A method and system for analyzing road green wave speed based on AI multi-objective

Abstract

The present invention provides a method and system for analyzing the green wave speed of a road based on AI multi-objectives, and relates to the field of urban safety and resilience traffic technology; the method comprises the following steps: obtaining a preset speed range of a target green wave road, and randomly generating an initial population according to the preset speed range, including multiple signal light timing strategies, each signal light timing strategy is for at least one intersection and covers control parameters of different time periods; after elite retention and selecting a number of first signal light timing strategies based on congestion, an approximate Pareto optimal solution set is obtained, the Pareto optimal solution set includes at least one alternative signal light timing strategy, traversing the Pareto optimal solution set to obtain a target speed range, and using the alternative signal light timing strategy corresponding to the target speed range as the target signal light timing strategy; controlling the signal light timing of each intersection of the target green wave road; the method accurately determines the speed range of the target green wave road and significantly improves traffic efficiency.

Description

Technical Field

[0001] The present invention relates to the field of urban safety and resilience transportation technology, and in particular to a method and system for analyzing road green wave speed based on AI multi-objective. Background Art

[0002] With the continuous development of cities and the continuous increase in the number of cars, traffic congestion is becoming increasingly serious. Traditional traffic light control methods often use fixed timing schemes, which are difficult to adapt to dynamically changing traffic flows, resulting in low road traffic efficiency and long waiting times for vehicles.

[0003] As an effective means of traffic management, the green wave belt aims to improve the traffic efficiency of a road by setting the signal light timing reasonably so that vehicles can pass through multiple intersections continuously without stopping. However, determining the appropriate green wave speed interval is the key to achieving the green wave belt. At present, the method of determining the green wave speed interval mainly relies on experience and simple calculations, which lacks scientificity and accuracy. Summary of the invention

[0004] In view of the above-mentioned defects or deficiencies in the prior art, the present invention aims to provide a method and system for analyzing the green wave speed of roads based on AI multi-objective analysis, so as to improve the scientificity and accuracy of determining the green wave speed interval.

[0005] In a first aspect, the present invention proposes a method for analyzing road green wave speed based on AI multi-objective, comprising the following steps:

[0006] Acquire a preset speed interval of a target green wave road, where the preset speed interval is used to represent a speed interval of vehicles traveling on the target green wave road within a target preset time period; the target green wave road includes a plurality of intersections;

[0007] According to the preset speed interval, an initial population is randomly generated, wherein the initial population includes a plurality of signal light timing strategies, each of which is for at least one intersection and covers control parameters of different time periods;

[0008] Among all the signal light timing strategies, a plurality of first signal light timing strategies are obtained after elite reservation and congestion-based selection;

[0009] Based on a plurality of the first signal light timing strategies, an approximate Pareto optimal solution set is obtained, wherein the Pareto optimal solution set includes at least one alternative signal light timing strategy, and each of the alternative signal light timing strategies has a corresponding alternative speed interval;

[0010] Traversing the Pareto optimal solution set to obtain a target speed interval, and using the alternative traffic light timing strategy corresponding to the target speed interval as the target traffic light timing strategy;

[0011] According to the target signal light timing strategy, the signal lights at each intersection of the target green wave road are controlled within the target preset time period; wherein the target speed range is the alternative speed range that best meets the specific range of urban traffic management targets.

[0012] According to the technical solution provided by the present invention, obtaining a plurality of first signal light timing strategies that have been elite-retained and selected based on congestion from among all the signal light timing strategies comprises at least the following steps:

[0013] Performing non-dominated sorting on all the signal light timing strategies to obtain a plurality of non-dominated signal light timing strategies, and classifying the non-dominated signal light timing strategies into different levels; the levels include at least a first level; the first level is a set of non-dominated signal light timing strategies at an optimal level after performing non-dominated sorting on all the signal light timing strategies;

[0014] All the non-dominated signal light timing strategies at the first level are used as the first signal light timing strategies.

[0015] According to the technical solution provided by the present invention, the hierarchy also includes remaining hierarchies except the first hierarchy;

[0016] After performing non-dominated sorting on all the signal light timing strategies to obtain a plurality of non-dominated signal light timing strategies, and classifying the non-dominated signal light timing strategies into different levels, the following steps are also included:

[0017] calculating the individual congestion degree of each of the non-dominated signal light timing strategies in each of the remaining levels;

[0018] The non-dominated signal light timing strategy for which the individual congestion degree is greater than or equal to a first preset threshold is used as the first signal light timing strategy.

