Intelligent traffic light scheduling method and system combined with edge calculation

Abstract

The invention provides an intelligent traffic light scheduling method and system combined with edge calculation, and relates to the field of traffic signal optimization scheduling, and the method comprises the steps: monitoring and obtaining traffic correlation data of a target region in a historical time zone; traffic light scheduling optimization analysis of a target area in a preset time zone is carried out, and a first optimization scheduling strategy is generated and comprises a plurality of control schemes of a plurality of traffic lights in the target area; and mapping and transmitting the plurality of control schemes to edge processing units of a plurality of traffic lights, and dynamically scheduling the plurality of traffic lights. The method aims to solve the problems of traffic jam, resource waste and the like caused by the fact that a traditional method is often based on a fixed time period and cannot be dynamically adjusted according to factors such as real-time traffic flow, road conditions and weather. The traffic light scheduling strategy is generated and optimized by combining cloud and edge calculation, dynamic adjustment can be performed according to the real-time flow, road condition and weather change of each intersection, and the operation efficiency of the whole traffic system is optimized.

Description

Technical Field

[0001] The present invention relates to the field of traffic signal optimization scheduling, and particularly to an intelligent traffic light scheduling method and system combined with edge computing. Background Art

[0002] Traditional traffic light scheduling systems have played an important role in urban traffic management, but their inherent limitations have also led to many problems. Most existing traffic light scheduling systems work based on fixed-time strategies, that is, preset the traffic light cycles and signal durations for different time periods (such as morning rush hour, evening rush hour, night, etc.). This scheduling method based on preset time periods does not take into account real-time factors such as actual traffic flow, road conditions, and weather. Therefore, in a complex traffic environment, it is unable to flexibly respond to changes in different situations.

[0003] During the traffic peak period, the traditional system may have problems such as too long or too short signal light cycles, resulting in traffic congestion or resource waste. For example, on a road section with a large traffic flow, a too short green light time may cause vehicles not to fully pass through, increasing the queuing waiting time; while on a road section with a small traffic flow, a too long green light time will waste time and energy, leading to the inefficient operation of the entire traffic system. In addition, climate change and weather factors (such as rain, snow, strong wind, etc.) also affect traffic flow and driving behavior, but the traditional scheduling system fails to consider these dynamic factors. Summary of the Invention

[0004] The purpose of the present invention is to provide an intelligent traffic light scheduling method and system combined with edge computing to solve the problems that traditional traffic light scheduling strategies are often based on fixed time periods and cannot be dynamically adjusted according to real-time traffic flow, road conditions, weather and other factors, resulting in traffic congestion, resource waste, etc., including:

[0005] In a first aspect, the present invention provides an intelligent traffic light scheduling method combined with edge computing, including: monitoring and obtaining traffic-related data in the target area within the historical time zone, and communicating and transmitting the traffic-related data to the cloud center platform; in the cloud center platform, performing optimization analysis on the traffic light scheduling in the target area within the preset time zone to generate a first optimized scheduling strategy, where the first optimized scheduling strategy includes a number of control schemes for a number of traffic lights in the target area; mapping and transmitting the number of control schemes to the edge processing units of the number of traffic lights to perform dynamic scheduling on the number of traffic lights within the preset time zone.

[0006] Preferably, the intelligent traffic light scheduling method combined with edge computing also includes: monitoring and obtaining a number of traffic flows and a number of road condition information of a number of streets in the target area within the historical time zone; monitoring and obtaining weather data of the target area within the historical time zone, and using the several traffic flows, several road condition information and weather data as traffic-related data.

[0007] Preferably, the intelligent traffic light scheduling method combined with edge computing also includes: obtaining several control parameter thresholds of several traffic lights in the target area, wherein the control parameter thresholds include a red light duration threshold and a green light duration threshold; randomly generating a first scheduling strategy based on the several red light duration thresholds and the several green light duration thresholds; predicting the red light waiting time of regional vehicles based on the several traffic flows, several road conditions information, weather data and the first scheduling strategy, and outputting the first waiting time; continuing to randomly select scheduling strategies and iteratively predict waiting times until a predetermined number of iterations is reached, the optimization converges, and outputs the scheduling strategy corresponding to the shortest waiting time as the first optimized scheduling strategy.

[0008] Preferably, the intelligent traffic light scheduling method combined with edge computing also includes: collecting multiple sample traffic flow sets, multiple sample road condition information sets, sample weather data sets and sample scheduling strategy sets, and counting the regional vehicle red light waiting time under different sample traffic flow, sample road condition information, sample weather data sets and sample scheduling strategies to obtain a sample waiting time set; using the multiple sample traffic flow sets, multiple sample road condition information sets, sample weather data sets and sample scheduling strategy sets as input, using the sample waiting time set as supervision, training the BP neural network until convergence, and obtaining a waiting time predictor; inputting the several traffic flow, several road condition information, weather data and the first scheduling strategy into the waiting time predictor, predicting the regional vehicle red light waiting time, and outputting the first waiting time.

[0009] Preferably, the intelligent traffic light scheduling method combined with edge computing also includes: mapping and transmitting the several control schemes to the edge processing units of several traffic lights; within the preset time zone, obtaining several real-time road traffic sets of the roads associated with the several traffic lights through fixed-point monitoring; and dynamically scheduling several traffic lights within the preset time zone based on the several control schemes and the several real-time road traffic sets.

[0010] Preferably, the intelligent traffic light scheduling method combined with edge computing further includes: randomly selecting any one of the several traffic lights as the first traffic light, and obtaining the first control scheme and the first real-time road traffic flow set of the first traffic light; performing real-time optimization and correction on the first control scheme according to the first real-time road traffic flow set, and outputting a first optimized control scheme; sequentially analyzing to obtain several optimized control schemes, and performing dynamic scheduling on the several traffic lights within a preset time period according to the several optimized control schemes.

[0011] Preferably, the intelligent traffic light scheduling method combined with edge computing further includes: inputting the first real-time road traffic flow set into a traffic flow-correction coefficient comparison table, and outputting a first correction coefficient, wherein the traffic flow-correction coefficient comparison table is mapped and constructed based on the sample traffic flow set and the sample correction coefficient set of the first traffic light, and the correction coefficient is the correction weight of the red light duration; performing real-time optimization and correction on the first control scheme according to the first correction coefficient, and outputting a first optimized control scheme.

[0012] In a second aspect, the present invention further provides an intelligent traffic light scheduling system combined with edge computing for executing the intelligent traffic light scheduling method combined with edge computing as described in the first aspect, including: a traffic-related data acquisition module for monitoring and acquiring traffic-related data in a target area within a historical time period, and communicating and transmitting the traffic-related data to a cloud center platform; a traffic light scheduling optimization analysis module for performing traffic light scheduling optimization analysis on the target area within a preset time period on the cloud center platform to generate a first optimized scheduling strategy, wherein the first optimized scheduling strategy includes several control schemes for several traffic lights in the target area; a traffic light dynamic regulation module for mapping and transmitting the several control schemes to the edge processing units of the several traffic lights to perform dynamic scheduling on the several traffic lights within a preset time period.

[0013] The embodiments of the present invention have the following advantages: by monitoring and acquiring traffic-related data in a target area within a historical time period, and communicating and transmitting the traffic-related data to a cloud center platform; then performing traffic light scheduling optimization analysis on the target area within a preset time period on the cloud center platform to generate a first optimized scheduling strategy, wherein the first optimized scheduling strategy includes several control schemes for several traffic lights in the target area; and then mapping and transmitting the several control schemes to the edge processing units of the several traffic lights to perform dynamic scheduling on the several traffic lights within a preset time period; that is to say, by combining cloud and edge computing, generating and optimizing traffic light scheduling strategies, ensuring that the strategies are not only optimized based on global traffic flow, but also can be dynamically adjusted according to the real-time traffic flow, road conditions and weather changes at each intersection, and optimizing the operation efficiency of the entire traffic system. Description of the Drawings

[0014] Figure 1 It is the flowchart of the steps of an intelligent traffic light scheduling method combining edge computing according to the present invention;

[0015] Figure 2 It is the structural schematic diagram of an intelligent traffic light scheduling system combining edge computing according to the present invention.

[0016] Explanation of the reference numerals:

[0017] Traffic-related data acquisition module 10, traffic light scheduling optimization analysis module 20, traffic light dynamic regulation module 30. Specific implementation manner

[0018] By providing an intelligent traffic light scheduling method and system combining edge computing, the present invention solves the problems that traditional traffic light scheduling strategies are often based on fixed time periods and cannot be dynamically adjusted according to real-time traffic flow, road conditions, weather and other factors, resulting in traffic congestion, resource waste, etc. By combining cloud and edge computing, traffic light scheduling strategies are generated and optimized to ensure that the strategies are not only optimized based on the global traffic flow, but also can be dynamically adjusted according to the real-time traffic flow, road conditions and weather changes at each intersection, optimizing the operation efficiency of the entire traffic system.

