A link travel time prediction method and system

Abstract

The application provides a road section travel time prediction method and system, the method comprising: obtaining two continuous data points in vehicle trajectory data, and determining path travel time based on the two continuous data points; determining path free flow travel time and path delay time based on the path travel time, wherein the path free flow travel time comprises at least one road section free flow travel time corresponding to at least one road section; determining at least one road section static delay time and at least one road section dynamic delay time corresponding to the at least one road section in the path delay time based on the scene where the two continuous data points are located; determining road section delay time of each road section based on the at least one road section static delay time and the at least one road section dynamic delay time; and determining road section travel time based on the road section delay time and the road section free flow travel time. The application achieves the purpose of obtaining more accurate road section travel time of each road section through low sampling frequency vehicle trajectory data.

Description

Technical Field

[0001] This invention relates to the field of traffic condition perception technology, specifically to a method and system for predicting road segment travel time. Background Technology

[0002] Route travel time estimation provides valuable travel information for travel decisions and also assists traffic management departments in alleviating congestion and developing smart cities.

[0003] However, due to the low sampling frequency of the vehicle network, the collected travel time based on the vehicle network trajectory may not cover the entire road segment (e.g., multiple road segments are crossed between two sampling points), that is, the travel time of each road segment cannot be known.

[0004] Therefore, there is an urgent need to propose a method and system for predicting travel time on road segments to solve the above-mentioned technical problems. Summary of the Invention

[0005] In view of this, it is necessary to provide a method and system for predicting road segment travel time to solve the technical problem in the prior art that the travel time of each road segment cannot be known.

[0006] On one hand, the present invention provides a method for predicting travel time for road segments, comprising:

[0007] Obtain two consecutive data points from the vehicle trajectory data, and determine the path travel time based on the two consecutive data points;

[0008] The path free-flow travel time and path delay time are determined based on the path travel time, wherein the path free-flow travel time includes at least one segment free-flow travel time corresponding to at least one road segment;

[0009] Based on the scenario where the two consecutive data points are located, determine the static delay time and the dynamic delay time of at least one road segment corresponding to at least one road segment in the path delay time;

[0010] The road segment delay time for each road segment is determined based on the static delay time and the dynamic delay time of the at least one road segment.

[0011] The segment travel time is determined based on the segment delay time and the segment free-flow travel time.

[0012] In some possible implementations, determining the static delay time and dynamic delay time of at least one road segment corresponding to the at least one road segment in the path delay time includes:

[0013] The static delay time of the at least one road segment is determined based on the static delay calculation model;

[0014] When the first data point and the second data point of the two consecutive data points are located in two adjacent road segments and are not in the queuing area, the dynamic delay time of the first road segment is determined based on the path delay time and the static delay time of the road segment.

[0015] When the first data point is located in the upstream queuing area and the second data point is not in the queuing area, a first scaling factor is determined, and the second segment dynamic delay time is determined based on the path delay time, the segment static delay time, and the first scaling factor.

[0016] When the second data point is located in the downstream queuing area, a second scaling factor is determined, and a multimodal graph prediction model is constructed. Based on the multimodal graph prediction model, the estimated travel time of the road segment is determined, and based on the second scaling factor, the estimated travel time of the road segment, the static delay time of the road segment, and the free travel time of the road segment, the dynamic delay time of the third road segment is determined.

[0017] In some possible implementations, the static delay calculation model is as follows:

[0018]

[0019] In the formula, For road section j Static delay time of the road section; For road section j The signal period; For road section j Effective green light time; For road section j Red light time; For road section j Number of waiting periods; For road section j Real-time flow rate; For road section k The signal period; n This represents the total number of road segments.

[0020] In some possible implementations, the dynamic delay time of the first road segment is:

[0021]

[0022] In the formula, This refers to the dynamic delay time for the first road segment; The path delay time; for j Static delay time of the road segment.

[0023] In some possible implementations, the dynamic delay time of the second road segment is:

[0024]

[0025] In the formula, This refers to the dynamic delay time for the second road segment; The path delay time; for j Static delay time of the road segment; This is the first scaling factor.