[0019] According to the technical solution provided by the present invention, the non-dominated sorting of all the signal light timing strategies specifically includes the following steps:

[0020] Performing multi-objective evaluation on all the signal light timing strategies respectively to obtain an evaluation result of each signal light timing strategy;

[0021] The evaluation result includes the target score corresponding to each target; the multiple targets include at least the target of reducing the waiting time of vehicles at the intersection, the target of increasing the overall hourly flow of vehicles at the intersection, the target of increasing the average speed of the road, and the target of reducing the road delay time ratio;

[0022] A dominance relationship is set, and based on the evaluation result of each of the traffic light timing strategies, a non-domination sort is performed according to the dominance relationship.

[0023] According to the technical solution provided by the present invention, the multiple objectives constitute the dimensions of the multi-objective space;

[0024] The calculating of the individual congestion degree of each of the non-dominated signal light timing strategies in each of the remaining levels specifically comprises the following steps:

[0025] Obtaining adjacent individuals of the non-dominated signal light timing strategy in each dimension of the multi-objective space;

[0026] Calculating the adjacent distance between the non-dominated signal light timing strategy and the adjacent individuals in each dimension;

[0027] Based on the adjacent distances between the non-dominated signal light timing strategy and the adjacent individuals in all dimensions, the individual congestion degree of the non-dominated signal light timing strategy is obtained.

[0028] According to the technical solution provided by the present invention, the step of obtaining the preset speed range of the target green wave road at least includes the following steps:

[0029] Acquire a historical driving data set of the target green wave road, the historical driving data set at least including multiple collection time periods and driving data corresponding to each collection time period, the driving data at least including vehicle speed data, vehicle flow data, and traffic congestion conditions between each intersection on the target green wave road;

[0030] Marking the collection period corresponding to the target preset period as the target collection period, and obtaining the average speed of each vehicle in the target collection period according to the vehicle speed data between each of the intersections corresponding to the target collection period;

[0031] According to the average speed of each vehicle, a reference speed of the target acquisition period is obtained, and vehicles corresponding to the average speed whose deviation from the reference speed is greater than or equal to a second preset threshold are eliminated to obtain a target vehicle;

[0032] The interval consisting of the maximum value and the minimum value of the average speeds corresponding to all the target vehicles is used as the preset speed interval.

[0033] According to the technical solution provided by the present invention, traversing the Pareto optimal solution set to obtain the target speed range specifically includes the following steps:

[0034] Obtaining the main objectives of urban traffic management within the target preset time period, and obtaining quantitative standards based on the main objectives;

[0035] Calculating an evaluation result of each of the candidate signal light timing strategies, and obtaining a degree of compliance of each of the candidate signal light timing strategies with the quantitative standard according to the evaluation result of each of the candidate signal light timing strategies;

[0036] The alternative speed interval corresponding to the alternative traffic light timing strategy with the highest degree of compliance is used as the target speed interval.

[0037] According to the technical solution provided by the present invention, the acquisition of the main target of urban traffic management within the target preset time period specifically includes the following steps:

[0038] Determine the road type of the target green wave road, the road type includes at least a regular road and a special road, and the special road includes at least a road with a special stop requirement in at least one special time period;

[0039] If the road type of the target green wave road is the conventional road or the special road and the target preset time period is not within the special time period, the universal target is used as the main target;

[0040] If the road type of the target green wave road is the special road and the target preset time period is within the special time period, the safety goal and the efficiency goal are combined with the universal goal to obtain the main goal.

[0041] According to the technical solution provided by the present invention, the control of the timing of the traffic lights at each intersection of the target green wave road within the target preset time period according to the target traffic light timing strategy specifically includes the following steps:

[0042] Acquire the real-time traffic condition of the target green wave road in real time, wherein the real-time traffic condition at least includes whether there is a traffic accident or temporary construction;

[0043] When the real-time road condition does not contain the traffic accident or the temporary construction, or the real-time road condition contains the traffic accident or the temporary construction and the impact on the road traffic stability is less than a third preset threshold, according to the target signal light timing strategy, the timing of the signal lights at each intersection of the target green wave road is controlled within the target preset time period;

[0044] After the real-time traffic condition of the target green wave road is acquired in real time, the following steps are also included:

[0045] When the real-time road condition has the traffic accident or the temporary construction and the impact of the temporary construction on the road traffic stability is greater than or equal to the third preset threshold, calculating an updated speed interval, wherein the updated speed interval is used to reduce vehicle congestion;

[0046] According to the updated timing strategy corresponding to the updated speed interval, the timing of the traffic lights at each intersection of the target green wave road is controlled within the target preset time period.

