of the set road section, including: performing numerical normalization processing on the set number, the time length of the fixed time interval, the respective copies of past data corresponding to the past days and the various road section parameters of the set road section respectively and then synchronously inputting them into the smart traffic prediction model, and the total number of left-turning vehicles, the total number of straight-moving vehicles and the total number of electric bicycles passing at the front end of the set road section within the fixed time interval on the day intelligently predicted are numerically expressed in the form of normalized values;

[0112] For example, the set number, the time length of the fixed time interval, the respective copies of past data corresponding to the past days, and the various section parameters of the set road section are numerically normalized and then synchronously input into the intelligent traffic prediction model, and the total number of left-turning vehicles, the total number of straight-moving vehicles, and the total number of electric bicycles passing at the front end of the set road section within the fixed time interval on the day of the intelligent prediction are numerically normalized. The numerical representation includes: the set number, the time length of the fixed time interval, the respective copies of past data corresponding to the past days, and the various section parameters of the set road section are binary-converted and then synchronously input into the intelligent traffic prediction model, and the total number of left-turning vehicles, the total number of straight-moving vehicles, and the total number of electric bicycles passing at the front end of the set road section within the fixed time interval on the day of the intelligent prediction are numerically represented after binary-conversion processing;

[0113] For example, the set number, the time length of the fixed time interval, the respective copies of past data corresponding to the past days, and the various section parameters of the set road section are numerically normalized and then synchronously input into the intelligent traffic prediction model, and the total number of left-turning vehicles, the total number of straight-going vehicles, and the total number of electric bicycles passing at the front end of the set road section within the fixed time interval on the day of intelligent prediction are numerically normalized numerical representations, and also include: using a numerical processing component and a synchronous drive interface to collaboratively complete the numerical normalization of the set number, the time length of the fixed time interval, the respective copies of past data corresponding to the past days, and the various section parameters of the set road section, and synchronously input into the intelligent traffic prediction model;

[0114] And wherein, performing multiple training operations on the BP neural network to obtain the BP neural network after performing the multiple training operations, and outputting the BP neural network after performing the multiple training operations as the intelligent traffic prediction model, the number of training operations performed by the BP neural network is proportional to the time length of the fixed time interval, including: using a numerical simulation mode to complete the testing and simulation of the data processing process of performing multiple training operations on the BP neural network;

[0115] Specifically, using the numerical simulation mode to complete the test and simulation of the data processing process of performing multiple training operations on the BP neural network includes: using the MATLAB toolbox to complete the test and simulation of the data processing process of performing multiple training operations on the BP neural network.

[0116] The foregoing description of the exemplary embodiments of the present invention is provided for the purpose of illustration and description. It is not intended to be exhaustive or to limit the present invention to the exact form disclosed. Obviously, many modifications and variations are obvious to those skilled in the art. The exemplary embodiments are selected and described in order to best illustrate the principles of the present invention and its practical application, so that other persons skilled in the art can understand the various embodiments and various variations of the present invention applicable to the specific purpose contemplated. It is intended that the scope of the present invention be limited by the appended claims and their equivalents.

Claims

1. A smart traffic dispatch optimization method based on cloud-edge collaborative computing, characterized in that: The method comprises: Using an edge monitoring device disposed above the front end of a set road section to monitor each set of past data corresponding to each past day, the single set of past data corresponding to each past day includes the total number of left-turning vehicles, the total number of straight-going vehicles, and the total number of electric bicycles passing at the front end of the set road section within a fixed time interval on that day, wherein the set road section includes a set number of lanes; Capturing various section parameters of the set section, wherein the various section parameters of the set section are the section length of the set section, the width of a single lane, the number of surrounding sections, the left turn enable mark of the front section, and the left turn enable mark of the rear section; Performing multiple training operations on the BP neural network to obtain the BP neural network after the multiple training operations are performed, and outputting the BP neural network after the multiple training operations as the intelligent traffic prediction model, wherein the number of training operations performed by the BP neural network is proportional to the time length of the fixed time interval; A cloud analysis device disposed at a remote end is used to use a smart traffic prediction model to intelligently predict the total number of left-turning vehicles, the total number of straight-going vehicles, and the total number of electric bicycles passing at the front end of the set road section within the fixed time interval on the day based on the set number, the time length of the fixed time interval, the respective copies of past data corresponding to the past days, and the various road section parameters of the set road section; The sum of the total number of left-turning vehicles at the front end of the set road section within the fixed time interval on the day and the total number of straight-going vehicles at the front end of the set road section within the fixed time interval on the day is taken as the predicted total number of vehicles, and the percentage of the total number of left-turning vehicles at the front end of the set road section within the fixed time interval on the day to the predicted total number of vehicles is calculated as a reference left-turn percentage, and when the total number of left-turning vehicles at the front end of the set road section within the fixed time interval on the day exceeds a set number threshold, a left-turn lane is configured for the set road section within the fixed time interval on the day; When the ratio of the total number of electric bicycles passing at the front end of a set road section within a fixed time interval on that day to the total number of straight-going vehicles at the front end of the set road section within a fixed time interval on that day exceeds a preset ratio limit, the electric bicycles passing at the front end of the set road section within the fixed time interval on that day are allowed to use the leftmost straight-going lane of the set road section; When the total number of left-turn vehicles at the front end of a set road section within a fixed time interval on the day exceeds a set number threshold, configuring a left-turn lane for the set road section within the fixed time interval on the day includes: configuring a matching number of left-turn lanes for the set road section within the fixed time interval on the day based on a reference left-turn percentage; Wherein, configuring a matching number of left-turn lanes for a set road section within a fixed time interval of the day based on the reference left-turn percentage includes: the number of left-turn lanes configured for the set road section within the fixed time interval of the day is monotonically positively correlated with the reference left-turn percentage; Wherein, in the set road section, the arrangement order of the left turn lane, the straight lane and the electric bicycle lane is from left to right; The cloud analysis device is connected to each edge monitoring device corresponding to each road section via a wireless network.

