Perception-driven intelligent multi-modal traffic signal optimization method and system

Abstract

This invention discloses a perception-driven intelligent multi-modal traffic signal optimization method and system, belonging to the field of traffic signal optimization technology. The method includes: collecting traffic flow data; constructing a multi-modal traffic flow model; connecting multiple traffic lights within a target area to obtain street light control signals; predicting traffic flow trends based on the traffic flow data and the street light control signals, using the multi-modal traffic flow model; and configuring a timing optimization scheme for multiple traffic lights within the target area based on the traffic flow trend prediction results. This invention solves the technical problem in the prior art where traffic light timing schemes cannot be accurately adjusted dynamically based on real-time traffic conditions and multi-modal traffic behavior interactions, leading to low traffic intersection efficiency. It achieves the technical effect of optimizing traffic light timing in a target area based on prediction results, effectively improving the traffic intersection efficiency.

Description

Technical Field

[0001] This invention relates to the field of traffic signal optimization technology, specifically to a multi-mode traffic signal optimization method and system driven by perception intelligence. Background Technology

[0002] With rapid urbanization, the urban population and number of motor vehicles have surged, making road traffic flow extremely complex. Traditional traffic signal timing systems have many shortcomings. They are usually set based on fixed time intervals or simple traffic flow statistics, which cannot adapt to complex and changing traffic conditions. They cannot reasonably allocate road resources during peak and off-peak hours, and they also ignore the travel needs of non-motorized vehicles and pedestrians, fail to consider the interaction between different modes of transportation, and the accuracy of existing traffic flow prediction methods is poor, thus failing to provide reliable support for traffic signal timing optimization.

[0003] Existing technologies suffer from the technical problem that traffic signal timing schemes cannot be dynamically adjusted accurately based on real-time traffic conditions and the interaction of multi-modal traffic behaviors, resulting in low traffic efficiency at intersections. Summary of the Invention

[0004] This application provides a perception-driven intelligent multi-mode traffic signal optimization method and system to address the technical problem that existing traffic signal timing schemes cannot be accurately adjusted dynamically based on real-time traffic conditions and the interaction between multiple traffic behaviors, resulting in low traffic efficiency at traffic intersections.

[0005] In view of the above problems, this application provides a multi-modal traffic signal optimization method and system driven by perception intelligence.

[0006] A first aspect of this application provides a perception-intelligent driven multi-modal traffic signal optimization method, the method comprising:

[0007] Traffic flow data is collected, including motor vehicle flow, non-motor vehicle flow, and pedestrian flow; a multi-modal traffic flow model is constructed to simulate the interactive behavioral relationships between motor vehicle behavior characteristics, non-motor vehicle behavior characteristics, and pedestrian behavior characteristics; multiple traffic lights within the target area are connected to obtain street light control signals; based on the traffic flow data and the street light control signals, the multi-modal traffic flow model is used to predict traffic flow trends; and based on the traffic flow trend prediction results, a timing optimization scheme for multiple traffic lights within the target area is configured.

[0008] A second aspect of this application provides a perception-intelligent driven multi-modal traffic signal optimization system, the system comprising:

[0009] The system includes a traffic flow data acquisition module for collecting traffic flow data, including motor vehicle flow, non-motor vehicle flow, and pedestrian flow; a multi-modal traffic flow model construction module for constructing a multi-modal traffic flow model to simulate the interactive behavioral relationships between motor vehicle behavior characteristics, non-motor vehicle behavior characteristics, and pedestrian behavior characteristics; a street light control signal acquisition module for connecting multiple traffic lights within the target area to obtain street light control signals; a traffic flow trend prediction module for predicting traffic flow trends based on the traffic flow data, the street light control signals, and the multi-modal traffic flow model; and a timing optimization scheme configuration module for configuring timing optimization schemes for multiple traffic lights within the target area based on the traffic flow trend prediction results.

[0010] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0011] Traffic flow data, including motor vehicle flow, non-motor vehicle flow, and pedestrian flow, is collected. A multi-modal traffic flow model is constructed to simulate the interactive behavioral relationships among motor vehicle behavior characteristics, non-motor vehicle behavior characteristics, and pedestrian behavior characteristics. Multiple traffic lights within a target area are connected to obtain street light control signals. Based on the traffic flow data and the street light control signals, the multi-modal traffic flow model is used to predict traffic flow trends. Based on the traffic flow trend prediction results, a timing optimization scheme for multiple traffic lights within the target area is configured. This achieves the technical effect of optimizing traffic light timing in a target area based on prediction results, effectively improving the traffic efficiency of intersections. Attached Figure Description

[0012] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0013] Figure 1 This is a schematic flowchart of a multi-mode traffic signal optimization method driven by perception intelligence provided in an embodiment of this application;

[0014] Figure 2 This is a schematic diagram of the structure of a perception-driven, intelligent-driven multi-mode traffic signal optimization system provided in an embodiment of this application.

