CN110803203B - Method and system for predicting evolution of high-speed railway running track - Google Patents
Method and system for predicting evolution of high-speed railway running track Download PDFInfo
- Publication number
- CN110803203B CN110803203B CN201911076711.6A CN201911076711A CN110803203B CN 110803203 B CN110803203 B CN 110803203B CN 201911076711 A CN201911076711 A CN 201911076711A CN 110803203 B CN110803203 B CN 110803203B
- Authority
- CN
- China
- Prior art keywords
- scene
- train
- time
- track
- operation scene
- 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
Images
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B61—RAILWAYS
- B61L—GUIDING RAILWAY TRAFFIC; ENSURING THE SAFETY OF RAILWAY TRAFFIC
- B61L27/00—Central railway traffic control systems; Trackside control; Communication systems specially adapted therefor
Abstract
The invention discloses an evolution prediction method and a system of a high-speed railway track, and a related method comprises the following steps: acquiring real-time change information of a railway running environment in real time, and acquiring a train state and an environment state through comprehensive processing so as to identify an operation scene; according to the operation scene recognition result and a pre-constructed scene probability transition matrix, performing rolling prediction on the operation scene of the train in a future period of time; and according to the rolling prediction result of the operation scene, carrying out track splicing on the train, and carrying out conflict identification on the spliced running track. The related scheme can accurately predict the driving scene in the actual dynamic operation environment and the discretized running track of the train in a future period of time.
Description
Technical Field
The invention relates to the technical field of rail transit, in particular to an evolution prediction method and system of a high-speed railway track.
Background
The evolution prediction of the driving track is the basis for formulating the train operation control and scheduling adjustment strategy and is also the basic condition for realizing intelligent scheduling.
In the actual operation commanding process, the operation scene evolution results of several seconds, several minutes and several hours in the future are simulated by the scheduling commanding personnel at any time and are used as the basis of scheduling decisions. Therefore, the prediction of the train running track in a future period of time, especially the prediction of the running track in an abnormal running scene, is a basic requirement of train running control and dispatching command. However, in the current dynamic environment of actual operation, under the condition that the driving order is interfered or is late, especially under the condition of an emergency, the prediction of the driving scene and the driving track in a certain time in the future and the formulation of the corresponding driving strategy on the basis mainly depend on the self experience of a dispatcher, so that the method has larger personal subjectivity and experience limitation, and is lack of a corresponding objective and comprehensive computer-aided system.
Disclosure of Invention
The invention aims to provide a method and a system for predicting the evolution of a high-speed railway running track, which can accurately predict a running scene in an actual dynamic operation environment and a discretized running track of a train in a future period of time.
The purpose of the invention is realized by the following technical scheme:
a method for predicting the evolution of a high-speed railway track comprises the following steps:
acquiring real-time change information of a railway running environment in real time, and acquiring a train state and an environment state through comprehensive processing so as to identify an operation scene;
according to the operation scene recognition result and a pre-constructed scene probability transition matrix, performing rolling prediction on the operation scene of the train in a future period of time;
and according to the rolling prediction result of the operation scene, carrying out track splicing on the train, and carrying out conflict identification on the spliced running track.
An evolution prediction system of a high-speed railway track comprises:
the scene recognition module is used for acquiring real-time change information of a railway driving environment in real time and acquiring a train state and an environment state through comprehensive processing so as to recognize an operation scene;
the scene prediction module is used for performing rolling prediction on the operation scene of the train in a future period of time according to the operation scene recognition result and a pre-constructed scene probability transition matrix;
and the train track prediction module is used for splicing the tracks of the trains according to the rolling prediction result of the operation scene and identifying conflicts of spliced running tracks.
The technical scheme provided by the invention can be seen that the evolution prediction of the train running track is realized through the identification and prediction of the operation scene, the prediction precision of the train running track of the high-speed railway train running dispatching command system in a period of time in the future, particularly an emergency, can be effectively improved, the potential train running conflict can be rapidly and accurately recognized, an auxiliary decision suggestion is given, and the decision quality and the working efficiency of dispatching command personnel can be greatly improved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on the drawings without creative efforts.
Fig. 1 is a flowchart of an evolution prediction method for a high-speed railway track according to an embodiment of the present invention;
fig. 2 is a schematic diagram of a scene classification recognition rule according to an embodiment of the present invention;
fig. 3 is a schematic structural diagram of an evolution prediction system of a high-speed railway track according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention are clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments of the present invention without making any creative effort, shall fall within the protection scope of the present invention.