[0026] In some possible implementations, the dynamic delay time of the third segment is:

[0027]

[0028]

[0029] In the formula, The dynamic delay time for the third section; The path delay time; This is the second scaling factor; The allocation coefficient; Free travel time for the route; Estimate travel time for a route segment; for j+k Static delay time of the road segment.

[0030] In some possible implementations, the multimodal graph prediction model includes a modal graph generation module, a spatiotemporal module, and a fusion module; the step of determining the estimated travel time for a road segment based on the multimodal graph prediction model includes:

[0031] Based on the modal diagram generation module, the intersection modal similarity matrix, intra-segment modal similarity matrix, and cross-modal matrix of the first and second road segments are determined;

[0032] Based on the spatiotemporal module, the spatiotemporal characteristics of the intersection modal similarity matrix, the spatiotemporal characteristics of the road segment modal similarity matrix, and the spatiotemporal characteristics of the intersection modal similarity matrix across modalities are determined.

[0033] The fusion module fuses the spatiotemporal features of the intersection, the spatiotemporal features within the road segment, and the spatiotemporal features of the intersection mode to obtain the estimated travel time of the road segment.

[0034] In some possible implementations, the modal diagram generation module includes an intersection modal diagram generation unit, a road segment modal diagram generation unit, and an intersection modal diagram generation unit; the step of determining the intersection modal similarity matrix, the road segment modal similarity matrix, and the cross-modal matrix of the first road segment and the second road segment based on the modal diagram generation module includes:

[0035] The intersection modal diagram generation unit constructs an intersection geometric complexity similarity matrix and a signal period similarity matrix, and determines the intersection modal similarity matrix based on the geometric complexity similarity matrix and the signal period similarity matrix;

[0036] Based on the modal diagram generation unit within the road segment, a congestion level similarity matrix and a road attribute similarity matrix are constructed, and the modal similarity matrix within the road segment is determined based on the congestion level similarity matrix and the road attribute similarity matrix;

[0037] The intersection modal similarity matrix and the intra-segment modal similarity matrix are combined by the intersection modal similarity matrix to obtain the spatial cross-modal matrix, and the cross-modal matrix is ​​determined based on the spatial cross-modal matrix.

[0038] In some possible implementations, the spatiotemporal module includes a first unimodal spatiotemporal unit, a second unimodal spatiotemporal unit, and a cross-modal spatiotemporal unit. The first unimodal spatiotemporal unit includes a first spatial graph convolutional layer, a first temporal attention layer, and a first normalization layer. The second unimodal spatiotemporal unit includes a second spatial graph convolutional layer, a second temporal attention layer, and a second normalization layer. The cross-modal spatiotemporal unit includes a spatiotemporal graph attention layer and a third normalization layer.

[0039] On the other hand, the present invention also provides a route travel time prediction system, comprising:

[0040] The path travel time determination unit is used to acquire two consecutive data points in the vehicle trajectory data and determine the path travel time based on the two consecutive data points.

[0041] A path travel time allocation unit is used to determine the path free-flow travel time and path delay time based on the path travel time, wherein the path free-flow travel time includes at least one road segment free-flow travel time corresponding to at least one road segment;

[0042] The path delay time classification unit is used to determine, based on the scenario where the two consecutive data points are located, at least one static delay time and at least one dynamic delay time of the path delay time corresponding to the at least one road segment.

[0043] A road segment delay time determination unit is used to determine the road segment delay time of each road segment based on the static delay time and the dynamic delay time of the at least one road segment;

[0044] A segment travel time determination unit is used to determine the segment travel time based on the segment delay time and the segment free-flow travel time.

[0045] The beneficial effects of the above embodiments are as follows: The road segment travel time prediction method provided by the present invention divides the path travel time into path free-flow travel time and path delay time, and further divides the path delay time into road segment static delay time and road segment dynamic delay time. This enables the accurate determination of road segment static delay time and path free-flow travel time using vehicle trajectory data with low sampling frequency, thereby determining a more accurate road segment dynamic delay time, and thus obtaining the road segment travel time for each road segment.

[0046] Furthermore, when determining the dynamic delay time of a road segment, the present invention considers the scenario of two consecutive data points, making the determined dynamic delay time of the road segment more consistent with the actual application scenario, thereby improving the accuracy of the determined dynamic delay time of the road segment and thus improving the accuracy of the determined travel time of the road segment. Attached Figure Description

[0047] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0048] Figure 1 A schematic flowchart of an embodiment of the road segment travel time prediction method provided by the present invention;

[0049] Figure 2 For the present invention Figure 1 A schematic diagram of an embodiment of S103;

[0050] Figure 3 This is a schematic diagram of an embodiment of the multimodal graph prediction model provided by the present invention;

[0051] Figure 4 For the present invention Figure 2 A schematic diagram of an embodiment of determining the travel time of a road segment in S204;

[0052] Figure 5 For the present invention Figure 4 A schematic diagram of an embodiment of S401;

[0053] Figure 6 A schematic diagram of an embodiment of the route travel time prediction system provided by the present invention;

[0054] Figure 7 A schematic diagram of an embodiment of the electronic device provided by the present invention. Detailed Implementation

[0055] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0056] It should be understood that the illustrative drawings are not drawn to scale. The flowcharts used in this invention illustrate operations implemented according to some embodiments of the invention. It should be understood that the operations in the flowcharts may not be implemented in sequence, and steps without logical contextual relationships may be reversed or performed simultaneously. Furthermore, those skilled in the art, guided by the content of this invention, may add one or more other operations to the flowcharts, or remove one or more operations from the flowcharts.

[0057] Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. These functional entities can be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor systems and / or microcontroller systems.

[0058] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a mutually exclusive, independent, or alternative embodiment. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0059] This invention provides a method and system for predicting road segment travel time, which will be described below.

[0060] Figure 1 This is a schematic flowchart of an embodiment of the road segment travel time prediction method provided by the present invention, as shown below. Figure 1 As shown, the methods for predicting travel time for road segments include:

[0061] S101. Obtain two consecutive data points from the vehicle trajectory data, and determine the path travel time based on the two consecutive data points;

[0062] S102. Determine the path free-flow travel time and path delay time based on the path travel time. The path free-flow travel time includes the free-flow travel time of at least one road segment corresponding to at least one road segment.

[0063] S103. Based on the scenario where two consecutive data points are located, determine the static delay time and dynamic delay time of at least one road segment corresponding to at least one road segment in the path delay time;

[0064] S104. Determine the segment delay time of each segment based on at least one segment's static delay time and at least one segment's dynamic delay time;

[0065] S105. Determine the segment travel time based on segment delay time and segment free-flow travel time.

[0066] Compared with the prior art, the road segment travel time prediction method provided in this embodiment of the invention divides the path travel time into path free-flow travel time and path delay time, and further divides the path delay time into road segment static delay time and road segment dynamic delay time. This enables the accurate determination of road segment static delay time and path free-flow travel time using vehicle trajectory data with low sampling frequency, thereby determining a more accurate road segment dynamic delay time, and thus obtaining the road segment travel time for each road segment.

[0067] Furthermore, in determining the dynamic delay time of a road segment, this embodiment of the invention considers the scenario of two consecutive data points, making the determined dynamic delay time of the road segment more consistent with the actual application scenario, thereby improving the accuracy of the determined dynamic delay time of the road segment and thus improving the accuracy of the determined travel time of the road segment.

[0068] Among them, route travel time Path free-flow travel time and route delay time The relationship between them is:

[0069]

[0070]

[0071] Among them, path delay time Including those passing through n Road delay time for each section .

[0072] Specifically, segment delay time and segment static delay time Dynamic delay time of road segment The relationship between them is:

[0073]

[0074] In some embodiments of the present invention, such as Figure 2 As shown, step S103 includes:

[0075] S201. Determine the static delay time of at least one road segment based on the static delay calculation model;

[0076] S202. When the first data point and the second data point in two consecutive data points are located in two adjacent road segments and are not in the queuing area, the dynamic delay time of the first road segment is determined based on the path delay time and the static delay time of the road segment.

[0077] S203. When the first data point is located in the upstream queuing area and the second data point is not in the queuing area, determine the first scaling factor, and determine the second segment dynamic delay time based on the path delay time, the segment static delay time and the first scaling factor.

[0078] S204. When the second data point is located in the downstream queuing area, determine the second scaling factor and construct a multimodal graph prediction model. Based on the multimodal graph prediction model, determine the estimated travel time of the road segment. Based on the second scaling factor, the estimated travel time of the road segment, the static delay time of the road segment, and the free travel time of the road segment, determine the dynamic delay time of the third road segment.

[0079] It should be noted that in step S204, when the second data point is located in the downstream queuing area, there are two situations: one is that only the second data point is located in the downstream queuing area, and the first data point is outside the queuing area; the other is that the first data point is located in the upstream queuing area, and the second data point is located in the downstream queuing area.

[0080] It should also be noted that: the first scaling factor is the proportion of the path length in the upstream queuing area, the second scaling factor is the proportion of the path length in the queuing area, when only the second data point is located in the downstream queuing area, the second scaling factor is the proportion of the path length in the downstream queuing area, and when the first data point is located in the upstream queuing area and the second data point is located in the downstream queuing area, the second scaling factor is the proportion of the path length in both the upstream and downstream queuing areas.

[0081] In this embodiment of the invention, when the second data point is located in the downstream queuing area, a multimodal graph prediction model is constructed. Based on the multimodal graph prediction model, the estimated travel time of the road segment is determined. The multimodal graph prediction model can extract more potential time correlations from the potential information between different modes, thereby improving the accuracy of the determined dynamic delay time of the third road segment, and thus improving the accuracy of the road segment travel time.

[0082] In a specific embodiment of the present invention, the static delay calculation model is as follows:

[0083]

[0084] In the formula, For road section j Static delay time of the road section; For road section j The signal period; For road section j Effective green light time; For road section j Red light time; For road section j Number of waiting periods; For road section j Real-time flow rate; For road section k The signal period; n This represents the total number of road segments.

[0085] In a specific embodiment of the present invention, the dynamic delay time of the first road segment is:

[0086]

[0087] In the formula, The dynamic delay time for the first segment; For path delay time; for j Static delay time of the road segment.

[0088] In a specific embodiment of the present invention, the dynamic delay time of the second road segment is:

[0089]

[0090] In the formula, The dynamic delay time for the second section; For path delay time; for j Static delay time of the road segment; This is the first scaling factor.

[0091] In a specific embodiment of the present invention, the dynamic delay time of the third segment is:

[0092]

[0093]

[0094] In the formula, The dynamic delay time for the third section; For path delay time; This is the second scaling factor; The allocation coefficient; Free travel time for the route; Estimate travel time for a route segment; for j+k Static delay time of the road segment.

[0095] In some embodiments of the present invention, such as Figure 3 As shown, the multimodal graph prediction model includes a modal graph generation module, a spatiotemporal module, and a fusion module; then, as... Figure 4 As shown, step S204, which determines the estimated travel time of a road segment based on a multimodal graph prediction model, includes:

[0096] S401. Based on the modal diagram generation module, determine the intersection modal similarity matrix, intra-segment modal similarity matrix, and cross-modal matrix of the first and second road segments;

[0097] S402. Based on the spatiotemporal module, determine the spatiotemporal characteristics of the intersection modal similarity matrix, the spatiotemporal characteristics of the road segment modal similarity matrix, and the spatiotemporal characteristics of the intersection modal similarity matrix across modal matrices.

[0098] S403. Based on the fusion module, the spatiotemporal features of the intersection, the spatiotemporal features within the road segment, and the spatiotemporal features of the intersection mode are fused to obtain the estimated travel time of the road segment.

[0099] The embodiments of the present invention determine the intersection modal similarity matrix, intra-segment modal similarity matrix, and cross-modal matrix of the first and second road segments based on the modal diagram generation module. It can consider the correlation of traffic attributes from multiple perspectives such as intersection modality, intra-segment modality, and cross-modality, thereby improving the accuracy of the obtained road segment estimated travel time and thus improving the accuracy of road segment travel time.

