CN111862595B - Speed prediction method, system, medium and device based on road network topological relation - Google Patents
Speed prediction method, system, medium and device based on road network topological relation Download PDFInfo
- Publication number
- CN111862595B CN111862595B CN202010513633.8A CN202010513633A CN111862595B CN 111862595 B CN111862595 B CN 111862595B CN 202010513633 A CN202010513633 A CN 202010513633A CN 111862595 B CN111862595 B CN 111862595B
- Authority
- CN
- China
- Prior art keywords
- road
- speed
- seed
- section
- road network
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 58
- 230000008859 change Effects 0.000 claims abstract description 19
- 230000006870 function Effects 0.000 claims description 14
- 238000004590 computer program Methods 0.000 claims description 13
- 238000007781 pre-processing Methods 0.000 claims description 7
- 239000004480 active ingredient Substances 0.000 claims description 6
- 238000012549 training Methods 0.000 claims description 5
- 239000003550 marker Substances 0.000 claims description 4
- 238000012163 sequencing technique Methods 0.000 claims description 4
- 238000011144 upstream manufacturing Methods 0.000 abstract description 7
- 238000012545 processing Methods 0.000 description 11
- 238000004891 communication Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 6
- 230000008569 process Effects 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 3
- 229910052739 hydrogen Inorganic materials 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 230000002159 abnormal effect Effects 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 206010039203 Road traffic accident Diseases 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 235000019800 disodium phosphate Nutrition 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000011478 gradient descent method Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
- G08G1/0125—Traffic data processing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/10—Geometric CAD
- G06F30/18—Network design, e.g. design based on topological or interconnect aspects of utility systems, piping, heating ventilation air conditioning [HVAC] or cabling
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
- G08G1/0137—Measuring and analyzing of parameters relative to traffic conditions for specific applications
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Geometry (AREA)
- Analytical Chemistry (AREA)
- Theoretical Computer Science (AREA)
- Chemical & Material Sciences (AREA)
- Mathematical Analysis (AREA)
- Computational Mathematics (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Computer Hardware Design (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Traffic Control Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a speed prediction method, a speed prediction system, a speed prediction medium and speed prediction equipment based on a road network topological relation, wherein the speed prediction method based on the road network topological relation comprises the following steps: establishing road section topology information for the road network of the selected area according to the end point information of all road sections in the road network; selecting a seed section and a non-seed section in a road network; the seed road sections are a plurality of road sections with the highest contribution value to the whole prediction in the road network; carrying out weight learning on the first-order prediction model of the non-seed road section in the road network to predict a speed change difference value; and overall predicting the speed value of the road network. On the basis of the existing two-step prediction model for predicting the speed by using the traffic tendency, the method predicts the speed of the non-seed road section by using the idea of predicting the speed of the seed road section, measures the speed influence degree of the adjacent road sections by using the topological property among the road sections and the historical data, and predicts the speed change value according to the assumption that the speed change value of each road section is determined by the upstream road section.
Description
Technical Field
The invention belongs to the technical field of intelligent traffic, and relates to a prediction method and a prediction system, in particular to a speed prediction method, a speed prediction system, a speed prediction medium and speed prediction equipment based on a road network topological relation.
Background
With the increasing production and use of private cars, traffic congestion and traffic accidents have become ubiquitous phenomena in urban traffic. And the traffic control and the timely and accurate path planning for vehicles can be implemented through road network data, real-time road condition information and the like provided by the intelligent traffic system, so that the congestion of urban traffic is relieved, and the public traffic safety is guaranteed. The traffic speed prediction can provide real-time speed basis on urban roads, and help an intelligent traffic system to carry out traffic control and plan a travel path.
Traffic speed predictions include predictions in the time dimension and predictions in the spatial dimension. The prediction of the time dimension is to predict the future speed by using the speed data of the same road section in the previous time spans, and the prediction of the space dimension is to predict the speeds of other road sections by using the speed information of the current known road section. The time-dimension speed prediction methods are more, and include time series regression, machine learning prediction, deep learning prediction and other methods, and among them, deep learning methods such as DBN, CNN and LSTM become the hot point of speed prediction research. For the whole road network, speed acquisition of urban roads is lost frequently, the cost for acquiring the speeds of all road sections in real time is high, and the speed prediction of the spatial dimension can predict the speeds of other road sections based on partial road section data.
Therefore, how to provide a speed prediction method, system, medium and device based on road network topological relation to solve the defect that the prior art cannot predict the speed of other road segments based on partial road segment data and the like is a technical problem to be solved urgently by those skilled in the art.
Disclosure of Invention
In view of the above-mentioned shortcomings of the prior art, the present invention provides a speed prediction method, system, medium and device based on road network topological relation, which is used to solve the problem that the prior art cannot predict the speed of other road segments based on partial road segment data.
To achieve the above and other related objects, an aspect of the present invention provides a speed prediction method based on road network topological relation, for improving a first-order prediction model of a single road segment; the speed prediction method based on the road network topological relation comprises the following steps: establishing road section topology information for the road network of the selected area according to the end point information of all road sections in the road network; selecting a seed road section and a non-seed road section in the road network; the seed road sections are a plurality of road sections with the highest contribution value to the overall prediction in the road network; carrying out weight learning on the first-order prediction models of all non-seed road sections in the road network to predict a speed change difference value; and overall predicting the speed value of the road network.
In an embodiment of the present invention, between the step of establishing topology information of all road segments, the speed prediction method based on road network topology relation further includes preprocessing historical speed data in a historical time period in the road network of the selected area.