The embodiment of the invention provides an evolution prediction method of a high-speed railway running track, which can realize the real-time prediction of a running scene and a train running space-time track in a future period of time under a dynamic operation environment. In the embodiment of the invention, the key influence factors of the train operation scene are firstly extracted, an identification model of the operation scene is constructed on the basis, then the operation track of the train is subjected to discrete eventing, the operation track of the train is subjected to micro deconstruction analysis from the perspective of space-time dimension on the basis of data driving, and the evolution prediction of the operation track of the train in a period of time in the future is realized by adopting a statistical analysis and machine learning method. As shown in fig. 1, the main steps include:
step 1, real-time change information of a railway running environment is obtained in real time, and a train state and an environment state are obtained through comprehensive processing, so that an operation scene is identified.
In the embodiment of the invention, the real-time change information of the railway driving environment is acquired in real time, and the method mainly comprises the following steps: line status information, weather information, signal status information, and train status information.
Wherein: the line state information mainly includes: the train distribution, the line blocking state, the speed limiting state and other operation states in the whole dispatching command area and the adjacent areas can be obtained through a dispatching centralized system; the weather information comprises the current natural weather states of rain, snow, wind and the like, and corresponding information can be obtained through an interface of the disaster prevention system and the dispatching centralized system; the signal state information includes: the states of signal facilities such as signal machines, turnouts, access roads, block partitions and the like can be acquired through an interlocking system and a train control system; the train state information includes: the train number, position, speed and other states of the train can be acquired through a train stage plan and a wireless block center system.
On the basis of obtaining real-time change information of the railway driving environment, the interlocking characteristic of the signal facility and the safe driving rule are combined for comprehensive processing, and key factors for identifying the operation scene can be obtained: train status and environmental status.
In the embodiment of the invention, on the basis of the spatial dimension characteristics of the driving events, the corresponding time characteristics are combined, and the classification idea is adopted, so that the operation scene can be defined as the combination of the train state and the environment state in a specific time period.
As shown in table 1, the train conditions mainly include: train clusters, train positions and late times; the train clusters are trains of the same train type, the same driving route and the same operation mode.
TABLE 1 train State
The starting late time is divided into two gears according to the interval running buffering time, and the late time is smaller than the interval buffering time and the late time is larger than the interval buffering time. When the departure delay is larger than the interval buffering time, the train runs at the maximum speed allowed by the environment. The arrival late time is divided into two grades, the late time is smaller than the station operation buffering time, the late time is larger than the station operation buffering time, and when the late time is smaller than the station operation buffering time, the train leaves the station at the right time according to the planned time; and when the late time is greater than the station operation buffering time, the train leaving late time is the difference between the arrival late time and the station operation buffering time.
As shown in table 2, the environmental conditions mainly include: temporary speed limit information and protection annunciator state information.
TABLE 2 environmental conditions
The real-time change information of the railway running environment is introduced, the influence on weather and the interference of a line can be finally embodied by a speed limit command and line blocking, and the blocking can be regarded as that the speed limit of a section is zero in order to simplify the information description.
In order to facilitate scene identification, in the embodiment of the present invention, a train state and an environment state are expressed in advance in a vector manner, and an operation scene classification rule is determined, as shown in fig. 2, an operation scene classification identification rule is exemplarily given.
When the operation scene is identified every time, the operation scene of the current train can be determined by combining the operation scene classification rule according to the inner product of the train state of the current train and the corresponding vector of the environment state, so that the operation scene can be rapidly identified.
And 2, performing rolling prediction on the operation scene of the train in a future period of time according to the operation scene identification result and a pre-constructed scene probability transition matrix.
Before introducing the operation scenario prediction mode, a scenario probability transition matrix is explained.
In the embodiment of the invention, the scene probability transition matrix is constructed by historical data and an operation scene identification result, specifically, on the basis of historical data training, statistical calculation is carried out on the identified operation scene, a scene probability transition matrix is constructed by combining driving constraint conditions, the requirement identification of a train scene is realized, and self-adaptive online correction is carried out through an actual driving track.
The method for constructing the scene probability transition matrix and adaptively correcting comprises the following steps:
1) and counting the number of the operation scenes in each scene category in the historical data according to the operation scene result.
2) Determining a transition relation between operation scenes of the historical data, if the next state of the operation scene alpha is an operation scene beta and the operation scenes accord with train operation constraints, enabling the alpha and the beta to have the transition relation and to be represented as alpha → beta, and recording the transition relation between the alpha and the beta.