[0100] In specific embodiments of the present invention, such as Figure 3 As shown, the modal diagram generation module includes an intersection modal diagram generation unit, a road segment modal diagram generation unit, and an intersection modal diagram generation unit; then, as follows... Figure 5 As shown, step S401 includes:

[0101] S501. Construct the intersection geometric complexity similarity matrix and signal period similarity matrix based on the intersection modal diagram generation unit, and determine the intersection modal similarity matrix based on the geometric complexity similarity matrix and signal period similarity matrix;

[0102] S502. Construct a congestion level similarity matrix and a road attribute similarity matrix based on the modal diagram generation unit within the road segment, and determine the modal similarity matrix within the road segment based on the congestion level similarity matrix and the road attribute similarity matrix;

[0103] S503. The cross-modal graph generation unit combines the intersection modal similarity matrix and the road segment modal similarity matrix to obtain the spatial cross-modal matrix, and determines the cross-modal matrix based on the spatial cross-modal matrix.

[0104] In determining the modal similarity matrix of an intersection, this invention considers the similarity of geometric complexity and signal cycle, focusing on intersection-specific traffic elements, namely, intersection structural information and cycle information. When determining the intra-segment modal similarity matrix, it considers congestion level and road attributes, focusing on segment traffic elements. When determining the cross-modal matrix, it considers the dynamic interaction between the segment and the intersection. In other words, by integrating multiple traffic elements and factors, the invention constructs the intersection modal similarity matrix, the intra-segment modal similarity matrix, and the cross-modal matrix, improving the fit between these matrices and actual application scenarios, thereby enhancing the accuracy of the determined segment travel time.

[0105] In a specific embodiment of the present invention, the intersection modal similarity matrix for:

[0106]

[0107]

[0108] In the formula, The similarity matrix represents the geometric complexity of the intersections; It is a similarity matrix of the signal periods; To set a threshold; and These are the weighting coefficients; I i and I j Indicates the first i and the j An intersection.

[0109] In a specific embodiment of the present invention, the modal similarity matrix within the road segment... for:

[0110]

[0111]

[0112]

[0113]

[0114] In the formula, The similarity matrix is ​​based on the level of crowding. This is a similarity matrix of road attributes; For spacetime warp operations; for and The neighbor determination operator; s i and s j Indicates the first i and the j Each section of road, Harmony For weights, M is the implicit code matrix; This is the Hadamard product operator.

[0115] In a specific embodiment of the present invention, the cross-modal matrix for:

[0116]

[0117]

[0118]

[0119] In the formula, ( ) T This is for the transpose operation.

[0120] This invention utilizes LSTM to mine spatial cross-modal matrices. The temporal correlation allows the obtained cross-modal matrix to integrate spatial and temporal features, thereby improving the accuracy of travel time estimation for determined road segments.

[0121] In specific embodiments of the present invention, such as Figure 3 As shown, the spatiotemporal module includes a first unimodal spatiotemporal unit, a second unimodal spatiotemporal unit, and a cross-modal spatiotemporal unit. The first unimodal spatiotemporal unit includes a first spatial graph convolutional layer, a first temporal attention layer, and a first normalization layer. The second unimodal spatiotemporal unit includes a second spatial graph convolutional layer, a second temporal attention layer, and a second normalization layer. The cross-modal spatiotemporal unit includes a spatiotemporal graph attention layer and a third normalization layer.

[0122] Since the similarity matrix density of the first and second unimodal spatiotemporal units is denser, the first unimodal spatiotemporal unit includes a first spatial convolutional layer, and the second unimodal spatiotemporal unit includes a second spatial convolutional layer to solve for spatial correlation. However, the cross-modal spatiotemporal unit is used to solve for the cross graph between road segments and intersections, which inherently possesses spatial correlation; therefore, no spatial convolutional layer is included in the cross-modal spatiotemporal unit.

[0123] Furthermore, a first unimodal spatiotemporal unit is configured to include a first temporal attention layer, and a second unimodal spatiotemporal unit is configured to include a second temporal attention layer to solve for temporal correlations. This effectively captures spatiotemporal features on the dynamic spatiotemporal graph while ensuring both accuracy and computational efficiency. Since the spatiotemporal correlations between related objects in the cross-modal spatiotemporal unit are relatively sparse, a spatiotemporal graph attention layer is used to directly solve for spatiotemporal correlations.

[0124] It should be understood that the workflow and principle of the first and second single-modal spatiotemporal units are the same.

[0125] In a specific embodiment of the present invention, the calculation process of the first spatial graph convolutional layer is as follows:

[0126]

[0127] In the formula, This is the output of the first spatial graph convolutional layer; Enter the historical travel time; The coefficients are those of a polynomial of order k.

[0128] The embodiments of the present invention can integrate the information of adjacent spatial nodes by setting a first spatial graph convolutional layer, thereby improving the prediction accuracy.

[0129] In a specific embodiment of the present invention, the calculation process of the first time attention layer is as follows:

[0130]

[0131]

[0132]

[0133]

[0134] In the formula, This is the output of the attention layer at the first moment; Attention score, used to represent v i and v j The correlation; W 1 、W 2 、W 3 、b These are trainable parameters.

[0135] In a specific embodiment of the present invention, the calculation process of the spatiotemporal graph attention layer is as follows:

[0136]

[0137]

[0138]

[0139]

[0140] In the formula, This is the output of the spatiotemporal graph attention layer; These are learnable weight vectors; for V j Historical travel time of the route.

[0141] It should be noted that the specific process in step S403, which involves fusing the spatiotemporal features of the intersection, the spatiotemporal features within the road segment, and the spatiotemporal features of the intersection modality based on the fusion module to obtain the estimated travel time of the road segment, is as follows:

[0142]

[0143] In the formula, Estimate travel time for a route segment; , , These are trainable parameters; For Hadama accumulation.

[0144] To better implement the segment travel time prediction method in the embodiments of the present invention, based on the segment travel time prediction method, the embodiments of the present invention also provide a segment travel time prediction system, such as... Figure 6 As shown, the segment travel time prediction system 600 includes:

[0145] The path travel time determination unit 601 is used to acquire two consecutive data points in the vehicle trajectory data and determine the path travel time based on the two consecutive data points.

[0146] The path travel time allocation unit 602 is used to determine the path free-flow travel time and path delay time based on the path travel time, wherein the path free-flow travel time includes at least one road segment free-flow travel time corresponding to at least one road segment.

[0147] The path delay time classification unit 603 is used to determine the static delay time and dynamic delay time of at least one road segment corresponding to at least one road segment based on the scenario where two consecutive data points are located.

[0148] The road segment delay time determination unit 604 is used to determine the road segment delay time of each road segment based on at least one road segment static delay time and at least one road segment dynamic delay time;

[0149] The segment travel time determination unit 605 is used to determine the segment travel time based on the segment delay time and the segment free-flow travel time.

[0150] The road segment travel time prediction system 600 provided in the above embodiments can realize the technical solutions described in the above road segment travel time prediction method embodiments. The specific implementation principles of each module or unit can be found in the corresponding content in the above road segment travel time prediction method embodiments, and will not be repeated here.

[0151] like Figure 7 As shown, the present invention also provides an electronic device 700. The electronic device 700 includes a processor 701, a memory 702, and a display 703. Figure 7 Only some components of the electronic device 700 are shown, but it should be understood that it is not required to implement all of the components shown, and more or fewer components may be implemented instead.

[0152] In some embodiments, processor 701 may be a central processing unit (CPU), microprocessor, or other data processing chip, used to run program code stored in memory 702 or process data, such as the segment travel time prediction method of the present invention.

[0153] In some embodiments, processor 701 may be a single server or a group of servers. The server group may be centralized or distributed. In some embodiments, processor 701 may be local or remote. In some embodiments, processor 701 may be implemented on a cloud platform. In one embodiment, the cloud platform may include a private cloud, public cloud, hybrid cloud, community cloud, distributed cloud, intranet, multi-cloud, etc., or any combination thereof.

[0154] In some embodiments, memory 702 may be an internal storage unit of electronic device 700, such as a hard disk or memory of electronic device 700. In other embodiments, memory 702 may also be an external storage device of electronic device 700, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc. equipped on electronic device 700.

[0155] Furthermore, the memory 702 may include both internal storage units of the electronic device 700 and external storage devices. The memory 702 is used to store application software and various types of data installed on the electronic device 700.

[0156] In some embodiments, display 703 may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, or an OLED (Organic Light-Emitting Diode) touchscreen. Display 703 is used to display information from electronic device 700 and to display a visual user interface. Components 701-703 of electronic device 700 communicate with each other via a system bus.

[0157] In some embodiments of the present invention, when the processor 701 executes the segment travel time prediction program in the memory 702, the following steps may be performed:

[0158] Obtain two consecutive data points from the vehicle trajectory data, and determine the path travel time based on the two consecutive data points;

[0159] The path free-flow travel time and path delay time are determined based on the path travel time. The path free-flow travel time includes the free-flow travel time of at least one road segment corresponding to at least one road segment.

[0160] Based on the scenario where two consecutive data points are located, determine the static delay time and dynamic delay time of at least one road segment corresponding to at least one road segment in the path delay time;

[0161] The segment delay time for each segment is determined based on at least one segment's static delay time and at least one segment's dynamic delay time;

[0162] Segment travel time is determined based on segment delay time and segment free-flow travel time.

[0163] It should be understood that when the processor 701 executes the segment travel time prediction program in the memory 702, in addition to the functions mentioned above, it can also perform other functions, as detailed in the description of the corresponding method embodiments above.

[0164] Furthermore, the embodiments of the present invention do not specifically limit the type of electronic device 700 mentioned. Electronic device 700 can be a mobile phone, tablet computer, personal digital assistant (PDA), wearable device, laptop computer, or other portable electronic device. Exemplary embodiments of portable electronic devices include, but are not limited to, portable electronic devices running iOS, Android, Microsoft, or other operating systems. The aforementioned portable electronic device can also be other portable electronic devices, such as a laptop computer with a touch-sensitive surface (e.g., a touch panel). It should also be understood that in some other embodiments of the present invention, electronic device 700 may not be a portable electronic device, but rather a desktop computer with a touch-sensitive surface (e.g., a touch panel).

[0165] Accordingly, this application also provides a computer-readable storage medium for storing a computer-readable program or instruction. When the program or instruction is executed by a processor, it can implement the steps or functions of the road segment travel time prediction method provided in the above-described method embodiments.

[0166] Those skilled in the art will understand that all or part of the processes of the methods described in the above embodiments can be implemented by a computer program instructing related hardware (such as a processor, controller, etc.), and the computer program can be stored in a computer-readable storage medium. The computer-readable storage medium may be a disk, optical disk, read-only memory, or random access memory, etc.

[0167] The present invention has provided a detailed description of a method and system for predicting travel time on road segments. Specific examples have been used to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. At the same time, those skilled in the art will recognize that there will be changes in the specific implementation methods and application scope based on the ideas of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for predicting travel time for a road segment, characterized in that, include: Obtain two consecutive data points from the vehicle trajectory data, and determine the path travel time based on the two consecutive data points; The path free-flow travel time and path delay time are determined based on the path travel time, wherein the path free-flow travel time includes at least one segment free-flow travel time corresponding to at least one road segment; Based on the scenario where the two consecutive data points are located, determine the static delay time and the dynamic delay time of at least one road segment corresponding to at least one road segment in the path delay time; The road segment delay time is determined based on the static delay time and the dynamic delay time of the at least one road segment. The segment travel time is determined based on the segment delay time and the segment free-flow travel time. The step of determining the static delay time and dynamic delay time of at least one road segment corresponding to at least one road segment in the path delay time based on the scenario where the two consecutive data points are located includes: The static delay time of the at least one road segment is determined based on the static delay calculation model; When the first data point and the second data point of the two consecutive data points are located in two adjacent road segments and are not in the queuing area, the dynamic delay time of the first road segment is determined based on the path delay time and the static delay time of the road segment. When the first data point is located in the upstream queuing area and the second data point is not in the queuing area, a first scaling factor is determined, and the second segment dynamic delay time is determined based on the path delay time, the segment static delay time, and the first scaling factor. When the second data point is located in the downstream queuing area, a second scaling factor is determined, and a multimodal graph prediction model is constructed. Based on the multimodal graph prediction model, the estimated travel time of the road segment is determined, and based on the second scaling factor, the estimated travel time of the road segment, the static delay time of the road segment, and the free-flow travel time of the road segment, the dynamic delay time of the third road segment is determined. The multimodal graph prediction model includes a modal graph generation module, a spatiotemporal module, and a fusion module; the step of determining the estimated travel time for a road segment based on the multimodal graph prediction model includes: Based on the modal diagram generation module, the intersection modal similarity matrix, intra-segment modal similarity matrix, and cross-modal matrix of the first and second road segments are determined; Based on the spatiotemporal module, the spatiotemporal characteristics of the intersection modal similarity matrix, the spatiotemporal characteristics of the road segment modal similarity matrix, and the spatiotemporal characteristics of the intersection modal similarity matrix across modalities are determined. Based on the fusion module, the spatiotemporal features of the intersection, the spatiotemporal features within the road segment, and the spatiotemporal features of the intersection mode are fused to obtain the estimated travel time of the road segment; The modal diagram generation module includes an intersection modal diagram generation unit, a road segment modal diagram generation unit, and an intersection modal diagram generation unit; the step of determining the intersection modal similarity matrix, the road segment modal similarity matrix, and the cross-modal matrix of the first road segment and the second road segment based on the modal diagram generation module includes: The intersection modal diagram generation unit constructs an intersection geometric complexity similarity matrix and a signal period similarity matrix, and determines the intersection modal similarity matrix based on the geometric complexity similarity matrix and the signal period similarity matrix; Based on the modal diagram generation unit within the road segment, a congestion level similarity matrix and a road attribute similarity matrix are constructed, and the modal similarity matrix within the road segment is determined based on the congestion level similarity matrix and the road attribute similarity matrix; The intersection modal similarity matrix and the intra-segment modal similarity matrix are combined by the intersection modal similarity matrix to obtain the spatial cross-modal matrix, and the cross-modal matrix is ​​determined based on the spatial cross-modal matrix.

2. The method for predicting travel time for road segments according to claim 1, characterized in that, The static delay calculation model is as follows: In the formula, For road section j Static delay time of the road section; For road section j The signal period; For road section j Effective green light time; For road section j Red light time; For road section j Number of waiting periods; For road section j Real-time flow rate; For road section k The signal period; n This represents the total number of road segments; The delay time for the stated path.

3. The method for predicting travel time for road segments according to claim 1, characterized in that, The dynamic delay time for the first road segment is: In the formula, This refers to the dynamic delay time for the first road segment; The path delay time; for j Static delay time of the road segment.

4. The method for predicting travel time of a road segment according to claim 1, characterized in that, The dynamic delay time for the second segment is: In the formula, This refers to the dynamic delay time for the second road segment; The path delay time; for j Static delay time of the road segment; This is the first scaling factor.

5. The method for predicting travel time for road segments according to claim 1, characterized in that, The dynamic delay time for the third road segment is: In the formula, The dynamic delay time for the third section; The path delay time; This is the second scaling factor; The allocation coefficient; Free-flow travel time for the road segment; Estimate travel time for a route segment; for j+k Static delay time of the road segment.

6. The method for predicting travel time of a road segment according to claim 1, characterized in that, The spatiotemporal module includes a first unimodal spatiotemporal unit, a second unimodal spatiotemporal unit, and a cross-modal spatiotemporal unit. The first unimodal spatiotemporal unit includes a first spatial graph convolutional layer, a first temporal attention layer, and a first normalization layer. The second unimodal spatiotemporal unit includes a second spatial graph convolutional layer, a second temporal attention layer, and a second normalization layer. The cross-modal spatiotemporal unit includes a spatiotemporal graph attention layer and a third normalization layer.

7. A route travel time prediction system, characterized in that, The system applicable to the segment travel time prediction method according to any one of claims 1-6, the system comprising: The path travel time determination unit is used to acquire two consecutive data points in the vehicle trajectory data and determine the path travel time based on the two consecutive data points. A path travel time allocation unit is used to determine the path free-flow travel time and path delay time based on the path travel time, wherein the path free-flow travel time includes at least one road segment free-flow travel time corresponding to at least one road segment; The path delay time classification unit is used to determine, based on the scenario where the two consecutive data points are located, at least one static delay time and at least one dynamic delay time of the path delay time corresponding to the at least one road segment. A road segment delay time determination unit is used to determine the road segment delay time of each road segment based on the static delay time and the dynamic delay time of the at least one road segment; A segment travel time determination unit is used to determine the segment travel time based on the segment delay time and the segment free-flow travel time.

Images