In an embodiment of the invention, the overall prediction contribution value includes a support feature contribution value for representing the number of seeds in the first-order prediction model and an overlay feature contribution value for representing the number of non-seed road segments adjacent to the seed road segment.
In an embodiment of the present invention, the step of selecting the seed road segment and the non-seed road segment in the road network includes: calculating a weighted sum of the support feature contribution value and the coverage feature contribution value of each road segment in the road network; determining the seed section by the maximum weighted sum; and the rest road sections except the seed road section in the road network are non-seed road sections.
In an embodiment of the present invention, after the step of selecting the seed road segments and the non-seed road segments in the road network is performed, the speed prediction method based on the road network topological relation further includes: the marker of the selected seed section is updated to a known marker, and the estimation level thereof is updated to 0.
In an embodiment of the present invention, the step of learning the weight of the first-order prediction models of all non-seed road segments in the road network includes: calculating an objective function according to historical speed data of the road section; the objective function is used for representing the mean value of the difference value square sum of the speed difference real value and the speed difference estimated value of the road section in a training time period; updating a weight value used for representing influence degree between road sections according to the target function and the gradient descent mode so as to realize weight value learning of a first-order prediction model of all non-seed road sections in the road network; and calculating the speed difference value estimated value of the seed road section according to the updated weight value so as to obtain the speed value of the seed road section.
In an embodiment of the present invention, the step of overall predicting the speed value of the road network includes: marking the non-seed road section as an unknown mark, and calculating a prediction index in a first-order prediction model; the prediction index is associated with an estimated level of an adjacent road segment within the model; and sequencing the non-seed road sections according to the prediction indexes, and updating the speed difference value of the sequenced non-seed road sections to predict the speed value of the non-seed road sections.
The invention provides a speed prediction system based on road network topological relation, which is used for improving a first-order prediction model of a single road section; the speed prediction system based on the road network topological relation comprises: the topology establishing module is used for establishing the topology information of the road sections for the road network of the selected area according to the endpoint information of all the road sections in the road network; the selection module is used for selecting a seed road section and a non-seed road section in the road network; the seed road sections are a plurality of road sections with the highest contribution value to the overall prediction in the road network; the learning module is used for learning the weight of the first-order prediction models of all non-seed road sections in the road network so as to predict the speed change difference; and the prediction module is used for integrally predicting the speed value of the road network.
Yet another aspect of the present invention provides a medium, on which a computer program is stored, which when executed by a processor implements the method for predicting speed based on road network topological relation.
A final aspect of the invention provides an apparatus comprising: a processor and a memory; the memory is used for storing computer programs, and the processor is used for executing the computer programs stored by the memory so as to enable the equipment to execute the speed prediction method based on the road network topological relation.
As described above, the speed prediction method, system, medium and apparatus based on road network topological relation according to the present invention have the following advantages:
the speed prediction method, the system, the medium and the equipment based on the road network topological relation reconstruct an original model on the basis of the existing two-step prediction model for predicting the speed by utilizing the traffic tendency, continue the idea of predicting the speed of a non-seed road section through a seed road section, utilize the topology among the road sections in the road network, measure the speed influence degree of adjacent road sections through historical data, and predict and obtain the speed change value according to the assumption that the speed change value of each road section is determined by the upstream road section of the road section. In addition, the invention also provides a thought of overall prediction of the road network, can predict all unknown road section speeds of the road network, and solves the problem of prediction deadlock in the overall prediction process.
Drawings
Fig. 1 shows a schematic diagram of a first-order prediction model of a single road segment in the prior art.
Fig. 2 is a flowchart illustrating a speed prediction method based on road network topology according to an embodiment of the present invention.
Fig. 3 shows a schematic diagram of the established topology of the present invention.
Fig. 4 is a schematic flow chart of S25 in the method for predicting speed based on road network topology relationship according to the present invention.
FIG. 5 is a schematic diagram illustrating deadlock prediction according to the present invention.
Fig. 6 is a schematic structural diagram of a speed prediction system based on road network topology according to an embodiment of the present invention.
Description of the element reference numerals
Detailed Description
The embodiments of the present invention are described below with reference to specific embodiments, and other advantages and effects of the present invention will be easily understood by those skilled in the art from the disclosure of the present specification. The invention is capable of other and different embodiments and of being practiced or of being carried out in various ways, and its several details are capable of modification in various respects, all without departing from the spirit and scope of the present invention. It is to be noted that the features in the following embodiments and examples may be combined with each other without conflict.
It should be noted that the drawings provided in the following embodiments are only for illustrating the basic idea of the present invention, and the components related to the present invention are only shown in the drawings rather than drawn according to the number, shape and size of the components in actual implementation, and the type, quantity and proportion of the components in actual implementation may be changed freely, and the layout of the components may be more complicated.
The technical principles of the speed prediction method, the speed prediction system, the speed prediction medium and the speed prediction equipment based on the road network topological relation are as follows:
the invention discovers that the adjacent road sections have similar traffic change trends through the speed data, reconstructs the original model on the basis of the existing two-step prediction model for predicting the speed by utilizing the traffic trend and continues the idea of predicting the speed of the non-seed road section through the seed road section. In addition, the model also provides the idea of overall prediction of the road network, all unknown road section speeds of the road network can be predicted, and the problem of prediction deadlock in the overall prediction process is solved.