[0015] Figure labeling: Traffic flow data acquisition module 10, multi-mode traffic flow model construction module 20, street light control signal acquisition module 30, traffic flow change trend prediction module 40, timing optimization scheme configuration module 50. Detailed Implementation

[0016] This application provides a perception-driven intelligent multi-mode traffic signal optimization method and system to address the technical problem that existing traffic signal timing schemes cannot be accurately adjusted dynamically based on real-time traffic conditions and the interaction of multi-mode traffic behaviors, resulting in low traffic efficiency at traffic intersections.

[0017] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0018] Example 1, as Figure 1 As shown, this application provides a perception-driven intelligent multi-modal traffic signal optimization method, the method comprising:

[0019] Step S100: Collect traffic flow data, which includes motor vehicle flow, non-motor vehicle flow, and pedestrian flow.

[0020] Specifically, the first step is to collect traffic flow data using various sensor devices, such as geomagnetic sensors, video detectors, and radar sensors, deployed at traffic intersections and surrounding roads. Geomagnetic sensors accurately detect the passage of motor vehicles, thus calculating vehicle flow; video detectors, using image recognition technology, can not only identify motor vehicles but also effectively monitor non-motorized vehicles and pedestrians, compiling their flow data; radar sensors further supplement this by detecting information such as the speed of traffic participants. The vehicle, non-motorized vehicle, and pedestrian flow data collected by these devices are transmitted in real-time and stably to the server at the traffic signal control center, providing a solid data foundation for subsequent analysis and processing.

[0021] Step S200: Construct a multimodal traffic flow model, which is used to simulate the interactive behavioral relationships between motor vehicle behavior characteristics, non-motor vehicle behavior characteristics, and pedestrian behavior characteristics.

[0022] Specifically, a neural network model is used to construct a multimodal traffic flow model that simulates the interactive relationships of the behavioral characteristics of motor vehicles, non-motor vehicles, and pedestrians. First, comprehensive historical traffic flow data for the target area is collected, covering different time periods, weather conditions, and seasons, including information such as the number, speed, and direction of motor vehicles, non-motor vehicles, and pedestrians. Next, data preprocessing is performed to remove noise and erroneous records, and the data is normalized to a uniform scale before extracting meaningful features such as flow rate of change and density. Then, a suitable neural network model is selected based on the requirements, such as a recurrent neural network (RNN) and its variants, long short-term memory networks (LSTM) and gated recurrent units (GRU), capable of handling sequential data. When designing the model architecture, the number of neurons in the input layer depends on the number of extracted features, with features of motor vehicles, non-motor vehicles, and pedestrians input into different branches. The hidden layer learns complex relationships through multiple nonlinear transformations, and interaction nodes are set between different branches to capture interactive behaviors. The output layer outputs the simulated traffic flow state. Then, the model is trained using preprocessed historical data. Appropriate loss functions and optimization algorithms are set, and weights and biases are continuously adjusted during training to enable the model to fit the data and learn the interaction patterns of multimodal traffic flows. Finally, a multimodal traffic flow model that meets the requirements is constructed.

[0023] Step S300: Connect multiple traffic lights in the target area to obtain street light control signals.

[0024] Specifically, this involves connecting multiple traffic lights within the target area to obtain street light control signals. In practice, a stable connection is established between the traffic signal control center's server and each traffic light in the target area using wired or wireless communication technology. Standardized communication protocols, such as NTCIP, are employed to ensure smooth and accurate data transmission and command exchange between the server and the traffic lights. Once the connection is successful, the server can obtain various control signals from the traffic lights in real time, including the current status of the light (red, green, or yellow), the signal cycle duration, and the duration allocation of each phase (such as the straight-ahead phase and the left-turn phase). These street light control signals are crucial for understanding the current traffic signal operation and play an indispensable role in subsequent analysis of traffic flow changes and optimization of traffic light timing.

[0025] Step S400: Based on the traffic flow data and the street light control signal, use the multi-modal traffic flow model to predict the traffic flow change trend.

[0026] Specifically, based on the previously collected traffic flow data and acquired street light control signals, a pre-constructed multi-modal traffic flow model is used to predict traffic flow trends. On the computing platform of the traffic signal control center, real-time traffic flow data and the latest acquired street light control signals are input parameters into the multi-modal traffic flow model. The model, based on various traffic behavior patterns and interactions learned during training, combined with current traffic conditions, simulates and predicts traffic flow trends over a future period. For example, it predicts the changes in motor vehicle traffic flow in different directions at a certain intersection over the next 15 minutes, whether congestion will occur, the possible locations of congestion, and the degree of congestion; the aggregation of non-motorized vehicles and pedestrians at different time periods and their impact on motor vehicle traffic, etc. To improve the accuracy and reliability of the predictions, time series analysis methods and advanced machine learning prediction algorithms, such as Long Short-Term Memory (LSTM) networks and Support Vector Regression (SVR), are used. These algorithms can deeply mine the time-series characteristics and potential patterns in traffic data. Taking into account the uncertainty of traffic demand and the impact of unforeseen events (such as traffic accidents and road construction) on traffic flow, they comprehensively predict traffic flow by setting random variables and using scenario analysis, thus making the prediction results more consistent with actual traffic conditions. The prediction results are presented to traffic management personnel in an intuitive and visual form, such as displaying traffic flow change curves for different time periods and directions through charts, and marking areas that may experience congestion on maps. This allows management personnel to understand the development trend of traffic flow in a timely manner and provides strong support for subsequent decision-making.

[0027] Step S500: Based on the traffic flow trend prediction results, configure the timing optimization scheme for multiple traffic lights in the target area.

[0028] Specifically, based on the predicted traffic flow trends, timing optimization schemes for multiple traffic lights within the target area are configured. At the traffic signal control center, edge computing nodes are first deployed at multiple traffic lights within the target area. These edge computing nodes possess certain computing and data processing capabilities, enabling preliminary analysis and processing of traffic data locally. Based on the local traffic data collected by the edge computing nodes and combined with the predicted traffic flow trends, local traffic light timing schemes are generated. Each edge computing node's generated local timing scheme is synchronized with the central node via the network, uploading its own scheme information to the central node. At the central node, motor vehicle traffic efficiency, non-motor vehicle traffic efficiency, and pedestrian traffic efficiency are jointly optimized as objectives. Based on real-time traffic flow data, the weights of these three objectives are dynamically allocated. For example, during morning and evening rush hours, when motor vehicle traffic is high, the weight of motor vehicle traffic efficiency is appropriately increased; during school dismissal times, when pedestrian traffic increases, the weight of pedestrian traffic efficiency is increased. By dynamically adjusting the weights, multiple different timing adaptation schemes are obtained. To select the optimal timing scheme, the crucial factor of pedestrian crossing demand is first introduced. Based on pedestrian flow data and actual pedestrian crossing needs, multiple timing adaptation schemes are prioritized, and a first screening criterion is set. For example, if a scheme can ensure pedestrians cross the street quickly and safely, reducing pedestrian waiting time, then that scheme will have a higher priority in the ranking. Simultaneously, to further optimize the screening process, high-flow and high-conflict areas within the target area are marked. Through in-depth analysis of traffic data and on-site observation, key traffic flow nodes are identified. These key nodes typically include road intersections, pedestrian crossings, and non-motorized vehicle lane junctions—locations prone to traffic congestion and conflicts. Traffic flow characteristics are analyzed for key nodes in different areas. In high-flow traffic flow key nodes, the focus is on analyzing potential conflicts between different modes of transportation, such as conflicts between motorized and non-motorized vehicles, and between pedestrians, and corresponding first feature vectors are configured. In high-conflict key nodes, detailed analysis is conducted on indicators such as queue length, waiting time, and queue dissipation speed for vehicles and pedestrians, and second feature vectors are configured. Based on the information contained in these two feature vectors, multiple timing adaptation schemes are subjected to secondary screening, with a second screening condition set. After these two rounds of screening, the most suitable timing optimization scheme for multiple traffic lights within the target area is finally determined. This optimization scheme is transmitted to the control systems of each traffic light via the network, automatically updating the timing parameters of the traffic lights to achieve real-time optimized control of traffic signals. Furthermore, during the operation of traffic signals, the actual traffic flow is continuously monitored, feedback data is collected, and the timing scheme is dynamically adjusted according to new traffic conditions to ensure that traffic signals can always adapt to constantly changing traffic demands, improve the overall traffic efficiency of intersections, alleviate traffic congestion, and ensure traffic safety.

[0029] In one possible implementation, step S200 further includes:

[0030] Step S210: Obtain historical traffic flow data within the target area.

[0031] Step S220: Based on the historical traffic flow data, decompose the historical traffic flow of motor vehicles, the historical traffic flow of non-motor vehicles, and the historical traffic flow of pedestrians, and set up a training data set.

[0032] Step S230: Based on the training dataset, train the multimodal traffic flow model and use a sliding window mechanism to dynamically update the traffic flow change trend prediction results.

[0033] Specifically, historical traffic flow data within the target area is acquired by extracting data from the traffic management department's database and the storage units of various traffic monitoring devices distributed throughout the target area. This data covers long-term periods, including information from various scenarios such as weekdays and non-weekdays, different seasons, morning and evening peak hours, and off-peak hours. It includes traffic flow data for motor vehicles, non-motor vehicles, and pedestrians, as well as corresponding timestamps, weather conditions, road construction, and other related information, providing a comprehensive basis for subsequent in-depth analysis.