[0047] In a second aspect, the present invention proposes a system for analyzing road green wave speed based on AI multi-objectives, which is used to implement the method for analyzing road green wave speed based on AI multi-objectives as described above, comprising:

[0048] A collection module, the collection module is configured to obtain a preset speed interval of a target green wave road, the preset speed interval is used to represent a speed interval of vehicles traveling on the target green wave road within a target preset time period; the target green wave road includes a plurality of intersections;

[0049] A calculation module, wherein the calculation module is configured to randomly generate an initial population according to the preset speed interval, wherein the initial population includes a plurality of signal light timing strategies, each of which is for at least one intersection and covers control parameters of different time periods;

[0050] The calculation module is further configured to obtain, from all the signal light timing strategies, a plurality of first signal light timing strategies that have been selected after elite reservation and based on congestion;

[0051] The calculation module is further configured to obtain an approximate Pareto optimal solution set based on a plurality of the first signal light timing strategies, wherein the Pareto optimal solution set includes at least one candidate signal light timing strategy, and each of the candidate signal light timing strategies has a corresponding candidate speed interval;

[0052] The calculation module is further configured to traverse the Pareto optimal solution set to obtain a target speed interval, and use the candidate signal light timing strategy corresponding to the target speed interval as the target signal light timing strategy;

[0053] A control module, wherein the control module is configured to control the timing of the traffic lights at each intersection of the target green wave road within the target preset time period according to the target traffic light timing strategy; wherein the target speed range is the alternative speed range that best meets the specific range of urban traffic management targets.

[0054] Compared with the prior art, the beneficial effect of the present invention is that the present invention can comprehensively consider multiple traffic management goals through the AI-based multi-objective analysis method, and more scientifically and accurately determine the speed range of the target green wave road. Compared with the traditional method, the rationality and effectiveness of the green wave speed range are improved, and the intersection signal lights are controlled according to the determined target signal light timing strategy, which can realize the continuous passage of vehicles on the green wave road, reduce the waiting time of vehicles, increase the overall hourly flow of vehicles at the intersection, increase the average speed of the road, and reduce the road delay time ratio, thereby significantly improving traffic efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 A flowchart of the steps of the method for analyzing road green wave speed based on AI multi-objective provided by the present invention;

[0056] Figure 2 A schematic diagram of a system for analyzing road green wave speed based on AI multi-objectives provided by the present invention. DETAILED DESCRIPTION

[0057] The present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the relevant inventions, rather than to limit the invention. It should also be noted that, for ease of description, only the parts related to the invention are shown in the accompanying drawings.

[0058] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0059] Example 1

[0060] As mentioned in the background technology, in view of the problems in the prior art, the present invention proposes a method for analyzing the road green wave speed based on AI multi-objective. Figure 1 As shown, the following steps are included:

[0061] S1. Obtain a preset speed interval of a target green wave road, where the preset speed interval is used to represent a speed interval of vehicles traveling on the target green wave road within a target preset time period; the target green wave road includes multiple intersections;

[0062] Furthermore, obtaining the preset speed range of the target green wave road at least includes the following steps:

[0063] Acquire a historical driving data set of the target green wave road, the historical driving data set at least including multiple collection time periods and driving data corresponding to each collection time period, the driving data at least including vehicle speed data, vehicle flow data, and traffic congestion conditions between each intersection on the target green wave road;

[0064] Marking the collection period corresponding to the target preset period as the target collection period, and obtaining the average speed of each vehicle in the target collection period according to the vehicle speed data between each of the intersections corresponding to the target collection period;

[0065] According to the average speed of each vehicle, a reference speed of the target acquisition period is obtained, and vehicles corresponding to the average speed whose deviation from the reference speed is greater than or equal to a second preset threshold are eliminated to obtain a target vehicle;

[0066] The interval consisting of the maximum value and the minimum value of the average speeds corresponding to all the target vehicles is used as the preset speed interval.

[0067] S2. randomly generating an initial population according to the preset speed range, wherein the initial population includes a plurality of signal light timing strategies, each of which is for at least one intersection and covers control parameters of different time periods;

[0068] Specifically, the random generation of the initial population and the continuous evolution method can adapt to the dynamic changes in traffic flow. In different time periods and traffic conditions, the system can automatically adjust the signal timing strategy and speed range to improve the adaptability and stability of the traffic system.

[0069] For example, the traffic management department collects a large amount of historical data through sensors and cameras installed on the road and from past traffic studies. These data include the speed of vehicles on the road during different time periods (such as morning and evening peaks, off-peak hours, and weekends on weekdays). After analyzing the historical data, it is found that during the morning peak hours on weekdays, the speed of vehicles on the road is usually 30-40 kilometers per hour; during the evening peak hours, due to the large traffic volume, the speed is slightly lower, at 25-35 kilometers per hour; during the off-peak hours, the speed is relatively fast, at 40-50 kilometers per hour; the speed distribution on weekends is more dispersed, but roughly 35-45 kilometers per hour. Based on these analysis results, the preset speed ranges for different time periods are determined.