Example one
The embodiment provides a speed prediction method based on a road network topological relation, which is used for improving a first-order prediction model of a single road section; the speed prediction method based on the road network topological relation comprises the following steps:
establishing road section topology information for the road network of the selected area according to the end point information of all road sections in the road network;
selecting a seed road section and a non-seed road section in the road network; the seed road sections are a plurality of road sections with the highest contribution value to the overall prediction in the road network;
carrying out weight learning on the first-order prediction models of all non-seed road sections in the road network to predict a speed change difference value;
and overall predicting the speed value of the road network.
The speed prediction method based on road network topological relation provided by the present embodiment will be described in detail with reference to the drawings. The speed prediction method based on the road network topological relation is used for improving a first-order prediction model of a single road section shown in figure 1, designing a model which is more suitable for actual traffic characteristics and more reasonable, and expanding the prediction model of each road section to the whole road network to realize the whole speed prediction of the road network. As shown in fig. 1, the neighboring road segments of the road segment F are the road segments A, D, H, which have linear correlation and are independent of each other in influence on the speed change of the road segment F, and the first-order prediction model of the single road segment has the following settings:
(1) the model only comprises a central road section and a neighbor road section, and opposite road sections of the same road do not belong to the neighbor road sections;
(2) the influence of the neighbour section on the central section is calculated as the difference of the speed with respect to the average speed, the speed difference being defined as
(3) The neighboring road segments of each road segment can have speed influence on the road segment, and the influence factor is a correlation coefficient, and the correlation coefficient is defined as the expression (1):
where i, j represents two link numbers, and days _ count represents the number of days of history data, and represents two link xiAnd xjIs representative of two road sections xiAnd xjThe average speed of the corresponding histories is simultaneously greater than or equal to or simultaneously less than the average speed of the corresponding histories in different days of the same time period t
The ratio of the number of the active ingredients to the total amount of the active ingredients,
denotes xiRoad section and xjThe road segments are also greater than or equal to the number of days of their historical average speed,
indicating the number of days that the two road segments are simultaneously less than their historical average speed.
(4) To reduce error propagation, an estimation level L is defined, a propagation factor t, L of the seed segment is 0, L of the center segment is max (L of the neighbor segment) +1, and the influence degree of the neighbor segment is exponential to t.
From this, a speed difference estimate for the road segment x may be obtained
Is expressed as formula (2):
wherein,
representing neighbour road sections xjEstimate of the velocity difference, wjDenotes the weight of the influence on x, corr (x, x)j) Representing road sections x and xjCorrelation coefficient of (1), LjRepresenting a road section xjThe estimated level of (c).
The flow of the overall prediction of the speed information of the road network based on the speed prediction method based on the road network topological relation in this embodiment is shown in fig. 2, and the speed prediction method based on the road network topological relation specifically includes the following steps:
s21, the historical speed data in the historical time period in the road network of the selected area is preprocessed.
In the present embodiment, the preprocessing of the historical speed data includes culling abnormal value data and filling missing value data.
For example, the speed value of the road section is not in the normal range, and the speed is cleared to 0 if the speed value is (5 km/h-100 km/h). And for the originally unrecorded data or the data (missing data) with the abnormal values removed from the front part and 0, filling by adopting a historical average filling method so as to ensure the data integrity of all road sections.
And S22, establishing road section topology information for the road network of the selected area according to the end point information of all road sections in the road network.
In this embodiment, with a known driving direction as a reference (for example, china drives right), topology information of all road segments is established, for example, three upstream road segments are provided at a road segment at an intersection, which represent three situations of straight driving, left turning and right turning, in this embodiment, a u-turn road segment is not considered, specifically, as shown in the topology diagram of fig. 3, two reverse road segments are provided in each of four directions, and for the road segment F, the upstream road segment, which is the traffic source of the road segment F, includes the road segment A, D, H, so that the road segment F can be predicted by using the prediction model of fig. 1.
And S23, selecting a seed road segment and a non-seed road segment in the road network. In this embodiment, the seed road segments are some road segments in the road network that have the highest contribution value to the overall prediction. The overall prediction contribution value includes a support feature contribution value representing a number of seeds in the first-order prediction model and an overlay feature contribution value representing a number of non-seed segments adjacent to the seed segment.
Specifically, it appears as a support feature contribution value in terms of non-seed segments, i.e., the number of seeds in the first-order prediction model, SUP is expressed by formula (3):
SUP(x)＝|A1(x) N-gate S | formula (3)
Expressed as an overlay characteristic contribution value, i.e., the number of non-seed segments adjacent to a seed segment, at the angle of the seed segment, COV is expressed by equations (4-1) and (4-2):
A-1(s)＝{xi|s∈A(xi) Equation (4-1)
COV(S)＝|∪s∈SA-1(s) | formula (4-2)
Wherein, subscript 1 represents that the neighbor road segment and the center road segment have only one step, S represents the seed set of the road network, A-1(s) represents a set of first order model center segments with neighboring seeds s, | · | represents a count notation, sup (x) represents the number of supports in a single model, for the number of supports in the entire road network
When sup (x) is 0 when there is no seed in the adjacent section of x, the support number of the whole road network can be expressed as
As shown in fig. 3, black arrows represent seed segments and white arrows represent non-seed segments. For the road segment F, the first-order prediction model thereof includes road segments A, D and H, where a and H are both seeds, so the supporting feature contribution value of the road segment F is 2, and similarly, the supporting feature contribution values of the road segment B, C, E, G are 2, 0, and 2, respectively. Therefore, SUP of the local road network is the sum of the support feature contribution values of the non-seed road segments 10; for the coverage feature contribution value, a set of coverage non-seed road segments of each seed road segment is obtained, the seed road segment a is in the first-order prediction model of the non-seed road segment C, F, G, so the coverage set of the seed road segment a is { C, F, G }, and the coverage sets of the seed road segments D and H are { B, F, G } and { B, C, F }. Therefore, if the coverage set of all the seed segments is { B, C, F, G }, the COV of the local road network is 4. In this embodiment, the seeds are selected such that the larger the number of supports and the number of coverages of the road network, the better.