2. The intelligent traffic dispatch optimization method based on cloud-edge collaborative computing according to claim 1, characterized in that: The edge monitoring device arranged at the front end of the set road section is used to monitor the past data corresponding to each day in the past, and the single past data corresponding to each day in the past includes the total number of left-turning vehicles, the total number of straight-going vehicles and the total number of electric bicycles passing at the front end of the set road section within a fixed time interval on that day, and the set road section includes a set number of lanes, including: the smaller the value of the set number, the fewer the days of the selected past days; Among them, various section parameters of the set road section are captured, and the various section parameters of the set road section are the section length of the set road section, the width of a single lane, the number of surrounding sections, the left turn enable flag of the front section and the left turn enable flag of the rear section, including: among the set number of lanes in the set road section, the width of each lane is the same, and the length of each lane is the same, and the length of each lane is equal to the section length of the set road section.

3. The intelligent traffic dispatch optimization method based on cloud-edge collaborative computing as claimed in claim 2 is characterized by: Performing multiple training operations on the BP neural network to obtain the BP neural network after the multiple training operations, and outputting the BP neural network after the multiple training operations as a smart traffic prediction model, wherein the number of training operations performed by the BP neural network is proportional to the time length of the fixed time interval, including: in each training operation performed on the BP neural network, the total number of left-turning vehicles, the total number of straight-going vehicles, and the total number of electric bicycles passing at the front end of a set road section within a fixed time interval on a certain day in history are used as output content of the BP neural network, and the set number, the time length of the fixed time interval, the respective copies of past data corresponding to the past days corresponding to the certain day in history, and the various road section parameters of the set road section are used as input content of the BP neural network to complete this training operation; Among them, in each training operation performed on the BP neural network, the total number of left-turning vehicles, the total number of straight-going vehicles and the total number of electric bicycles passing at the front end of a set road section within a fixed time interval on a certain historical day are used as the output content of the BP neural network, and the set number, the time length of the fixed time interval, the various past data corresponding to the past days corresponding to the certain historical day and the various section parameters of the set road section are used as the input content of the BP neural network. Completing this training operation includes: the various past data corresponding to the past days corresponding to the certain historical day are the various past data corresponding to the past days before the certain historical day.

4. The intelligent traffic dispatch optimization method based on cloud-edge collaborative computing as claimed in claim 3 is characterized in that: When the ratio of the total number of electric bicycles passing at the front end of a set road section within a fixed time interval on a day to the total number of straight-going vehicles at the front end of the set road section within a fixed time interval on a day exceeds a preset ratio limit, after allowing the electric bicycles passing at the front end of the set road section within the fixed time interval on a day to use the leftmost straight-going lane of the set road section, the method further includes: A projection mechanism disposed above the front end of the set road section is used to perform a projection operation of a standard pattern of an electric bicycle on the leftmost straight lane of the set road section.

5. The intelligent traffic dispatch optimization method based on cloud-edge collaborative computing as claimed in claim 3 is characterized in that: The sum of the total number of left-turning vehicles at the front end of the set road section within the fixed time interval on the day and the total number of straight-going vehicles at the front end of the set road section within the fixed time interval on the day is used as the predicted total number of vehicles, and the percentage of the total number of left-turning vehicles at the front end of the set road section within the fixed time interval on the day to the predicted total number of vehicles is calculated as a reference left-turn percentage. When the total number of left-turning vehicles at the front end of the set road section within the fixed time interval on the day exceeds a set number threshold, after configuring a left-turn lane for the set road section within the fixed time interval on the day, the method further includes: A lane traffic signboard arranged at the front end of a set road section is used to mark the respective traffic directions of each left turn lane and each straight lane of the set road section.

6. The intelligent traffic dispatch optimization method based on cloud-edge collaborative computing as claimed in claim 3 is characterized in that: Performing multiple training operations on the BP neural network to obtain the BP neural network after the multiple training operations are performed, and outputting the BP neural network after the multiple training operations as the intelligent traffic prediction model, after the number of training operations performed by the BP neural network is proportional to the time length of the fixed time interval, the method further includes: A data storage mechanism is used to complete the model storage of the smart traffic prediction model by storing various model parameters of the smart traffic prediction model.

7. The intelligent traffic dispatch optimization method based on cloud-edge collaborative computing as claimed in claim 3 is characterized in that: The edge monitoring device disposed above the front end of the set road section is used to monitor the respective past data corresponding to each past day, and the single past data corresponding to each past day includes the total number of left-turning vehicles, the total number of straight-going vehicles, and the total number of electric bicycles passing at the front end of the set road section within a fixed time interval on that day, and the set road section includes a set number of lanes before the lane, and the method further includes: A wireless communication link is used to establish a wireless network connection between the cloud analysis device and the edge monitoring device of the set road section.

8. The intelligent traffic dispatch optimization method based on cloud-edge collaborative computing according to any one of claims 3 to 7, characterized in that: The edge monitoring device disposed above the front end of a set road section is used to monitor the respective sets of past data corresponding to each past day, wherein the single set of past data corresponding to each past day includes the total number of left-turning vehicles, the total number of straight-moving vehicles, and the total number of electric bicycles passing at the front end of the set road section within a fixed time interval on that day, wherein the set road section includes a set number of lanes and further includes: the edge monitoring device has a built-in camera mechanism, a counting mechanism, and a timing mechanism, which are used to monitor the total number of left-turning vehicles, the total number of straight-moving vehicles, and the total number of electric bicycles passing at the front end of the set road section within a fixed time interval on each day; Among them, the edge monitoring equipment has a built-in camera mechanism, a counting mechanism and a timing mechanism, which are used to monitor the total number of left-turning vehicles, the total number of straight-moving vehicles and the total number of electric bicycles passing at the front end of a set road section within a fixed time interval every day. The counting mechanism is respectively connected to the camera mechanism and the timing mechanism.

9. The intelligent traffic dispatch optimization method based on cloud-edge collaborative computing according to any one of claims 3 to 7, characterized in that: Capturing various section parameters of the set section, wherein the various section parameters of the set section are the section length of the set section, the width of a single lane, the number of surrounding sections, the left turn enable flag of the front section and the left turn enable flag of the rear section, and further comprising: the left turn enable flag of the front section of the set section is used to indicate whether the front section of the set section allows the vehicle to turn left, and the left turn enable flag of the rear section of the set section is used to indicate whether the rear section of the set section allows the vehicle to turn left; The left-turn enabling flag of the road section ahead of the set road section is used to indicate whether the road section ahead of the set road section allows the vehicle to turn left, and the left-turn enabling flag of the road section behind the set road section is used to indicate whether the road section behind the set road section allows the vehicle to turn left, including: when the left-turn enabling flag of the road section ahead of the set road section is 0B0001, it indicates that the road section ahead of the set road section allows the vehicle to turn left, and when the left-turn enabling flag of the road section ahead of the set road section is 0B0000, it indicates that the road section ahead of the set road section prohibits the vehicle from turning left; Among them, the left turn enable flag of the front section of the set road section is used to indicate whether the vehicle is allowed to turn left on the front section of the set road section, and the left turn enable flag of the rear section of the set road section is used to indicate whether the vehicle is allowed to turn left on the rear section of the set road section, including: when the left turn enable flag of the rear section of the set road section is 0B0011, it indicates that the vehicle is allowed to turn left on the rear section of the set road section; when the left turn enable flag of the rear section of the set road section is 0B0010, it indicates that the vehicle is prohibited from turning left on the rear section of the set road section.

10. The intelligent traffic dispatch optimization method based on cloud-edge collaborative computing according to any one of claims 3 to 7, characterized in that: Performing multiple training operations on the BP neural network to obtain the BP neural network after the multiple training operations, and outputting the BP neural network after the multiple training operations as the intelligent traffic prediction model, wherein the number of training operations performed by the BP neural network is proportional to the time length of the fixed time interval, further comprising: using a content mapping formula to represent a content mapping relationship in which the number of training operations performed by the BP neural network is proportional to the time length of the fixed time interval; Among them, performing multiple training operations on the BP neural network to obtain the BP neural network after performing the multiple training operations, and outputting the BP neural network after performing the multiple training operations as a smart traffic prediction model, the number of training operations performed by the BP neural network is proportional to the time length of the fixed time interval, and also includes: in the content mapping formula, the number of training operations performed by the BP neural network is output data, and the time length of the fixed time interval is input data.

Images