[0019] Next, the technical solutions in the present invention will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments of the present invention. It should be understood that the present invention is not limited by the example embodiments described herein. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the protection scope of the present invention. In addition, it should be noted that for the sake of description, only the parts related to the present invention are shown in the accompanying drawings rather than all of them.

[0020] Embodiment 1. Please refer to the attached Figure 1 The present invention provides an intelligent traffic light scheduling method combining edge computing, which is applied to an intelligent traffic light scheduling system combining edge computing, and specifically includes the following steps:

[0021] Step 101: Monitor and acquire traffic-related data in the target area within the historical time zone, and communicate and transmit the traffic-related data to the cloud center platform; Step 102: In the cloud center platform, perform traffic light scheduling optimization analysis of the target area within the preset time zone to generate a first optimized scheduling strategy, where the first optimized scheduling strategy includes a number of control schemes for a number of traffic lights in the target area; Step 103: Map and transmit the number of control schemes to the edge processing units of the number of traffic lights to perform dynamic scheduling of the number of traffic lights within the preset time zone.

[0022] Specifically, first, various monitoring devices deployed in the target area (such as traffic sensors, cameras, road flow meters, etc.) are used to collect data related to traffic flow, road conditions, weather, etc. These data include traffic flow (recording traffic volume, vehicle speed, etc. through sensors or cameras), road condition information (such as factors affecting road conditions like traffic congestion, accidents, construction, etc.), and weather data (such as meteorological conditions like rain, snow, haze, etc. that may affect traffic flow and driving behavior); the collected data is transmitted to the cloud center platform in real time as input data for traffic light scheduling optimization analysis.

[0023] Then, at the cloud center platform, these traffic-related data in historical time zones (such as historical traffic flow, weather, road conditions, etc.) are used for traffic light scheduling optimization analysis. Through big data analysis and intelligent algorithms in the cloud, according to factors such as traffic flow, road conditions, and weather, the cloud will generate several control schemes for traffic lights. Each scheme includes control parameters such as signal light cycles, durations, and timings of each traffic light in the time zone, generating a preliminary traffic light scheduling strategy, obtaining the first optimized scheduling strategy, where the first optimized scheduling strategy includes several control schemes for several traffic lights in the target area.

[0024] The optimized traffic light scheduling control scheme will be transmitted to the edge computing unit where each traffic light is located. The edge computing unit is usually located at the intersection site and is a node of distributed computing, capable of processing the scheduling scheme from the cloud nearby. After each intersection's edge processing unit receives the control scheme transmitted from the cloud, it fine-tunes or dynamically adjusts the traffic light control scheme according to real-time traffic flow data (such as information on traffic volume and vehicle speed collected in real time through sensors). For example, when the traffic volume at a certain intersection suddenly increases, the edge computing unit can extend the green light time and reduce the red light waiting time.

[0025] The whole process realizes a distributed traffic light scheduling system. The cloud is responsible for global optimization analysis and formulating the overall scheduling strategy, while the edge computing unit is responsible for local real-time scheduling adjustment at specific intersections; in this way, while ensuring global optimization, it can quickly respond to changes in traffic flow, optimize traffic light control, improve traffic efficiency, and reduce congestion and resource waste.

[0026] Furthermore, step 101 of the present invention further includes:

[0027] Monitoring and obtaining the traffic flow and road condition information of several streets in the target area in historical time zones; monitoring and obtaining the weather data of the target area in historical time zones, and using the several traffic flows, several road condition information, and weather data as traffic-related data.

[0028] Specifically, several traffic flows and several road condition information of several streets in the target area within the historical time zone are monitored and obtained. Traffic flow data refers to the vehicle flow information of each street within a certain historical period. These data are usually collected through traffic monitoring devices (such as geomagnetic sensors, cameras, radars, etc.). The flow data includes information such as the number of vehicles passing through a certain intersection, vehicle speed, and lane occupancy rate. These data can reflect the traffic density and flow state of each street at different time periods. For example, during the morning rush hour, there may be a large vehicle flow on some streets, while during non-peak hours, the vehicle flow is less. Road condition information refers to the data reflecting the road traffic conditions, mainly including: traffic congestion (such as whether certain sections are congested at certain times), traffic accidents (certain accidents or obstacles may affect the normal traffic capacity of the section), and road construction (road construction areas usually affect traffic fluency). These information are crucial for judging the traffic capacity of each street. Usually, road condition information is collected through traffic cameras, road surface sensors or manual reports, etc.

[0029] Monitor and obtain the weather data in the target area within the historical time zone. Weather data refers to the meteorological factors that affect traffic flow and road conditions. For example, precipitation conditions (such as rain, snow, frost, etc.) will make the road slippery and the visibility poor, thus affecting the driving speed and vehicle flow of drivers. The data collected from multiple sources (traffic flow, road conditions, and weather) together constitute traffic-related data. This data combination can comprehensively reflect the traffic conditions in the target area and the impact of external factors on traffic flow.

[0030] Furthermore, step 102 of the present invention further includes:

[0031] Obtain several control parameter thresholds of several traffic lights in the target area, where the control parameter thresholds include red light duration thresholds and green light duration thresholds; randomly generate a first scheduling strategy based on the several red light duration thresholds and several green light duration thresholds; predict the red light waiting duration of regional vehicles according to the several traffic flows, several road condition information, weather data and the first scheduling strategy, and output the first waiting duration; continue to randomly select scheduling strategies and waiting durations for iterative prediction until the predetermined number of iterations is reached, then optimize and converge, and output the scheduling strategy corresponding to the shortest waiting duration as the first optimized scheduling strategy.

[0032] Specifically, obtain several control parameter thresholds of several traffic lights within the target area. Among them, the control parameter thresholds include the red light duration threshold and the green light duration threshold. The red light duration threshold refers to the maximum duration of the specified red light signal. There is an upper limit for the red light duration at each intersection to ensure that traffic in each direction has sufficient time to pass. The green light duration threshold refers to the maximum duration of the specified green light signal. This parameter controls the time length during which vehicles in each direction can pass. Obtaining these control parameter thresholds is to set a reasonable time range when generating the scheduling strategy later, to avoid the red light control time being too long or too short.

[0033] Next, based on multiple red light duration thresholds and green light duration thresholds, randomly select any red light duration and any green light duration within the red light duration thresholds and green light duration thresholds as the first control scheme of the first traffic light, and obtain multiple first control schemes of several traffic lights, and construct the first scheduling strategy. This scheduling strategy includes the duration and control cycle of each traffic light. Then on the cloud platform, the system will predict the waiting duration of vehicles in the red light period within the area based on traffic flow, road conditions information, weather data, and the first scheduling strategy. The prediction of the waiting duration is very important because it directly affects traffic fluency and the average queuing time of vehicles. Through the comprehensive analysis of these data, predict the first waiting duration, that is, the average waiting time of vehicles in the red light period based on the first scheduling strategy.

[0034] Further, through randomly selecting new scheduling strategies and predicting new waiting durations, perform iterative prediction until the preset number of iterations is reached. In each iteration, the scheduling strategy will be fine-tuned, and the duration of the traffic light will be adjusted according to the result of the previous iteration. The purpose is to reduce the waiting duration. Each adjustment will be based on the predicted waiting duration data to ensure that the scheduling strategy can more effectively reduce the waiting time of vehicles at red lights. After several iterations, the scheduling strategy will be gradually optimized. Finally, the system will select the scheduling strategy corresponding to the shortest waiting duration as the final scheduling scheme, that is, the first optimized scheduling strategy.

[0035] This process realizes a method for generating a traffic light scheduling strategy based on random generation and iterative optimization. By introducing factors such as traffic flow, road conditions, and weather, the system can continuously optimize the control duration and signal cycle of traffic lights, and finally output an optimal traffic light scheduling strategy, thereby minimizing the waiting duration of vehicles and improving the overall traffic efficiency.

[0036] Furthermore, the present invention further includes the following steps:

[0037] Collect multiple sample traffic flow sets, multiple sample road condition information sets, sample weather data sets, and sample scheduling strategy sets, and count the red light waiting times of regional vehicles under different sample traffic flows, sample road condition information, sample weather data, and sample scheduling strategies to obtain a sample waiting time set; use the multiple sample traffic flow sets, multiple sample road condition information sets, sample weather data sets, and sample scheduling strategy sets as inputs, and use the sample waiting time set as supervision to train a BP neural network until convergence to obtain a waiting time predictor; input the several traffic flows, several road condition information, weather data, and the first scheduling strategy into the waiting time predictor to predict the red light waiting time of regional vehicles and output a first waiting time.

[0038] Specifically, first, collect multiple sample traffic flow sets, multiple sample road condition information sets, sample weather data sets, and sample scheduling strategy sets, and count the red light waiting times of regional vehicles under different sample traffic flows, sample road condition information, sample weather data, and sample scheduling strategies to obtain a sample waiting time set. Then, based on the BP neural network, construct a waiting time predictor, and use the multiple sample traffic flow sets, multiple sample road condition information sets, sample weather data sets, and sample scheduling strategy sets as inputs, and use the sample waiting time set as supervision to perform supervised training on the waiting time predictor. First, pass the input data through the neural network and calculate the output of the network; then, compare the output of the neural network with the actual sample waiting time and calculate the error; then, adjust the weights and biases of the network according to the error to minimize the output error; finally, through multiple iterative trainings until the network converges, that is, the error drops to an acceptable level. Through training, the BP neural network will learn the relationship between traffic flow, road conditions, weather, and scheduling strategy and waiting time. Finally, this network model (i.e., the waiting time predictor) can predict new input data.

[0039] Finally, input the several traffic flows, several road condition information, weather data, and the first scheduling strategy into the waiting time predictor to predict the red light waiting time of regional vehicles and output a first waiting time. By training the BP neural network, the system can establish a relationship model between traffic light scheduling and waiting time based on traffic flow, road conditions, weather, and scheduling strategy in historical sample data. After training, the system can predict the waiting time of vehicles at red lights under specific conditions. This prediction result can help the system continuously optimize the traffic light scheduling strategy, reduce waiting time, and improve traffic fluency.

[0040] Furthermore, step 103 of the present invention further includes:

[0041] Map and transmit the several control schemes to the edge processing units of several traffic lights; within the preset time period, obtain several real-time road traffic flow sets of the roads associated with the several traffic lights through fixed-point monitoring; according to the several control schemes and the several real-time road traffic flow sets, perform dynamic scheduling of the several traffic lights within the preset time period.

[0042] Specifically, the optimal or preliminary traffic light control strategies (including the durations of red lights and green lights, cycles, switching methods, etc.) generated in the cloud need to be transmitted to the edge processing units at actual intersections. The edge processing units are local devices distributed at each intersection and have the ability to process data, enabling calculations and controls to be performed nearby. Edge processing units usually have real-time control over parameters such as traffic flow and signal light durations. Then, within the preset time period, obtain several real-time road traffic flow sets of the roads associated with the several traffic lights through fixed-point monitoring. The edge processing unit of each traffic light will regularly monitor the real-time traffic flow data of the road it is associated with. These data are generally collected through traffic sensors, video surveillance, geomagnetic sensors, radar, or other devices. These data sets may include information such as the number of vehicles, vehicle speeds, and traffic densities at each intersection or street during the current period. Then, by combining the control scheme (the optimized traffic light scheduling strategy in the cloud) and the real-time road traffic flow data (such as an increase or decrease in the traffic volume at a certain intersection), the edge processing unit will dynamically adjust each traffic light signal. Its main goal is to optimize the traffic light switching timing in real time and reduce traffic congestion.

[0043] By transmitting the optimized traffic light control strategy to the edge processing unit at the intersection and combining the real-time traffic flow data, the durations and cycles of the traffic light signals are adjusted in real time, thereby reducing traffic congestion and improving road traffic efficiency; in this way, the system can quickly respond to changes in traffic conditions and achieve more flexible and intelligent traffic signal control.

[0044] Furthermore, the present invention further includes the following steps:

[0045] Randomly select any one of the several traffic lights as the first traffic light, and obtain the first control scheme and the first real-time road traffic flow set of the first traffic light; perform real-time optimization and correction on the first control scheme according to the first real-time road traffic flow set, and output the first optimized control scheme; sequentially analyze and obtain several optimized control schemes, and perform dynamic scheduling of the several traffic lights within the preset time period according to the several optimized control schemes.

[0046] Specifically, first, randomly select any one of the several traffic lights as the first traffic light, and obtain the first control plan and the first real-time road traffic flow set of the first traffic light; then, perform real-time optimization and correction on the first control plan according to the first real-time road traffic flow set, that is, fine-tune the duration of the traffic light according to the current traffic conditions to maximize the traffic flow. For example, if the traffic flow in a certain direction is large, the system may extend the green light duration in that direction; if the traffic flow in a certain direction is small, the system may reduce the green light duration or quickly switch to other directions, and output the first optimized control plan. Then, analyze the control plans of all other traffic lights one by one, and use the same optimization algorithm to optimize each traffic light according to the real-time traffic flow and road conditions. Analyze them in turn to obtain several optimized control plans. Each traffic light will generate an optimized control plan, and the optimized plans at different intersections may be different, and the signal cycle is adjusted according to the respective traffic flow, road conditions and environmental conditions. Finally, perform dynamic scheduling on the several traffic lights within a preset time period according to the several optimized control plans.

[0047] Further, the present invention further includes the following steps:

[0048] Input the first real-time road traffic flow set into the traffic flow - correction coefficient comparison table, and output the first correction coefficient, where the traffic flow - correction coefficient comparison table is mapped and constructed based on the sample traffic flow set and the sample correction coefficient set of the first traffic light, and the correction coefficient is the correction weight of the red light duration; perform real-time optimization and correction on the first control plan according to the first correction coefficient, and output the first optimized control plan.

[0049] Specifically, first, obtain the real-time road traffic flow data of the section where the first traffic light is located. These data may include traffic flow, vehicle speed, traffic density, etc., reflecting the current traffic conditions of the section. Then, input the first real-time road traffic flow set into the traffic flow - correction coefficient comparison table. This table is a pre-constructed mapping relationship table that associates different traffic flow data (such as traffic flow, vehicle speed, etc.) with the corresponding correction coefficients. Each traffic flow condition corresponds to a correction coefficient, which represents the correction weight of the red light duration. The correction coefficient reflects the impact of traffic flow changes on traffic light scheduling. When the traffic flow increases, the system may need to increase the red light duration to reduce congestion. The larger the correction coefficient, the greater the adjustment range of the red light duration. The purpose of the correction coefficient is to dynamically adjust the red light duration according to the traffic flow changes to optimize traffic control. By inputting the real-time road traffic flow set, find the first correction coefficient that matches the current traffic flow from the traffic flow - correction coefficient comparison table. This coefficient determines the adjustment range of the red light duration at the current time period.

[0050] Then, the first control scheme is optimized and corrected in real time according to the first correction coefficient. If the traffic flow is large, the system will extend the duration of the red light to avoid vehicle congestion; if the traffic flow is small, the duration of the red light may be reduced to shorten the waiting time, and the first optimized control scheme is output.

[0051] In summary, the intelligent traffic light scheduling method combining edge computing provided by the present invention has the following technical effects:

[0052] By monitoring and obtaining traffic correlation data in the target area within the historical time zone, and communicating and transmitting the traffic correlation data to the cloud center platform; then, in the cloud center platform, traffic light scheduling optimization analysis of the target area within the preset time zone is performed to generate a first optimized scheduling strategy, where the first optimized scheduling strategy includes several control schemes for several traffic lights in the target area; then, the several control schemes are mapped and transmitted to the edge processing units of several traffic lights to perform dynamic scheduling of several traffic lights within the preset time zone; that is to say, by combining cloud and edge computing, a traffic light scheduling strategy is generated and optimized to ensure that the strategy is not only optimized based on the global traffic flow, but also can be dynamically adjusted according to the real-time traffic flow, road conditions and weather changes at each intersection, optimizing the operation efficiency of the entire traffic system.

[0053] Embodiment 2, based on the same inventive concept as the intelligent traffic light scheduling method combining edge computing in the foregoing embodiment, the present invention also provides an intelligent traffic light scheduling system combining edge computing. Please refer to the appendix Figure 2 , including: a traffic correlation data acquisition module 10, configured to monitor and obtain traffic correlation data in the target area within the historical time zone, and communicate and transmit the traffic correlation data to the cloud center platform; a traffic light scheduling optimization analysis module 20, configured to perform traffic light scheduling optimization analysis of the target area within the preset time zone in the cloud center platform to generate a first optimized scheduling strategy, where the first optimized scheduling strategy includes several control schemes for several traffic lights in the target area; a traffic light dynamic regulation module 30, configured to map and transmit the several control schemes to the edge processing units of several traffic lights to perform dynamic scheduling of several traffic lights within the preset time zone.

[0054] Further, the intelligent traffic light scheduling system combining edge computing is further configured to: monitor and obtain the traffic flows and road condition information of several streets in the target area within the historical time zone; monitor and obtain weather data in the target area within the historical time zone, and use the several traffic flows, road condition information and weather data as traffic correlation data.

[0055] Furthermore, the intelligent traffic light scheduling system combined with edge computing is also used for: obtaining a number of control parameter thresholds of a number of traffic lights in a target area, where the control parameter thresholds include a red light duration threshold and a green light duration threshold; randomly generating a first scheduling strategy based on the number of red light duration thresholds and the number of green light duration thresholds; predicting the red light waiting duration of regional vehicles according to the number of traffic flows, the number of road conditions information, weather data, and the first scheduling strategy, and outputting a first waiting duration; continuing to randomly select a scheduling strategy and waiting duration for iterative prediction until a predetermined number of iterations is reached, then optimizing and converging, and outputting the scheduling strategy corresponding to the shortest waiting duration as the first optimized scheduling strategy.

[0056] Furthermore, the intelligent traffic light scheduling system combined with edge computing is also used for: collecting a number of sample traffic flow sets, a number of sample road condition information sets, a sample weather data set, and a sample scheduling strategy set, and statistically obtaining the red light waiting duration of regional vehicles under different sample traffic flows, sample road condition information, sample weather data, and sample scheduling strategies to obtain a sample waiting duration set; using the number of sample traffic flow sets, the number of sample road condition information sets, the sample weather data set, and the sample scheduling strategy set as inputs, and using the sample waiting duration set as supervision to train a BP neural network until convergence to obtain a waiting duration predictor; inputting the number of traffic flows, the number of road condition information, weather data, and the first scheduling strategy into the waiting duration predictor to predict the red light waiting duration of regional vehicles and output a first waiting duration.

[0057] Furthermore, the intelligent traffic light scheduling system combined with edge computing is also used for: mapping and transmitting the number of control schemes to the edge processing units of the number of traffic lights; within the preset time zone, fixedly monitoring and obtaining a number of real-time road flow sets of the roads associated with the number of traffic lights; dynamically scheduling the number of traffic lights within the preset time zone according to the number of control schemes and the number of real-time road flow sets.

[0058] Furthermore, the intelligent traffic light scheduling system combined with edge computing is also used for: randomly selecting any one of the number of traffic lights as the first traffic light, and obtaining a first control scheme and a first real-time road flow set of the first traffic light; performing real-time optimization and correction on the first control scheme according to the first real-time road flow set, and outputting a first optimized control scheme; sequentially analyzing to obtain a number of optimized control schemes, and dynamically scheduling the number of traffic lights within the preset time zone according to the number of optimized control schemes.

[0059] Further, the intelligent traffic light scheduling system combined with edge computing is further configured to: input the first real-time road traffic flow set into a traffic flow - correction coefficient comparison table, and output a first correction coefficient, where the traffic flow - correction coefficient comparison table is mapped and constructed based on the sample traffic flow set and the sample correction coefficient set of the first traffic light, and the correction coefficient is the correction weight of the red light duration; perform real-time optimization and correction on the first control scheme according to the first correction coefficient, and output a first optimized control scheme.

[0060] The various embodiments in this specification are described in a progressive manner. The key point of each embodiment is the difference from other embodiments. The intelligent traffic light scheduling method and specific examples in the foregoing Embodiment 1 are equally applicable to the intelligent traffic light scheduling system in this embodiment. Through the foregoing detailed description of the intelligent traffic light scheduling method combined with edge computing, those skilled in the art can clearly know the intelligent traffic light scheduling system combined with edge computing in this embodiment. Therefore, for the sake of brevity of the specification, it will not be described in detail here. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple. For related parts, please refer to the description of the method part.

[0061] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present invention. Various modifications to these embodiments will be obvious to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but will be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0062] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the present invention and its equivalent technologies, the present invention is also intended to include these changes and variations.

Claims

1. An intelligent traffic light scheduling method combined with edge computing, characterized in that the method include: Monitor and obtain traffic-related data of the target area in the historical time zone, and transmit the traffic-related data to the cloud center platform; On the cloud center platform, a traffic light scheduling optimization analysis is performed in a target area within a preset time zone to generate a first optimization scheduling strategy, wherein the first optimization scheduling strategy includes a plurality of control schemes for a plurality of traffic lights in the target area; The control schemes are mapped and transmitted to edge processing units of a plurality of traffic lights, and the plurality of traffic lights are dynamically scheduled within a preset time zone.

2. The intelligent traffic light scheduling method combined with edge computing according to claim 1, wherein Monitor and obtain traffic-related data of the target area in the historical time zone, including: Monitor and obtain a number of traffic flows and a number of road condition information of a number of streets in a target area within a historical time zone; The weather data of the target area in the historical time zone is monitored and acquired, and the plurality of traffic flows, the plurality of road condition information and the weather data are used as traffic-related data.

3. The intelligent traffic light scheduling method combined with edge computing according to claim 2, wherein On the cloud center platform, traffic light scheduling optimization analysis is performed in the target area within the preset time zone to generate the first optimization scheduling strategy, including: Acquire a plurality of control parameter thresholds of a plurality of traffic lights in a target area, wherein the control parameter thresholds include a red light duration threshold and a green light duration threshold; Randomly generating a first scheduling strategy based on a number of red light duration thresholds and a number of green light duration thresholds; Predicting the waiting time of regional vehicles at red lights according to the plurality of traffic flows, the plurality of road condition information, the weather data and the first dispatching strategy, and outputting a first waiting time; Continue to randomly select scheduling strategies and wait time iterative predictions until the predetermined number of iterations is reached, then the optimization converges and outputs the scheduling strategy corresponding to the shortest wait time as the first optimized scheduling strategy.

4. The intelligent traffic light scheduling method combining edge computing according to claim 3, wherein Predicting the waiting time of regional vehicles at red lights according to the plurality of traffic flows, the plurality of road condition information, the weather data and the first dispatching strategy, and outputting the first waiting time, including: Collect multiple sample traffic flow sets, multiple sample road condition information sets, sample weather data sets and sample scheduling strategy sets, and count the red light waiting time of regional vehicles under different sample traffic flow, sample road condition information, sample weather data and sample scheduling strategies to obtain a sample waiting time set; Using the multiple sample traffic flow sets, multiple sample road condition information sets, sample weather data sets and sample scheduling strategy sets as inputs, using the sample waiting time sets as supervision, training the BP neural network until convergence, and obtaining a waiting time predictor; The plurality of traffic flows, the plurality of road condition information, the weather data and the first dispatching strategy are input into the waiting time predictor to predict the waiting time of regional vehicles at red lights and output a first waiting time.

5. The intelligent traffic light scheduling method combined with edge computing according to claim 1, characterized in that Transmitting the plurality of control schemes to edge processing units of a plurality of traffic lights, and dynamically scheduling the plurality of traffic lights within a preset time zone, including: Transmitting the plurality of control scheme mappings to the edge processing units of the plurality of traffic lights; In the preset time zone, fixed-point monitoring is performed to obtain a plurality of real-time road flow sets of roads associated with a plurality of traffic lights; According to the plurality of control schemes and the plurality of real-time road flow sets, a plurality of traffic lights are dynamically scheduled within a preset time zone.

6. The intelligent traffic light scheduling method combined with edge computing according to claim 5, wherein Performing dynamic scheduling within a preset time zone for a plurality of traffic lights according to the plurality of control schemes and the plurality of real-time road traffic flow sets, including: Randomly selecting any one of the plurality of traffic lights as the first traffic light, and obtaining the first control scheme and the first real-time road traffic flow set of the first traffic light; Performing real-time optimization and correction on the first control scheme according to the first real-time road traffic flow set, and outputting a first optimized control scheme; Sequentially analyzing to obtain a plurality of optimized control schemes, and performing dynamic scheduling within a preset time zone for the plurality of traffic lights according to the plurality of optimized control schemes.

7. The intelligent traffic light scheduling method combined with edge computing according to claim 6, characterized in that Performing real-time optimization and correction on the first control scheme according to the first real-time road traffic flow set, and outputting a first optimized control scheme, including: Inputting the first real-time road traffic flow set into a traffic flow - correction coefficient look-up table, and outputting a first correction coefficient, wherein the traffic flow - correction coefficient look-up table is mapped and constructed based on the sample traffic flow set and the sample correction coefficient set of the first traffic light, and the correction coefficient is the correction weight of the red light duration; Performing real-time optimization and correction on the first control scheme according to the first correction coefficient, and outputting a first optimized control scheme.

8. The intelligent traffic light scheduling system combined with edge computing is characterized in that Steps for implementing the intelligent traffic light scheduling method combined with edge computing according to any one of claims 1 to 7, including: A traffic - related data acquisition module, configured to monitor and acquire traffic - related data of a target area within a historical time zone, and communicate and transmit the traffic - related data to a cloud center platform; A traffic light scheduling optimization analysis module, configured to perform traffic light scheduling optimization analysis of a target area within a preset time zone on the cloud center platform, and generate a first optimized scheduling strategy, wherein the first optimized scheduling strategy includes a plurality of control schemes for a plurality of traffic lights within the target area; A traffic light dynamic regulation module, configured to map and transmit the plurality of control schemes to edge processing units of the plurality of traffic lights, and perform dynamic scheduling within a preset time zone for the plurality of traffic lights.

Images