3) On the basis, respectively counting the transition relation among the scene categories to make the scene set in the scene category X as { X1,x2,…,xnThe scene set in the scene category Y is { Y }1,y2,…,ykIf xi→yj,xi∈X,yjE, Y, the scene type X and the scene type Y have a transition relation, a scene type probability transition matrix is constructed, and matrix elements
Where count represents the total number, then aX,YAs the transition probability of scene class X to class Y.
4) According to the actual driving track, aX,YAnd (6) correcting. That is, according to the current generated driving track, updating the transition data of each scene in the matrix, such as scene x in the running process of the trainiHas an actual transition scenario of yjUpdating the count (x), wherein the transition relation of the current scene has a larger weight due to a smaller mutation probability of the external environment in the actual operation process, so that the count (x) is corrected according to the transition relation of the online scenei→yj)＝count(xi→yj)old+ w, where count (x)i→yj)oldFor the last calculation, w is the weight coefficient, w > 1, update
In the embodiment of the invention, from the perspective of discretization characteristics in the spatial dimension of the driving events, by combining the time continuity of train operation, the operation of the train on each traveling unit can be regarded as an event constrained by time, and the transition between adjacent events can be embodied by the signal state change associated with the traveling unit, namely the change of the train space-time trajectory and the scene is driven by the change of the event.
In the operation scene prediction, an operation scene identified by a scene identification module and a pre-constructed scene probability transition matrix are combined with operation requirements (driving related decisions made by scheduling commanders in advance, including phase plans, scheduling commands, construction plans and the like), an event driving and rolling prediction mode is adopted according to a signal and driving constraint relation, and scene prediction and evolution of each train in a future period are realized without considering train group constraints, wherein the main process is as follows:
1) assume that the currently identified operational scenario is piOperating scenario piThe scene category to which it belongs is P.
2) And determining the next position loc of train operation according to the real-time driving plan.
3) According to the scene type P, the scene type P is determined by combining the scene probability transition matrix, and the position of the maximum value of the transition probability in the scene probability transition matrix, namely max (a), is determined under the constraint condition that the next position of the train is metX,YX ═ P, x.loc ═ loc), the next driving scene class Q ═ Y, a is determined according to the maximum probability principleX,YIs the transition probability of scene category X to category Y, where x.loc refers to the location of the train in scene category X, i.e., matching the scene category according to location attributes.
4) If the operation scene P exists in the operation scene P, the operation scene P is compared with the operation scene PiCompletely matched operation scenes, and an operation scene p exists in the scene class QiScene q of transition relationsjAnd the next driving operation scene is qjAnd otherwise, the next driving operation scene is made to be the operation scene with the highest occurrence frequency in the scene class Q.
Here, the number of the first and second electrodes,judging whether an operation scene P exists or notiThe purpose of the perfectly matched operational scenario is: the scene category is a large category of scenes, and under the large category of scenes, there are still more detailed small categories of scenes, mainly the difference of the late time, so as to prevent the scene matching error under extreme conditions, improve the success rate, and increase the steps of complete matching.
5) And updating the current operation scene according to the result of the previous step, and repeating the steps 1) to 4) until the time limit of scene prediction is met.
And 3, according to the rolling prediction result of the operation scene, carrying out track splicing on the train, and carrying out conflict identification on the spliced running track.
In the embodiment of the invention, the tracks of the trains are spliced according to the rolling prediction result of the operation scene, and the spliced travelling tracks are identified in a collision way by combining the train running sequence, arrival and departure time and the tracking interval constraint of the trains specified by the travelling plan, so that the travelling conflicts which are probably generated at the earliest time are identified and the early warning prompt is given, wherein the relevant steps are as follows:
1) the train mark is used as the only mark of the train running track, the stay time of the train on each signal unit is determined according to the prediction result of the operation scene, the space-time track of each train running in the operation scene is spliced, and the prediction of the train running space-time track in a period of time in the future is formed.
2) And determining the arrival time and departure time of the train at the station in a future period of time according to the predicted train track, and giving a corresponding late early warning prompt if the arrival time or the departure time is later than the scheduled specified time.
3) And calculating whether the interval tracking interval of the tracked train meets the minimum tracking time or not, and giving out a collision early warning prompt of the tracking interval if the interval tracking interval does not meet the minimum tracking time.
4) If the minimum tracking interval of the interval is met, whether the arrival interval and departure interval of the tracked train meet the minimum interval constraint is calculated, and if not, a collision early warning prompt of the arrival interval is given; if the predicted train arrival sequence and the predicted train departure sequence at the station are not consistent with the train sequence specified by the train plan, giving out an early warning prompt of conflict of the running sequence, and taking the predicted running sequence as a conflict solution suggested by the system.