Specifically, the S23 includes the following steps:
in the road network, calculating a weighted sum of the support feature contribution value and the coverage feature contribution value of each road segment x.
The calculation formula of the weighted sum of the support feature contribution value and the coverage feature contribution value of each road segment x is shown in formula (5):
where β represents the influence ratio of the support number. The seed selection problem can be solved by a greedy algorithm, and a road section which enables the increment of the support characteristic contribution value and the coverage characteristic contribution value of the whole road network to be maximum is selected and added into the seed road section set each time until a preset seed number (a proportion number of 10-20%) is reached, namely the goal of adding the jth seed is maximized. The target maximization of the jth seed is expressed as formula (6):
determining the seed section by calculating a maximum weighted sum by formula (5); and the rest road sections except the seed road section in the road network are non-seed road sections.
And S24, updating the mark of the selected seed road segment to a known mark, and updating the estimation level to 0, wherein the mark represents the road segment which is known without the iteration of the prediction process. In this embodiment, the estimated level of the seed is 0, because the prediction algorithm is not needed for estimation, and the estimated level of the road segment that needs one iteration of prediction is 1, and the estimated level of the road segment that needs two iterations is 2.
And S25, performing weight learning on the first-order prediction models of all non-seed road sections in the road network to predict the speed change difference.
Please refer to fig. 4, which shows a flowchart of S25. As shown in fig. 4, the S25 specifically includes the following steps:
s251, calculating an objective function according to historical speed data of the road section; the objective function is used for representing the mean value of the difference value square sum of the speed difference value real value and the speed difference value estimated value of the road section in a training time period.
The objective function is shown in equation (7):
where w represents the set of ownership values in the first order predictive model centered on the road segment x, and N represents the number of days in the training set.
And S252, updating the weight value used for representing the influence degree between the road sections according to the objective function and the gradient descent mode so as to realize weight value learning of the first-order prediction model of all non-seed road sections in the road network.
In this embodiment, the weight value w is solved by using a gradient descent methodjIs as in equation (8):
integrated velocity difference estimation
In the formula (2), α represents the gradient decreasing rate, and the weight wjIs shown in equation (9):
s253, calculating the speed difference value estimated value of the seed road section according to the updated weight value
To obtain the speed value of the seed section
S26, the speed values of the road network are predicted as a whole.
In this embodiment, the S26 includes:
marking the non-seed road section as an unknown mark, and calculating a prediction index in a first-order prediction model; the prediction index is associated with an estimated level of an adjacent road segment within the model.
In this embodiment, after weight learning is completed, the non-seed road segments may be predicted, but each first-order prediction model does not include a seed road segment, and under the condition that the seed road segments are limited, the whole prediction of the road network needs to be planned.
Firstly, marking a seed road section and a non-seed road section respectively, marking the seed road section as known and the non-seed road section as unknown, then sequencing the unknown road sections according to a prediction index E in a corresponding model, wherein the calculation of the E has a relation with the estimation grade of the adjacent road sections in the model, and the specific calculation formula of the specific prediction index E is as follows:
the non-seed links are ranked according to the prediction indexes, and the speed difference values of the ranked non-seed links are updated to predict the speed values of the non-seed links, as shown in fig. 3, the first-order prediction model of the link F includes the seed link A, D, H, so its prediction index E is 1+1+1 is 3, and compared to other non-seed links, the first-order prediction models thereof have at least 1 non-seed link neighbor, i.e., the estimation level >0, and the prediction indexes decrease as the estimation level increases, so their prediction indexes are necessarily less than 3, and thus, the link F is predicted first.
In practice, a situation in which the prediction is stalled (i.e., the number of predicted road segments for two adjacent iterations is not changed) may occur because when there are unknown road segments in the model, if there are unknown road segments, the estimation of the next predicted exponential order road segment is skipped, and finally, the loop is repeated several times, which may solve such a situation. However, a deadlock may be generated due to the interdependence between the road segments, as shown in fig. 5, where the road segment I, J, K, L is a non-seed road segment, and the prediction of each road segment depends on the upstream road segment to complete the prediction first, the predictions of the four road segments are forced to be stalled, and the deadlock requires that a certain road segment is predicted first through special processing, so that the situation of prediction stalling can be solved.
The prediction stagnation is expressed in that the number of the unknown road sections left after two adjacent prediction cycles is the same, the unknown road sections under the stagnation condition are sorted according to the number marked as the known road sections in the model, because the less the known road sections in the model are, the smaller the prediction influence of the first-order prediction model on the unknown road sections is, the road sections arranged at the top are processed according to the independent road sections, namely the speed average value of the road sections is used as the predicted value, the speed difference value prediction is set to be 0 and the updating mark is set, and then the prediction is continued in the previous prediction cycle.
The speed prediction method based on the road network topological relation reconstructs an original model and continues the idea of predicting the speed of a non-seed road section through a seed road section on the basis of an existing two-step prediction model for predicting the speed by using a traffic trend, the topological property among the road sections in the road network is used, the speed influence degree of adjacent road sections is measured through historical data, and the speed change value is predicted according to the assumption that the speed change value of each road section is determined by the upstream road section of the road section. In addition, the embodiment also provides a thought of overall prediction of the road network, all unknown road section speeds of the road network can be predicted, and the problem of prediction deadlock in the overall prediction process is solved.
The present embodiment also provides a medium (also referred to as a computer-readable storage medium) on which a computer program is stored, wherein the computer program is executed by a processor to implement the above speed prediction method based on road network topological relation.
One of ordinary skill in the art will appreciate that the computer-readable storage medium is: all or part of the steps for implementing the above method embodiments may be performed by hardware associated with a computer program. The aforementioned computer program may be stored in a computer readable storage medium. When executed, the program performs steps comprising the method embodiments described above; and the aforementioned storage medium includes: various media that can store program codes, such as ROM, RAM, magnetic or optical disks.
The embodiment further provides a speed prediction system based on road network topological relation, which is used for improving a first-order prediction model of a single road section; the speed prediction system based on the road network topological relation comprises:
the topology establishing module is used for establishing the topology information of the road sections for the road network of the selected area according to the endpoint information of all the road sections in the road network;
the selection module is used for selecting a seed road section and a non-seed road section in the road network; the seed road sections are a plurality of road sections with the highest contribution value to the overall prediction in the road network;
the learning module is used for learning the weight of the first-order prediction models of all non-seed road sections in the road network so as to predict the speed change difference;
and the prediction module is used for integrally predicting the speed value of the road network.
The speed prediction system based on the road network topological relation provided by the embodiment will be described in detail below by integrating the figures. Please refer to fig. 6, which is a schematic structural diagram of a speed prediction system based on road network topology relationship in an embodiment. As shown in fig. 6, the speed prediction system 6 based on road network topological relation includes a preprocessing module 61, a topology establishing module 62, a selecting module 63, an updating module 64, a learning module 65 and a prediction module 66.
The preprocessing module 61 is configured to preprocess historical speed data in a historical time period in a road network of a selected area.
In this embodiment, the preprocessing module 61 preprocesses the historical speed data including culling outlier data and filling missing value data.
The topology establishing module 62 coupled to the preprocessing module 61 is configured to establish topology information of road segments for the road network in the selected area according to the end point information of all road segments in the road network.
A selection module 63 coupled to the topology establishment module 62 is configured to select seed road segments and non-seed road segments in the road network. In this embodiment, the seed road segments are some road segments in the road network that have the highest contribution value to the overall prediction. The overall prediction contribution value includes a support feature contribution value representing a number of seeds in the first-order prediction model and an overlay feature contribution value representing a number of non-seed segments adjacent to the seed segment.
Specifically, the selection module 63 is configured to calculate a weighted sum of the support feature contribution value and the coverage feature contribution value of each road segment in the road network; determining the seed section by the maximum weighted sum; and the rest road sections except the seed road section in the road network are non-seed road sections.
An updating module 64 coupled to the topology establishment module 62 and the selection module 63, respectively, is configured to update the label of the selected seed road segment to a known label, the estimation level of which is updated to 0.
The selection module 63, the update module 64 and the learning module 65 are respectively used for learning the weight of the first-order prediction model of all non-seed road segments in the road network to predict the speed variation difference.
Specifically, the learning module 65 is configured to calculate an objective function according to historical speed data of the road segment; the objective function is used for representing the mean value of the difference value square sum of the speed difference real value and the speed difference estimated value of the road section in a training time period; updating a weight value used for representing influence degree between road sections according to the target function and the gradient descent mode so as to realize weight value learning of a first-order prediction model of all non-seed road sections in the road network; and calculating the speed difference value estimated value of the seed road section according to the updated weight value so as to obtain the speed value of the seed road section.
A prediction module 66 coupled to the learning module 65 is adapted to predict the velocity values of the road network as a whole.
Specifically, the prediction module 66 marks the non-seed road segment as an unknown mark, and calculates a prediction index in the first-order prediction model; the prediction index is associated with an estimated level of an adjacent road segment within the model; and sequencing the non-seed road sections according to the prediction indexes, and updating the speed difference value of the sequenced non-seed road sections to predict the speed value of the non-seed road sections.
It should be noted that the division of the modules of the above system is only a logical division, and the actual implementation may be wholly or partially integrated into one physical entity, or may be physically separated. And the modules can be realized in a form that all software is called by the processing element, or in a form that all the modules are realized in a form that all the modules are called by the processing element, or in a form that part of the modules are called by the hardware. For example: the x module can be a separately established processing element, and can also be integrated in a certain chip of the system. In addition, the x-module may be stored in the memory of the system in the form of program codes, and may be called by one of the processing elements of the system to execute the functions of the x-module. Other modules are implemented similarly. All or part of the modules can be integrated together or can be independently realized. The processing element described herein may be an integrated circuit having signal processing capabilities. In implementation, each step of the above method or each module above may be implemented by an integrated logic circuit of hardware in a processor element or an instruction in the form of software. These above modules may be one or more integrated circuits configured to implement the above methods, such as: one or more Application Specific Integrated Circuits (ASICs), one or more microprocessors (DSPs), one or more Field Programmable Gate Arrays (FPGAs), and the like. When a module is implemented in the form of a Processing element scheduler code, the Processing element may be a general-purpose processor, such as a Central Processing Unit (CPU) or other processor capable of calling program code. These modules may be integrated together and implemented in the form of a System-on-a-chip (SOC).
EXAMPLE III
This embodiment provides an apparatus, the apparatus comprising: a processor, memory, transceiver, communication interface, or/and system bus; the memory and the communication interface are connected with the processor and the transceiver through a system bus and are used for completing mutual communication, the memory is used for storing a computer program, the communication interface is used for communicating with other equipment, and the processor and the transceiver are used for operating the computer program to enable the equipment to execute the steps of the speed prediction method based on the road network topological relation.
The above-mentioned system bus may be a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, or the like. The system bus may be divided into an address bus, a data bus, a control bus, and the like. For ease of illustration, only one thick line is shown, but this does not mean that there is only one bus or one type of bus. The communication interface is used for realizing communication between the database access device and other equipment (such as a client, a read-write library and a read-only library). The Memory may include a Random Access Memory (RAM), and may further include a non-volatile Memory (non-volatile Memory), such as at least one disk Memory.
The Processor may be a general-purpose Processor, and includes a Central Processing Unit (CPU), a Network Processor (NP), and the like; the Integrated Circuit may also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA) or other Programmable logic device, discrete Gate or transistor logic device, or discrete hardware components.
The protection scope of the speed prediction method based on road network topological relation in the present invention is not limited to the execution sequence of the steps listed in this embodiment, and all the schemes of adding, subtracting, and replacing the steps in the prior art according to the principle of the present invention are included in the protection scope of the present invention.
The invention also provides a speed prediction system based on the road network topological relation, which can realize the speed prediction method based on the road network topological relation, but the realization device of the speed prediction method based on the road network topological relation comprises but is not limited to the structure of the speed prediction system based on the road network topological relation listed in the embodiment, and all structural modifications and replacements in the prior art made according to the principle of the invention are included in the protection scope of the invention.
In summary, the speed prediction method, system, medium and device based on the road network topological relation reconstruct the original model based on the existing two-step prediction model for predicting the speed by using the traffic tendency, and use the idea of predicting the speed of the non-seed road section by using the seed road section, and use the topology among the road sections in the road network to measure the speed influence degree of the adjacent road sections by using the historical data, and predict the speed change value according to the assumption that the speed change value of each road section is determined by the upstream road section. In addition, the invention also provides a thought of overall prediction of the road network, can predict all unknown road section speeds of the road network, and solves the problem of prediction deadlock in the overall prediction process. The invention effectively overcomes various defects in the prior art and has high industrial utilization value.
The foregoing embodiments are merely illustrative of the principles and utilities of the present invention and are not intended to limit the invention. Any person skilled in the art can modify or change the above-mentioned embodiments without departing from the spirit and scope of the present invention. Accordingly, it is intended that all equivalent modifications or changes which can be made by those skilled in the art without departing from the spirit and technical spirit of the present invention be covered by the claims of the present invention.
Claims (10)
1. A speed prediction method based on road network topological relation is characterized by being used for improving a first-order prediction model of a single road section;
wherein the first order prediction model for a single road segment has the following settings:
the model only comprises a central road section and a neighbor road section, and opposite road sections of the same road do not belong to the neighbor road sections;
the influence of the neighbour section on the central section is calculated as the difference of the speed with respect to the average speed, the speed difference being defined as
The neighboring road sections of each road section can have speed influence on the road section, and the influence factor is represented by a correlation coefficient defined as
Where i, j represents two link numbers, days _ count represents the number of days of history data, and the correlation coefficient represents two link xiAnd xjThe average speed of the corresponding histories is simultaneously greater than or equal to or simultaneously less than the average speed of the corresponding histories in different days of the same time period t
The ratio of the number of the active ingredients to the total amount of the active ingredients,
denotes xiRoad section and xjThe road segments are also greater than or equal to the number of days of their historical average speed,
represents the number of days that the two road segments are simultaneously less than their historical average speed;
in order to reduce error propagation, an estimation level L is defined, a propagation factor t is defined, L of a seed road section is 0, L of a central road section is max (L of a neighbor road section) +1, and the influence degree of the neighbor road section is in an exponential relation with t;
the speed prediction method based on the road network topological relation comprises the following steps:
establishing road section topology information for the road network of the selected area according to the end point information of all road sections in the road network;
selecting a seed road section and a non-seed road section in the road network; the seed road sections are a plurality of road sections with the highest contribution value to the overall prediction in the road network;
carrying out weight learning on the first-order prediction models of all non-seed road sections in the road network to predict the speed change difference of the seed road sections;
and overall predicting the speed value of the road network.
2. The method of claim 1 wherein between the steps of establishing topology information for all road segments, said method further comprises preprocessing historical speed data over historical time periods in a road network in a selected area.
3. The road network topology relationship-based speed prediction method according to claim 1,
the overall prediction contribution value includes a support feature contribution value representing a number of seeds in the first-order prediction model and an overlay feature contribution value representing a number of non-seed segments adjacent to the seed segment.
4. The road network topology relationship-based speed prediction method according to claim 3,
the step of selecting seed road segments and non-seed road segments in the road network comprises:
calculating a weighted sum of the support feature contribution value and the coverage feature contribution value of each road segment in the road network;
determining the seed section by the maximum weighted sum; and the rest road sections except the seed road section in the road network are non-seed road sections.
5. The road network topology relationship-based speed prediction method according to claim 4,
after the step of selecting the seed road segments and the non-seed road segments in the road network is performed, the speed prediction method based on the road network topological relation further comprises the following steps:
the marker of the selected seed section is updated to a known marker, and the estimation level thereof is updated to 0.
6. The method for predicting speed based on road network topological relation according to claim 5, wherein said step of learning weight of first order prediction model of all non-seed road segments in road network comprises:
calculating an objective function according to historical speed data of the road section; the objective function is used for representing the mean value of the difference value square sum of the speed difference real value and the speed difference estimated value of the road section in a training time period;
updating a weight value used for representing influence degree between road sections according to the target function and the gradient descent mode so as to realize weight value learning of a first-order prediction model of all non-seed road sections in the road network;
and calculating the speed difference value estimated value of the seed road section according to the updated weight value so as to obtain the speed value of the seed road section.
7. The method for predicting speeds based on road network topological relation according to claim 6, wherein said step of overall predicting speed values of said road network comprises:
marking the non-seed road section as an unknown mark, and calculating a prediction index in a first-order prediction model; the prediction index is associated with an estimated level of an adjacent road segment within the model;
and sequencing the non-seed road sections according to the prediction indexes, and updating the speed difference value of the sequenced non-seed road sections to predict the speed value of the non-seed road sections.
8. A speed prediction system based on road network topological relation is characterized in that a first-order prediction model of a single road section is improved;
wherein the first order prediction model for a single road segment has the following settings:
the model only comprises a central road section and a neighbor road section, and opposite road sections of the same road do not belong to the neighbor road sections;
the influence of the neighbour section on the central section is calculated as the difference of the speed with respect to the average speed, the speed difference being defined as
The neighboring road sections of each road section can have speed influence on the road section, and the influence factor is represented by a correlation coefficient defined as
Where i, j represents two link numbers, days _ count represents the number of days of history data, and the correlation coefficient represents two link xiAnd xjThe average speed of the corresponding histories is simultaneously greater than or equal to or simultaneously less than the average speed of the corresponding histories in different days of the same time period t
The ratio of the number of the active ingredients to the total amount of the active ingredients,
denotes xiRoad section and xjThe road segments are also greater than or equal to the number of days of their historical average speed,
represents the number of days that the two road segments are simultaneously less than their historical average speed;
in order to reduce error propagation, an estimation level L is defined, a propagation factor t is defined, L of a seed road section is 0, L of a central road section is max (L of a neighbor road section) +1, and the influence degree of the neighbor road section is in an exponential relation with t;
the speed prediction system based on the road network topological relation comprises:
the topology establishing module is used for establishing the topology information of the road sections for the road network of the selected area according to the endpoint information of all the road sections in the road network;
the selection module is used for selecting a seed road section and a non-seed road section in the road network; the seed road sections are a plurality of road sections with the highest contribution value to the overall prediction in the road network;
the learning module is used for learning the weight of the first-order prediction models of all non-seed road sections in the road network so as to predict the speed change difference of the seed road sections;
and the prediction module is used for integrally predicting the speed value of the road network.
9. A medium having stored thereon a computer program, characterized in that the computer program, when being executed by a processor, implements a road network topology relationship based speed prediction method according to any one of claims 1 to 7.
10. An apparatus, comprising: a processor and a memory;
the memory is used for storing a computer program, and the processor is used for executing the computer program stored in the memory to make the device execute the speed prediction method based on road network topological relation according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010513633.8A CN111862595B (en) | 2020-06-08 | 2020-06-08 | Speed prediction method, system, medium and device based on road network topological relation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010513633.8A CN111862595B (en) | 2020-06-08 | 2020-06-08 | Speed prediction method, system, medium and device based on road network topological relation |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111862595A CN111862595A (en) | 2020-10-30 |
| CN111862595B true CN111862595B (en) | 2021-12-31 |
Family
ID=72986361
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010513633.8A Active CN111862595B (en) | 2020-06-08 | 2020-06-08 | Speed prediction method, system, medium and device based on road network topological relation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111862595B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112801399B (en) * | 2021-02-08 | 2022-08-12 | 腾讯科技（深圳）有限公司 | Path generation method and device, terminal equipment and storage medium |
| CN114023074B (en) * | 2022-01-10 | 2022-04-15 | 佛山市达衍数据科技有限公司 | Traffic jam prediction method, device and medium based on multiple signal sources |
| CN115410386B (en) * | 2022-09-05 | 2024-02-06 | 同盾科技有限公司 | Short-time speed prediction method and device, computer storage medium and electronic equipment |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104217579A (en) * | 2013-07-23 | 2014-12-17 | 同济大学 | Method for searching critical road sections of traffic networks on basis of redundancy of road section sub-networks |
| CN106096302A (en) * | 2016-06-22 | 2016-11-09 | 江苏迪纳数字科技股份有限公司 | Based on time and the data recovery method of section dependency |
| CN106251625A (en) * | 2016-08-18 | 2016-12-21 | 上海交通大学 | Three-dimensional urban road network global state Forecasting Methodology under big data environment |
| CN106781489A (en) * | 2016-12-29 | 2017-05-31 | 北京航空航天大学 | A kind of road network trend prediction method based on recurrent neural network |
| CN107689152A (en) * | 2016-08-04 | 2018-02-13 | 西门子公司 | Method and apparatus for carrying out traffic processing in a road network |
| CN108648457A (en) * | 2018-06-28 | 2018-10-12 | 苏州大学 | A kind of method, apparatus and computer readable storage medium of prediction of speed |
| CN109118014A (en) * | 2018-08-30 | 2019-01-01 | 浙江工业大学 | A kind of traffic flow speed prediction technique based on time recurrent neural network |
| CN109285346A (en) * | 2018-09-07 | 2019-01-29 | 北京航空航天大学 | A kind of city road net traffic state prediction technique based on key road segment |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102110365B (en) * | 2009-12-28 | 2013-11-06 | 日电（中国）有限公司 | Road condition prediction method and road condition prediction system based on space-time relationship |
| US9536424B2 (en) * | 2014-02-10 | 2017-01-03 | Here Global B.V. | Adaptive traffic dynamics prediction |
| US10215579B2 (en) * | 2016-12-14 | 2019-02-26 | Samsung Electronics Co., Ltd. | Method and apparatus for estimating driving information |
| CN111080029B (en) * | 2019-12-26 | 2022-09-06 | 山东大学 | Urban traffic road speed prediction method and system based on multi-path segment space-time correlation |
-
2020
- 2020-06-08 CN CN202010513633.8A patent/CN111862595B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104217579A (en) * | 2013-07-23 | 2014-12-17 | 同济大学 | Method for searching critical road sections of traffic networks on basis of redundancy of road section sub-networks |
| CN106096302A (en) * | 2016-06-22 | 2016-11-09 | 江苏迪纳数字科技股份有限公司 | Based on time and the data recovery method of section dependency |
| CN107689152A (en) * | 2016-08-04 | 2018-02-13 | 西门子公司 | Method and apparatus for carrying out traffic processing in a road network |
| CN106251625A (en) * | 2016-08-18 | 2016-12-21 | 上海交通大学 | Three-dimensional urban road network global state Forecasting Methodology under big data environment |
| CN106781489A (en) * | 2016-12-29 | 2017-05-31 | 北京航空航天大学 | A kind of road network trend prediction method based on recurrent neural network |
| CN108648457A (en) * | 2018-06-28 | 2018-10-12 | 苏州大学 | A kind of method, apparatus and computer readable storage medium of prediction of speed |
| CN109118014A (en) * | 2018-08-30 | 2019-01-01 | 浙江工业大学 | A kind of traffic flow speed prediction technique based on time recurrent neural network |
| CN109285346A (en) * | 2018-09-07 | 2019-01-29 | 北京航空航天大学 | A kind of city road net traffic state prediction technique based on key road segment |
Non-Patent Citations (5)
| Title |
|---|
| A Traffic Speed Imputation Method Based on Self-adaption and Clustering;Xiayu Qiu Yaying Zhang;《2019 the 4th IEEE International Conference on Big Data Analytics》;20191231;26-31 * |
| Short-Term Road Speed Forecasting Based on Hybrid RBF Neural Network With the Aid of Fuzzy System-Based Techniques in Urban Traffic Flow;CHUN AI 等;《SPECIAL SECTION ON BIG DATA TECHNOLOGY AND APPLICATIONS IN INTELLIGENT TRANSPORTATION》;20200407;69461-69470 * |
| 一种基于FCD的城市道路车流速度估计算法;殷伟;《中国科学技术大学学报》;20080930;1113-1117 * |
| 城市交通智能路网的关键技术及应用;蒋昌俊 等;《自动化博览》;20150531;76-78 * |
| 基于时空相似性的城市交通路网分区及速度预测研究;李拓;《中国优秀硕士学位论文全文数据库工程科技Ⅱ辑》;20200115;C034-989 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111862595A (en) | 2020-10-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111862595B (en) | Speed prediction method, system, medium and device based on road network topological relation | |
| Moilanen | On the limitations of graph-theoretic connectivity in spatial ecology and conservation | |
| Ding et al. | Complex network theory applied to the growth of Kuala Lumpur’s public urban rail transit network | |
| Haghighi et al. | Robotic swarm control from spatio-temporal specifications | |
| CN104464304A (en) | Urban road vehicle running speed forecasting method based on road network characteristics | |
| CN106652441A (en) | Urban road traffic condition prediction method based on spatial-temporal data | |
| Tsagkis et al. | Analysing urban growth using machine learning and open data: An artificial neural network modelled case study of five Greek cities | |
| CN105701571A (en) | Short-term traffic flow prediction method based on nerve network combination model | |
| CN106202335A (en) | A kind of big Data Cleaning Method of traffic based on cloud computing framework | |
| CN105160889B (en) | More source points collaboration leading method of road network traffic flow under a kind of urban waterlogging scene | |
| CN107392307A (en) | The Forecasting Methodology of parallelization time series data | |
| Meng et al. | Microscopic traffic simulation model-based optimization approach for the contraflow lane configuration problem | |
| CN114399125B (en) | Motorcade optimal trajectory control method and device, electronic equipment and storage medium | |
| CN105844334B (en) | A kind of temperature interpolation method based on radial base neural net | |
| CN112797995A (en) | Vehicle emergency navigation method with space-time characteristic situation information | |
| CN109931903A (en) | A kind of cylindricity assessment method based on improvement whale optimization algorithm | |
| CN108985518A (en) | A kind of traffic congestion charging policy optimization method based on VISSIM emulation | |
| CN109033181B (en) | Wind field geographic numerical simulation method for complex terrain area | |
| CN110689174A (en) | Personnel route planning method and device based on public transport | |
| CN111008730B (en) | Crowd concentration prediction model construction method and device based on urban space structure | |
| CN115617034A (en) | Multi-agent environment exploration method and device, electronic equipment and storage medium | |
| CN110866051A (en) | System for representing space-time evolution of regional logistics enterprise | |
| CN115657166A (en) | Method and device for determining weather forecast value across multiple climate areas | |
| CN118072522B (en) | Urban traffic control method and terminal equipment | |
| Lin et al. | A systematic cooperation method for in-car navigation based on future time windows |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |