Traffic signal light switching point prediction method and device, electronic equipment and storage medium

Abstract

The disclosure provides a traffic signal light switching point prediction method and device, electronic equipment and storage medium, relates to the technical field of data processing, specifically relates to big data, intelligent transportation and cloud computing and the like, and can be applied to the technical field of navigation, Internet of Vehicles, intelligent cockpit and automatic driving, and comprises the following steps: acquiring vehicle trajectory data of a target intersection; calculating switching point prediction correlation information of a target traffic signal light in the target intersection according to the vehicle trajectory data; wherein the target traffic signal light comprises a straight signal light and / or a non-straight signal light; the switching point prediction correlation information comprises signal light switching rule information, signal light phase difference information and signal light real-time switching point information; and predicting switching point information of the target traffic signal light according to the switching point prediction correlation information of the target traffic signal light. The embodiment of the disclosure can improve the accuracy and efficiency of traffic signal light switching point prediction.

Description

Technical Field

[0001] This disclosure relates to the field of data processing technology, specifically to technologies such as big data, intelligent transportation, and cloud computing, and can be applied to the fields of navigation, vehicle networking, smart cockpit, and autonomous driving technology. Background Technology

[0002] Traffic lights are crucial infrastructure for traffic management departments to control vehicle traffic order and adjust road traffic flow. They also represent the most complex and user-centric traffic scenario for navigation. Utilizing big data and cloud computing technologies to mine traffic light switching point information is beneficial for intelligent transportation construction. It allows for precise calculation of intersection costs, serving as auxiliary features to predict dynamic traffic conditions such as congestion and congestion dissipation. This has significant application value in various related fields, including navigation, vehicle-to-everything (V2X) technology, smart cockpits, and autonomous driving. Furthermore, displaying traffic light switching point information to users can alleviate anxiety caused by the unknown duration of red lights. Therefore, accurately identifying traffic light switching points is of paramount importance. Summary of the Invention

[0003] This disclosure provides a method, apparatus, electronic device, and storage medium for predicting traffic light switching points, which can improve the accuracy and efficiency of traffic light switching point prediction.

[0004] In a first aspect, embodiments of this disclosure provide a method for predicting traffic light switching points, including:

[0005] Obtain vehicle trajectory data at the target intersection;

[0006] The switching point prediction association information of the target traffic lights at the target intersection is calculated based on the vehicle trajectory data; wherein, the target traffic lights include straight-ahead traffic lights and / or non-straight-ahead traffic lights; the switching point prediction association information includes traffic light switching pattern information, traffic light phase difference information, and real-time traffic light switching point information;

[0007] Predict the switching point information of the target traffic light based on the switching point prediction association information.

[0008] Secondly, embodiments of this disclosure provide a traffic signal switching point prediction device, comprising:

[0009] The vehicle trajectory data acquisition module is used to acquire vehicle trajectory data at the target intersection.

[0010] The switching point prediction association information calculation module is used to calculate the switching point prediction association information of the target traffic lights at the target intersection based on the vehicle trajectory data; wherein, the target traffic lights include straight-ahead traffic lights and / or non-straight-ahead traffic lights; the switching point prediction association information includes traffic light switching pattern information, traffic light phase difference information, and real-time switching point information of traffic lights;

[0011] The switching point information prediction module is used to predict the switching point information of the target traffic light based on the switching point prediction association information of the target traffic light.

[0012] Thirdly, embodiments of this disclosure provide an electronic device, including:

[0013] At least one processor; and

[0014] A memory communicatively connected to the at least one processor; wherein,

[0015] The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the traffic light switching point prediction method provided in the first aspect embodiment.

[0016] Fourthly, embodiments of this disclosure also provide a non-transitory computer-readable storage medium storing computer instructions for causing the computer to execute the traffic light switching point prediction method provided in the first aspect embodiment.

[0017] Fifthly, this disclosure also provides a computer program product, including a computer program that, when executed by a processor, implements the traffic light switching point prediction method provided in the first aspect embodiment.

[0018] This embodiment of the disclosure acquires vehicle trajectory data for target intersections, including straight-ahead traffic lights and / or non-straight-ahead traffic lights. Based on the acquired vehicle trajectory data, it calculates switching point prediction correlation information for the target traffic lights at the target intersection, including traffic light switching patterns, traffic light phase differences, and real-time switching points. Then, based on the aforementioned switching point prediction correlation information, it predicts the switching point information of the target traffic lights, thus solving the problem of low accuracy in existing traffic light switching point prediction methods and improving the accuracy and efficiency of traffic light switching point prediction.

[0019] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this disclosure, nor is it intended to limit the scope of this disclosure. Other features of this disclosure will become readily apparent from the following description. Attached Figure Description

[0020] The accompanying drawings are provided to better understand this solution and do not constitute a limitation of this disclosure. Wherein:

[0021] Figure 1 This is a flowchart of a traffic light switching point prediction method provided in an embodiment of this disclosure;

[0022] Figure 2 This is a flowchart of a traffic light switching point prediction method provided in an embodiment of this disclosure;

[0023] Figure 3 This is a flowchart illustrating a traffic light switching point prediction method provided in an embodiment of this disclosure;

[0024] Figure 4 This is a structural diagram of a traffic signal switching point prediction device provided in an embodiment of this disclosure;

[0025] Figure 5 This is a schematic diagram of the structure of an electronic device used to implement the traffic signal switching point prediction method of the present disclosure. Detailed Implementation

[0026] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0027] During driving navigation, there is a more than 90% chance that users will not see the countdown timer for traffic lights when waiting at a traffic light. Furthermore, users are mostly unaware of the remaining time on the traffic light as they approach the intersection. This lack of information about the countdown timer at the intersection creates uncertainty and anxiety for the user, and the inability to determine whether to proceed quickly or slowly when approaching the intersection affects their psychological expectation.

[0028] By predicting the countdown timer information of traffic lights and sending this prediction to users, users gain complete knowledge of the countdown status of traffic lights, which helps improve overall road traffic efficiency and reduce traffic accidents. Furthermore, the predicted traffic light information can be used to accurately calculate the cost of passing through intersections, serving as an auxiliary feature to predict dynamic traffic conditions such as road congestion and congestion dissipation. Simultaneously, the predicted traffic light information has a positive impact on the user's driving experience, driving safety, and the surrounding environment. For example, knowing the remaining time of a red light while waiting prevents users from becoming distracted and starting late; knowing that the green light is about to end when crossing prevents users from braking suddenly or even rear-ending others; and knowing that the red light has a long time left allows users to turn off their engines while waiting at a red light, reducing air pollution and urban noise.

[0029] Understandably, in the prediction process of non-straight-ahead traffic lights, due to the influence of non-straight-ahead vehicle trajectories, both the technical implementation difficulty and prediction accuracy are currently difficult to achieve the desired results. Currently, the main methods for predicting switching points of traffic lights for different directions are as follows:

[0030] (1) The platform cooperates with the traffic signal management department to obtain the traffic signal switching times and timing patterns of certain intersections or areas through which the other party passes, and then predicts the switching times of traffic signals for different turns at the corresponding intersections based on the official traffic signal switching times and timing patterns.

[0031] (2) The countdown and other features of the traffic lights excavated in the straight direction can be shared with the traffic lights in other turning directions. For example, the switching point information of traffic lights in other turning directions such as left turn or U-turn can be predicted by using the traffic light feature information excavated in the direction.

[0032] (3) Since there are few available vehicle trajectories in non-straight directions, vehicle trajectories in different directions are applied in the same way during the actual prediction process. For example, some straight vehicle trajectories are directly used as left-turn vehicle trajectories to predict the traffic lights in the left-turn direction. We try to realize the countdown function of traffic lights for different directions by using the switching rules of the light states and the start and stop characteristics of the user's online trajectory.

[0033] However, each of the above methods has its own problems in application:

[0034] Regarding scheme (1), when the platform and relevant departments cooperate, the number of intersections and lights that can be covered is limited, resulting in significant communication and information costs, and the ability to identify erroneous information is weak. Predicting traffic light switching information through scheme (1) largely depends on mutual assistance and cooperation between the two parties, has a long cooperation cycle, involves many dependent parties, and is prone to instability and uncertainty, and is also severely affected by policy.

[0035] Regarding scheme (2), if traffic lights are installed at the same intersection for both straight and other directions, there will be phase differences between different traffic lights at the same intersection. For example, left turns may be allowed before straight traffic, and some intersections may have left-turn waiting areas. Therefore, it is unreasonable to directly share the countdown and other features of the traffic lights in the straight direction with the traffic lights in other turning directions. The accuracy of the prediction information for the switching points of traffic lights for non-straight traffic is also low.

[0036] Regarding scheme (3), it is understandable that vehicle trajectories in different directions will vary depending on the traffic light's release strategy, resulting in different vehicle trajectory performances for users turning in different directions. For example, vehicles turning left often start ahead of vehicles going straight. That is, there are fewer and more chaotic vehicle trajectories for non-straight-going traffic, and the matching time between non-straight-going vehicle trajectories and the traffic light's guidance direction (such as the left-turn time) is much shorter, which will lead to inaccurate detection of non-straight-going traffic light cycles. For example, due to the shared lanes, there are abnormal start and stop phenomena in the trajectories of vehicles traveling in multiple directions. Therefore, applying the same method to predict traffic light switching point information for vehicle trajectories traveling in different directions cannot guarantee the accuracy of switching point prediction.

[0037] In one example Figure 1 This is a flowchart of a traffic light switching point prediction method provided in this embodiment. This embodiment is applicable to situations where traffic light switching point information is predicted using multi-dimensional switching point prediction correlation information, including traffic light phase differences. This method can be executed by a traffic light switching point prediction device, which can be implemented in software and / or hardware, and is generally integrated into an electronic device. This electronic device can be a server device, used in conjunction with an in-vehicle terminal or navigation client. Correspondingly, such as... Figure 1 As shown, the method includes the following operations:

[0038] S110. Obtain vehicle trajectory data at the target intersection.

[0039] The target intersection can be any intersection whose traffic light switching point information needs to be predicted. The type of target intersection can be cross-shaped, T-shaped, etc., and there can be multiple such intersections, as long as they match traffic lights whose switching point information needs to be predicted. This embodiment does not limit the type of target intersection. Vehicle trajectory data, i.e., the trajectory data generated when a vehicle passes through the target intersection, can be offline or online trajectory data, as long as it matches the target intersection. This embodiment does not limit the trajectory type or acquisition method of the vehicle trajectory data.

[0040] In this embodiment of the disclosure, the server can utilize vehicle trajectory data collected from the target intersection to mine trajectory features, and then use these features to predict the traffic light switching points at the target intersection. Optionally, the vehicle trajectory data at the target intersection can be trajectory data uploaded to the server in real time as vehicles pass through the target intersection.

[0041] Understandably, to ensure the accuracy of switching point information prediction, the vehicle trajectory data for the target intersection can be the full trajectory data of the target intersection, that is, it can include the driving trajectories of vehicles in all directions within the target intersection. Vehicle trajectory data can be distinguished into straight-going vehicle trajectories and non-straight-going vehicle trajectories, etc.

[0042] Meanwhile, the vehicle trajectory data at the target intersection can be data collected statistically over a period of time, or data collected in real time at the current moment. When the vehicle trajectory data at the target intersection is offline data, this offline data can be locally stored data or reliable ground truth data provided by a third party, etc.

[0043] S120. Calculate the switching point prediction association information of the target traffic lights at the target intersection based on the vehicle trajectory data; wherein, the target traffic lights include straight-ahead traffic lights and / or non-straight-ahead traffic lights; the switching point prediction association information includes traffic light switching pattern information, traffic light phase difference information, and real-time traffic light switching point information.

[0044] The target traffic lights can be all traffic lights involved in the target intersection, including straight-ahead and non-straight-ahead traffic lights, or only some of the traffic lights in the target intersection, such as straight-ahead or non-straight-ahead traffic lights. It can also be some straight-ahead and some non-straight-ahead traffic lights; this embodiment does not limit the content of the target traffic lights. Straight-ahead traffic lights are those used to guide vehicles to travel straight in the target intersection, and non-straight-ahead traffic lights are those used to guide vehicles to travel non-straight in the target intersection. The switching point prediction association information can be reference information used to calculate the switching point information of the target traffic lights, and it can reflect multiple dimensions of attribute information of the target traffic lights. Traffic light switching pattern information can be, for example, the switching cycle of the traffic lights. Switching point information can be the switching time between traffic lights, such as the time when green turns red, red turns yellow, and yellow turns green. Traffic light phase difference information, also known as traffic light "offset" or "time difference," specifically refers to the difference in the start time of the green or red light of a traffic light phase at an intersection. It can be divided into two types: absolute phase difference and relative phase difference. The former refers to the difference in the start time of the green or red light between the signal phase at each intersection and the reference intersection signal phase in the control system. The latter refers to the difference in the start time of the green or red light between two adjacent intersections. Real-time traffic light switching point information can be the switching point information calculated in real-time for the target traffic light.

[0045] In this embodiment, the server can utilize the acquired vehicle trajectory data to calculate multi-dimensional prediction and correlation information for switching points, including traffic light switching patterns, phase differences, and real-time switching points at the target intersection. To ensure driving safety, the switching information can optionally be categorized into two types: red-to-green and green-to-red. The duration of green-to-red switching can include the durations of green-to-yellow and yellow-to-red switching. Correspondingly, the switching point information can include the switching times of red-to-green and green-to-red switching.

[0046] For example, regarding traffic light switching pattern information, the server can calculate the time difference between the first vehicle's start sample and the start samples of other vehicles under the target traffic light a at intersection A, based on vehicle trajectory data. Then, by analyzing the differences in the target traffic light a at different times over a day and multiple days, the server can obtain the red-green light cycle of the target traffic light a at intersection A, which serves as the traffic light switching pattern information for the target traffic light a at intersection A. Alternatively, the server can also use vehicle trajectory data to calculate the traffic light switching pattern information for all traffic lights at intersection A simultaneously.

[0047] For example, regarding the signal light phase difference information at intersection A, the server can calculate the matching signal light switching pattern information for all target traffic lights at intersection A, and then use the matching signal light switching pattern information calculated for each target traffic light to deduce the switching relationship between each target traffic light, thereby obtaining the signal light phase difference information between each target traffic light at intersection A.

[0048] For example, regarding real-time traffic light switching point information, the server can calculate the specific switching status of the relevant target traffic lights at intersection A at 09:02 based on the vehicle trajectory data associated with intersection A uploaded in real time between 09:00 and 09:05 AM. For instance, at 09:02 AM, the straight-ahead traffic light in the north-south direction at intersection A might switch from green to red. Furthermore, if most vehicles using navigation software smoothly passed light a at intersection A between 9:00 AM and 9:01 AM, and in the following half-minute, most vehicles braked and reached zero speed before light a at intersection A, a reasonable estimate of the real-time countdown information for light a can be obtained. Based on this real-time countdown information, the server can calculate the real-time traffic light switching point information.

[0049] S130. Predict the switching point information of the target traffic light based on the switching point prediction association information of the target traffic light.

[0050] Accordingly, once the server calculates the switching pattern information, phase difference information, and real-time switching point information of the target traffic lights at the target intersection, it can use this multi-dimensional switching point prediction association information to predict the switching point information of the target traffic lights at the target intersection. Optionally, the switching point information of the target traffic lights may include switching time and countdown information.

[0051] For example, if the target intersection only includes straight-ahead traffic lights, a baseline straight-ahead traffic light's switching point information for a future period can be calculated based on the traffic light switching pattern information and real-time switching point information of the traffic lights at the target intersection. Then, using the traffic light phase difference information between the other traffic lights at the target intersection and the baseline straight-ahead traffic light, the switching point information of the other traffic lights at the target intersection for a future period can be calculated based on the baseline straight-ahead traffic light's switching point information for a future period.

[0052] For example, if the target intersection includes both straight-ahead traffic lights and non-straight-ahead traffic lights, the switching point information of the straight-ahead traffic lights in the future can be calculated first from the traffic light switching pattern information and the real-time switching point information of the traffic lights at the target intersection. Then, based on the phase difference information between the straight-ahead traffic lights and the non-straight-ahead traffic lights, the switching point information of the non-straight-ahead traffic lights in the future can be calculated.

[0053] For example, the switching point information of the traffic lights in target intersection A in the future can be calculated based on the traffic light switching pattern information and the real-time switching point information of the traffic lights. Furthermore, based on the traffic light phase difference information 'a' between the traffic lights in target intersection A and target intersection B, the switching point information of the traffic lights in target intersection B in the future can be calculated by combining the traffic light phase difference information 'a' with the switching point information of the traffic lights in target intersection A in the future.

[0054] It is understandable that traffic light phase difference information can accurately reflect the switching patterns between different traffic lights. Therefore, for traffic lights with low reliability of trajectory data (such as those with fewer or more irregular trajectories), the switching point information can be predicted by combining it with that of traffic lights with more numerous and more obvious trajectories. Using the phase difference information between these two traffic lights, the switching point information of the traffic lights with low reliability of trajectory data can be calculated. For example, the phase difference information between the straight-ahead light and the left-turn light at intersection A is: the straight-ahead green light changes to the left-turn green light after 1 minute. When the switching point of the straight-ahead light at intersection A is calculated to be 10:00 (from red to green) based on the switching pattern information associated with the straight-ahead light and the real-time switching point information, the switching point of the left-turn light at intersection A can be calculated to be 10:01 (from red to green). Therefore, the above traffic light switching point prediction method, utilizing traffic light phase difference information, can effectively improve the prediction accuracy of traffic light switching point information.

[0055] Meanwhile, using traffic light phase difference information to predict traffic light switching points can also reduce the time and computational costs associated with parsing and processing vehicle trajectory data. For example, when determining the phase difference between the traffic lights at intersection A and B, the switching patterns and real-time switching points of the traffic lights at intersection A can be calculated solely based on the vehicle trajectory data. After predicting the switching points at intersection A, the phase difference between the traffic lights at intersection A and B can be added to the switching point information at intersection A to obtain the switching point information at intersection B, without needing to recalculate the switching point information at intersection B using the vehicle trajectory data. Therefore, the above-mentioned traffic light switching point prediction method, utilizing traffic light phase difference information, can effectively improve the prediction efficiency of traffic light switching point information.

[0056] This embodiment of the disclosure acquires vehicle trajectory data for target intersections, including straight-ahead traffic lights and / or non-straight-ahead traffic lights. Based on the acquired vehicle trajectory data, it calculates switching point prediction correlation information for the target traffic lights at the target intersection, including traffic light switching patterns, traffic light phase differences, and real-time switching points. Then, based on the aforementioned switching point prediction correlation information, it predicts the switching point information of the target traffic lights, thus solving the problem of low accuracy in existing traffic light switching point prediction methods and improving the accuracy and efficiency of traffic light switching point prediction.

[0057] In one example Figure 2 This is a flowchart of a traffic light switching point prediction method provided in an embodiment of this disclosure. Figure 3 This is a flowchart illustrating a traffic light switching point prediction method provided in this embodiment. Based on the technical solutions of the above embodiments, this embodiment has been optimized and improved, and provides various specific optional implementation methods for calculating the switching point prediction association information and predicting the switching point information of the target traffic light based on the switching point prediction association information of the target traffic light.

[0058] like Figure 2 A traffic signal switching point prediction method is shown, comprising:

[0059] S210. Obtain vehicle trajectory data at the target intersection.

[0060] S220. Perform data preprocessing on the vehicle trajectory data to obtain preprocessed vehicle trajectory data.

[0061] Preprocessed vehicle trajectory data refers to trajectory data that has been preprocessed to obtain more accurate, reliable, and usable data. In other words, preprocessed vehicle trajectory data is high-quality vehicle trajectory data.

[0062] In one optional embodiment of this disclosure, the step of preprocessing the vehicle trajectory data to obtain preprocessed vehicle trajectory data may include: filtering out abnormal trajectory data from the vehicle trajectory data to obtain normal vehicle trajectory data; determining target trajectory filtering conditions; and matching the normal vehicle trajectory data with the target trajectory filtering conditions to obtain the preprocessed vehicle trajectory data.

[0063] The target trajectory filtering condition can be to further filter normal vehicle trajectory data to obtain vehicle trajectory data that meets the requirements and can be used to predict traffic light switching points.

[0064] Specifically, during the preprocessing of vehicle trajectory data, abnormal trajectory data can be filtered out first to obtain normal vehicle trajectory data, thus ensuring the accuracy and reliability of the vehicle trajectory data. For example, abnormal trajectory data may include, but is not limited to, parking trajectories, non-driving trajectories, suspected red-light running trajectories, and abnormal speed trajectories. Based on the obtained normal vehicle trajectory data, further filtering can be performed according to target trajectory filtering conditions to obtain the desired vehicle trajectory data, such as vehicle trajectories that reflect the switching characteristics of the target traffic light.

[0065] By preprocessing vehicle trajectory data, a trajectory set for predicting traffic light switching points can be obtained.

[0066] In one optional embodiment of this disclosure, the target trajectory filtering conditions may include target trajectory type and target user trajectory behavior; the step of matching the normal vehicle trajectory data with the target trajectory filtering conditions to obtain the preprocessed vehicle trajectory data may include: filtering normal vehicle trajectory data including the target trajectory type as candidate vehicle trajectory data; and filtering candidate vehicle trajectory data including the target user trajectory behavior from the candidate vehicle trajectory data as the preprocessed vehicle trajectory data.

[0067] The target trajectory type can be, for example, a trajectory type that is significantly affected by the switching process of the target traffic light. For instance, based on the road network structure information and vehicle location information of the target traffic light, trajectories whose vehicle trajectory positions are within a certain distance range from the target traffic light positions can be selected as the target trajectory type. The target user's trajectory behavior can be a trajectory behavior that directly reflects the switching characteristics of the target traffic light.

[0068] In this embodiment of the disclosure, to further improve the usability of normal vehicle trajectory data, the normal vehicle trajectory data can be further filtered according to target trajectory filtering conditions. For example, target trajectory types (i.e., vehicle trajectories under specific scenarios) can be selected from the normal vehicle trajectory data as candidate vehicle trajectory data, such as vehicle trajectories within a certain range of the target intersection, to ensure a high degree of consistency between the vehicle trajectories and the guiding trend of the target traffic light at the target intersection. Furthermore, the candidate vehicle trajectory data can be further filtered according to the target trajectory filtering conditions to select vehicle trajectory data with special vehicle driving behaviors (i.e., target user trajectory behaviors) as preprocessed vehicle trajectory data, such as vehicle trajectory data showing specific parking or acceleration / deceleration behaviors at traffic lights.

[0069] After obtaining the preprocessed vehicle trajectory data after the above filtering, the quality of the preprocessed vehicle trajectory data can be evaluated, and high-quality vehicle trajectory data can be further selected as the final preprocessed vehicle trajectory data. For example, based on the preprocessed vehicle trajectory data, data uploaded by specific navigation software or data with sufficient high-quality tracking points can be selected as the final preprocessed vehicle trajectory data.

[0070] Therefore, the above-mentioned preprocessing process for vehicle trajectory data can improve the accuracy, reliability, and usability of the preprocessed vehicle trajectory data, and further improve the accuracy of traffic light switching point prediction.

[0071] S230. Extract the multi-dimensional trajectory features of the preprocessed vehicle trajectory data.

[0072] The multi-dimensional trajectory features can be trajectory-related features in multiple dimensions, as well as trajectory profiles. Optionally, trajectory-related features can include not only the features of the trajectory itself, but also user features associated with the trajectory and traffic light-related features, as long as they can reflect information related to the vehicle trajectory. This disclosure does not limit the feature types or feature content of the multi-dimensional trajectory features.

[0073] In a specific example, such as Figure 3 As shown, after the user navigation system sends vehicle trajectory data back to the server, the server can extract trajectory-related features and trajectory profiles based on the preprocessed high-quality vehicle trajectory data. For example, multi-dimensional trajectory features may include, but are not limited to, information about the relationship between the user's vehicle location and the light's location, the duration of the user's vehicle parking, different user types, different posterior trajectory directions, road attribute information, the time of the user's last trajectory passage before the light and the time of the user's first trajectory passage after the light, as well as the significant differences between the trajectory profiles and feature vectors before and after the trajectory starting point.

[0074] The information includes the relationship between the user's vehicle location and the traffic light location, which can include information on how the user's trajectory is affected by traffic light changes. The user's vehicle parking time information can include waiting at the light, the red light's affected area, and the red light duration. Different user types can include passenger car users, bus users, private car users, taxi users, truck users, users familiar with the route, users unfamiliar with the route, and user driving profiles. User driving profiles can be derived from historical big data to extract driving styles such as aggressive or conservative. Correspondingly, user type can reflect the magnitude of the deviation between user reaction time and start time. Different posterior trajectory directions can distinguish different turning information for a single traffic light. Road attribute information, such as, but not limited to, the number of lanes, road width, road speed limit level, combined with current traffic flow and road conditions, can reflect the correlation between the user's acceleration / deceleration behavior and the timing of the traffic light change. The time of the user's last trajectory before the light and the time of the user's first trajectory after the light can represent the length of the green light. The magnitude of the significant difference between the trajectory profile and feature characterization vector before and after the trajectory start point can reflect the difference between the start time and the actual red-green switching time, as well as the confidence level.

[0075] S240. Calculate the predicted association information of the switching point of the target traffic light in the target intersection based on the multi-dimensional trajectory features.

[0076] Correspondingly, after extracting feature-rich multi-dimensional trajectory features from vehicle trajectory data, the switching point prediction association information of the target traffic light at the target intersection can be calculated based on these multi-dimensional trajectory features. Because multi-dimensional trajectory features refine and distinguish trajectory features from multiple dimensions, the switching point prediction association information calculated based on these multi-dimensional trajectory features has high accuracy and reliability.

[0077] S250. Calculate the predicted association information of the switching point of the target traffic light in the target intersection based on the multi-dimensional trajectory features.

[0078] In an optional embodiment of this disclosure, the vehicle trajectory data may include offline vehicle trajectory data; the step of calculating the switching point prediction association information of the target traffic light at the target intersection based on the multi-dimensional trajectory features may include: determining a traffic light switching cycle mining model for calculating the traffic light switching pattern; inputting the multi-dimensional trajectory features matched by the offline vehicle trajectory data into the traffic light switching cycle mining model to output the traffic light switching pattern information through the traffic light switching cycle mining model; determining a phase difference information mining model for calculating the traffic light phase difference information; inputting the multi-dimensional trajectory features matched by the offline vehicle trajectory data into the phase difference information mining model to output the traffic light phase difference information through the phase difference information mining model.

[0079] Among them, the traffic light switching cycle mining model can be a model used to mine and calculate the traffic light switching pattern based on vehicle trajectory data, and the phase difference information mining model can be a model used to mine and calculate the traffic light phase difference information based on vehicle trajectory data.

[0080] In this embodiment of the disclosure, vehicle trajectory data may include two types: offline vehicle trajectory data and online vehicle trajectory data. For example... Figure 3 As shown, for offline vehicle trajectory data collected by the server, a traffic light switching cycle mining model can be used. Figure 3 This is called a periodic mining algorithm, which mines and calculates the switching patterns of each traffic light and the traffic light switching intervals for each different turn. Simultaneously, the server can also utilize phase difference information to mine model information. Figure 3 This is called the phase difference mining algorithm, which mines and calculates the phase difference information between different directions of each traffic light, such as how many seconds before the left turn signal turns green.

[0081] By utilizing traffic light switching cycle mining models and phase difference information mining models to mine and calculate offline vehicle trajectory data, the traffic light switching patterns and phase difference information of target traffic lights at target intersections can be quickly extracted, providing a calculation basis for subsequent prediction of the switching points of target traffic lights at target intersections.

[0082] In an optional embodiment of this disclosure, the vehicle trajectory data may include online vehicle trajectory data; the step of calculating the switching point prediction association information of the target traffic light at the target intersection based on the multi-dimensional trajectory features may include: determining a traffic light switching point calculation model for calculating the real-time switching point information of the traffic light; inputting the multi-dimensional trajectory features matched by the online vehicle trajectory data into the traffic light switching point calculation model, so as to output the real-time switching point information of the traffic light through the traffic light switching point calculation model.

[0083] Among them, the traffic light switching point calculation model can be a model used to mine and calculate real-time traffic light switching point information based on vehicle trajectory data.

[0084] In this embodiment of the disclosure, the online vehicle trajectory data collected by the server can be used to mine and calculate real-time traffic light switching point information using a traffic light switching point calculation model. For example... Figure 3As shown, the traffic light switching cycle mining model can also be called the turn-by-turn single-switching-point optimization algorithm. The traffic light switching cycle mining model can calculate the switching time (also called a single switching point) and confidence information of a single traffic light based on a single trajectory point. Finally, based on multiple single switching points calculated for the same single traffic light over a period of time, it generates the switching points that the single traffic light can use.

[0085] Accordingly, after calculating the real-time switching point information of traffic lights based on the traffic light switching point calculation model, this information is sent to the traffic light switching point prediction module on the server. This module then utilizes this information to offline mine the switching pattern information of individual traffic lights, allowing it to predict the future red-to-green switching time of that single traffic light. Furthermore, based on the future red-to-green switching time of a single traffic light, combined with the phase difference information between it and other individual traffic lights, the traffic light switching point prediction module can predict the future red-to-green switching times of other individual traffic lights. Moreover, based on the red-to-green switching times of each traffic light, the current light status and countdown information of each traffic light can be calculated, thus flexibly implementing the countdown function output for multi-turn traffic lights.

[0086] The above technical solution utilizes a traffic light switching cycle mining model, a phase difference information mining model, and a traffic light switching point calculation model to quickly mine and calculate the prediction and correlation information of various switching points of the target traffic lights at the target intersection from offline and online vehicle trajectory data obtained by the server, thereby improving the calculation efficiency and accuracy of the switching point prediction and correlation information.

[0087] Understandably, before using the traffic light switching cycle mining model, phase difference information mining model, and traffic light switching point calculation model to calculate the switching point prediction association information, it is necessary to train and optimize the model using ground value samples of vehicle trajectory data in advance. The model should be put into practical application only after successful training, and the required switching point prediction association information should be mined and calculated based on the collected vehicle trajectory data.

[0088] Taking the traffic light switching point calculation model as an example, optionally, trajectory features can be extracted and analyzed from the acquired vehicle trajectory sample data. This includes, but is not limited to, extracting single trajectory-related features for single switching point identification and features of other related trajectories within a short period. Feature vectors of trajectory segments before and after the vehicle's starting point can be characterized separately. During model training, deep learning algorithms, including CNN (Convolutional Neural Network) and DNN (Deep Nueral Network), can be used to learn the time difference between the actual traffic light switching time and the user vehicle's start-stop time. Simultaneously, methods such as focal-loss (a loss function that handles imbalanced sample classification) can be used to reduce the weight of excessively biased samples in the loss function, thereby reducing the usefulness of sample data that is not very useful for red-green switching identification in the model. Furthermore, to prevent overfitting, methods such as Dropout and Early stopping can be introduced to improve the model's generalization ability under conditions of sparse and noisy trajectories.

[0089] After training, the model can be used for prediction. When using the model for prediction, computation can be performed using an online GPU (graphics processing unit). To support model prediction, a real-time trajectory flow feature construction system can be built on the navigation server to perform real-time predictions. For example, during the prediction process, switching points with low confidence predicted by the traffic light switching point calculation model are filtered out, and the model is used to jointly predict multiple switching points within the same time period. This allows for joint modeling of a massive number of historical switching points from the same period, enabling a switching point prediction algorithm that is flexible and effective for both short and long time periods, including stable switching patterns and stable switching points.

[0090] S260. Based on the signal switching pattern information and real-time switching point information of the straight-ahead signal light in the target traffic signal light, predict the switching point information of the straight-ahead signal light.

[0091] Understandably, due to the large amount of vehicle trajectory data and obvious trajectory characteristics, the prediction of switching point information for straight-ahead traffic lights is usually quite accurate and efficient. Therefore, when predicting the switching point information of a target traffic light based on its associated switching point prediction information, it is possible to first identify the straight-ahead traffic lights within the target traffic light system and obtain their switching pattern information and real-time switching point information. Correspondingly, the switching point information for the straight-ahead traffic lights within the target traffic light system can be predicted first using the switching pattern information and real-time switching point information obtained through data mining and calculation.

[0092] S270. Predict the switching point information of the non-straight-ahead traffic light based on the switching point information of the straight-ahead traffic light and the phase difference information of the traffic light.

[0093] It is understandable that the phase difference between the straight-ahead traffic lights and non-straight-ahead traffic lights at a target intersection usually follows a certain pattern. Furthermore, the phase difference between traffic lights at different intersections can mostly be identified through the straight-ahead traffic lights. Therefore, once the switching point information of the straight-ahead traffic lights is predicted, the switching point information of the non-straight-ahead traffic lights can be predicted based on the switching point information of the straight-ahead traffic lights and the traffic light phase difference information. Optionally, the non-straight-ahead traffic lights can be any of the non-straight-ahead traffic lights included in the target intersection where the straight-ahead traffic light is located.

[0094] The above technical solution improves the accuracy and efficiency of predicting the switching point information of target traffic lights at the same target intersection by first predicting the switching point information of the straight-ahead traffic lights and then predicting the switching point information of the non-straight-ahead traffic lights at the same target intersection based on the switching point information of the straight-ahead traffic lights and the phase difference information of the traffic lights.

[0095] In one optional embodiment of this disclosure, the number of real-time switching point information for the same target traffic light is multiple; the method may further include: verifying the real-time switching point information for each signal light of the target traffic light; and determining the target real-time switching point information for the target traffic light based on the verification results of the real-time switching point information for each signal light of the target traffic light.

[0096] Among them, the target traffic light real-time switching point information can be the only available real-time switching point information that can be uniquely determined for the target traffic light.

[0097] It is understandable that when a traffic light switching point calculation model predicts the real-time switching point information of a target traffic light based on a single trajectory, multiple different real-time switching point information may be calculated for the same target traffic light. In this case, a relatively accurate red-green switching time can be obtained by cross-validating multiple trajectories as the target traffic light's real-time switching point information. Optionally, validating the real-time switching point information of each signal of the target traffic light can be done by using extracted trajectory features to verify whether the real-time switching point information conforms to trajectory feature patterns, etc.

[0098] For example, assuming that multiple different real-time red-to-green switching points are calculated for the same target traffic light based on different online trajectories, such as 10:00, 10:01, 9:50, 10:00, and 10:02, then 10:00 can be selected as one of the real-time red-to-green switching points for that target traffic light. Alternatively, the average of the switching points can be calculated as one of the real-time red-to-green switching points for that target traffic light. Alternatively, a model prediction algorithm can be used to comprehensively identify and determine the switching points of each red-to-green signal, considering multiple influencing factors, thereby obtaining one of the real-time red-to-green switching points for that target traffic light.

[0099] The above technical solution, when verifying the real-time switching information of multiple calculated traffic lights for the same target traffic light to determine the actual usable target traffic light, can improve the availability, reliability and accuracy of the target traffic light.

[0100] In an optional embodiment of this disclosure, the target traffic lights are traffic lights at different target intersections; the step of predicting the switching point information of the target traffic lights based on the switching point prediction association information of the target traffic lights may include: determining a reference target traffic light from each of the target traffic lights; predicting the switching point information of the target traffic lights based on the switching point prediction association information of the reference target traffic lights; wherein, the switching point prediction association information of the reference target traffic lights includes the signal switching pattern information of the reference target traffic lights, the intersection-level signal light phase difference information (i.e., the phase difference between different intersections), and the real-time switching point information of the reference target traffic lights.

[0101] The reference target traffic light can be a type of traffic light that can be used to calculate traffic light switching point information at other intersections. Intersection-level traffic light phase difference information can be the phase difference information between traffic lights at different intersections.

[0102] In this embodiment, the switching point information of traffic lights at other intersections can be predicted based on the traffic lights at one intersection, thus avoiding the need to predict the switching point information of traffic lights for all intersections using the vehicle trajectories of that intersection. Specifically, a reference target traffic light can be determined first from the target traffic lights. Optionally, the reference target traffic light can be determined using big data analysis to mine and calculate traffic lights with reference value from each target traffic light. Furthermore, the switching point information of traffic lights at different target intersections can be predicted based on the switching pattern information of the reference target traffic light, the phase difference information of intersection-level traffic lights, and the real-time switching point information of traffic lights, among other switching point prediction correlation information.

[0103] For example, the timing information of the straight-ahead traffic light at intersection A is typically used as reference information for traffic lights at other related intersections. If the timing information of the straight-ahead traffic light at intersection B needs to be configured based on the timing information of the straight-ahead traffic light at intersection A, then the straight-ahead traffic light at intersection A can be used as the reference target traffic light. Correspondingly, when calculating the phase difference information of the reference target traffic light, not only can the phase difference information of the reference target traffic light relative to other traffic lights at its corresponding intersection be calculated, but the phase difference information of the reference target traffic light with traffic lights at other intersections can also be calculated as intersection-level traffic light phase difference information. Then, using the switching pattern information of the reference target traffic light, the intersection-level traffic light phase difference information, and the real-time switching point information of the traffic lights, the switching point information of traffic lights at different target intersections can be predicted. This target intersection can be the intersection where the reference target traffic light is located, or it can be an intersection that does not include the reference target traffic light, such as the intersection preceding or following the intersection where the reference target traffic light is located.

[0104] The above technical solution predicts the traffic light switching information of other intersections by using the switching point information of traffic lights at one intersection. This not only reduces the time and computing power costs associated with parsing and processing vehicle trajectory data and improves the efficiency of traffic light switching point prediction, but also reduces the prediction calculation process of related traffic light switching point prediction information, thus improving the accuracy of traffic light switching information.

[0105] like Figure 3 As shown, when there is a conflict between the real-time traffic light switching point information calculated by the server through the traffic light switching point calculation model, the server can also identify the conflicting target traffic light and, through the built-in offline model, take the conflicting target traffic light offline, so that relevant clients such as navigation clients no longer display the switching point information of the conflicting target traffic light, thereby ensuring the accuracy of the traffic light switching point information display and improving the user experience.

[0106] S280. Determine the target vehicle that meets the conditions for issuing the target traffic signal switching point information.

[0107] The conditions for sending switching point information may include at least one of the following: the vehicle is about to pass through a fixed road segment; or it requests intersection queue shape point data, including the target traffic light, from the server. The target vehicle is the vehicle that can receive and display the switching point information of the target traffic light sent by the server.

[0108] S290. The switching point information of the target traffic light is sent to the target vehicle for display.

[0109] In this embodiment of the disclosure, when the server predicts the switching point information of the target traffic light at the target intersection, it can send the switching point information of the target traffic light to the target vehicles that meet the conditions for sending the switching point information in real time. For example, when the navigation route of the target vehicle involves the target traffic light, the switching point information of the target traffic light can be sent to the target vehicle to display the switching point information of the target traffic light on the navigation side of the target vehicle, so that the user can understand the real-time light status and switching point status of each target traffic light in the route, and provide a reference for the user's driving.

[0110] The above technical solution improves the accuracy and efficiency of traffic light switching point prediction by utilizing the phase difference information between traffic lights to predict the switching point information of traffic lights.

[0111] Understandably, applying the aforementioned traffic light switching point prediction method to relevant products can significantly improve the user experience. For example, taking navigation applications as an example, applying the traffic light switching point prediction method to navigation products can greatly improve the user's real-time navigation experience, reach millions of users daily and tens of millions of page views (PVs), reduce traffic accidents, and enhance users' overall reputation for map navigation.

[0112] The collection, storage, use, processing, transmission, provision, and disclosure of user personal information (such as user vehicle trajectory data) in this technical solution comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0113] It should be noted that any arrangement or combination of the technical features in the above embodiments also falls within the protection scope of this disclosure.

[0114] In one example Figure 4This is a structural diagram of a traffic signal switching point prediction device provided in an embodiment of this disclosure. This embodiment is applicable to situations where multi-dimensional switching point prediction correlation information, including traffic light phase differences, is used to predict traffic light switching point information. The device is implemented through software and / or hardware and is specifically configured in an electronic device. This electronic device can be a server device, used in conjunction with an in-vehicle terminal or navigation client.

[0115] like Figure 4 A traffic signal switching point prediction device 400, as shown, includes: a vehicle trajectory data acquisition module 410, a switching point prediction correlation information calculation module 420, and a switching point information prediction module 430. Among them,

[0116] The vehicle trajectory data acquisition module 410 is used to acquire vehicle trajectory data at the target intersection.

[0117] The switching point prediction association information calculation module 420 is used to calculate the switching point prediction association information of the target traffic lights at the target intersection based on the vehicle trajectory data; wherein, the target traffic lights include straight-ahead traffic lights and / or non-straight-ahead traffic lights; the switching point prediction association information includes traffic light switching pattern information, traffic light phase difference information, and real-time switching point information of traffic lights;

[0118] The switching point information prediction module 430 is used to predict the switching point information of the target traffic light based on the switching point prediction association information of the target traffic light.

[0119] This embodiment of the disclosure acquires vehicle trajectory data for target intersections, including straight-ahead traffic lights and / or non-straight-ahead traffic lights. Based on the acquired vehicle trajectory data, it calculates switching point prediction correlation information for the target traffic lights at the target intersection, including traffic light switching patterns, traffic light phase differences, and real-time switching points. Then, based on the aforementioned switching point prediction correlation information, it predicts the switching point information of the target traffic lights, thus solving the problem of low accuracy in existing traffic light switching point prediction methods and improving the accuracy and efficiency of traffic light switching point prediction.

[0120] Optionally, the switching point prediction association information calculation module 420 is further configured to: perform data preprocessing on the vehicle trajectory data to obtain preprocessed vehicle trajectory data; extract multi-dimensional trajectory features from the preprocessed vehicle trajectory data; and calculate the switching point prediction association information of the target traffic light at the target intersection based on the multi-dimensional trajectory features.

[0121] Optionally, the switching point prediction association information calculation module 420 is further configured to: filter out abnormal trajectory data from the vehicle trajectory data to obtain normal vehicle trajectory data; determine target trajectory filtering conditions; and match the normal vehicle trajectory data with the target trajectory filtering conditions to obtain the preprocessed vehicle trajectory data.

[0122] Optionally, the target trajectory filtering conditions include target trajectory type and target user trajectory behavior; the switching point prediction association information calculation module 420 is further configured to: filter normal vehicle trajectory data including the target trajectory type as candidate vehicle trajectory data; and filter candidate vehicle trajectory data including the target user trajectory behavior from the candidate vehicle trajectory data as the preprocessed vehicle trajectory data.

[0123] Optionally, the vehicle trajectory data includes offline vehicle trajectory data; the switching point prediction association information calculation module 420 is further configured to: determine a signal light switching cycle mining model for calculating the signal light switching pattern; input the multi-dimensional trajectory features matched by the offline vehicle trajectory data into the signal light switching cycle mining model, so as to output the signal light switching pattern information through the signal light switching cycle mining model; determine a phase difference information mining model for calculating the signal light phase difference information; input the multi-dimensional trajectory features matched by the offline vehicle trajectory data into the phase difference information mining model, so as to output the signal light phase difference information through the phase difference information mining model.

[0124] Optionally, the vehicle trajectory data includes online vehicle trajectory data; the switching point prediction association information calculation module 420 is further configured to: determine a signal light switching point calculation model for calculating the real-time switching point information of the traffic lights; input the multi-dimensional trajectory features matched by the online vehicle trajectory data into the signal light switching point calculation model, so as to output the real-time switching point information of the traffic lights through the signal light switching point calculation model.

[0125] Optionally, the switching point information prediction module 430 is further configured to: predict the switching point information of the straight-ahead signal light based on the signal light switching pattern information and real-time switching point information of the signal light in the target traffic signal light; and predict the switching point information of the non-straight-ahead signal light based on the switching point information of the straight-ahead signal light and the signal light phase difference information.

[0126] Optionally, the number of real-time switching point information for the same target traffic light can be multiple; the traffic light switching point prediction device further includes a target traffic light real-time switching point information determination module, used to: verify the real-time switching point information of each signal light of the target traffic light; and determine the target traffic light real-time switching point information of the target traffic light based on the verification results of the real-time switching point information of each signal light of the target traffic light.

[0127] Optionally, the target traffic lights are traffic lights at different target intersections; the switching point information prediction module 430 is further configured to: determine a reference target traffic light from each of the target traffic lights; predict the switching point information of the target traffic lights based on the switching point prediction association information of the reference target traffic lights; wherein, the switching point prediction association information of the reference target traffic lights includes the signal switching pattern information of the reference target traffic lights, the intersection-level signal light phase difference information, and the real-time switching point information of the reference target traffic lights.

[0128] Optionally, the traffic signal switching point prediction device further includes a switching point information sending module, used to: determine a target vehicle that meets the conditions for sending the switching point information of the target traffic signal; and send the switching point information of the target traffic signal to the target vehicle for display.

[0129] The traffic light switching point prediction device described above can execute the traffic light switching point prediction method provided in any embodiment of this disclosure, and has the corresponding functional modules and beneficial effects of the method. Technical details not described in detail in this embodiment can be found in the traffic light switching point prediction method provided in any embodiment of this disclosure.

[0130] Since the traffic light switching point prediction device described above is an apparatus capable of executing the traffic light switching point prediction method in the embodiments of this disclosure, those skilled in the art can understand the specific implementation methods and various variations of the traffic light switching point prediction device in this embodiment based on the traffic light switching point prediction method described in the embodiments of this disclosure. Therefore, how the traffic light switching point prediction device implements the traffic light switching point prediction method in the embodiments of this disclosure will not be described in detail here. Any apparatus used by those skilled in the art to implement the traffic light switching point prediction method in the embodiments of this disclosure falls within the scope of protection intended by this disclosure.

[0131] In one example, this disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0132] Figure 5A schematic block diagram of an example electronic device 500 that can be used to implement embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0133] like Figure 5 As shown, device 500 includes a computing unit 501, which can perform various appropriate actions and processes based on a computer program stored in read-only memory (ROM) 502 or a computer program loaded from storage unit 508 into random access memory (RAM) 503. RAM 503 may also store various programs and data required for the operation of device 500. The computing unit 501, ROM 502, and RAM 503 are interconnected via bus 504. Input / output (I / O) interface 505 is also connected to bus 504.

[0134] Multiple components in device 500 are connected to I / O interface 505, including: input unit 506, such as keyboard, mouse, etc.; output unit 507, such as various types of monitors, speakers, etc.; storage unit 508, such as disk, optical disk, etc.; and communication unit 509, such as network card, modem, wireless transceiver, etc. Communication unit 509 allows device 500 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0135] The computing unit 501 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 501 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 501 performs the various methods and processes described above, such as the traffic light switching point prediction method. For example, in some embodiments, the traffic light switching point prediction method may be implemented as a computer software program tangibly contained in a machine-readable medium, such as storage unit 508. In some embodiments, part or all of the computer program may be loaded and / or installed on device 500 via ROM 502 and / or communication unit 509. When the computer program is loaded into RAM 503 and executed by the computing unit 501, one or more steps of the traffic light switching point prediction method described above may be performed. Alternatively, in other embodiments, the computing unit 501 may be configured to perform a traffic light switching point prediction method by any other suitable means (e.g., by means of firmware).

[0136] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0137] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0138] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0139] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0140] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0141] Computer systems can include clients and servers. Clients and servers are generally geographically separated and typically interact via communication networks. The client-server relationship is established by computer programs running on the respective computers and having a client-server relationship with each other. Servers can be cloud servers, also known as cloud computing servers or cloud hosts, which are hosting products within the cloud computing service ecosystem to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability. Servers can also be servers for distributed systems or servers integrated with blockchain technology.

[0142] This embodiment of the disclosure acquires vehicle trajectory data for target intersections, including straight-ahead traffic lights and / or non-straight-ahead traffic lights. Based on the acquired vehicle trajectory data, it calculates switching point prediction correlation information for the target traffic lights at the target intersection, including traffic light switching patterns, traffic light phase differences, and real-time switching points. Then, based on the aforementioned switching point prediction correlation information, it predicts the switching point information of the target traffic lights, thus solving the problem of low accuracy in existing traffic light switching point prediction methods and improving the accuracy and efficiency of traffic light switching point prediction.

[0143] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this disclosure can be achieved, and this is not limited herein.

[0144] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A method for predicting traffic light switching points, comprising: Obtain vehicle trajectory data at the target intersection; Based on the vehicle trajectory data, the predicted association information of the switching points of the target traffic lights at the target intersection is calculated; wherein, the target traffic lights include straight-ahead traffic lights and / or non-straight-ahead traffic lights; the predicted association information of the switching points includes traffic light switching pattern information, traffic light phase difference information, and real-time switching point information of traffic lights; the traffic light phase difference information is the difference between the start time of the green light or red light of the intersection signal phase; Predict the switching point information of the target traffic light based on the switching point prediction association information; The step of predicting the switching point information of the target traffic light based on the switching point prediction association information of the target traffic light includes: Based on the signal switching pattern information and real-time switching point information of the straight-ahead signal light in the target traffic signal light, predict the switching point information of the straight-ahead signal light; The switching point information of the non-straight-ahead traffic lights is predicted based on the switching point information of the straight-ahead traffic lights and the phase difference information of the traffic lights.

2. The method according to claim 1, wherein, The step of calculating the predicted association information of the switching point of the target traffic light at the target intersection based on the vehicle trajectory data includes: The vehicle trajectory data is preprocessed to obtain preprocessed vehicle trajectory data; Extract multi-dimensional trajectory features from the preprocessed vehicle trajectory data; Based on the multi-dimensional trajectory features, the predicted association information of the switching point of the target traffic light at the target intersection is calculated.

3. The method according to claim 2, wherein, The step of preprocessing the vehicle trajectory data to obtain preprocessed vehicle trajectory data includes: Abnormal vehicle trajectory data is filtered out from the vehicle trajectory data to obtain normal vehicle trajectory data; Determine the target trajectory filtering conditions; The normal vehicle trajectory data is matched with the target trajectory filtering conditions to obtain the preprocessed vehicle trajectory data.

4. The method according to claim 3, wherein, The target trajectory filtering conditions include the target trajectory type and the target user trajectory behavior; The step of matching the normal vehicle trajectory data with the target trajectory filtering conditions to obtain the preprocessed vehicle trajectory data includes: Filter normal vehicle trajectory data, including the target trajectory type, as candidate vehicle trajectory data; The candidate vehicle trajectory data that includes the target user's trajectory behavior is selected as the preprocessed vehicle trajectory data.

5. The method according to claim 2, wherein, The vehicle trajectory data includes offline vehicle trajectory data; The step of calculating the predicted association information of the switching point of the target traffic light at the target intersection based on the multi-dimensional trajectory features includes: Determine the signal light switching cycle mining model for calculating the signal light switching pattern; The multi-dimensional trajectory features matched by the offline vehicle trajectory data are input into the traffic light switching cycle mining model, so as to output the traffic light switching pattern information through the traffic light switching cycle mining model; Determine the phase difference information mining model for calculating the phase difference information of the traffic lights; The multi-dimensional trajectory features matched by the offline vehicle trajectory data are input into the phase difference information mining model, so as to output the traffic light phase difference information through the phase difference information mining model.

6. The method according to claim 2, wherein, The vehicle trajectory data includes online vehicle trajectory data; The step of calculating the predicted association information of the switching point of the target traffic light at the target intersection based on the multi-dimensional trajectory features includes: Determine the traffic light switching point calculation model for calculating the real-time switching point information of the traffic lights; The multi-dimensional trajectory features matched by the online vehicle trajectory data are input into the traffic light switching point calculation model, so as to output the real-time switching point information of the traffic lights through the traffic light switching point calculation model.

7. The method according to claim 1, wherein, The number of real-time switching points for the same target traffic light is multiple; the method further includes: The real-time switching point information of each traffic light of the target traffic light is verified; The target traffic light's real-time switching point information is determined based on the verification results of the real-time switching point information of each traffic light in the target traffic light.

8. The method according to claim 1, wherein, The target traffic lights are traffic lights at different target intersections; the prediction of the switching point information of the target traffic lights based on the switching point prediction association information includes: Determine a reference target traffic light from among the aforementioned target traffic lights; Predict the switching point information of the target traffic light based on the switching point prediction association information of the benchmark target traffic light; The reference target traffic light switching point prediction association information includes the traffic light switching pattern information of the reference target traffic light, the intersection-level traffic light phase difference information, and the real-time switching point information of the reference target traffic light.

9. The method according to claim 1, further comprising: Identify target vehicles that meet the conditions for receiving the target traffic light switching point information; The switching point information of the target traffic light is sent to the target vehicle for display.

10. A traffic signal switching point prediction device, comprising: The vehicle trajectory data acquisition module is used to acquire vehicle trajectory data at the target intersection. The switching point prediction association information calculation module is used to calculate the switching point prediction association information of the target traffic lights at the target intersection based on the vehicle trajectory data; wherein, the target traffic lights include straight-ahead traffic lights and / or non-straight-ahead traffic lights; the switching point prediction association information includes traffic light switching pattern information, traffic light phase difference information, and real-time switching point information of traffic lights; the traffic light phase difference information is the difference between the start time of the green light or red light of the intersection signal phase; The switching point information prediction module is used to predict the switching point information of the target traffic light based on the switching point prediction association information of the target traffic light; The switching point information prediction module is further used for: Based on the signal switching pattern information and real-time switching point information of the straight-ahead signal light in the target traffic signal light, predict the switching point information of the straight-ahead signal light; The switching point information of the non-straight-ahead traffic lights is predicted based on the switching point information of the straight-ahead traffic lights and the phase difference information of the traffic lights.

11. The apparatus according to claim 10, wherein, The switching point prediction association information calculation module is also used for: The vehicle trajectory data is preprocessed to obtain preprocessed vehicle trajectory data; Extract multi-dimensional trajectory features from the preprocessed vehicle trajectory data; Based on the multi-dimensional trajectory features, the predicted association information of the switching point of the target traffic light at the target intersection is calculated.

12. The apparatus according to claim 11, wherein, The switching point prediction association information calculation module is also used for: Abnormal vehicle trajectory data is filtered out from the vehicle trajectory data to obtain normal vehicle trajectory data; Determine the target trajectory filtering conditions; The normal vehicle trajectory data is matched with the target trajectory filtering conditions to obtain the preprocessed vehicle trajectory data.

13. The apparatus according to claim 12, wherein, The target trajectory filtering conditions include target trajectory type and target user trajectory behavior; the switching point prediction association information calculation module is also used for: Filter normal vehicle trajectory data, including the target trajectory type, as candidate vehicle trajectory data; The candidate vehicle trajectory data that includes the target user's trajectory behavior is selected as the preprocessed vehicle trajectory data.

14. The apparatus according to claim 11, wherein, The vehicle trajectory data includes offline vehicle trajectory data; the switching point prediction and association information calculation module is also used for: Determine the signal light switching cycle mining model for calculating the signal light switching pattern; The multi-dimensional trajectory features matched by the offline vehicle trajectory data are input into the traffic light switching cycle mining model, so as to output the traffic light switching pattern information through the traffic light switching cycle mining model; Determine the phase difference information mining model for calculating the phase difference information of the traffic lights; The multi-dimensional trajectory features matched by the offline vehicle trajectory data are input into the phase difference information mining model, so as to output the traffic light phase difference information through the phase difference information mining model.

15. The apparatus according to claim 11, wherein, The vehicle trajectory data includes online vehicle trajectory data; the switching point prediction and association information calculation module is also used for: Determine the traffic light switching point calculation model for calculating the real-time switching point information of the traffic lights; The multi-dimensional trajectory features matched by the online vehicle trajectory data are input into the traffic light switching point calculation model, so as to output the real-time switching point information of the traffic lights through the traffic light switching point calculation model.

16. The apparatus according to claim 10, wherein, The number of real-time switching point information for the same target traffic light is multiple; the device also includes a target traffic light real-time switching point information determination module, used for: The real-time switching point information of each traffic light of the target traffic light is verified; The target traffic light's real-time switching point information is determined based on the verification results of the real-time switching point information of each traffic light in the target traffic light.

17. The apparatus according to claim 10, wherein, The target traffic lights are traffic lights at different target intersections; the switching point information prediction module is also used for: Determine a reference target traffic light from among the aforementioned target traffic lights; Predict the switching point information of the target traffic light based on the switching point prediction association information of the benchmark target traffic light; The reference target traffic light switching point prediction association information includes the traffic light switching pattern information of the reference target traffic light, the intersection-level traffic light phase difference information, and the real-time switching point information of the reference target traffic light.

18. The apparatus according to claim 10, further comprising a switching point information sending module, used for: Identify target vehicles that meet the conditions for receiving the target traffic light switching point information; The switching point information of the target traffic light is sent to the target vehicle for display.

19. An electronic device comprising: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the traffic signal switching point prediction method according to any one of claims 1-9.

20. A non-transitory computer-readable storage medium storing computer instructions for causing a computer to perform the traffic signal switching point prediction method according to any one of claims 1-9.

21. A computer program product, comprising a computer program / instructions, characterized in that, When the computer program / instruction is executed by the processor, it implements the traffic signal switching point prediction method as described in any one of claims 1-9.

Images