[0070] Specifically, the control parameters include at least the green light duration of each signal light and the phase difference; taking the morning rush hour as an example, according to the preset speed range of 30-40 km / h, the system randomly generates multiple signal light timing strategies as the initial population. For example, a signal light timing strategy may be that the green light duration in the east-west direction of intersection A is 40 seconds and the green light duration in the north-south direction is 30 seconds during the period from 8 to 9 in the morning, and the green light duration in the east-west direction of intersection B is 35 seconds and the green light duration in the north-south direction is 35 seconds during this period, etc., covering the control parameters of different intersections and time periods. Another signal light timing strategy may be that the green light duration in the east-west direction of intersection A is 45 seconds and the green light duration in the north-south direction is 25 seconds, and the green light duration in the east-west direction of intersection A is 30 seconds and the green light duration in the north-south direction is 40 seconds, etc. By randomly generating multiple such signal light timing strategies, an initial population is formed, which provides a basis for the subsequent optimization process.

[0071] S3. From all the signal light timing strategies, a plurality of first signal light timing strategies selected through elite reservation and based on congestion are obtained;

[0072] Specifically, through elite retention and congestion-based selection methods, excellent traffic light timing strategies are ensured to be retained and inherited, while the diversity of the population is increased, avoiding the algorithm from converging to the local optimal solution too early, and improving the algorithm's search ability and optimization effect.

[0073] Furthermore, obtaining a plurality of first signal light timing strategies that are selected through elite reservation and based on congestion from among all the signal light timing strategies comprises at least the following steps:

[0074] Performing non-dominated sorting on all the signal light timing strategies to obtain a plurality of non-dominated signal light timing strategies, and classifying the non-dominated signal light timing strategies into different levels; the levels include at least a first level; the first level is a set of non-dominated signal light timing strategies at an optimal level after performing non-dominated sorting on all the signal light timing strategies;

[0075] All the non-dominated signal light timing strategies at the first level are used as the first signal light timing strategies.

[0076] Further, the hierarchy includes remaining hierarchies except the first hierarchy;

[0077] After performing non-dominated sorting on all the signal light timing strategies to obtain a plurality of non-dominated signal light timing strategies, and classifying the non-dominated signal light timing strategies into different levels, the following steps are also included:

[0078] calculating the individual congestion degree of each of the non-dominated signal light timing strategies in each of the remaining levels;

[0079] The non-dominated signal light timing strategy for which the individual congestion degree is greater than or equal to a first preset threshold is used as the first signal light timing strategy.

[0080] Specifically, elite retention means selecting all individuals in the first level to enter the next generation, which means that individuals classified as the first level in the non-dominated sorting have a special status. These individuals perform best on all objectives because they are not dominated by any other individuals. Directly retaining all individuals in the first level to the next generation population ensures that these excellent solutions will not be lost in the evolutionary process. This can speed up the convergence of the algorithm because these elite individuals provide a good starting point for subsequent optimization.

[0081] For non-dominated layers other than the first layer (subsequent non-dominated layers, i.e., the remaining layers), a selection method based on crowding is adopted. The crowding reflects the relative position relationship between an individual and its neighboring individuals in the multi-target space. An individual with high crowding means that it is more different from other individuals in the target space and has higher diversity and exploration value. Having a higher probability of being selected means that when selecting individuals in subsequent non-dominated layers to enter the next generation of population, individuals with high crowding are more likely to be selected. This can increase the diversity of the population while maintaining excellent solutions, and prevent the algorithm from converging to the local optimal solution too early.

[0082] S4. Based on the plurality of the first traffic light timing strategies, an approximate Pareto optimal solution set is obtained, wherein the Pareto optimal solution set includes at least one alternative traffic light timing strategy, and each of the alternative traffic light timing strategies has a corresponding alternative speed interval;

[0083] Specifically, genetic operations are performed on the selected plurality of the first traffic light timing strategies, including crossover and mutation operations to generate new individuals; the above-mentioned sorting, congestion calculation, selection and genetic operation steps are repeated until the algorithm converges to obtain an approximate Pareto optimal solution set.

[0084] S5, traversing the Pareto optimal solution set to obtain a target speed interval, and using the alternative traffic light timing strategy corresponding to the target speed interval as the target traffic light timing strategy;

[0085] S6. According to the target signal light timing strategy, control the timing of the signal lights at each intersection of the target green wave road within the target preset time period; wherein the target speed range is the alternative speed range that best meets the specific range of the urban traffic management target.

[0086] Specifically, this method makes full use of artificial intelligence technology to realize the intelligent control of traffic lights, providing an efficient and scientific means for urban traffic management, which helps to improve the intelligence level of urban traffic management, alleviate traffic congestion and improve the urban traffic environment.

[0087] In a preferred embodiment, the non-dominated sorting of all the signal light timing strategies specifically includes the following steps:

[0088] Performing multi-objective evaluation on all the signal light timing strategies respectively to obtain an evaluation result of each signal light timing strategy;

[0089] The evaluation result includes the target score corresponding to each target; the multiple targets include at least the target of reducing the waiting time of vehicles at the intersection, the target of increasing the overall hourly flow of vehicles at the intersection, the target of increasing the average speed of the road, and the target of reducing the road delay time ratio;

[0090] Specifically, a multi-objective function is used to evaluate each objective separately. Objective function 1 is the waiting time of vehicles at each intersection. The smaller the objective function 1 is, the greater the score of this objective is. Objective function 2 is the overall hourly flow rate at each intersection. The larger the objective function 2 is, the greater the score of this objective is. Objective function 3 is the average road speed target. The larger the objective function 3 is, the greater the score of this objective is. Objective function 4 is the road delay time ratio. The smaller the objective function 4 is, the greater the score of this objective is.

[0091] A dominance relationship is set, and based on the evaluation result of each of the traffic light timing strategies, a non-domination sort is performed according to the dominance relationship.

[0092] Optionally, the dominance relationship is that if traffic light timing strategy C is not worse than traffic light timing strategy D in all objectives and is better than traffic light timing strategy D in at least one objective, then traffic light timing strategy C is said to dominate traffic light timing strategy D. Then, sorting is performed: the first step is to find all individuals that are not dominated by other individuals, and these individuals constitute the first-level non-dominated solution. The second step is to continue to find individuals that are not dominated by other remaining individuals from the remaining individuals to form the second-level non-dominated solution, and so on. For example, suppose there are three individuals: traffic light timing strategy X, traffic light timing strategy Y, and traffic light timing strategy Z. The performance on the four objectives is as follows:

[0093] Traffic light timing strategy X: waiting time is 10 minutes, traffic volume is 1000 vehicles / hour, average vehicle speed is 40km / h, and delay time ratio is 0.3.

[0094] Traffic light timing strategy Y: waiting time is 12 minutes, traffic volume is 1,100 vehicles / hour, average vehicle speed is 45 km / h, and delay time ratio is 0.25.

[0095] Traffic light timing strategy Z: waiting time is 8 minutes, traffic volume is 900 vehicles / hour, average vehicle speed is 35km / h, and delay time ratio is 0.4.

[0096] By comparison, we can find that signal light timing strategy X is not dominated by signal light timing strategy Y and signal light timing strategy Z, signal light timing strategy Y is not dominated by signal light timing strategy Z but is dominated by individual signal light timing strategies, and signal light timing strategy Z is dominated by signal light timing strategy X and signal light timing strategy Y. Therefore, signal light timing strategy X belongs to the first-level non-dominated solution, signal light timing strategy Y belongs to the second-level non-dominated solution, and signal light timing strategy Z belongs to the dominated subsequent level.

[0097] In a preferred embodiment, the multiple objectives constitute the dimensions of a multi-objective space;

[0098] The calculating of the individual congestion degree of each of the non-dominated signal light timing strategies in each of the remaining levels specifically comprises the following steps:

[0099] Obtaining adjacent individuals of the non-dominated signal light timing strategy in each dimension of the multi-objective space;

[0100] Specifically, adjacent individuals usually refer to two individuals that are in the same non-dominated level as the individual in the multi-target space and are closest to the individual in the target dimension.

[0101] Calculating the adjacent distance between the non-dominated signal light timing strategy and the adjacent individuals in each dimension;

[0102] Specifically, the distance calculation method can adopt algorithms such as Euclidean distance and Manhattan distance. For example, for the distance calculation of one of the individuals on a certain target dimension, if Euclidean distance is adopted, assuming that the value of the individual on a certain target dimension is A1, the value of the adjacent individual on the left side on the dimension is B1, and the value of the adjacent individual on the right side on the dimension is C1, then the distance between the individual and the adjacent individual on this target dimension can be calculated as |(A1 - B1) + (A1 - C1)| / 2.

[0103] Based on the adjacent distances between the non-dominated signal light timing strategy and the adjacent individuals in all dimensions, the individual congestion degree of the non-dominated signal light timing strategy is obtained.

[0104] Specifically, by adding the distances between adjacent individuals of an individual in each target dimension, the crowding degree of the individual is obtained.

[0105] In a preferred embodiment, traversing the Pareto optimal solution set to obtain the target speed range specifically includes the following steps:

[0106] Obtaining the main objectives of urban traffic management within the target preset time period, and obtaining quantitative standards based on the main objectives;

[0107] Furthermore, the obtaining of the main target of urban traffic management within the target preset time period specifically includes the following steps:

[0108] Determine the road type of the target green wave road, the road type includes at least a regular road and a special road, and the special road includes at least a road with a special stop requirement in at least one special time period;

[0109] Optionally, the special road may be a road including a school section or a bus stop section. For a school section, the special time period may be the school and after-school time from Monday to Friday, and other time periods are regular time periods. For a bus stop section, the special time period is a time period set according to the stop times of the bus line, and other normal bus travel times are regular time periods.

[0110] If the road type of the target green wave road is the conventional road or the special road and the target preset time period is not within the special time period, the universal target is used as the main target;

[0111] Specifically, if the road is a regular road, or a special road during normal hours, then special needs need to be considered and only ordinary goals need to be met. Ordinary goals can be selected from universal factors such as traffic flow, vehicle speed control, reducing congestion, and improving traffic efficiency.

[0112] If the road type of the target green wave road is the special road and the target preset time period is within the special time period, the safety goal and the efficiency goal are combined with the universal goal to obtain the main goal.

[0113] If it is during a special period of time on a special road, then we need to consider not only the above-mentioned general factors, but also the time required for students to pass safely or the time required for passengers to get on the bus safely.

[0114] Specifically, the evaluation results of the alternative traffic light timing strategies can be obtained through model training.

[0115] Calculating an evaluation result of each of the candidate signal light timing strategies, and obtaining a degree of compliance of each of the candidate signal light timing strategies with the quantitative standard according to the evaluation result of each of the candidate signal light timing strategies;

[0116] The alternative speed interval corresponding to the alternative traffic light timing strategy with the highest degree of compliance is used as the target speed interval.

[0117] In a preferred embodiment, the control of the timing of the traffic lights at each intersection of the target green wave road within the target preset time period according to the target traffic light timing strategy specifically includes the following steps:

[0118] Acquire the real-time traffic condition of the target green wave road in real time, wherein the real-time traffic condition at least includes whether there is a traffic accident or temporary construction;

[0119] Specifically, the situation of temporary construction can be obtained through electronic eyes or government feedback, and whether there are traffic accidents under real-time road conditions can be obtained through cloud sharing of electronic eyes or vehicle navigation systems.

[0120] When the real-time road condition does not contain the traffic accident or the temporary construction, or the real-time road condition contains the traffic accident or the temporary construction and the impact on the road traffic stability is less than a third preset threshold, controlling the timing of the traffic lights at each intersection of the target green wave road within the target preset time period;

[0121] If there are no such emergencies, or the impact of these emergencies is not high and road traffic remains stable, then the traffic lights can still be controlled according to the target traffic light timing strategy calculated previously to ensure that the vehicle speed during that period remains within the target speed range.

[0122] After the real-time traffic condition of the target green wave road is acquired in real time, the following steps are also included:

[0123] When the real-time road condition has the traffic accident or the temporary construction and the impact of the temporary construction on the road traffic stability is greater than or equal to the third preset threshold, calculating an updated speed interval, wherein the updated speed interval is used to reduce vehicle congestion;

[0124] According to the updated timing strategy corresponding to the updated speed interval, the timing of the traffic lights at each intersection of the target green wave road is controlled within the target preset time period.

[0125] If an emergency occurs on the target filtering road and the impact of these emergencies is very high, the road traffic is unstable, and the speed range needs to be updated according to the calculation;

[0126] Optionally, a traffic flow model is used for prediction. By inputting information such as current traffic flow, accidents or construction conditions, and road parameters, the model can predict vehicle congestion and travel time under different speed ranges. By adjusting the speed range parameters, the vehicle congestion output by the model is minimized, thereby determining the updated speed range.

[0127] When the updated speed range is obtained, the traffic light timing is controlled according to the updated timing strategy area corresponding to the updated speed range, so that vehicles currently on the target green wave road can pass more smoothly, thereby improving traffic flow.

[0128] Example 2

[0129] Based on Example 1, this example proposes a system for analyzing road green wave speed based on AI multi-objectives, which is used to implement the method for analyzing road green wave speed based on AI multi-objectives as described above. Figure 2 As shown, including:

[0130] A collection module, the collection module is configured to obtain a preset speed interval of a target green wave road, the preset speed interval is used to represent a speed interval of vehicles traveling on the target green wave road within a target preset time period; the target green wave road includes a plurality of intersections;

[0131] A calculation module, wherein the calculation module is configured to randomly generate an initial population according to the preset speed interval, wherein the initial population includes a plurality of signal light timing strategies, each of which is for at least one intersection and covers control parameters of different time periods;

[0132] The calculation module is further configured to obtain, from all the signal light timing strategies, a plurality of first signal light timing strategies that have been selected after elite reservation and based on congestion;

[0133] The calculation module is further configured to obtain an approximate Pareto optimal solution set based on a plurality of the first signal light timing strategies, wherein the Pareto optimal solution set includes at least one candidate signal light timing strategy, and each of the candidate signal light timing strategies has a corresponding candidate speed interval;

[0134] The calculation module is further configured to traverse the Pareto optimal solution set to obtain a target speed interval, and use the candidate signal light timing strategy corresponding to the target speed interval as the target signal light timing strategy;

[0135] A control module, wherein the control module is configured to control the timing of the traffic lights at each intersection of the target green wave road within the target preset time period according to the target traffic light timing strategy; wherein the target speed range is the alternative speed range that best meets the specific range of urban traffic management targets.

[0136] This article uses specific examples to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core ideas. The above is only a preferred implementation method of the present invention. It should be pointed out that due to the limitations of textual expression, there are objectively infinite specific structures. For ordinary technicians in this technical field, without departing from the principles of the present invention, several improvements, modifications or changes can be made, and the above technical features can be combined in an appropriate manner; these improvements, modifications, changes or combinations, or the direct application of the inventive concept and technical solution to other occasions without improvement, should be regarded as the protection scope of the present invention.

Claims

1. A method for analyzing road green wave speed based on AI multi-objective, characterized in that: The following steps are involved: Acquire a preset speed interval of a target green wave road, where the preset speed interval is used to represent a speed interval of vehicles traveling on the target green wave road within a target preset time period; the target green wave road includes a plurality of intersections; According to the preset speed interval, an initial population is randomly generated, wherein the initial population includes a plurality of signal light timing strategies, each of which is for at least one intersection and covers control parameters of different time periods; Among all the signal light timing strategies, a plurality of first signal light timing strategies are obtained after elite reservation and congestion-based selection; Based on a plurality of the first signal light timing strategies, an approximate Pareto optimal solution set is obtained, wherein the Pareto optimal solution set includes at least one alternative signal light timing strategy, and each of the alternative signal light timing strategies has a corresponding alternative speed interval; Traversing the Pareto optimal solution set to obtain a target speed interval, and using the alternative traffic light timing strategy corresponding to the target speed interval as the target traffic light timing strategy; According to the target signal light timing strategy, the signal lights at each intersection of the target green wave road are controlled within the target preset time period; wherein the target speed range is the alternative speed range that best meets the specific range of the urban traffic management target; The step of obtaining a plurality of first signal light timing strategies selected through elite reservation and based on congestion from among all the signal light timing strategies comprises at least the following steps: Performing non-dominated sorting on all the signal light timing strategies to obtain a plurality of non-dominated signal light timing strategies, and classifying the non-dominated signal light timing strategies into different levels; the levels include at least a first level; the first level is a set of non-dominated signal light timing strategies at an optimal level after performing non-dominated sorting on all the signal light timing strategies; using all the non-dominated signal light timing strategies at the first level as the first signal light timing strategy; The hierarchy also includes remaining hierarchies except the first hierarchy; After performing non-dominated sorting on all the signal light timing strategies to obtain a plurality of non-dominated signal light timing strategies, and classifying the non-dominated signal light timing strategies into different levels, the following steps are also included: calculating the individual congestion degree of each of the non-dominated signal light timing strategies in each of the remaining levels; The non-dominated signal light timing strategy for which the individual congestion degree is greater than or equal to a first preset threshold is used as the first signal light timing strategy; The non-dominated sorting of all the signal light timing strategies specifically includes the following steps: Performing multi-objective evaluation on all the signal light timing strategies respectively to obtain an evaluation result of each signal light timing strategy; The evaluation result includes the target score corresponding to each target; the multiple targets include at least the target of reducing the waiting time of vehicles at the intersection, the target of increasing the overall hourly flow of vehicles at the intersection, the target of increasing the average speed of the road, and the target of reducing the road delay time ratio; Setting a dominance relationship, and performing a non-domination sorting according to the dominance relationship based on the evaluation result of each of the signal light timing strategies; The traversing the Pareto optimal solution set to obtain the target speed range specifically includes the following steps: Obtaining the main objectives of urban traffic management within the target preset time period, and obtaining quantitative standards based on the main objectives; Calculating an evaluation result of each of the candidate signal light timing strategies, and obtaining a degree of compliance of each of the candidate signal light timing strategies with the quantitative standard according to the evaluation result of each of the candidate signal light timing strategies; The alternative speed interval corresponding to the alternative traffic light timing strategy with the highest degree of compliance is used as the target speed interval.

2. The method for analyzing road green wave speed based on AI multi-objective according to claim 1 is characterized in that: Multiple targets constitute the dimension of the multi-target space; The calculating of the individual congestion degree of each of the non-dominated signal light timing strategies in each of the remaining levels specifically comprises the following steps: Obtaining adjacent individuals of the non-dominated signal light timing strategy in each dimension of the multi-objective space; Calculating the adjacent distance between the non-dominated signal light timing strategy and the adjacent individuals in each dimension; Based on the adjacent distances between the non-dominated signal light timing strategy and the adjacent individuals in all dimensions, the individual congestion degree of the non-dominated signal light timing strategy is obtained.

3. The method for analyzing road green wave speed based on AI multi-objective according to claim 1 is characterized in that: The step of obtaining the preset speed range of the target green wave road includes at least the following steps: Acquire a historical driving data set of the target green wave road, the historical driving data set at least including multiple collection time periods and driving data corresponding to each collection time period, the driving data at least including vehicle speed data, vehicle flow data, and traffic congestion conditions between each intersection on the target green wave road; Marking the collection period corresponding to the target preset period as the target collection period, and obtaining the average speed of each vehicle in the target collection period according to the vehicle speed data between each of the intersections corresponding to the target collection period; According to the average speed of each vehicle, a reference speed of the target acquisition period is obtained, and vehicles corresponding to the average speed whose deviation from the reference speed is greater than or equal to a second preset threshold are eliminated to obtain a target vehicle; The interval consisting of the maximum value and the minimum value of the average speeds corresponding to all the target vehicles is used as the preset speed interval.

4. The method for analyzing road green wave speed based on AI multi-objective according to claim 1 is characterized in that: The obtaining of the main target of urban traffic management within the target preset time period specifically includes the following steps: Determine the road type of the target green wave road, the road type includes at least a regular road and a special road, and the special road includes at least a road with a special stop requirement in at least one special time period; If the road type of the target green wave road is the conventional road or the special road and the target preset time period is not within the special time period, the universal target is used as the main target; If the road type of the target green wave road is the special road and the target preset time period is within the special time period, the safety goal and the efficiency goal are combined with the universal goal to obtain the main goal.

5. The method for analyzing road green wave speed based on AI multi-objective according to claim 1 is characterized in that: The method of controlling the timing of the traffic lights at each intersection of the target green wave road within the target preset time period according to the target traffic light timing strategy specifically includes the following steps: Acquire the real-time traffic condition of the target green wave road in real time, wherein the real-time traffic condition at least includes whether there is a traffic accident or temporary construction; When the real-time road condition does not contain the traffic accident or the temporary construction, or the real-time road condition contains the traffic accident or the temporary construction and the impact on the road traffic stability is less than a third preset threshold, according to the target signal light timing strategy, the timing of the signal lights at each intersection of the target green wave road is controlled within the target preset time period; After the real-time traffic condition of the target green wave road is acquired in real time, the following steps are also included: When the real-time road condition has the traffic accident or the temporary construction and the impact of the temporary construction on the road traffic stability is greater than or equal to the third preset threshold, calculating an updated speed interval, wherein the updated speed interval is used to reduce vehicle congestion; According to the updated timing strategy corresponding to the updated speed interval, the timing of the traffic lights at each intersection of the target green wave road is controlled within the target preset time period.

6. A system for analyzing road green wave speed based on AI multi-objectives, used to implement the method for analyzing road green wave speed based on AI multi-objectives as claimed in any one of claims 1 to 5, characterized in that: include: A collection module, the collection module is configured to obtain a preset speed interval of a target green wave road, the preset speed interval is used to represent a speed interval of vehicles traveling on the target green wave road within a target preset time period; the target green wave road includes a plurality of intersections; A calculation module, wherein the calculation module is configured to randomly generate an initial population according to the preset speed interval, wherein the initial population includes a plurality of signal light timing strategies, each of which is for at least one intersection and covers control parameters of different time periods; The calculation module is further configured to obtain, from all the signal light timing strategies, a plurality of first signal light timing strategies that have been selected after elite reservation and based on congestion; The calculation module is further configured to obtain an approximate Pareto optimal solution set based on a plurality of the first signal light timing strategies, wherein the Pareto optimal solution set includes at least one candidate signal light timing strategy, and each of the candidate signal light timing strategies has a corresponding candidate speed interval; The calculation module is further configured to traverse the Pareto optimal solution set to obtain a target speed interval, and use the candidate signal light timing strategy corresponding to the target speed interval as the target signal light timing strategy; A control module, wherein the control module is configured to control the timing of the traffic lights at each intersection of the target green wave road within the target preset time period according to the target traffic light timing strategy; wherein the target speed range is the alternative speed range that best meets the specific range of urban traffic management targets.

Images