According to the scheme of the embodiment of the invention, the prediction precision of the train running track of the high-speed railway running dispatching command system in a future period of time, especially in an emergency, can be effectively improved, the potential running conflict can be rapidly and accurately identified, an auxiliary decision suggestion is given, and the decision quality and the working efficiency of dispatching command personnel can be greatly improved.
Another embodiment of the present invention further provides an evolution prediction system for a driving trajectory of a high-speed railway, as shown in fig. 3, which mainly includes:
the scene recognition module is used for acquiring real-time change information of a railway driving environment in real time and acquiring a train state and an environment state through comprehensive processing so as to recognize an operation scene;
the scene prediction module is used for performing rolling prediction on the operation scene of the train in a future period of time according to the operation scene recognition result and a pre-constructed scene probability transition matrix;
and the train track prediction module is used for splicing the tracks of the trains according to the rolling prediction result of the operation scene and identifying conflicts of spliced running tracks.
In the embodiment of the invention, the acquiring real-time change information of the railway driving environment in real time, and acquiring the train state and environment state information through comprehensive processing, so as to identify the operation scene comprises the following steps:
representing the train state and the environment state in a vector mode in advance, and determining an operation scene classification and identification rule;
the real-time change information of the railway driving environment is obtained in real time, and the method comprises the following steps: line status information, weather information, signal status information, and train status information;
the real-time change information of the railway running environment is comprehensively processed by combining the interlocking characteristics of the signal facilities and the safe running rules to obtain the state of the train and the state of the environment; wherein, the train state includes: train clusters, train positions and late times; the train clusters are trains of the same train type, the same driving route and the same operation mode; the environmental states include: temporary speed limit information and protection annunciator state information;
and determining the operation scene of the current train by combining the operation scene classification and identification rules according to the inner product of the train state of the current train and the corresponding vector of the environment state.
In the embodiment of the invention, the scene probability transition matrix is obtained by constructing historical data and an operation scene recognition result, and self-adaptive online correction is carried out through an actual driving track;
as shown in fig. 3, the scene probability transition matrix construction and adaptive correction are implemented by the scene learning module in the following manner:
counting the number of operation scenes in each scene category in the historical data according to the operation scene result;
determining a transition relation between operation scenes in historical data, if the next state of an operation scene alpha is an operation scene beta and the operation scenes accord with train operation constraints, enabling the alpha and the beta to have a transition relation, namely alpha → beta, and recording the transition relation between the alpha and the beta;
on the basis, respectively counting the transition relation among the scene categories to make the scene set in the scene category X as { X1,x2,…,xnThe scene set in the scene category Y is { Y }1,y2,…,ykIf xi→yj,xi∈X,yjE, Y, the scene type X and the scene type Y have a transition relation, a scene type probability transition matrix is constructed, and matrix elements
Where count represents the total number, then aX,YAs a transition probability of the scene class X to the class Y;
according to the actual driving track, aX,YAnd (6) correcting.
In the embodiment of the present invention, the performing rolling prediction on the operation scenario of the train in a future period of time according to the operation scenario identification result and the scenario probability transition matrix includes:
the currently identified operating scenario is piOperating scenario piThe scene category to which the method belongs is P;
determining the next position loc of train operation according to the real-time driving plan;
according to the scene type P, the scene type P is determined by combining the scene probability transition matrix, and the position of the maximum value of the transition probability in the scene probability transition matrix, namely max (a), is determined under the constraint condition that the next position of the train is metX,YX ═ P, x.loc ═ loc), the next driving scene class Q ═ Y, a is determined according to the maximum probability principleX,YX.loc refers to the position of the train in the scene category X, i.e. the scene category is matched according to the position attribute;
if the operation scene P exists in the operation scene P, the operation scene P is compared with the operation scene PiCompletely matched operation scenes, and an operation scene p exists in the scene class QiScene q of transition relationsjAnd the next driving operation scene is qjOtherwise, the next driving operation scene is made to be the operation scene with the highest occurrence frequency in the scene class Q;
and updating the current operation scene according to the result of the previous step, and repeating the previous steps until the time limit of scene prediction is met.
In the embodiment of the invention, the tracks of the trains are spliced according to the rolling prediction result of the operation scene, and the spliced travelling tracks are identified in a collision way by combining the train running sequence, arrival and departure time and the tracking interval constraint of the trains specified by the travelling plan, so that the travelling conflicts which are probably generated at the earliest time are identified and the early warning prompt is given, wherein the relevant steps are as follows:
the train mark is used as the only mark of the train running track, the stay time of the train on each signal unit is determined according to the prediction result of the operation scene, the splicing of the space-time track of the running of each train in the operation scene is realized, and the prediction of the space-time track of the running of the train in a period of time in the future is formed;
according to the predicted train track, determining the arrival time and departure time of the train at the station in a future period of time, and if the arrival time or the departure time is later than the scheduled time, giving a corresponding late early warning prompt;
calculating whether an interval tracking interval of the tracked train meets the minimum tracking time or not, and if not, giving a collision early warning prompt of the tracking interval;
if the minimum tracking interval of the interval is met, whether the arrival interval and departure interval of the tracked train meet the minimum interval constraint is calculated, and if not, a collision early warning prompt of the arrival interval is given; if the predicted train arrival sequence and the predicted train departure sequence at the station are not consistent with the train sequence specified by the train plan, giving out an early warning prompt of conflict of the running sequence, and taking the predicted running sequence as a conflict solution suggested by the system.
In the embodiment of the invention, the scene recognition module, the scene learning module, the scene prediction module and the train track prediction module can adopt a centralized mode or a distributed mode in which part of modules are used as independent subsystems in a physical implementation mode,
it should be noted that, the specific implementation manner of the functions implemented by the functional modules included in the system is described in detail in the foregoing embodiments, and therefore, the detailed description is omitted here.
It will be clear to those skilled in the art that, for convenience and simplicity of description, the foregoing division of the functional modules is merely used as an example, and in practical applications, the above function distribution may be performed by different functional modules according to needs, that is, the internal structure of the system is divided into different functional modules to perform all or part of the above described functions.
Through the above description of the embodiments, it is clear to those skilled in the art that the above embodiments can be implemented by software, and can also be implemented by software plus a necessary general hardware platform. With this understanding, the technical solutions of the embodiments can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a usb disk, a removable hard disk, etc.), and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods according to the embodiments of the present invention.
The above description is only for the preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present invention are included in the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Claims (8)
1. A method for predicting the evolution of a high-speed railway track is characterized by comprising the following steps:
acquiring real-time change information of a railway running environment in real time, and acquiring a train state and an environment state through comprehensive processing so as to identify an operation scene;
according to the operation scene recognition result and a pre-constructed scene probability transition matrix, performing rolling prediction on the operation scene of the train in a future period of time;
according to the rolling prediction result of the operation scene, carrying out track splicing on the train, and carrying out conflict identification on the spliced running track;
the method comprises the following steps of acquiring real-time change information of a railway running environment in real time, and acquiring train state and environment state information through comprehensive processing, so that the operation scene identification comprises the following steps:
representing the train state and the environment state in a vector mode in advance, and determining an operation scene classification and identification rule;
the real-time change information of the railway driving environment is obtained in real time, and the method comprises the following steps: line status information, weather information, signal status information, and train status information;
the real-time change information of the railway running environment is comprehensively processed by combining the interlocking characteristics of the signal facilities and the safe running rules to obtain the state of the train and the state of the environment; wherein, the train state includes: train clusters, train positions and late times; the train clusters are trains of the same train type, the same driving route and the same operation mode; the environmental states include: temporary speed limit information and protection annunciator state information;
and determining the operation scene of the current train by combining the operation scene classification and identification rules according to the inner product of the train state of the current train and the corresponding vector of the environment state.
2. The method for predicting the evolution of the traffic track of the high-speed railway according to claim 1, wherein the scene probability transition matrix is constructed by historical data and an operation scene recognition result, and is adaptively corrected on line by an actual traffic track;
the method for constructing the scene probability transition matrix and adaptively correcting comprises the following steps:
counting the number of operation scenes in each scene category in the historical data according to the operation scene result;
determining a transition relation between operation scenes of the historical data, if the next state of the operation scene alpha is an operation scene beta and the operation scenes accord with train operation constraints, enabling the alpha and the beta to have the transition relation, namely alpha → beta, and recording the transition relation between the alpha and the beta;
on the basis, respectively counting the transition relation among the scene categories to make the scene set in the scene category X as { X1,x2,…,xnThe scene set in the scene category Y is { Y }1,y2,…,ykIf xi→yj,xi∈X,yjE, Y, the scene type X and the scene type Y have a transition relation, a scene type probability transition matrix is constructed, and matrix elements
Where count represents the total number, then aX,YAs a transition probability of the scene class X to the class Y;
according to the actual driving track, aX,YAnd (6) correcting.
3. The method for predicting the evolution of the driving track of the high-speed railway according to claim 1, wherein the rolling prediction of the operation scene of the train in a future period according to the operation scene recognition result and the scene probability transition matrix comprises:
the currently identified operating scenario is piOperating scenario piThe scene category to which the method belongs is P;
determining the next position loc of train operation according to the real-time driving plan;
according to the scene type P, the scene type P is determined by combining the scene probability transition matrix, and the position of the maximum value of the transition probability in the scene probability transition matrix, namely max (a), is determined under the constraint condition that the next position of the train is metX,YX ═ P, x.loc ═ loc), the next driving scene class Q ═ Y, a is determined according to the maximum probability principleX,YX.loc refers to the position of the train in the scene category X, i.e. the scene category is matched according to the position attribute;
if the operation scene P exists in the operation scene P, the operation scene P is compared with the operation scene PiCompletely matched operation scenes, and an operation scene p exists in the scene class QiScene q of transition relationsjAnd the next driving operation scene is qjOtherwise, the next driving operation scene is made to be the operation scene with the highest occurrence frequency in the scene class Q;
and updating the current operation scene according to the result of the previous step, and repeating the previous steps until the time limit of scene prediction is met.
4. The method for predicting the evolution of the traffic track of the high-speed railway according to claim 1, wherein the tracks of the trains are spliced according to a rolling prediction result of an operation scene, and collision recognition is performed on the spliced traffic track by combining a train running sequence, arrival and departure times and tracking interval constraints of the trains, which are specified by a traffic plan, so that traffic collisions which are possibly generated earliest are recognized and an early warning prompt is given, wherein the method comprises the following steps:
the train mark is used as the only mark of the train running track, the stay time of the train on each signal unit is determined according to the prediction result of the operation scene, the splicing of the space-time track of the running of each train in the operation scene is realized, and the prediction of the space-time track of the running of the train in a period of time in the future is formed;
according to the predicted train track, determining the arrival time and departure time of the train at the station in a future period of time, and if the arrival time or the departure time is later than the scheduled time, giving a corresponding late early warning prompt;
calculating whether an interval tracking interval of the tracked train meets the minimum tracking time or not, and if not, giving a collision early warning prompt of the tracking interval;
if the minimum tracking interval of the interval is met, whether the arrival interval and departure interval of the tracked train meet the minimum interval constraint is calculated, and if not, a collision early warning prompt of the arrival interval is given; if the predicted train arrival sequence and the predicted train departure sequence at the station are not consistent with the train sequence specified by the train plan, giving out an early warning prompt of conflict of the running sequence, and taking the predicted running sequence as a conflict solution suggested by the system.
5. An evolution prediction system of a high-speed railway track is characterized by comprising:
the scene recognition module is used for acquiring real-time change information of a railway driving environment in real time and acquiring a train state and an environment state through comprehensive processing so as to recognize an operation scene;
the scene prediction module is used for performing rolling prediction on the operation scene of the train in a future period of time according to the operation scene recognition result and a pre-constructed scene probability transition matrix;
and the train track prediction module is used for splicing the tracks of the trains according to the rolling prediction result of the operation scene and identifying conflicts of spliced running tracks.
The method comprises the following steps of acquiring real-time change information of a railway running environment in real time, and acquiring train state and environment state information through comprehensive processing, so that the operation scene identification comprises the following steps:
representing the train state and the environment state in a vector mode in advance, and determining an operation scene classification and identification rule;
the real-time change information of the railway driving environment is obtained in real time, and the method comprises the following steps: line status information, weather information, signal status information, and train status information;
the real-time change information of the railway running environment is comprehensively processed by combining the interlocking characteristics of the signal facilities and the safe running rules to obtain the state of the train and the state of the environment; wherein, the train state includes: train clusters, train positions and late times; the train clusters are trains of the same train type, the same driving route and the same operation mode; the environmental states include: temporary speed limit information and protection annunciator state information;
and determining the operation scene of the current train by combining the operation scene classification and identification rules according to the inner product of the train state of the current train and the corresponding vector of the environment state.
6. The system for predicting the evolution of the traffic track of the high-speed railway according to claim 5, wherein the scene probability transition matrix is constructed by historical data and an operation scene recognition result, and is adaptively corrected on line by an actual traffic track;
the scene probability transition matrix construction and the self-adaptive correction are realized by a scene learning module in the following mode:
counting the number of operation scenes in each scene category in the historical data according to the operation scene result;
determining a transition relation between operation scenes in historical data, if the next state of an operation scene alpha is an operation scene beta and the operation scenes accord with train operation constraints, enabling the alpha and the beta to have a transition relation, namely alpha → beta, and recording the transition relation between the alpha and the beta;
on the basis, respectively counting the transition relation among the scene categories to make the scene set in the scene category X as { X1,x2,…,xnThe scene set in the scene category Y is { Y }1,y2,…,ykIf xi→yj,xi∈X,yjE, Y, the scene type X and the scene type Y have a transition relation, a scene type probability transition matrix is constructed, and matrix elements
Where count represents the total number, then aX,YAs a transition probability of the scene class X to the class Y;
according to the actual driving track, aX,YAnd (6) correcting.
7. The system according to claim 5, wherein the rolling prediction of the operation scenario of the train in a future period of time according to the operation scenario recognition result and the scenario probability transition matrix comprises:
the currently identified operating scenario is piOperating scenario piThe scene category to which the method belongs is P;
determining the next position loc of train operation according to the real-time driving plan;
according to the scene type P, the scene type P is determined by combining the scene probability transition matrix, and the position of the maximum value of the transition probability in the scene probability transition matrix, namely max (a), is determined under the constraint condition that the next position of the train is metX,YX ═ P, x.loc ═ loc), the next driving scene class Q ═ Y, a is determined according to the maximum probability principleX,YX.loc refers to the position of the train in the scene category X, i.e. the scene category is matched according to the position attribute;
if the operation scene P exists in the operation scene P, the operation scene P is compared with the operation scene PiCompletely matched operation scenes, and an operation scene p exists in the scene class QiScene q of transition relationsjAnd the next driving operation scene is qjOtherwise, the next driving operation scene is made to be the operation scene with the highest occurrence frequency in the scene class Q;
and updating the current operation scene according to the result of the previous step, and repeating the previous steps until the time limit of scene prediction is met.
8. The system for predicting the evolution of the traffic track of the high-speed railway according to claim 5, wherein the tracks of the trains are spliced according to a rolling prediction result of an operation scene, and collision recognition is performed on the spliced traffic track by combining a train running sequence, arrival and departure times and tracking interval constraints of the trains specified by a traffic plan, so that traffic collisions which are possibly generated earliest are recognized and an early warning prompt is given, and the method comprises the following steps:
the train mark is used as the only mark of the train running track, the stay time of the train on each signal unit is determined according to the prediction result of the operation scene, the splicing of the space-time track of the running of each train in the operation scene is realized, and the prediction of the space-time track of the running of the train in a period of time in the future is formed;
according to the predicted train track, determining the arrival time and departure time of the train at the station in a future period of time, and if the arrival time or the departure time is later than the scheduled time, giving a corresponding late early warning prompt;
calculating whether an interval tracking interval of the tracked train meets the minimum tracking time or not, and if not, giving a collision early warning prompt of the tracking interval;
if the minimum tracking interval of the interval is met, whether the arrival interval and departure interval of the tracked train meet the minimum interval constraint is calculated, and if not, a collision early warning prompt of the arrival interval is given; if the predicted train arrival sequence and the predicted train departure sequence at the station are not consistent with the train sequence specified by the train plan, giving out an early warning prompt of conflict of the running sequence, and taking the predicted running sequence as a conflict solution suggested by the system.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911076711.6A CN110803203B (en) | 2019-11-06 | 2019-11-06 | Method and system for predicting evolution of high-speed railway running track |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911076711.6A CN110803203B (en) | 2019-11-06 | 2019-11-06 | Method and system for predicting evolution of high-speed railway running track |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110803203A CN110803203A (en) | 2020-02-18 |
| CN110803203B true CN110803203B (en) | 2021-11-26 |
Family
ID=69501363
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911076711.6A Active CN110803203B (en) | 2019-11-06 | 2019-11-06 | Method and system for predicting evolution of high-speed railway running track |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110803203B (en) |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1729465A (en) * | 2002-12-20 | 2006-02-01 | 联合开关及信号公司 | Dynamic optimizing traffic planning method and system |
| CN103021191A (en) * | 2012-11-22 | 2013-04-03 | 浙江理工大学 | Intelligent traffic control device and control method |
| JP2014168978A (en) * | 2013-03-01 | 2014-09-18 | Hitachi Ltd | Simulator for automatically generating train running performance data |
| CN104331953A (en) * | 2014-10-29 | 2015-02-04 | 云南大学 | Car behavior data identification and management method based on internet of things |
| CN105023451A (en) * | 2014-04-15 | 2015-11-04 | 福特全球技术公司 | Driving scenario prediction and automatic vehicle setting adjustment |
| CN106938655A (en) * | 2015-03-31 | 2017-07-11 | 江苏理工学院 | Subway transportation conflict method for early warning |
| CN107742193A (en) * | 2017-11-28 | 2018-02-27 | 江苏大学 | A kind of driving Risk Forecast Method based on time-varying state transition probability Markov chain |
| CN107958269A (en) * | 2017-11-28 | 2018-04-24 | 江苏大学 | A kind of driving risk factor Forecasting Methodology based on hidden Markov model |
| CN109740839A (en) * | 2018-11-23 | 2019-05-10 | 北京交通大学 | Train Dynamic method of adjustment and system under a kind of emergency event |
-
2019
- 2019-11-06 CN CN201911076711.6A patent/CN110803203B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1729465A (en) * | 2002-12-20 | 2006-02-01 | 联合开关及信号公司 | Dynamic optimizing traffic planning method and system |
| CN103021191A (en) * | 2012-11-22 | 2013-04-03 | 浙江理工大学 | Intelligent traffic control device and control method |
| JP2014168978A (en) * | 2013-03-01 | 2014-09-18 | Hitachi Ltd | Simulator for automatically generating train running performance data |
| CN105023451A (en) * | 2014-04-15 | 2015-11-04 | 福特全球技术公司 | Driving scenario prediction and automatic vehicle setting adjustment |
| CN104331953A (en) * | 2014-10-29 | 2015-02-04 | 云南大学 | Car behavior data identification and management method based on internet of things |
| CN106938655A (en) * | 2015-03-31 | 2017-07-11 | 江苏理工学院 | Subway transportation conflict method for early warning |
| CN107742193A (en) * | 2017-11-28 | 2018-02-27 | 江苏大学 | A kind of driving Risk Forecast Method based on time-varying state transition probability Markov chain |
| CN107958269A (en) * | 2017-11-28 | 2018-04-24 | 江苏大学 | A kind of driving risk factor Forecasting Methodology based on hidden Markov model |
| CN109740839A (en) * | 2018-11-23 | 2019-05-10 | 北京交通大学 | Train Dynamic method of adjustment and system under a kind of emergency event |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110803203A (en) | 2020-02-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Marković et al. | Analyzing passenger train arrival delays with support vector regression | |
| US8082071B2 (en) | System and method of multi-generation positive train control system | |
| Corman et al. | A review of online dynamic models and algorithms for railway traffic management | |
| US8433461B2 (en) | Method of planning the movement of trains using pre-allocation of resources | |
| Lee et al. | A delay root cause discovery and timetable adjustment model for enhancing the punctuality of railway services | |
| Medeossi et al. | A method for using stochastic blocking times to improve timetable planning | |
| CN112070325B (en) | Road network train optimization method, device, equipment and storage medium under abnormal event | |
| Wang et al. | Robust control for dynamic train regulation in fully automatic operation system under uncertain wireless transmissions | |
| Cheng et al. | A Fuzzy Petri Nets approach for railway traffic control in case of abnormality: Evidence from Taiwan railway system | |
| CN112232552A (en) | Emergency uncertainty-oriented train operation plan adjustment risk control method | |
| CN115384586B (en) | Railway parallel scheduling system, method and application thereof | |
| Wang et al. | Hierarchical model predictive control for on-line high-speed railway delay management and train control in a dynamic operations environment | |
| CN110803203B (en) | Method and system for predicting evolution of high-speed railway running track | |
| Guo et al. | Near real-time timetabling for metro system energy optimization considering passenger flow and random delays | |
| García-Ródenas et al. | A mixed-integer linear program for real-time train platforming management | |
| CN115909729A (en) | Simulation prediction method and system based on digital twin under condition of subway line network operation interruption | |
| Ursarova et al. | MULTI-CRITERIA EVALUATION OF PROFESSIONAL QUALITIES OF RAILWAY DISPATCHING PERSONNEL USING COMPUTER SIMULATIONS. | |
| US20230281532A1 (en) | Service plan change assistance apparatus, decision model creation apparatus, decision model creation program, and train traffic management system | |
| Zhao | Railway traffic flow optimisation with differing control systems | |
| Caropreso et al. | A long-term analysis of passenger flows on a regional rail line | |
| CN113935581A (en) | Freight train scheduling method and system based on rolling time domain event influence analysis | |
| US20220335351A1 (en) | Resource Operation Planning Assistance Apparatus and Resource Operation Planning Assistance Method | |
| Zhang et al. | A conditional Bayesian delay propagation model for large-scale railway traffic networks | |
| Zhang et al. | A railway train number tracking method using a prediction approach | |
| Yue et al. | Reinforcement learning for online dispatching policy in real-time train timetable rescheduling |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |