Multi-intersection traffic signal control method, device and medium

Abstract

The embodiment of the invention discloses a traffic signal control method and device for multiple intersections and a medium, belongs to the technical field of intelligent traffic, and solves the problem of low real-time performance and accuracy of traffic control for multiple intersections. Obtaining driving data of vehicles passing through a specified intersection in the to-be-detected multi-intersection so as to determine a driving data set based on a time sequence corresponding to the specified intersection; dividing the driving data set into a plurality of groups, and inputting the plurality of groups of driving data sets into different traffic state prediction models to obtain traffic data corresponding to the to-be-detected intersection at the next moment; determining a traffic state change type corresponding to the to-be-detected intersection based on the traffic data at the current moment and the traffic data at the next moment; and determining a target state corresponding to each specified intersection based on the traffic state change type, the traffic change data and the road network diagram of the to-be-detected multi-intersection, generating a traffic signal lamp instruction based on the target state, and controlling the traffic signal of the to-be-detected multi-intersection.

Description

Technical Field

[0001] The present application relates to the field of intelligent transportation technology, and in particular to a method, device and medium for controlling traffic signals at multiple intersections. Background Art

[0002] With the accelerated advancement of urbanization, traffic congestion has become a major bottleneck restricting the development of modern cities. In a busy urban traffic network, the traffic conditions at multiple intersections are particularly complex and changeable, which places extremely high demands on traffic management and signal control.

[0003] However, most traditional multi-intersection traffic light control systems use fixed signal cycles and fixed green and red light time configurations. During peak hours, when traffic volume surges, fixed signal cycles are often difficult to respond to in a timely manner, resulting in longer queues and more serious traffic congestion. During non-peak hours, although traffic volume decreases, traffic lights still operate according to fixed cycles, which not only wastes road resources but also may cause unnecessary traffic delays. Therefore, due to the low efficiency of generating and executing traditional traffic light control commands, it is difficult to achieve real-time traffic control, resulting in low real-time and accuracy of traffic control at multiple intersections. Summary of the invention

[0004] The embodiments of the present application provide a traffic signal control method, device and medium for multiple intersections, which are used to solve the following technical problems: the generation and execution efficiency of traditional traffic light control instructions is low, and it is difficult to achieve real-time traffic control, resulting in low real-time and accuracy of traffic control at multiple intersections.

[0005] The present application embodiment adopts the following technical solutions:

[0006] An embodiment of the present application provides a traffic signal control method for multiple intersections, including: obtaining vehicle driving data passing through a designated intersection among the multiple intersections to be tested, processing the vehicle driving data of the designated intersection to determine a time-series-based driving data set corresponding to the designated intersection; dividing the driving data set into multiple groups, and inputting the multiple groups of driving data sets into different traffic state prediction models to predict the traffic data at the next moment corresponding to the intersection to be tested; determining a traffic state change type corresponding to the intersection to be tested based on the traffic data at a current moment and the traffic data at a next moment; determining a target state corresponding to each designated intersection based on the traffic state change type, the corresponding change data between the traffic data at a current moment and the traffic data at a next moment, and a road network diagram of the multiple intersections to be tested; and generating a traffic signal light instruction corresponding to the multiple intersections to be tested based on the target state to control the traffic signals at the multiple intersections to be tested.

[0007] The embodiment of the present application improves the accuracy and robustness of the prediction by dividing the driving data set into multiple groups and inputting different traffic state prediction models. Different models capture different traffic characteristics, so the combined use can improve the overall prediction performance. By predicting the traffic data at the next moment, it is helpful to make decisions in advance, thereby effectively alleviating traffic congestion and improving road capacity. Determine the type of traffic state change, such as increased congestion, relief, sudden changes, etc., and formulate targeted traffic management strategies to ensure the smooth operation of traffic flow. By generating traffic light instructions to control traffic signals and dynamically adjusting the time of traffic lights, traffic flow is effectively managed, congestion and waiting time are reduced, road traffic efficiency is improved, and the real-time and accuracy of traffic control at intersections are improved.

[0008] In one implementation of the present application, vehicle driving data at a designated intersection is processed to determine a time-series-based driving data set corresponding to the designated intersection, specifically including: arranging the driving data corresponding to the same vehicle acquired according to time sequence to obtain time-series data corresponding to each vehicle; arranging the time-series data of different vehicles corresponding to each designated intersection to construct a vehicle traffic density distribution map corresponding to the designated intersection; wherein the vehicle traffic density distribution map includes multiple first traffic time periods, and also includes vehicle traffic data corresponding to multiple first traffic time periods; based on a preset density division threshold, merging and splitting multiple adjacent first traffic time periods to obtain multiple second traffic time periods; based on the multiple second traffic time periods corresponding to each designated intersection, and the vehicle traffic data corresponding to the multiple second traffic time periods, obtaining a time-series-based driving data set corresponding to each designated intersection.

[0009] In one implementation of the present application, a driving data set is divided into multiple groups, and the multiple groups of driving data sets are input into different traffic state prediction models to predict the traffic data corresponding to the intersection to be tested at the next moment, specifically including: comparing the driving data set with the historical driving data set, and dividing the driving data set once based on the comparison difference and the first preset difference threshold; comparing adjacent data in the driving data set, and dividing the driving data set twice based on the comparison difference and the second preset difference threshold; dividing the driving data set into a first category data set and a second category data set based on the first division and the second division; inputting the first category data set into a preset decision tree model to obtain first traffic prediction data; and inputting the second category data set into a preset naive Bayes model to obtain second traffic prediction data; inputting the first traffic prediction data and the second traffic prediction data into a preset meta-model to output the traffic data corresponding to the multiple intersections to be tested at the next moment through the meta-model.

[0010] In one implementation of the present application, based on the traffic data at the current moment and the traffic data at the next moment, the type of traffic state change corresponding to the intersection to be tested is determined, specifically including: based on the traffic data at the current moment, determining the current vehicle traffic density corresponding to the current multiple second traffic time periods; based on the vehicle traffic data corresponding to the second traffic time period at the current moment, determining the current vehicle traffic time; comparing the current vehicle traffic density with the reference traffic density value, and determining the first traffic congestion data based on the comparison result; comparing the current vehicle traffic time with the reference traffic time, and determining the second traffic congestion data based on the comparison result. ; Based on the traffic data at the next moment, determine the vehicle traffic density at the next moment corresponding to multiple second traffic time periods at the next moment; based on the traffic data corresponding to the second traffic time period at the next moment, determine the vehicle travel time at the next moment; compare the vehicle traffic density at the next moment with the reference traffic density value, and determine the third traffic congestion data based on the comparison result; compare the vehicle travel time at the next moment with the reference travel time, and determine the fourth traffic congestion data based on the comparison result; based on the first traffic congestion data, the second traffic congestion data, the third traffic congestion data and the fourth traffic congestion data, obtain the traffic state change type.

[0011] In one implementation of the present application, based on the first traffic congestion data, the second traffic congestion data, the third traffic congestion data and the fourth traffic congestion data, a traffic state change type is obtained, specifically including: weighting the first traffic congestion data and the second traffic congestion data to obtain the traffic congestion rate at the current moment; weighting the third traffic congestion data and the fourth traffic congestion data to obtain the traffic congestion rate at the next moment; when the traffic congestion rate at the current moment is greater than the traffic congestion rate at the next moment, determining that the designated intersection corresponds to the congestion relief type; when the traffic congestion rate at the current moment is not greater than the traffic congestion rate at the next moment, determining that the designated intersection corresponds to the congestion enhancement type.

[0012] In one implementation of the present application, based on the traffic state change type, the corresponding change data between the traffic data at the current moment and the traffic data at the next moment, and the road network diagram of the multiple intersections to be tested, the target state corresponding to each designated intersection is determined, specifically including: performing difference processing on the traffic congestion rate at the current moment and the traffic congestion rate at the next moment to obtain the traffic congestion change rate; constructing a traffic change data set based on the traffic congestion change rates corresponding to each designated intersection of the multiple intersections to be tested, and the identifiers of each designated intersection; matching the traffic change data set with the road network diagram of the multiple intersections to be tested to obtain a first matching diagram; and matching the traffic data corresponding to the traffic data at the next moment with the road network diagram of the multiple intersections to be tested to obtain a second matching diagram; inputting the first matching diagram and the second matching diagram corresponding to the multiple intersections to be tested into a preset traffic state adjustment model to obtain the target state corresponding to the multiple intersections to be tested.

[0013] In one implementation of the present application, traffic light instructions corresponding to the multiple intersections to be tested are generated based on the target state, specifically including: obtaining historical traffic data corresponding to the multiple intersections to be tested in the same period of time, and determining the lane queue length corresponding to the intersection to be tested based on the historical traffic data; determining the first flow direction saturation data corresponding to each road of the multiple intersections to be tested based on the lane queue length and the maximum capacity value of the road; after an interval of a preset time, re-determining the second flow direction saturation data corresponding to each road of the multiple intersections to be tested; determining the traffic light control efficiency corresponding to the multiple intersections to be tested based on the preset time interval, the first flow direction saturation data and the second flow direction saturation data; and generating traffic light instructions corresponding to the multiple intersections to be tested based on the traffic light control efficiency and the target state corresponding to the designated intersection.

[0014] In one implementation of the present application, based on the traffic light control efficiency and the target state corresponding to the designated intersection, traffic light instructions corresponding to the multiple intersections to be tested are generated, specifically including: comparing the current lane queue length corresponding to the multiple intersections to be tested with the lane queue length corresponding to the target state to determine the difference in the number of vehicles; determining different road adjustment areas based on the difference in the number of vehicles, and setting corresponding initial adjustment times for different adjustment areas based on the difference in the number of vehicles; dynamically adjusting the initial adjustment time based on the traffic light control efficiency; generating traffic light instructions corresponding to the multiple intersections to be tested based on the adjustment area and the dynamically adjusted adjustment time.

[0015] An embodiment of the present application provides a traffic signal control device for multiple intersections, comprising: at least one processor; and a memory in communication with the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can: obtain vehicle driving data passing through a designated intersection among the multiple intersections to be tested, process the vehicle driving data of the designated intersection to determine a time-series-based driving data set corresponding to the designated intersection; divide the driving data set into multiple groups, and input the multiple groups of driving data sets into different traffic state prediction models to predict the traffic data at the next moment corresponding to the intersection to be tested; determine the type of traffic state change corresponding to the intersection to be tested based on the traffic data at the current moment and the traffic data at the next moment; determine the target state corresponding to each designated intersection based on the traffic state change type, the corresponding change data between the traffic data at the current moment and the traffic data at the next moment, and the road network diagram of the multiple intersections to be tested; generate traffic signal light instructions corresponding to the multiple intersections to be tested based on the target state to control the traffic signals of the multiple intersections to be tested.

[0016] A non-volatile computer storage medium provided in an embodiment of the present application stores computer executable instructions, wherein the computer executable instructions are configured to: obtain vehicle driving data passing through a designated intersection among multiple intersections to be tested, process the vehicle driving data at the designated intersection to determine a time-series-based driving data set corresponding to the designated intersection; divide the driving data set into multiple groups, and input the multiple groups of driving data sets into different traffic state prediction models to predict the traffic data at the next moment corresponding to the intersection to be tested; determine the type of traffic state change corresponding to the intersection to be tested based on the traffic data at the current moment and the traffic data at the next moment; determine the target state corresponding to each designated intersection based on the traffic state change type, the corresponding change data between the traffic data at the current moment and the traffic data at the next moment, and the road network diagram of the multiple intersections to be tested; generate a traffic signal light instruction corresponding to the multiple intersections to be tested based on the target state to control the traffic signal at the multiple intersections to be tested.

[0017] At least one of the above technical solutions adopted in the embodiments of the present application can achieve the following beneficial effects: The embodiments of the present application improve the accuracy and robustness of the prediction by dividing the driving data set into multiple groups and inputting different traffic state prediction models. Different models capture different traffic characteristics, so the combined use can improve the overall prediction performance. By predicting the traffic data at the next moment, it is helpful to make decisions in advance, thereby effectively alleviating traffic congestion and improving road capacity. Determine the type of traffic state change, such as increased congestion, relief, sudden changes, etc., and formulate targeted traffic management strategies to ensure the smooth operation of traffic flow. By generating traffic light instructions to control traffic signals and dynamically adjusting the time of traffic lights, traffic flow is effectively managed, congestion and waiting time are reduced, road traffic efficiency is improved, and the real-time and accuracy of traffic control at intersections are also improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative labor. In the drawings:

[0019] Figure 1 A flow chart of a traffic signal control method for multiple intersections provided in an embodiment of the present application;

[0020] Figure 2 A schematic diagram of the structure of a traffic signal control device for multiple intersections provided in an embodiment of the present application.

[0021] Reference numerals:

[0022] 200: Traffic signal control device for multiple intersections, 201: Processor, 202: Memory. DETAILED DESCRIPTION

[0023] The embodiments of the present application provide a method, device and medium for controlling traffic signals at multiple intersections.

[0024] In order to enable those skilled in the art to better understand the technical solutions in this application, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the drawings in the embodiments of this application. Obviously, the described embodiments are only part of the embodiments of this application, not all of them. Based on the embodiments of this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this application.

[0025] The technical solution proposed in the embodiment of the present invention is described in detail below with reference to the accompanying drawings.

[0026] Figure 1 A flow chart of a traffic signal control method for multiple intersections provided in an embodiment of the present application is as follows: Figure 1 As shown, the traffic signal control method for multiple intersections includes the following steps:

[0027] Step 101: Acquire the driving data of vehicles passing through a designated intersection among the multiple intersections to be tested, and process the driving data of the vehicles at the designated intersection to determine a time-series-based driving data set corresponding to the designated intersection.

[0028] In one embodiment of the present application, the acquired driving data corresponding to the same vehicle are arranged according to the time sequence to obtain the time series data corresponding to each vehicle. The time series data of different vehicles corresponding to each designated intersection are arranged to construct a vehicle traffic density distribution map corresponding to the designated intersection; wherein the vehicle traffic density distribution map includes multiple first traffic time periods, and also includes vehicle traffic data corresponding to multiple first traffic time periods. Based on the preset density division threshold, multiple adjacent first traffic time periods are merged and split to obtain multiple second traffic time periods. Based on the multiple second traffic time periods corresponding to each designated intersection, and the vehicle traffic data corresponding to the multiple second traffic time periods, a time series-based driving data set corresponding to each designated intersection is obtained.

[0029] Specifically, the driving data of vehicles passing through the designated intersection among the multiple intersections to be tested is obtained, and the driving data of the same vehicle at different time points are arranged according to the time sequence. These data may include the time, speed, direction, etc. of the vehicle passing through the designated intersection. These time series data are classified according to vehicles and intersections, and a vehicle traffic density distribution map of the designated intersection is constructed. The vehicle traffic density distribution map in the embodiment of the present application is based on time, showing the number or density of vehicles passing through the intersection in different time periods. For example, for a designated intersection, if multiple vehicles pass through at different time points, a vehicle traffic density distribution map is constructed, in which the horizontal axis is time (such as hours) and the vertical axis is vehicle density (such as the number of vehicles passing per hour). Assuming that 100 vehicles pass through from 7:00 to 8:00 in the morning and 80 vehicles pass through from 8:00 to 9:00 in the morning, these two time periods will be represented by different heights in the figure.

[0030] Furthermore, in the vehicle traffic density distribution map, multiple first traffic time periods are divided, each of which corresponds to a specific time period and the vehicle traffic data in the time period. For example, a day can be divided into 24 first traffic time periods, each of which corresponds to one hour, and the vehicle traffic data in each time period is recorded. According to the preset vehicle density division threshold, adjacent first traffic time periods are merged or split to form a more reasonable second traffic time period. If the preset density division threshold is 50 vehicles per hour, then for the two time periods of 7:00-8:00 and 8:00-9:00 in the above example, since the vehicle density of 7:00-8:00 is higher than the threshold, and 8:00-9:00 is lower than the threshold, the system may use 7:00-8:00 as a second traffic time period, and if 8:00-9:00 is still lower than the threshold after being merged with other time periods, it may continue to be merged with other time periods. Based on the multiple second traffic time periods corresponding to each designated intersection, and the vehicle traffic data in these time periods, a time-series-based driving data set for each designated intersection is constructed.

[0031] Step 102: Divide the driving data set into multiple groups, and input the multiple groups of driving data sets into different traffic state prediction models to predict the traffic data corresponding to the intersection to be tested at the next moment.

[0032] In one embodiment of the present application, a driving data set is compared with a historical driving data set, and the driving data set is divided once based on the comparison difference and a first preset difference threshold. Adjacent data in the driving data set is compared, and the driving data set is divided twice based on the comparison difference and a second preset difference threshold. Based on the first division and the second division, the driving data set is divided into a first category data set and a second category data set. The first category data set is input into a preset decision tree model to obtain first traffic prediction data. And, the second category data set is input into a preset naive Bayes model to obtain second traffic prediction data. The first traffic prediction data and the second traffic prediction data are input into a preset meta-model to output the next moment traffic data corresponding to the multiple intersections to be tested through the meta-model.

[0033] Specifically, the currently acquired driving data set is compared with the stored historical driving data set. The content of the comparison may include key indicators such as the number of vehicles, average speed, and traffic flow in the same time period. The difference between the current data set and the historical data set is calculated, and the difference reflects the degree of change in traffic conditions. If the comparison difference corresponding to the current data exceeds the first preset difference threshold, it means that the current traffic conditions have changed significantly compared with the historical data. At this time, the data currently being compared is divided into a data set with significant changes. If the comparison difference does not exceed the threshold, the data currently being compared is divided into a data set with insignificant changes.

[0034] Furthermore, the adjacent data in the current data set are compared to detect short-term fluctuations in traffic conditions. If the comparison difference between the adjacent data exceeds the second preset difference threshold, it means that the traffic conditions have fluctuated significantly in a short period of time, and the data involved in the comparison are divided into a data set with significant fluctuations. If the comparison difference does not exceed the threshold, the data currently being compared is divided into a data set with insignificant fluctuations.

[0035] Furthermore, data sets with significant changes and significant fluctuations are classified as first-category data sets, and data sets with insignificant changes or insignificant fluctuations are classified as second-category data sets.

[0036] Furthermore, the first category data set is input into a preset decision tree model, wherein the preset decision tree model is used to process data with significant changes and fluctuations to obtain first traffic prediction data. The second category data set is input into a preset naive Bayes model, and the preset naive Bayes model performs well when processing relatively stable and predictable data to obtain second traffic prediction data. Finally, the first traffic prediction data and the second traffic prediction data are input into a preset meta-model, wherein the preset meta-model in the embodiment of the present application can be a weighted average, a voting mechanism or a stacking model in an integrated learning method, etc. The results of the two prediction models are integrated through the preset meta-model, and the next moment traffic data corresponding to the multiple intersections to be tested is output. The next moment traffic data in the embodiment of the present application includes key indicators such as the number of vehicles, average speed, and traffic flow.

[0037] Furthermore, the training method of the preset decision tree in the embodiment of the present application is: the collected data set features for training the decision tree are used as nodes of the decision tree, and the best feature splitting node is selected in a recursive manner until the stopping condition is met. During the construction process, the quality of the features is measured by indicators such as information gain and Gini index, so as to select the optimal splitting feature. In order to prevent the decision tree from overfitting, the complexity of the tree can also be reduced by pruning technology. The performance of the decision tree is evaluated using a test data set, including indicators such as accuracy and recall rate. According to the evaluation results, the parameters of the decision tree are adjusted to optimize the performance of the model.

[0038] The training process of the preset naive Bayes model in the embodiment of the present application is as follows: collect a data set for training naive Bayes, select features that have an important impact on classification from the data set, and for each category, calculate the conditional probability of each feature in the category. Use Bayes' theorem to calculate the probability of a given category for a given input feature value, and use a test data set to evaluate the performance of naive Bayes, including indicators such as accuracy and recall.

[0039] Step 103: Determine the traffic state change type corresponding to the intersection to be tested based on the current traffic data and the next traffic data.

[0040] In one embodiment of the present application, based on the traffic data at the current moment, the current vehicle traffic density corresponding to the current multiple second traffic periods is determined. Based on the vehicle traffic data corresponding to the second traffic period at the current moment, the current vehicle traffic duration is determined. The current vehicle traffic density is compared with the reference traffic density value, and the first traffic congestion data is determined based on the comparison result. The current vehicle traffic duration is compared with the reference traffic duration, and the second traffic congestion data is determined based on the comparison result. Based on the traffic data at the next moment, the vehicle traffic density at the next moment corresponding to the multiple second traffic periods at the next moment is determined. Based on the traffic data corresponding to the second traffic period at the next moment, the vehicle traffic duration at the next moment is determined. The vehicle traffic density at the next moment is compared with the reference traffic density value, and the third traffic congestion data is determined based on the comparison result. The vehicle traffic duration at the next moment is compared with the reference traffic duration, and the fourth traffic congestion data is determined based on the comparison result. Based on the first traffic congestion data, the second traffic congestion data, the third traffic congestion data and the fourth traffic congestion data, the traffic state change type is obtained.

[0041] Specifically, based on the time-series driving data sets corresponding to each designated intersection, the current vehicle traffic density corresponding to the current multiple second traffic time periods is determined, the current vehicle traffic density can be determined through the vehicle traffic density map, and the current vehicle traffic duration corresponding to the vehicle passing through multiple intersections can be determined through the driving data set.

[0042] Furthermore, the current vehicle traffic density is compared with a preset reference traffic density value. If the current density is higher than the reference density, it is considered that there is traffic congestion during this period, and the congestion rate is determined based on the ratio between the two and recorded as the first traffic congestion data. The current vehicle travel time is compared with the reference travel time. If the current time exceeds the reference time, it is considered that there is traffic congestion, and the congestion rate is determined based on the ratio between the two and recorded as the second traffic congestion data.

[0043] Furthermore, based on the traffic data obtained at the next moment, a vehicle traffic density map at the next moment is drawn to obtain the vehicle traffic density at the next moment and the travel time at the next moment. The vehicle traffic density at the next moment is compared with the preset reference traffic density value. If the current density is higher than the reference density, it is considered that there is traffic congestion during this period, and the congestion rate is determined based on the ratio between the two and recorded as the third traffic congestion data. The vehicle travel time at the next moment is compared with the reference travel time. If the next moment duration exceeds the reference duration, it is considered that there is traffic congestion, and the congestion rate is determined based on the ratio between the two and recorded as the fourth traffic congestion data.

[0044] In one embodiment of the present application, the first traffic congestion data and the second traffic congestion data are weighted to obtain the traffic congestion rate at the current moment. The third traffic congestion data and the fourth traffic congestion data are weighted to obtain the traffic congestion rate at the next moment. When the traffic congestion rate at the current moment is greater than the traffic congestion rate at the next moment, it is determined that the designated intersection corresponds to the congestion relief type. When the traffic congestion rate at the current moment is not greater than the traffic congestion rate at the next moment, it is determined that the designated intersection corresponds to the congestion enhancement type.

[0045] Specifically, the first traffic congestion data is obtained based on the vehicle traffic density, reflecting the density of vehicles on the road or intersection. The second traffic congestion data is obtained based on the vehicle travel time, reflecting the time required for vehicles to pass through the road or intersection. By weighting the two, the impact of vehicle density and traffic efficiency on traffic congestion can be comprehensively considered. Similarly, the traffic congestion data at the next moment is weighted to obtain the traffic congestion rate at the next moment. The traffic congestion rates obtained at the current moment and the next moment are compared. If the traffic congestion rate at the current moment is greater than the traffic congestion rate at the next moment, it means that the traffic congestion situation at the designated intersection is easing, and it can be classified as a "congestion relief type". If the traffic congestion rate at the current moment is not greater than the traffic congestion rate at the next moment, it means that the traffic congestion situation at the designated intersection has not improved or may worsen, and it can be classified as a "congestion enhancement type".

[0046] Step 104: Determine the target state corresponding to each designated intersection based on the traffic state change type, the corresponding change data between the current traffic data and the next traffic data, and the road network diagram of the multiple intersections to be tested.

[0047] In one embodiment of the present application, the traffic congestion rate at the current moment is differenced with the traffic congestion rate at the next moment to obtain the traffic congestion change rate. Based on the traffic congestion change rates corresponding to the designated intersections of the multiple intersections to be tested, and the identifiers of the designated intersections, a traffic change data set is constructed. The traffic change data set is matched with the road network diagram of the multiple intersections to be tested to obtain a first matching graph. And, the traffic data corresponding to the traffic data at the next moment is matched with the road network diagram of the multiple intersections to be tested to obtain a second matching graph. The first matching graph and the second matching graph corresponding to the multiple intersections to be tested are input into a preset traffic state adjustment model to obtain a target state corresponding to the multiple intersections to be tested.

[0048] Specifically, the traffic congestion rate at the current moment is processed with the traffic congestion rate at the next moment to obtain the traffic congestion change rate, which reflects the temporal change trend of the traffic congestion situation. Based on the traffic congestion change rate corresponding to each designated intersection in the multi-intersection to be tested, and the identification of these intersections (such as intersection number, location information, etc.), a traffic change data set is constructed. The traffic change data set is matched with the road network diagram of the multi-intersection to be tested to obtain a first matching graph. Through the matching process, the spatial distribution of the traffic congestion change rate is mapped to the road network diagram, which facilitates intuitive understanding of the change trend of traffic congestion. At the same time, the traffic data at the next moment, such as vehicle traffic density, travel time, etc., are also matched to the road network diagram to obtain a second matching graph, which reflects the traffic conditions at the next moment.

[0049] Furthermore, the first matching graph and the second matching graph are used as inputs to a preset traffic state adjustment model, and the target state corresponding to the multi-intersection to be tested is obtained through calculation and analysis of the model. The target state in the embodiment of the present application is the optimal traffic condition that can be achieved after alleviating the current traffic data, which is related to the degree of alleviating traffic congestion and the degree of improving traffic efficiency.

[0050] Step 105 : Generate traffic signal light instructions corresponding to the multiple intersections to be tested based on the target state, so as to control the traffic signals at the multiple intersections to be tested.

[0051] In one embodiment of the present application, historical traffic data corresponding to the multiple intersections to be tested in the same period of time is obtained, and based on the historical traffic data, the lane queue length corresponding to the intersection to be tested is determined. Based on the lane queue length and the maximum capacity value of the road, the first flow direction saturation data corresponding to each road of the multiple intersections to be tested is determined. After an interval of a preset time length, the second flow direction saturation data corresponding to each road of the multiple intersections to be tested is re-determined. Based on the preset interval time length, the first flow direction saturation data and the second flow direction saturation data, the traffic light control efficiency corresponding to the multiple intersections to be tested is determined. Based on the traffic light control efficiency and the target state corresponding to the designated intersection, a traffic light instruction corresponding to the multiple intersections to be tested is generated.

[0052] Specifically, first, collect historical traffic data of multiple intersections to be tested in the same period of time. These data may include vehicle flow, vehicle speed, lane occupancy, etc. Based on the collected historical traffic data, analyze and determine the lane queue length of each intersection to be tested. Use the lane queue length and the maximum capacity value of the road (that is, the maximum number of vehicles that the road can accommodate) to calculate the first flow direction saturation data of each road in each intersection to be tested. After a preset time interval, collect data again and calculate the second flow direction saturation data of each road in each intersection to be tested. Evaluate the control efficiency of the traffic light based on the preset time interval, the first flow direction saturation data, and the second flow direction saturation data.

[0053] Furthermore, the flow saturation data at two time points are compared, and it is analyzed whether the traffic light adjustment effectively reduces the lane queue length or improves the road capacity. That is, the control efficiency of the traffic light is obtained by calculating the ratio of the difference of the flow saturation data after comparison and the time interval. Finally, based on the control efficiency of the traffic light and the target state of the specified intersection, traffic light instructions suitable for the multi-intersection to be tested are generated. These instructions may include adjusting the green light time, red light time or yellow light time of the traffic light, and changing the synchronization strategy of the traffic light.

[0054] In one embodiment of the present application, the current lane queue length corresponding to the multi-intersection to be tested is compared with the lane queue length corresponding to the target state to determine the difference in the number of vehicles. Different road adjustment areas are determined based on the difference in the number of vehicles, and corresponding initial adjustment durations are set for different adjustment areas based on the difference in the number of vehicles. The initial adjustment duration is dynamically adjusted based on the control efficiency of the traffic light. Based on the adjustment area and the dynamically adjusted adjustment duration, a traffic light instruction corresponding to the multi-intersection to be tested is generated.

[0055] Specifically, the current queue length of each lane of the multi-intersection to be tested is measured, and these current queue lengths are compared with the target state, that is, the lane queue length under the expected optimal queue length. Through the comparison, the difference in the number of vehicles is determined, that is, the difference between the current queue length and the target queue length. According to the size and distribution of the difference in the number of vehicles, the road is divided into different adjustment areas, for example, these adjustment areas include high congestion areas (large difference in the number of vehicles), medium congestion areas (medium difference in the number of vehicles) and low congestion areas (small difference in the number of vehicles).

[0056] Furthermore, an initial adjustment duration is set for each adjustment area. For example, a high-congestion area requires a longer green light time to evacuate vehicles, while a low-congestion area requires a shorter green light time to maintain smooth traffic. According to the calculated traffic light control efficiency of the previous period adjacent to the current period, the initial adjustment duration is adjusted by a coefficient to dynamically adjust the green and red light durations of each adjustment area. For example, if the historical traffic light control efficiency is high, it means that the signal light adjustment has a significant effect on improving traffic flow. At this time, the green light duration can be appropriately increased or the red light duration can be reduced to further improve the traffic capacity. Based on the adjustment area determined in the above steps and the dynamically adjusted adjustment duration, traffic light instructions suitable for the multiple intersections to be tested are generated. These instructions should be able to adjust the timing of the signal lights in real time to adapt to changes in traffic flow, thereby optimizing traffic flow, reducing congestion, and improving road capacity.

[0057] Figure 2 The following is a schematic diagram of a traffic signal control device for multiple intersections provided in an embodiment of the present application. Figure 2 As shown, a traffic signal control device 200 for multiple intersections includes: at least one processor 201; and a memory 202 in communication with the at least one processor 201; wherein the memory 202 stores instructions that can be executed by the at least one processor 201, and the instructions are executed by the at least one processor 201 so that the at least one processor 201 can: obtain vehicle driving data passing through a designated intersection in the multiple intersections to be tested, process the vehicle driving data of the designated intersection, so as to determine a time-series-based driving data set corresponding to the designated intersection; divide the driving data set into multiple groups, and input the multiple groups of driving data sets into different traffic state prediction models to predict the traffic data corresponding to the intersection to be tested at the next moment; determine the traffic state change type corresponding to the intersection to be tested based on the traffic data at the current moment and the traffic data at the next moment; determine the target state corresponding to each designated intersection based on the traffic state change type, the corresponding change data between the traffic data at the current moment and the traffic data at the next moment, and the road network diagram of the multiple intersections to be tested; generate a traffic signal light instruction corresponding to the multiple intersections to be tested based on the target state, so as to control the traffic signal of the multiple intersections to be tested.

[0058] A non-volatile computer storage medium provided in an embodiment of the present application stores computer executable instructions, wherein the computer executable instructions are configured to: obtain vehicle driving data passing through a designated intersection among multiple intersections to be tested, process the vehicle driving data at the designated intersection to determine a time-series-based driving data set corresponding to the designated intersection; divide the driving data set into multiple groups, and input the multiple groups of driving data sets into different traffic state prediction models to predict the traffic data at the next moment corresponding to the intersection to be tested; determine the type of traffic state change corresponding to the intersection to be tested based on the traffic data at the current moment and the traffic data at the next moment; determine the target state corresponding to each designated intersection based on the traffic state change type, the corresponding change data between the traffic data at the current moment and the traffic data at the next moment, and the road network diagram of the multiple intersections to be tested; generate a traffic signal light instruction corresponding to the multiple intersections to be tested based on the target state to control the traffic signal at the multiple intersections to be tested.

[0059] Each embodiment in this application is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device, equipment, and non-volatile computer storage medium embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0060] The above is only an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the embodiments of the present application may have various changes and variations. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present application.

Claims

1. A traffic signal control method for multiple intersections, characterized in that: The method comprises: Acquire vehicle driving data passing through a designated intersection among the multiple intersections to be tested, and process the vehicle driving data of the designated intersection to determine a time-series-based driving data set corresponding to the designated intersection; Dividing the driving data set into multiple groups, and inputting the multiple groups of driving data sets into different traffic state prediction models to predict the traffic data corresponding to the intersection to be tested at the next moment; Determine the traffic state change type corresponding to the intersection to be tested based on the current traffic data and the next traffic data; Determine the target state corresponding to each of the designated intersections based on the traffic state change type, the corresponding change data between the current traffic data and the next traffic data, and the road network diagram of the multiple intersections to be tested; A traffic light instruction corresponding to the multi-intersection to be tested is generated based on the target state to control the traffic signal of the multi-intersection to be tested.

2. A traffic signal control method for multiple intersections according to claim 1, characterized in that: The processing of the vehicle driving data at the designated intersection to determine a time-series-based driving data set corresponding to the designated intersection specifically includes: Arrange the acquired driving data corresponding to the same vehicle according to the time sequence to obtain the time series data corresponding to each vehicle; Arranging the time series data of different vehicles corresponding to each of the designated intersections to construct a vehicle traffic density distribution map corresponding to the designated intersection; wherein the vehicle traffic density distribution map includes a plurality of first traffic time periods, and also includes a plurality of vehicle traffic data corresponding to the first traffic time periods respectively; Based on a preset density division threshold, merging and splitting a plurality of adjacent first travel time periods to obtain a plurality of second travel time periods; Based on the plurality of second traffic time periods corresponding to the designated intersections and the vehicle traffic data corresponding to the plurality of second traffic time periods, a time-series-based driving data set corresponding to the designated intersections is obtained.

3. The traffic signal control method for multiple intersections according to claim 1, characterized in that: The step of dividing the driving data set into a plurality of groups and inputting the plurality of driving data sets into different traffic state prediction models to predict the traffic data corresponding to the intersection to be tested at the next moment specifically includes: Comparing the driving data set with the historical driving data set, and dividing the driving data set once based on the comparison difference and a first preset difference threshold; Comparing adjacent data in the driving data set, and dividing the driving data set into two parts based on the comparison difference and a second preset difference threshold; Based on the first division and the second division, the driving data set is divided into a first category data set and a second category data set; Inputting the first category data set into a preset decision tree model to obtain first traffic prediction data; And, inputting the second category data set into a preset naive Bayes model to obtain second traffic prediction data; The first traffic prediction data and the second traffic prediction data are input into a preset meta-model, so as to output the traffic data at the next moment corresponding to the multi-intersection to be tested through the meta-model.

4. The traffic signal control method for multiple intersections according to claim 2, characterized in that: The determining, based on the current traffic data and the next traffic data, the type of traffic state change corresponding to the intersection to be tested specifically includes: Based on the current traffic data, determining the current vehicle traffic density corresponding to each of the second traffic time periods; Determine the current vehicle travel time based on the vehicle travel data corresponding to the second travel period at the current moment; Comparing the current vehicle traffic density with a reference traffic density value, and determining first traffic congestion data based on the comparison result; Comparing the current vehicle travel time with the reference travel time, and determining second traffic congestion data based on the comparison result; Based on the traffic data at the next moment, determining the vehicle traffic density at the next moment corresponding to the plurality of the second traffic time periods at the next moment; Determine the vehicle travel time at the next moment based on the travel data corresponding to the second travel period at the next moment; Comparing the vehicle traffic density at the next moment with a reference traffic density value, and determining third traffic congestion data based on the comparison result; Comparing the vehicle travel time at the next moment with the reference travel time, and determining fourth traffic congestion data based on the comparison result; A traffic state change type is obtained based on the first traffic congestion data, the second traffic congestion data, the third traffic congestion data, and the fourth traffic congestion data.

5. The traffic signal control method for multiple intersections according to claim 4, characterized in that: The obtaining of the traffic state change type based on the first traffic congestion data, the second traffic congestion data, the third traffic congestion data and the fourth traffic congestion data specifically includes: Performing weighted processing on the first traffic congestion data and the second traffic congestion data to obtain a current traffic congestion rate; Performing weighted processing on the third traffic congestion data and the fourth traffic congestion data to obtain a traffic congestion rate at a next moment; When the traffic congestion rate at the current moment is greater than the traffic congestion rate at the next moment, determining that the designated intersection corresponds to a congestion relief type; When the traffic congestion rate at the current moment is not greater than the traffic congestion rate at the next moment, it is determined that the designated intersection corresponds to an increased congestion type.

6. A traffic signal control method for multiple intersections according to claim 5, characterized in that: The method of determining the target state corresponding to each of the designated intersections based on the traffic state change type, the corresponding change data between the current traffic data and the next traffic data, and the road network diagram of the multiple intersections to be tested specifically includes: Performing difference processing on the traffic congestion rate at the current moment and the traffic congestion rate at the next moment to obtain a traffic congestion change rate; Constructing a traffic change data set based on the traffic congestion change rates corresponding to the designated intersections of the multiple intersections to be tested and the identifiers of the designated intersections; Matching the traffic change data set with the road network graph of the multiple intersections to be tested to obtain a first matching graph; and matching the traffic data corresponding to the traffic data at the next moment with the road network graph of the multiple intersections to be tested to obtain a second matching graph; The first matching graph and the second matching graph corresponding to the multi-intersection to be tested are input into a preset traffic state adjustment model to obtain a target state corresponding to the multi-intersection to be tested.

7. The traffic signal control method for multiple intersections according to claim 1, characterized in that: The generating of the traffic light instruction corresponding to the multiple intersections to be tested based on the target state specifically includes: Acquire historical traffic data corresponding to the multiple intersections to be tested in the same period of time, and determine the lane queue length corresponding to the intersections to be tested based on the historical traffic data; Based on the lane queue length and the maximum capacity value of the road, determining the first flow direction saturation data corresponding to each road of the multi-intersection to be tested; After a preset time interval, re-determine the second flow direction saturation data corresponding to each road of the multi-intersection to be tested; Determining the traffic light control efficiency corresponding to the multi-intersection to be tested based on the preset interval time, the first flow direction saturation data and the second flow direction saturation data; Based on the traffic light control efficiency and the target state corresponding to the designated intersection, a traffic light instruction corresponding to the multi-intersection to be tested is generated.

8. The method for controlling traffic signals at multiple intersections according to claim 1, characterized in that: The step of generating a traffic light instruction corresponding to the multi-intersection to be tested based on the traffic light control efficiency and the target state corresponding to the designated intersection specifically includes: Comparing the current lane queue length corresponding to the multi-intersection to be tested with the lane queue length corresponding to the target state to determine the difference in the number of vehicles; Determine different road adjustment areas based on the vehicle quantity difference, and set corresponding initial adjustment durations for the different adjustment areas based on the vehicle quantity difference; Based on the traffic light control efficiency, dynamically adjusting the initial adjustment duration; Based on the adjustment area and the dynamically adjusted adjustment duration, a traffic light instruction corresponding to the multiple intersections to be tested is generated.

9. A traffic signal control device for multiple intersections, characterized in that: The device comprises a memory for storing computer program instructions and a processor for executing the program instructions, wherein when the computer program instructions are executed by the processor, the device is triggered to execute the method according to any one of claims 1 to 8.

10. A non-volatile computer storage medium storing computer executable instructions, characterized in that: The computer executable instructions can execute the method according to any one of claims 1 to 8.

Images