[0034] Based on the acquired historical traffic flow data, historical traffic flow for motor vehicles, non-motor vehicles, and pedestrians is decomposed and used to establish a training dataset. Data processing algorithms are employed to segment the traffic flow data of different traffic participants according to dimensions such as time series, road segment, and traffic conditions. For example, traffic flow data for motor vehicles, non-motor vehicles, and pedestrians is extracted for each road segment at 5-minute intervals. Then, based on different traffic scenario labels, such as "peak-hour congestion" and "off-peak smooth traffic," the segmented data is categorized and organized, and representative data samples are selected to form the training dataset. This process ensures that the training data covers various typical traffic conditions, improving the comprehensiveness and accuracy of model training.

[0035] A multimodal traffic flow neural network model will be trained based on a prepared training dataset, and a sliding window mechanism will be used to dynamically update the traffic flow trend prediction results. First, the training dataset is input into the neural network model. The input layer receives various feature data from motor vehicles, non-motor vehicles, and pedestrians. Through complex nonlinear transformations by multiple neurons in the hidden layer, the model learns the interaction relationships between the behavioral characteristics of different traffic subjects. During training, an appropriate loss function (such as the mean squared error loss function) is used to measure the difference between the model's predicted values ​​and the actual values. The stochastic gradient descent algorithm is used to continuously adjust the model's weights and bias parameters to minimize the loss function value and improve the model's prediction accuracy. Simultaneously, to enable the model to adapt to constantly changing traffic conditions, a sliding window mechanism is introduced. With a fixed time interval as the window length, as new traffic flow data is continuously generated, new data is added to the window while the oldest data in the window is removed, forming an updated window of data. By using this updated window data to incrementally train or fine-tune the model, the model can learn the latest traffic flow characteristics and patterns in real time, thereby dynamically updating the prediction results of traffic flow change trends, ensuring the timeliness and accuracy of the prediction, and providing a reliable basis for the subsequent optimization of traffic signal timing.

[0036] In one possible implementation, step S500 further includes:

[0037] Step S510: Deploy edge computing nodes based on multiple traffic lights within the target area.

[0038] Step S520: Based on the edge computing node, generate a local traffic signal timing scheme and synchronize the data with the central node.

[0039] Step S530: Determine the timing optimization scheme for multiple traffic lights in the target area by uploading multiple local traffic light timing schemes to the central node.

[0040] Specifically, edge computing nodes are deployed based on multiple traffic lights within the target area. A detailed survey of the traffic layout in the target area is conducted, taking into account factors such as the distribution of traffic lights, the traffic volume of surrounding roads, and network coverage. For areas with high traffic volume and complex intersections, high-performance edge computing nodes are prioritized; while in areas with relatively low traffic volume and simpler road conditions, lighter nodes are deployed. These edge computing nodes connect to the traffic lights via wired or wireless communication technologies, enabling them to acquire real-time status information of the traffic lights and surrounding traffic data.

[0041] Local traffic light timing schemes are generated based on deployed edge computing nodes and synchronized with the central node. The edge computing nodes, leveraging their data acquisition capabilities, collect detailed traffic data in real time around the traffic lights in their area, covering information such as flow, speed, and direction of motor vehicles, non-motorized vehicles, and pedestrians. Then, using a pre-installed genetic algorithm, combined with historical traffic data, current real-time traffic conditions, and specific traffic rules, this data is deeply analyzed and processed to generate a suitable traffic light timing scheme for the local area. This scheme aims to maximize traffic efficiency and reduce waiting times for vehicles and pedestrians. After generating the scheme, the edge computing nodes upload the local timing scheme and related real-time traffic data to the central node via a stable and reliable communication network. Simultaneously, they receive instructions from the central node, traffic information from other areas, and overall traffic planning strategies, completing data synchronization to ensure that the local timing scheme is consistent with and coordinated with the overall traffic management goals of the target area.

[0042] By uploading multiple local traffic light timing schemes to the central node, an optimized timing scheme for multiple traffic lights within the target area is determined. After receiving the local timing schemes uploaded by each edge computing node, the central node performs a comprehensive analysis of these schemes. Factors considered include the traffic correlation between different areas, the balance of traffic flow at each intersection, and the improvement effect on overall traffic congestion. The central node uses more advanced optimization algorithms to integrate and adjust multiple local schemes, balancing the interests and needs of various parties, and ultimately determines a timing optimization scheme that can both meet the traffic needs of each local area and maximize the overall traffic benefits of the target area. This scheme is then distributed to each traffic light to achieve optimized traffic signal control.

[0043] In one possible implementation, step S530 further includes:

[0044] Step S531: Take the efficiency of motor vehicle traffic, the efficiency of non-motor vehicle traffic, and the efficiency of pedestrian traffic as a joint optimization objective.

[0045] Step S532: Based on real-time traffic flow data, dynamically assign weights to the joint optimization objective to obtain multiple timing adaptation schemes.

[0046] Step S533: Filter the multiple timing adaptation schemes to determine the timing optimization schemes for multiple traffic lights in the target area.

[0047] Specifically, the efficiency of motor vehicle traffic, non-motorized vehicle traffic, and pedestrian traffic are set as joint optimization goals. This means that when formulating timing plans, the focus is no longer solely on a single mode of transportation, but rather on comprehensively considering the needs of all parties to strive for a more rational allocation of road resources. Motor vehicle traffic efficiency is reflected in the number of motor vehicles passing through the intersection per unit time and their average delay time; non-motorized vehicle traffic efficiency can be measured by indicators such as waiting time and changes in riding speed of non-motorized vehicles at the intersection; pedestrian traffic efficiency focuses on the time required for pedestrians to cross the street and the waiting time. Using these three as joint optimization goals can comprehensively improve the overall traffic capacity of the intersection and ensure the travel experience of different traffic participants.

[0048] Based on real-time traffic flow data, dynamic weights are allocated to the joint optimization objectives, resulting in multiple timing adaptation schemes. Edge computing nodes continuously collect real-time traffic flow data at intersections, including changes in the flow of motor vehicles, non-motor vehicles, and pedestrians at different times and in different directions. Based on this data, a weight allocation algorithm is used to flexibly adjust the weights of the three optimization objectives according to real-time changes in traffic flow. For example, during weekday morning rush hour, motor vehicle flow increases significantly, so the weight of motor vehicle efficiency is appropriately increased; while during school dismissal time, pedestrian flow surges, so the weight of pedestrian efficiency is increased accordingly. Through dynamic adjustment of weights, combined with various possible combinations of traffic light timings, multiple timing adaptation schemes are generated to meet different traffic conditions and adapt to complex and ever-changing traffic scenarios.

[0049] Multiple timing adaptation schemes were screened to determine the optimal timing schemes for multiple traffic lights within the target area. The screening process comprehensively considered various factors, such as pedestrian crossing demand. Based on pedestrian flow data and actual pedestrian crossing needs, such as the presence of large numbers of students, the elderly, and other relatively slow-moving groups, multiple timing adaptation schemes were initially screened and prioritized. Simultaneously, high-flow and high-conflict areas were marked, and key traffic flow nodes, such as road intersections, pedestrian crossings, and non-motorized vehicle lane junctions, were identified. For these key nodes, potential conflicts between different modes of transportation, as well as characteristics such as queue length, waiting time, and queue dissipation speed for vehicles and pedestrians, were analyzed. Based on these detailed analysis results, the timing adaptation schemes were screened again, removing those that might exacerbate traffic congestion and increase conflicts. After multiple rounds of screening, the optimal timing adaptation scheme that maximizes the overall traffic efficiency of the intersection and ensures traffic safety under current traffic conditions was selected from multiple timing adaptation schemes, achieving precise control of traffic lights and optimizing urban traffic order.

[0050] In one possible implementation, step S533 further includes:

[0051] Step S5331: Introduce the need for pedestrians to cross the street.

[0052] Step S5332: Based on pedestrian traffic data and pedestrian crossing needs, prioritize the multiple timing adaptation schemes and set the first filtering condition.

[0053] Specifically, pedestrian crossing demand is incorporated into the process of determining multiple traffic light timing optimization schemes within the target area. Pedestrians, as a crucial component of traffic participants, have significant crossing needs that cannot be ignored. This not only concerns pedestrian safety but also impacts the operational efficiency of the entire traffic system. In real-world traffic scenarios, pedestrian crossing demand is influenced by various factors, such as the distribution of functional areas around intersections (whether they are near schools, shopping malls, residential areas, etc.) and pedestrian congestion during specific time periods (e.g., morning and evening rush hours, school arrival and departure times). Information on pedestrian crossing demand, including expected crossing time intervals and acceptable maximum waiting times, is collected through on-site surveys, historical data analysis, and real-time monitoring.

[0054] Based on pedestrian flow data and pedestrian crossing demand, multiple timing adaptation schemes are prioritized and a first screening criterion is set. Real-time pedestrian flow data is combined with previously collected pedestrian crossing demand information to evaluate each timing adaptation scheme. If a timing scheme can provide sufficient green time for pedestrians during periods of high pedestrian flow, ensuring their safe and smooth crossing while minimizing the impact on motor vehicle and non-motor vehicle traffic, then that scheme will be given a higher priority. For example, at intersections near schools, pedestrian flow is high during dismissal time, and most pedestrians are students. Timing schemes that guarantee students sufficient time to cross safely without causing prolonged traffic congestion are given priority. By prioritizing multiple timing adaptation schemes based on pedestrian-related factors and setting this as the first screening criterion, schemes that better meet pedestrian needs are initially selected, laying the foundation for further determination of the final timing optimization scheme. This reflects the protection of pedestrian rights and the emphasis on humane management of the traffic system.

[0055] In one possible implementation, step S533 further includes:

[0056] Step S5333: Mark high-flow areas and high-conflict areas.

[0057] Step S5334: Based on the high-flow area and high-conflict area, identify key traffic flow nodes and determine the traffic flow characteristics corresponding to the key traffic flow nodes.

[0058] Step S5335: Based on the traffic flow characteristics corresponding to the key traffic flow nodes, perform a second screening of the multiple timing adaptation schemes and set a second screening condition.

[0059] Specifically, high-traffic and high-conflict areas are marked. Traffic flow monitoring equipment, such as geomagnetic sensors and high-definition cameras, continuously collects real-time traffic flow data for each road segment within the target area. Data analysis algorithms compare the traffic flow data of each area with historical average traffic flow. When the traffic flow in a certain area is significantly higher than the average level, it is marked as a high-traffic area. Simultaneously, combined with a traffic conflict monitoring system, the frequency and severity of conflicts between different modes of transportation (motor vehicles, non-motor vehicles, and pedestrians) are assessed, and areas where conflicts frequently occur are marked as high-conflict areas. For example, during rush hour, intersections and surrounding sections of urban main roads are prone to congestion and conflicts due to the surge in traffic volume and the weaving of vehicles from different directions; these areas are typically marked as high-traffic and high-conflict areas.

[0060] Based on the identified high-volume and high-conflict areas, key traffic flow nodes are identified, and their corresponding traffic flow characteristics are determined. Within these areas, locations such as road intersections, pedestrian crossings, and non-motorized vehicle lane junctions are identified as key traffic flow nodes due to the convergence and dispersion of traffic flows and the complex traffic conditions. For these key nodes, traffic flow characteristics are analyzed from different dimensions. In high-volume areas, the focus is on analyzing potential conflicts between different modes of transportation, such as the probability of conflict between turning motor vehicles and straight-going non-motorized vehicles / pedestrians, thus configuring the first feature vector corresponding to the key traffic flow node. In high-conflict areas, the focus is on indicators such as queue length, waiting time, and queue dissipation speed for vehicles and pedestrians, thus configuring the second feature vector corresponding to the key traffic flow node. Combining the first and second feature vectors, the traffic flow characteristics corresponding to the key traffic flow nodes are comprehensively determined, providing an accurate basis for subsequent screening.

[0061] Based on the traffic flow characteristics corresponding to key traffic flow nodes, a secondary screening process is conducted on multiple timing adaptation schemes, with a second screening criterion set. Each timing adaptation scheme is then simulated and evaluated in the actual scenario of key traffic flow nodes. Schemes that can effectively alleviate traffic pressure in high-volume areas and reduce the frequency of conflicts in high-conflict areas are given higher priority. For example, if a timing scheme can significantly shorten vehicle queue lengths and reasonably reduce pedestrian waiting times at key nodes in high-conflict areas, then that scheme is more likely to be retained; conversely, if a scheme causes increased traffic congestion and conflicts at key nodes, it is eliminated. Through this secondary screening process based on the characteristics of key traffic flow nodes, and by setting more stringent second screening criteria, better schemes are further selected from multiple timing adaptation schemes. This ensures that the final determined traffic light timing optimization scheme better meets actual traffic needs, effectively improves traffic efficiency, and guarantees traffic safety.

[0062] In one possible implementation, step S5334 further includes:

[0063] Key traffic flow nodes are identified, including road intersections, pedestrian crossings, and non-motorized vehicle lane junctions. In high-traffic areas, potential conflicts between different modes of transportation are analyzed at these key traffic flow nodes, and a first feature vector is configured for each node. In high-conflict areas, queue lengths, waiting times, and queue dissipation speeds for vehicles and pedestrians are analyzed at these key traffic flow nodes, and a second feature vector is configured for each node. Based on the first and second feature vectors, the traffic flow characteristics corresponding to the key traffic flow nodes are determined.

[0064] Specifically, to accurately grasp traffic conditions and rationally optimize traffic signal timing, it is necessary to identify key traffic flow nodes and analyze their characteristics. Firstly, through a comprehensive survey and data analysis of the target area's traffic network, road intersections, pedestrian crossings, and non-motorized vehicle lane junctions are identified as key traffic flow nodes. These locations are areas where different modes of transportation converge and transition, resulting in complex and variable traffic conditions that significantly impact overall traffic efficiency.

[0065] At key traffic flow nodes in high-traffic areas, in-depth analysis is conducted to identify potential conflicts between different modes of transportation. For example, at road intersections, conflicts may arise between motor vehicles turning right and non-motorized vehicles and pedestrians going straight; at pedestrian crossings, there may also be conflicts between pedestrians crossing the street and right-turning motor vehicles. To address these situations, relevant data, such as traffic flow, speed, and direction of travel for different modes of transportation, are collected to configure the first feature vector corresponding to the key traffic flow nodes. This vector quantifies the probability and severity of potential conflicts between different modes of transportation, providing crucial information for assessing traffic conditions.

[0066] At key traffic flow nodes in high-conflict areas, the focus is on analyzing queue lengths, waiting times, and queue dissipation speeds for vehicles and pedestrians. This data is acquired in real-time using traffic monitoring equipment such as cameras and sensors. Prolonged queues and waiting not only reduce traffic efficiency but can also trigger more conflicts; queue dissipation speed reflects traffic management capabilities. By analyzing these indicators, a second feature vector is configured for each key traffic flow node. This vector directly reflects the degree of traffic congestion and the ability to restore smooth traffic flow in high-conflict areas.

[0067] For the potential conflicts between different modes of transportation at key nodes in high-flow traffic areas, reflected by the first feature vector, and the congestion at key nodes in high-conflict traffic areas, reflected by the second feature vector, the values ​​of various indicators are calculated respectively. Then, based on different traffic scenarios and actual needs, appropriate weights are assigned to the first and second feature vectors. For example, during weekday morning and evening rush hours, because traffic congestion has a greater impact on overall traffic, the second feature vector, reflecting congestion, can be given a higher weight; while during normal times when traffic flow is relatively stable, the weights of the two vectors can be appropriately balanced. Then, the indicator values ​​in each feature vector are multiplied by their corresponding weights and summed to obtain the weighted scores of the first and second feature vectors respectively. Finally, these two weighted scores are summed again to obtain a preliminary comprehensive score. To make this comprehensive score more accurately reflect traffic flow characteristics, it can be further revised and adjusted by combining historical traffic data and expert experience. For example, if a key traffic flow node has historically experienced a frequent occurrence of specific types of traffic conflicts or congestion, the influence of relevant factors can be appropriately increased in the comprehensive score. After such processing, the final values ​​and the information they represent are the traffic flow characteristics corresponding to key traffic flow nodes. They comprehensively consider potential conflicts and congestion, and can provide a reliable decision-making basis for tasks such as traffic signal timing optimization.

[0068] Example 2, based on the same inventive concept as the perception-driven multi-mode traffic signal optimization method in the foregoing examples, such as... Figure 2 As shown, this application provides a perception-intelligent driven multi-mode traffic signal optimization system. The system and method embodiments in this application are based on the same inventive concept. The system includes:

[0069] The traffic flow data acquisition module 10 is used to collect traffic flow data, which includes motor vehicle flow, non-motor vehicle flow, and pedestrian flow.

[0070] The multimodal traffic flow model construction module 20 is used to construct a multimodal traffic flow model, which is used to simulate the interactive relationship between motor vehicle behavior characteristics, non-motor vehicle behavior characteristics, and pedestrian behavior characteristics.

[0071] The street light control signal acquisition module 30 is used to connect to multiple traffic lights within the target area to obtain street light control signals.

[0072] The traffic flow change trend prediction module 40 is used to predict the traffic flow change trend based on the traffic flow data and the street light control signal, using the multi-mode traffic flow model.

[0073] The timing optimization scheme configuration module 50 is used to configure the timing optimization scheme of multiple traffic lights in the target area based on the traffic flow change trend prediction results.

[0074] Furthermore, the system is also used to implement the following functions:

[0075] Historical traffic flow data within the target area is obtained; based on the historical traffic flow data, the historical traffic flow of motor vehicles, non-motor vehicles, and pedestrians is decomposed, and a training data set is set; based on the training data set, the multi-modal traffic flow model is trained, and the traffic flow trend prediction results are dynamically updated using a sliding window mechanism.

[0076] Furthermore, the system is also used to implement the following functions:

[0077] Based on multiple traffic lights within the target area, edge computing nodes are deployed; based on the edge computing nodes, local traffic light timing schemes are generated and synchronized with the central node; by uploading multiple local traffic light timing schemes to the central node, an optimized timing scheme for multiple traffic lights within the target area is determined.

[0078] Furthermore, the system is also used to implement the following functions:

[0079] Motor vehicle traffic efficiency, non-motor vehicle traffic efficiency, and pedestrian traffic efficiency are used as joint optimization objectives. Based on real-time traffic flow data, the joint optimization objectives are dynamically weighted to obtain multiple timing adaptation schemes. The multiple timing adaptation schemes are then screened to determine the timing optimization schemes for multiple traffic lights within the target area.

[0080] Furthermore, the system is also used to implement the following functions:

[0081] Pedestrian crossing needs are introduced; based on pedestrian flow data and pedestrian crossing needs, the multiple timing adaptation schemes are prioritized and a first filtering condition is set.

[0082] Furthermore, the system is also used to implement the following functions:

[0083] High-volume areas and high-conflict areas are marked; based on the high-volume areas and high-conflict areas, key traffic flow nodes are identified, and the traffic flow characteristics corresponding to the key traffic flow nodes are determined; according to the traffic flow characteristics corresponding to the key traffic flow nodes, the multiple timing adaptation schemes are further filtered, and a second filtering condition is set.

[0084] Furthermore, the system is also used to implement the following functions:

[0085] Key traffic flow nodes are identified, including road intersections, pedestrian crossings, and non-motorized vehicle lane junctions. In high-traffic areas, potential conflicts between different modes of transportation are analyzed at these key traffic flow nodes, and a first feature vector is configured for each node. In high-conflict areas, queue lengths, waiting times, and queue dissipation speeds for vehicles and pedestrians are analyzed at these key traffic flow nodes, and a second feature vector is configured for each node. Based on the first and second feature vectors, the traffic flow characteristics corresponding to the key traffic flow nodes are determined.

[0086] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0087] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0088] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application intends to include such modifications and variations.

Claims

1. A multi-mode traffic signal optimization method driven by perception intelligence, characterized in that, The method includes: Collect traffic flow data, which includes motor vehicle traffic, non-motor vehicle traffic, and pedestrian traffic; A multimodal traffic flow model is constructed to simulate the interactive behavioral relationships among motor vehicle behavior characteristics, non-motor vehicle behavior characteristics, and pedestrian behavior characteristics. Connect multiple traffic lights within the target area to obtain street light control signals; Based on the traffic flow data and the street light control signals, the multi-modal traffic flow model is used to predict traffic flow trends. Based on the traffic flow trend prediction results, configure the timing optimization scheme of multiple traffic lights in the target area; The method for constructing a multimodal traffic flow model includes: Obtain historical traffic flow data within the target area; Based on the historical traffic flow data, the historical traffic flow of motor vehicles, non-motor vehicles, and pedestrians is decomposed, and a training data set is set up. Based on the training dataset, the multi-modal traffic flow model is trained, and the traffic flow trend prediction results are dynamically updated using a sliding window mechanism; the timing optimization schemes for multiple traffic lights within the target area are configured, the method including: Deploy edge computing nodes based on multiple traffic lights within the target area; Based on the edge computing nodes, a local traffic signal timing scheme is generated and synchronized with the central node. By uploading multiple local traffic light timing schemes to the central node, an optimized timing scheme for multiple traffic lights within a target area is determined; the method for determining the optimized timing scheme for multiple traffic lights within the target area includes: The efficiency of motor vehicle traffic, the efficiency of non-motor vehicle traffic, and the efficiency of pedestrian traffic are taken as joint optimization objectives; Based on real-time traffic flow data, the joint optimization objective is dynamically weighted to obtain multiple timing adaptation schemes. The multiple timing adaptation schemes are screened to determine the timing optimization schemes for multiple traffic lights within the target area.

2. The method as described in claim 1, characterized in that, The method involves filtering the multiple timing adaptation schemes to determine the optimal timing schemes for multiple traffic lights within the target area. Introducing the need for pedestrian crossings; Based on pedestrian traffic data and pedestrian crossing needs, the multiple timing adaptation schemes are prioritized and a first filtering condition is set.

3. The method as described in claim 2, characterized in that, The method includes: Mark high-traffic areas and high-conflict areas; Based on the high-flow and high-conflict areas, key traffic flow nodes are identified, and the traffic flow characteristics corresponding to the key traffic flow nodes are determined. Based on the traffic flow characteristics corresponding to the key traffic flow nodes, the multiple timing adaptation schemes are further screened, and a second screening condition is set.

4. The method as described in claim 3, characterized in that, The method for determining traffic flow characteristics corresponding to key traffic flow nodes further includes: Identify key traffic flow nodes, including road intersections, pedestrian crossings, and non-motorized vehicle lane junctions. At key traffic flow nodes in the high-traffic area, potential conflicts between different modes of transportation are analyzed, and a first feature vector corresponding to the key traffic flow node is configured. At key traffic flow nodes in the high-conflict area, analyze the queue length, waiting time, and queue dissipation speed of vehicles and pedestrians, and configure the second feature vector corresponding to the key traffic flow nodes. Based on the first feature vector and the second feature vector, the traffic flow characteristics corresponding to the key traffic flow nodes are determined.

5. A perception-driven, intelligent-driven multi-mode traffic signal optimization system, characterized in that, The system is used to implement the perception-intelligent driven multi-mode traffic signal optimization method according to any one of claims 1-4, the system comprising: The traffic flow data acquisition module is used to collect traffic flow data, which includes motor vehicle flow, non-motor vehicle flow, and pedestrian flow. A multimodal traffic flow model construction module is used to construct a multimodal traffic flow model, which is used to simulate the interactive relationship between motor vehicle behavior characteristics, non-motor vehicle behavior characteristics, and pedestrian behavior characteristics. The street light control signal acquisition module is used to connect to multiple traffic lights within the target area to obtain street light control signals; The traffic flow change trend prediction module is used to predict the traffic flow change trend based on the traffic flow data and the street light control signal, using the multi-mode traffic flow model. The timing optimization scheme configuration module is used to configure timing optimization schemes for multiple traffic lights within the target area based on traffic flow trend prediction results.

Images