EP3771615A1

EP3771615A1 - Method and apparatus for providing dynamic train control

Info

Publication number: EP3771615A1
Application number: EP19191212.0A
Authority: EP
Inventors: Suk Mun Oh; Ho Chan KWAK; Ji Young Song
Original assignee: Korea Railroad Research Institute KRRI
Current assignee: Korea Railroad Research Institute KRRI
Priority date: 2019-07-31
Filing date: 2019-08-12
Publication date: 2021-02-03
Also published as: JP2021024542A; JP6948361B2; KR20210016100A

Abstract

A dynamic train control method and apparatus based on autonomous collaboration are disclosed. The present disclosure in some embodiments provides an autonomous collaboration-based dynamic train control method and apparatus for enabling trains to autonomously travel based on exceptional circumstances determined by train-to-train communication and terrestrial communication by calculating a severity and a cascaded delay time of the exceptional circumstances and offering the optimal alternative.

Description

Field of the Invention

The present disclosure in some embodiments relates to a dynamic train control method and apparatus based on autonomous collaboration.

Background of the Invention

The statements in this section merely provide background information related to the present disclosure and may not constitute prior art.
In order to operate a train, a plurality of trains is generally operated on predetermined routes on a predetermined schedule. Occurrence of a delay in the preceding train affects the schedule of the trailing trains, so there may be a train conflict or track contention between the trains or chaos ensued.
A train running in the same direction as the preceding train's when delayed should operate to overtake the preceding train to abide by the scheduled arrival time. Trains arriving at train intersection a junction where multiple lines are merged into one track, need to be operated to rearrange the trains' sequence or their schedule.
Train schedules are typically generated based on the experience of the railway operator rather than based on optimization algorithms or simulations. When a railroad operator prepares a schedule for operating trains in advance, drivers and controllers follow the train operation schedule to the best of their ability.
As a driver or a controller operates a train according to a schedule, the operator runs into an unexpected situation such as a train fault, a track failure, a missed schedule, or a platform delay event. Therefore, train conflicts may occur including a same track occupancy at the same time, the occurrence of deadlock between up and down trains, and the failure to follow the headway constraints.
When a train conflict occurs, the control center resets the train route and schedule to ensure the punctuality and safety of the trains, thereby eliminating the train conflicts. Generally, train scheduling methods are determined by collecting all train related information from the control center. The control center delivers the determined train schedule to each train. There is a risk of human errors in the process of receiving and communicating train schedules between train controllers and drivers.
However, the control center needs to collect all train information in real time in order to carry out train scheduling for all the trains. There is a real difficulty in collecting train information from all trains in real time.
Since the control center has to analyze large-scale complex railway networks when a train conflict occurs, it takes a long time to resolve the contention. In other words, the controller adjusting the train schedule inside the control center has a limit in determining the appropriate action to resolve the train conflict.
A common train scheduling method relies on the controller's experience to resolve train conflicts. The risk of human error arises because the controller personally rearranges the train route and schedule to resolve train conflicts. In the process of solving the train conflicts, the approach is based on a uniform standard without considering the current running state of the trains, which degrades the overall train operating efficiency.

SUMMARY OF THE INVENTION

The present disclosure according to some embodiments seeks to provide an autonomous collaboration-based dynamic train control method and apparatus for enabling a train to autonomously travel based on exceptional circumstances determined by train-to-train or inter-train communication and terrestrial communication by calculating a severity and a cascaded delay time of the exceptional circumstances and thus offering the optimal alternative.
According to an aspect of the present disclosure, there is provided an apparatus for providing a dynamic train control, including a train condition recognition unit, a severity determining unit, a cascaded delay time prediction unit, an alternatives selection unit, and a train operation unit. The train condition recognition unit is configured to check whether there is an occurrence of an exceptional circumstance of a preceding train based on a train operation state and a track operation state. The severity determining unit is configured to calculate a severity of the exceptional circumstance. The cascaded delay time prediction unit is configured to predict, based on the severity, a cascaded delay time on trains sequentially affected by the exceptional circumstance among trains located on the same path as the preceding train. The alternatives selection unit is configured to select an optimal alternative among a plurality of pre-stored alternatives to minimize the cascaded delay time. The train operation unit is configured to perform a control so that a train trailing the preceding train travels in the course following the optimal alternative.
According to another aspect of the present disclosure, a method of providing a dynamic train control includes checking whether there is an occurrence of an exceptional circumstance of a preceding train based on a train operation state and a track operation state, and calculating a severity of the exceptional circumstance, and predicting, based on the severity, a cascaded delay time on trains sequentially affected by the exceptional circumstance among trains located on the same path as the preceding train, and selecting an optimal alternative among a plurality of pre-stored alternatives to minimize the cascaded delay time, and performing a control so that a train trailing the preceding train travels in the course following the optimal alternative.
As described above, the present disclosure in some embodiments allows a train to autonomously travel based on exceptional circumstances determined by inter-train communication and terrestrial communication by calculating a severity and a cascaded delay time of the exceptional circumstances and thus offering the optimal alternative.

BRIEF DESCRIPTION OF DRAWINGS

The above and other objects, features and advantages of the present disclosure will be more apparent from the following detailed description taken in conjunction with the accompanying drawings, in which:

FIG. 1 is a schematic diagram of an autonomous train navigation system according to at least one embodiment.
FIG. 2A is a schematic block diagram of a train condition recognition module according to at least one embodiment.
FIG. 2B is a schematic block diagram of a dynamic train control module according to at least one embodiment.
FIGS. 3A, 3B, 3C and 3D are diagrams for explaining the inter-train communication interface according to at least one embodiment.
FIG. 4 is a diagram for explaining the exceptional circumstance classification according to at least one embodiment.
FIG. 5 is a diagram for explaining the coefficients necessary for the autonomous train navigation according to at least one embodiment.
FIG. 6 is a diagram for explaining a method of classifying exceptional circumstances by scenario according to at least one embodiment.
FIG. 7 is a schematic block diagram of a train control learning server according to at least one embodiment.
FIG. 8 is a flowchart for explaining a train situation recognition method according to at least one embodiment.
FIG. 9 is a flowchart for explaining a method of learning an exceptional circumstances determination threshold by a train control server according to at least one embodiment.
FIG. 10 is a flowchart for explaining a dynamic train control method according to at least one embodiment.
FIG. 11 is a flowchart for explaining an alternatives application method according to at least one embodiment.
FIG. 12 is a flowchart for explaining the severity determination method according to at least one embodiment.
FIG. 13 is an exemplary view showing cascaded delay time according to at least one embodiment.
FIG. 14 is an exemplary diagram showing dynamic path setting according to at least one embodiment.

DESCRIPTION OF EMBODIMENTS

Hereinafter, some embodiments of the present disclosure will be described in detail with reference to the accompanying drawings.
Automatic Train Operation (ATO) as used in some embodiments means an automatic operation module mounted in a train for autonomous train navigation. Automatic Train Protection (ATP) is an automatic train stop module installed in a train for autonomous train navigation. Resource Manager (RM) refers to a resource management module installed in a train for autonomous train navigation, and Operation Control (OC) refers to an operation control module installed in a train for autonomous train driving.
Automatic Train Supervisor (ATS) refers to a ground system that periodically manages train conditions, and Train Control Management System (TCMS) refers to an integrated train control system that manages train fault and congestion.
FIG. 1 is a schematic diagram of the autonomous train navigation system according to at least one embodiment.
The autonomous train navigation system according to at least one embodiment of the present disclosure includes at least one train control apparatus 110, a relay apparatus 120, a train control learning server 130, and an ATS 140 but not necessarily limited to these components mentioned.
The train control apparatus 110 performs data communication with neighboring trains and ground infrastructure facilities (e.g., ATS 140) via a network.
The train control apparatus 110 includes a memory for storing a program or protocol for communicating with neighboring trains and ground infrastructure facilities (e.g., the ATS 140) via a network, a microprocessor for executing the relevant program to perform calculation and control, and the like.
The train control apparatus 110 is various apparatuses each including: among others (i) a communication apparatus such as a communication modem and the like for performing communications with various types of devices or wired/wireless networks, (ii) a memory for storing various programs and data, and (iii) a microprocessor for executing a program so as to perform calculation and controlling. According to at least one embodiment, the memory includes a computer-readable recording/storage medium such as a random access memory (RAM), a read only memory (ROM), a flash memory, an optical disk, a magnetic disk, a solid-state disk, and the like. According to at least one embodiment, the microprocessor is programmed for performing one or more of operations and/or functions described herein. According to at least one embodiment, the microprocessor is implemented, in whole or in part, by hardware specifically configured (one or more application specific integrated circuits or ASICs).
Relevant data and programs are stored in the memory, and the processor reads and processes the relevant data from the memory. The processor may be implemented such that one processor can perform each of the above functions, but multiple processors divide processing. The processor may be implemented in a general-purpose processor, but may be implemented as a dedicated chip manufactured to perform the function.
The train control apparatus 110 is implemented as a separate apparatus from the ATS 140 and the train control learning server 130, and it refers to an apparatus operating in a stand-alone mode. The train control apparatus 110 can be operated by an autonomous train navigation program embedded or installed.
The train control apparatus 110 includes ATO, ATS, RM, and OC, and recognizes train conditions based on autonomous collaboration with neighboring trains.
The train control apparatus 110 is mounted on a train for autonomous train navigation and is implemented in a compact size for controlling each train. The train control apparatus 110 performs autonomous train navigation by inter-train collaboration based on communication with neighboring trains. The individual installation and operation of the train control apparatus 110 on each train can minimize the ground infrastructure equipment required for direct communication between the train and the ATS 140.
The train control apparatus 110 receives neighboring train information through inter-train communication. The train control apparatus 110 compares the driving information of the subject train with the neighboring train information to determine the running state. The train control apparatus 110 distinguishes whether it is an exceptional circumstance based on the operation state. The train control apparatus 110 classifies the exceptional circumstance into a certain event.
The train control apparatus 110 according to at least one embodiment checks whether exceptional circumstances occur in the preceding train based on the train operation state and the railway or track operation state. The train control apparatus 110 calculates severity of exceptional circumstances. The train control apparatus 110 predicts, based on the severity, a cascaded delay time on a train that is sequentially affected by exceptional circumstances among the trains located on the same path or course as the preceding train's. The train control apparatus 110 selects an optimal alternative that minimizes the cascaded delay time among a plurality of pre-stored alternatives. The train control apparatus 110 controls the following trains to be operated after the preceding train in the course according to the optimal alternative.
The relay apparatus 120 transmits and receives data between the train control apparatus 110, the train control learning server 130, and the ATS 140 by using various wired and wireless communication technologies such as a mobile communication network, a local area network, an Internet network, an intranet network.
Hardware-wise, the train control learning server 130 includes a hardware module that is the same as a conventional Web server or a network server. The train control learning server 130 typically communicates with a plurality of train control apparatuses 110 via an open computer network such as the Internet.
The train control learning server 130 refers to a computer system or a computer software (web server program) for deriving and providing a task result corresponding to a task execution request of the train control apparatus 110. The train control learning server 130 includes a series of application programs running on a web server or various databases built in the apparatus, in addition to the above-described web server program.
The train control learning server 130 transmits the collected information to the ATS 140 when the train enters the train depot, and it receives, from the ATS 140, an exceptional circumstance occurrence threshold and an exceptional circumstances criterion threshold which are updated based on the collected information.
The train control learning server 130 updates the exceptional circumstance occurrence threshold and the exceptional circumstance classification threshold based on the information collected from the train control apparatus 110, and transmits the updated threshold to the train control apparatus 110.
The train control learning server 130 receives subject train information and neighboring train information from the train control apparatus mounted in a plurality of trains when they enter the train depot. The train control learning server 130 updates the exceptional circumstance occurrence threshold by comparing the subject train information and the neighboring train information with the scheduled operating information. The train control learning server 130 transmits the exceptional circumstance occurrence thresholds to the subject train. The train control learning server 130 updates the exceptional circumstance classification thresholds by learning the exceptional circumstance classification threshold based on the operation state information with respect to the subject train information and the neighboring train information. The train control learning server 130 transmits the exceptional circumstance classification threshold to the subject train.
The ATS 140 is implemented as a separate server independent of the trains. The ATS 140 communicates with each train and processes received big data to manage train-related information.
FIG. 2A is a schematic block diagram of a train condition recognition module according to at least one embodiment.
The train control apparatus 110 according to at least one embodiment includes a train condition recognition module 210 and a dynamic train control module 220. The train condition recognition module 210 includes, but not limited to, a communication unit 212, an exceptional circumstance determining unit 214, and a train condition recognition unit 216. The train condition recognition module 210 may include other components.
The respective components included in the train condition recognition module 210 may be connected to a communication path connecting software modules or hardware modules in the apparatus and may operate organically with each other. These components communicate by using one or more communication buses or signal lines.
Each component of the train condition recognition module 210 shown in FIG. 2A means a unit for processing at least one function or operation, and may be implemented as a software module, a hardware module, or a combination of software and hardware.
The communication unit 212 receives neighboring train information (real-time position, speed, course, event, train fault occurrence information, cabin congestion or on-board congestion) from a neighboring train (for example, preceding train) through train-to-train (T2T) communication, and receives track state information (track fault occurrence information and platform congestion or ground congestion) from the ATS 140.
The communication unit 212 extracts, from the neighboring train information, a scheduled subject train location value (x _i(t)), a subject train location value (x_i(t)), a set allowable train delay limit (X₁), the preceding train location value (x_i-1(t)), and a set allowable train interval limit (X₂).
The exceptional circumstance determining unit 214 checks whether exceptional circumstances occur based on at least one of neighboring train information and track state information.
The exceptional circumstance determining unit 214 extracts, from the neighboring train information, the scheduled subject train location value (x _i(t)), subject train location value (x_i(t)), set allowable train delay limit (X₁), preceding train location value (x_i-1(t)), and a set allowable train interval limit (X₂). The exceptional circumstance determining unit 214 confirms that exceptional circumstances occur when it determines x _i-1(t)-x_i-1(t)≥X₁ that the value obtained by subtracting the subject train location value (x_i(t)) from the scheduled subject train location value (x _i(t)) is equal to or greater than the set allowable train delay limit (X₁). Or the exceptional circumstance determining unit 214 confirms that exceptional circumstances occur when it determines x_i-1(t)-x_i(t)≤X₂ that the value obtained by subtracting the subject train location value (x_i(t)) from the location value x_i-1(t) of the preceding train is equal to or less than the set allowable train interval limit (X₂).
The exceptional circumstance determining unit 214 extracts the train fault occurrence information included in the neighboring train information. The exceptional circumstance determining unit 214 determines that exceptional circumstances occur, based on at least one of the position of the faulty train, the fault occurrence time, and the fault type included in the train fault occurrence information.
The exceptional circumstance determining unit 214 extracts the passenger information in the train cabin, which is detected when the subject train included in the neighboring train information starts from the train stop. The exceptional circumstance determining unit 214 calculates the cabin (on-board) congestion based on the passenger information in the train cabin. The exceptional circumstance determining unit 214 confirms that exceptional circumstances occur when the cabin (on-board) congestion of the train exceeds the predetermined threshold.
The exceptional circumstance determining unit 214 extracts the track fault occurrence information included in track state information. The exceptional circumstance determining unit 214 confirms that exceptional circumstances occur, based on the track fault occurrence location, fault occurrence time, and fault type included in the track fault occurrence information.
The exceptional circumstance determining unit 214 extracts the waiting passenger information on the next scheduled train stop of the subject train, which is included in the track state information. The exceptional circumstance determining unit 214 calculates the platform congestion or ground congestion based on the waiting passenger information of the platform at the next scheduled stop. The exceptional circumstance determining unit 214 confirms that exceptional circumstances occur when the platform congestion or ground congestion exceeds a preset threshold.
The train condition recognition unit 216 is responsive to when the exceptional circumstance determining unit 214 determines that the exceptional circumstance occurs, for classifying the exceptional circumstance into a specific exceptional circumstance of the predetermined exceptional circumstances, and recognizing the train situation based on the specific exceptional circumstance.
The train condition recognition unit 216 extracts, from the neighboring train information, a preceding train event (e_i-1(t)), subject train event (e_i(t)), arrival (Dpt_s) at a station (s), arrival event (Arr_s), preceding train course (p_i-1(t)), subject train course (pi(t)), preceding train speed (v_i-1(t)), subject train speed (v_i(t)), scheduled preceding train location value (x _i-1(t)), scheduled subject train location value (x _i(t)) preceding train location value (x_i-1(t)), subject train location value (x_i(t)), and end station arrival event (End_s).
The train condition recognition unit 216 classifies the exceptional circumstances based on information extracted from the neighboring train information, into a specific exceptional circumference among scenario 1 that is a detour route determination scenario for the trailing train when the preceding train is obstructed, scenario 2 that is a pass sequence adjustment scenario due to track resource contention, and scenario 3 that is a reverse train line adjustment scenario against train reverse delay.
The train condition recognition unit 216 checks whether e_i-1(t) = Dpt_s that preceding train event e_i-1(t) arrives (Dpt_s) at station s. The train condition recognition unit 216 checks whether e_i(t) = Arr_s that subject train event e_i(t) is an arrival event Arr_s. The train condition recognition unit 216 checks whether p_i(t) = p_i-1(t) that subject train path p_i(t) and preceding train path p_i-1(t) are the same. The train condition recognition unit 216 checks whether v_i-1(t) = 0 that preceding train speed v_i-1(t) is 0. The train condition recognition unit 216 checks whether x _i-1(t)-x_i-1(t)≥X₁ that the value obtained by subtracting the preceding train location value (x_i-1(t)) from the scheduled preceding train location value (x _i-1(t)) is equal to or greater than the set allowable train delay limit (X₁)). The train condition recognition unit 216 checks whether x_i-1(t)-x_i(t)≤X₂ that the value obtained by subtracting the subject train location value (x_i(t)) from the location value x_i-1(t) of the preceding train is equal to or less than the set allowable train interval limit (X₂). The train condition recognition unit 216 determines, when the aforementioned conditions are satisfied, that it is scenario 1: the detour route determination scenario for the trailing train when the preceding train is obstructed.
The train condition recognition unit 216 checks whether e_i-1(t) = Dpt_s that the preceding train event e_i-1(t) arrives (Dpt_s) at station s. The train condition recognition unit 216 whether checks e_i(t) = Dpt_s that the subject train event e_i(t) arrives (Dpt_s) at station s. The train condition recognition unit 216 checks whether p_i(t) ≠ p_i-1(t) that subject train path p_i(t) and preceding train path p_i-1(t) are unequal. The train condition recognition unit 216 checks whether v_i-1(t) = v_i(t) = 0 that preceding train speed v_i-1(t) and subject train speed v_i(t) are 0. The train condition recognition unit 216 determines whether x _i-1(t)-x_i-1(t)≥x₁ that the value obtained by subtracting the preceding train location value (x_i-1(t)) from the scheduled preceding train location value (x _i-1(t)) is equal to or greater than the set allowable train delay limit (X₁). The train condition recognition unit 216 determines, when the above conditions are all satisfied, that it is scenario 2: the pass sequence adjustment scenario due to track resource contention.
The train condition recognition unit 216 checks whether e_i(t) = End_s that subject train event e_i(t) is an end station arrival event (End_s). The train condition recognition unit 216 checks whether x _i-1(t)-x_i-1(t)≥x₁ that the value obtained by subtracting the subject train location value (x_i(t)) from the scheduled subject train location value (x _i(t)) is equal to or greater than the set allowable train delay limit (X₁). The train condition recognition unit 216 determines, when all of the above conditions are satisfied, that it is scenario 3: the reverse train line adjustment scenario against train reverse delay.
The dynamic train control module 220 calculates the severity of exceptional circumstances, and predicts, based on the severity, the cascaded delay time on trains sequentially affected by exceptional circumstances among the trains located on the same path (course) as the preceding train's. The dynamic train control module 220 selects an optimal alternative that minimizes the cascaded delay time among the plurality of pre-stored alternatives. The dynamic train control module 220 controls the trains trailing behind the preceding train to travel along the path according to the optimal alternative.
FIG. 2B is a schematic block diagram of a dynamic train control module according to at least one embodiment.
The dynamic train control module 220 according to at least one embodiment includes, but not necessarily limited to, a severity determining unit 222, a cascaded delay time prediction unit 224, an alternative selection unit 226, and a train operating unit 228. The dynamic train control module 220 may include other components.
The respective components included in the dynamic train control module 220 may be connected to a communication path connecting software modules or hardware modules in the apparatus so that they can operate organically with each other. These components communicate by using one or more communication buses or signal lines.
Each component of the dynamic train control module 220 shown in FIG. 2B means a unit for processing at least one function or operation, and may be implemented as a software module, a hardware module, or a combination of software and hardware.
The severity determining unit 222 calculates the severity of the exceptional circumstances received from the train condition recognition module 210.
The severity determining unit 222 receives an exceptional circumstance duration ti-i, an external notification (input) end time τ_i-1, and pre-learned thresholds T_i-1 with respect to the preceding train. The severity determining unit 222 determines that the severity is critical when exceptional circumstance duration ti-i and external notification (input) end time L_i-1 (max (t_i-1, τ_i-1)) are equal to or greater than pre-learned thresholds T_i-1. The severity determining unit 222 determines that the severity is minor when exceptional circumstance duration ti-i and external notification (input) end time τ_i-1 (max (t_i-1, τ_i-1)) are less than pre-learned thresholds T_i-1.
When no external notification (input) end time (τ_i-1) is received, the severity determining unit 222 sets an integer multiple of the exceptional circumstance duration ti-i up to the present time as the estimated notification end time.
The severity determining unit 222 determines that the severity is critical when the estimated notification end time is equal to or greater than pre-learned threshold T_i-1. The severity determining unit 222 determines that the severity is minor when the estimated notification end time is less than pre-learned threshold T_i-1.
The severity determining unit 222 calculates the severity by using pre-learned thresholds T_i which are learned differently for each of scenario 1 that is a detour route determination scenario, scenario 2 that is a pass sequence adjustment scenario, and scenario 3 that is a reverse train line adjustment scenario.
The cascaded delay time prediction unit 224 predicts, based on the severity, cascaded delay time on trains sequentially affected by exceptional circumstances among the trains located on the same route (course) as the preceding train's.
The cascaded delay time prediction unit 224 predicts the cascaded delay time when the severity of the exceptional circumstance is determined to be critical as of the current time point. When the cascaded delay time prediction unit 224 determines that the severity of the exceptional circumstance is minor as of the current time point, it leaves a separate task of predicting the cascaded delay time unperformed.
The cascaded delay time prediction unit 224 calculates delayed arrival times for all trains sequentially influenced by exceptional circumstances among the trains located on the same path (course) as the preceding train's, and calculates the congestion corresponding to the delayed arrival time. The cascaded delay time prediction unit 224 calculates a detour arrival time (cascaded delay time) which occurs when the train trailing the preceding train is sent on ahead by the auxiliary main track or detour (dynamically re-routed) among all the trains sequentially influenced by the exceptional circumstances, and calculates the congestion corresponding to the detour arrival time.
The alternative selection unit 226 selects an optimal alternative that minimizes the cascaded delay time among the plurality of pre-stored alternatives. The alternative selection unit 226 selects, based on the cascaded delay time, one of the course hold offer (alternative 1) and the course change offer (alternative 2) for the train trailing the preceding train.
The alternative selection unit 226 calculates the number of sequentially affected trains (the number of delayed trains) that occur when exceptional circumstances are maintained following the course hold offer (alternative 1). The alternative selection unit 226 calculates the number of sequentially influenced trains that occur when the train trailing the preceding train is sent on ahead by the auxiliary main track or detour (dynamically re-routed) following the course change offer (alternative 2). The alternative selection unit 226 compares the number of sequentially affected trains (the number of delayed trains) that occur when exceptional circumstances are maintained following the course hold offer (alternative 1) with the number of sequentially affected trains (the number of delayed trains) that occur when the train trailing the preceding train is sent on ahead by the auxiliary main track or detour (dynamically re-routed) following the course change offer (alternative 2). The alternative selection unit 226 selects one of the course hold offer (alternative 1) and the course change offer (alternative 2) according to the comparison result.
The alternative selection unit 226 selects the course change offer (alternative 2) when the number of sequentially affected trains (the number of delayed trains) that occur when the train trailing the preceding train is sent on ahead by the auxiliary main track or detour (dynamically re-routed) falls below the number of sequentially affected trains (the number of delayed trains) that occur when exceptional circumstances are maintained by a drop threshold or more.
The alternative selection unit 226 selects the course hold offer (alternative 1) when the number of sequentially affected trains that occur when the train trailing the preceding train is sent on ahead by the auxiliary main track or detour (dynamically re-routed) equals to or falls below the number of sequentially affected trains that occur when exceptional circumstances are maintained by less than the drop threshold.
The alternative selection unit 226 selects the course change offer (alternative 2) when the delayed arrival time and congestion that occur when the train trailing the preceding train is sent on ahead by the auxiliary main track or detour (dynamically re-routed) fall below the delayed arrival time and congestion that occur when exceptional circumstances are maintained, by predetermined drop thresholds or more.
The alternative selection unit 226 selects the course hold offer (alternative 1) when the delayed arrival time and congestion that occur when the train trailing the preceding train is sent on ahead by the auxiliary main track or detour (dynamically re-routed) are equal to or fall below the delayed arrival time and congestion that occur when exceptional circumstances are maintained, by less than the predetermined drop thresholds.
The alternative selection unit 226 selects one of the course hold offer (alternative 1) and course change offer (alternative 2) by using thresholds employed differently for each of scenario 1 that is a detour route determination scenario, scenario 2 that is a pass sequence adjustment scenario, and scenario 3 that is a reverse train line adjustment scenario.
The train operation unit 228 controls the train to be operated after the preceding train in the path (course) according to the optimal alternative.
FIGS. 3A, 3B, 3C and 3D are diagrams for explaining the inter-train communication interface according to at least one embodiment.
ATO, ATP, RM, and RM refer to modules mounted in the train for autonomous train navigation. ATS and TCMS refer to infrastructure installed on the ground to periodically manage train conditions.
The train control apparatus 110 according to at least one embodiment communicates directly with the train or via the terrestrial control system with the train, as shown in FIGS. 3A and 3B.
The train control apparatus 110 may communicate with other trains by establishing a direct communication path as shown in (a) of FIG. 3A, or by establishing an indirect communication path through a ground control system as shown in FIG. 3A at (b).
The train control apparatus 110 performs inter-train communications by using the dynamic path interface between the trains.
The ATS transmits schedule information including neighboring train information to the ATO. The schedule information includes the schedule information of each train centered on the repeating units of the rolling stocks (formation). A tag ID set (k) included in the schedule information means the conversion relation of the course information. The ATS sends the ATO additional information about its own event information management and counterpart train events. A train delay when occurred between ATS and ATO is designated as a failure code type for communication. A congestion related communication method is defined and communicated between ATS and ATO. A plurality of ATOs (ATO_(i) to ATO_{(i + 1)}) and a plurality of ATPs (ATP_(i) to ATP_{(i + 1)}) communicate with each other for communication between neighboring trains. A resource may be assigned an ID, or may be defined as a range according to the order of time between a fixed time and a relative time.
FIG. 3B shows a general train communication interface. The ATS sends the RM a resource control request, a resource cancellation request, and an authority request for point machine. The RM sends the ATS a resource control response, a resource cancellation response, and a response to authority request for point machine. The RM sends the ATP a version information request, a DB request, a resource status request, and a resource request/release request. The ATP sends the RM a version information response, a DB request response, a resource status response, and a resource request/release response. The ATP sends the OC a request for point machine direction control and a lock-of-switch request. The OC transmits to the ATP, a response to point machine direction control and a lock-of-switch response.
The ATS sends the ATP train status information (periodic) reports (including fault code) and a response to train control request. The ATP sends a train (emergency) control request to the ATS. The ATP sends a course assurance request to the ATO. The ATO sends ATP status information and a course assurance response to the ATP. The ATS sends schedule information, a response to schedule information modification, and a train (emergency) control request to the ATO. The ATO sends the ATS a schedule information request, a schedule information modification request, and a response to train (emergency) control request.
The FIG. 3C describes the train communication interface according to at least one embodiment. The ATS sends the RM a resource control request, a resource cancellation request, and an authority request for point machine. The RM sends the ATS the resource control response, resource cancellation response, and response to authority request for point machine. The RM sends the ATP the version information request, DB request, resource status request, and resource request/release request. The ATP sends the RM the version information response, DB request response, resource status response, and a resource request/release response. The ATP sends the OC the request for point machine direction control, and lock-of-switch request. The OC transmits to the ATP, the response to point machine direction control, and lock-of-switch response.
The ATS sends ATP_(i-1) the train status information (periodic) reports (including fault code), and response to train control request. The ATP_(i-1) sends a train (emergency) control request to the ATS. The ATS sends ATO_(i-1) the schedule information, response to schedule information modification, and train (emergency) control request. The response to schedule information modification includes neighbor schedules and forward station-centered schedules. The train (emergency) control request includes (forward station) platform congestion information and train dispatch procedure instructions. The ATO_(i-1) transmits to the ATS, a schedule information request to neighboring train, a schedule information modification request to neighboring train, the response to train (emergency) control request, and a report of train termination and train reverse.
The ATO_(i) sends the course assurance request to the ATO. The ATP_(i) sends the ATO_(i) the ATP status information including the current speed information, course assurance response, fault code, and in-vehicle congestion information. The ATO_(i-1) and ATI_(i+1) transmit the next event and a real-time delay time to the ATO . The ATP_(i+1) sends the ATO_(i) the ATP status information including the current speed information, course assurance response, and fault code.
The interface between the ATS and the ATO is shown in FIG. 3D. The protocols and messages communicating between the ATS and the ATO use the same format. The ATS and the ATO store and analyze multiple messages by using predefined protocols. The ATS and the ATO receive and store the message in multiple paths, and analyze it.
FIG. 4 is a diagram for explaining the exceptional circumstance classification according to at least one embodiment.
The communication unit 212 receives neighboring train information (real-time position, speed, course, event, train fault occurrence information, cabin congestion or on-board congestion) from a neighboring train through train-to-train communication, and receives track state information (track fault occurrence information and platform congestion or ground congestion) from the ATS 140.
The exceptional circumstance determining unit 214 monitors and transmits in real time the subject train information (real-time location, speed, course, event, train fault occurrence information, and cabin congestion or on-board congestion) to the neighboring train. The exceptional circumstance determining unit 214 determines the train operation state of the neighboring train based on the neighboring train information. The exceptional circumstance determining unit 214 determines the track operation state based on the track state information.
The exceptional circumstance determining unit 214 extracts the train fault occurrence information included in the neighboring train information. The exceptional circumstance determining unit 214 determines the train operation state of the neighboring train as a train fault based on the location of the faulty train, the time of occurrence of the fault, and the type of the fault which are included in the train fault occurrence information.
The exceptional circumstance determining unit 214 extracts the passenger information on the passengers in the train cabin, which is detected when the subject train included in the neighboring train information starts from the train stop. The exceptional circumstance determining unit 214 calculates the cabin (on-board) congestion based on the passenger information on the passengers in the train cabin. The exceptional circumstance determining unit 214 determines, when the cabin (on-board) congestion in the train exceeds a predetermined threshold, that the train operation state as a train congestion.
The exceptional circumstance determining unit 214 extracts the track fault occurrence information included in the track state information. The exceptional circumstance determining unit 214 determines the track operation state as a track fault based on the track fault occurrence location, the fault occurrence time, and the fault type which are included in the line track occurrence information. The exceptional circumstance determining unit 214 extracts the waiting passenger information on the waiting passengers of the platform at the next stop for the subject train included in the track state information. The exceptional circumstance determining unit 214 calculates the platform congestion or ground congestion based on the waiting passenger information of the platform at the next stop. The exceptional circumstance determining unit 214 determines, when the platform congestion or ground congestion exceeds a preset threshold, that the track operation state as a platform congestion.
The process for the exceptional circumstance determining unit 214 to determine the neighboring train operation state based on the neighboring train information is as follows.

① A separate fault detection unit, neither Train Control Management System (TCMS) nor Automatic Train Control System (ATCS) installed on each train, detects whether the train is faulty.
② When the fault detection unit detects a fault, it generates fault occurrence information and delivers it in real time to the ATO which is responsible for the dynamic routing of the ATCS.
③ The ATO receives the fault occurrence information (in recognition of the fault in real time) and then distributes the fault occurrence information including the fault condition of the subject train to neighboring trains around through train-to-train (T2T) communication in real time.
The real-time direct transmission of the fault occurrence information by the very faulty train to neighboring trains through T2T communication is quicker than involving the ground control system, for the neighboring trains around to receive and respond to the fault occurrence information.
④ The subject train determines the operation state due to the occurrence of a fault on the neighboring train when the fault occurrence information is received from the faulty train in real time.
The exceptional circumstance determining unit 214 in the subject train determines the operation state of the faulty train based on the location of the faulty train, the time of the fault occurrence, and the type of fault.
⑤ As a result, the train control apparatus 110 can quickly recognize whether a fault has occurred in a neighboring train and use it for a dynamic routing. The train control apparatus 110 evokes the subsequent process without delay when the path is dynamically reestablished.

The process for the exceptional circumstance determining unit 214 to determine the track operation state based on the track state information is as follows.

① The ATS constantly operates to detect whether or not a fault occurs in the ground track equipment (including track obstruction).
② When the ATS detects that a fault occurs in the ground track equipment, it selects the subsequent trains to be affected by the faulty ground track equipment.
The ATS sends track state information including fault occurrence information to ATOs in subsequent trains that are affected by the fault.
③ The train control apparatus 110 in the train receives the track state information including the fault occurrence information from the ATS in real time, and then determines the track operation state.
The train control apparatus 110 determines the track operation state based on the fault occurrence location, the fault occurrence time, and the fault type based on the fault occurrence information included in the track state information.
④ As a result, the train control apparatus 110 can quickly recognize the failure occurring on the track, and use it for dynamic routing. The train control apparatus 110 evokes the subsequent process without delay when the path is dynamically reestablished.

The process of determining the in-vehicle congestion information flow by the exceptional circumstance determining unit 214 is as follows.

① The TCMS detects cabin passenger information (passenger-on-board information) when the train leaves the station.
② The ATO requests the TCMS to give cabin passenger information (passenger-on-board information) in order to ascertain the operation state.
③ The ATO calculates the train cabin (on-board) congestion by using the cabin passenger information (passenger-on-board information) received from the TCMS.

The exceptional circumstance determining unit 214 determines the train operation state based on the train cabin (on-board) congestion.
The procedure of determining the platform congestion information flow by the exceptional circumstance determining unit 214 is as follows.

① The ATS constantly monitors platform waiting passenger information.
② The ATO requests the ATS to give waiting passenger information for the next platform to monitor the operation state.
③ The ATO calculates the platform congestion or ground congestion by using the platform waiting passenger information received from the ATS. The ATO uses platform congestion or ground congestion to predict the stopping time at the next station.
④ The ATO predicts the stopping time by using both the train cabin (on-board) congestion and the platform congestion or ground congestion.

The exceptional circumstance determining unit 214 determines the track operation state based on the platform congestion or ground congestion.
The exceptional circumstance determining unit 214 checks whether exceptional circumstances occur based on the train operation state and the track operation state. The exceptional circumstance determining unit 214 determines whether the train is normal or abnormal based on the train operation state and the track operation state.
The exceptional circumstance determining unit 214 checks whether exceptional circumstances occur based on the train operation state and the track operation state. The exceptional circumstance determining unit 214 determines that exceptional circumstances occur when the train operation state is determined to be a train fault or a train congestion. The exceptional circumstance determining unit 214 determines that exceptional circumstances occur when the track operation state is determined as a track fault or a platform congestion.
The exceptional circumstance determining unit 214 checks whether exceptional circumstances occur as follows.

① The exceptional circumstance determining unit 214 confirms that exceptional circumstances have occurred when the fault information is received from another train in operation.
② The exceptional circumstance determining unit 214 confirms that exceptional circumstances have occurred when it receives the fault occurrence information of the ground equipment from the ATS 140.

The train condition recognition unit 216 is responsive to when the exceptional circumstance determining unit 214 checks and concludes that an exceptional circumstance has occurred, for classifying the exceptional circumstance into a specific exceptional circumstance out of predetermined exceptional circumstances, and recognizing the train situation based on the specific exceptional circumstance.
The train condition recognition unit 216 classifies the exceptional circumstances based on information extracted from the neighboring train information, into a specific exceptional circumference among scenario 1: detour route determination scenario for the trailing train when the preceding train is obstructed; scenario 2: pass sequence adjustment scenario due to track resource contention; and scenario 3: reverse train line adjustment scenario against train reverse delay.
In other words, the train condition recognition unit 216 classifies the exceptional circumstances as follows.

① The train condition recognition unit 216 checks whether x _i (t) - x_i (t) ≥ X ₁ that a current train delay (difference of real-time location from scheduled location) is equal to or greater than a preset threshold.
② The train condition recognition unit 216 checks whether C_i(t)≥C₁ that the in-vehicle congestion detected in real time is equal to or greater than a preset threshold.
③ The train condition recognition unit 216 checks whether C_si(t)≥C₂ that the platform congestion detected in real time is equal to or greater than a preset threshold.

When the exceptional circumstances are confirmed, the train control apparatus 110 evokes the operation of classifying the response scenarios to exceptional circumstances, to define the exceptional circumstances.
FIG. 5 is a diagram for explaining the coefficients necessary for the autonomous train navigation according to at least one embodiment.
'i' shown in the table of FIG. 5 denotes train order. Location (x_i(t)) is set at a predetermined interval (for example, about 60 m-intervals). Each train is equipped with a tachometer, which is used to measure the number of turns of the wheel and thereby measure the distance from the sensor to the front of the train.
i ∈ I = {1,2, ..., N} is assumed to operate in the order of a train number set (integer) of train numbers, (i-1) → (i) → (i + 1). N is the maximum number of subject trains considered.
't' denotes a (present) time code.
g E G means a tag number set (integer) of tags installed on the ground for detecting the train locations.
s ∈ S = {1,2, ..., M} means a set of station numbers (integers).
M is the maximum number of stations considered. For convenience of explanation, it is assumed that the first stop is the start station of the train in any one direction, and that the M-th stop is the end station that the train finally arrives at. The train is supposed to reverse at both the start and end stops and retravel unless stated otherwise.
ξ∈π_s refers to a set of paths that can be set according to the rules of interworking within a station [s].
p _i = ξ₁ × ξ₂ × ··· × ξ _M refers to the entire course preset in the scheduling step, through which a train [i] travels from the start station to the end station.
p_i(t) is the path dynamically selected by train [i] at (current) time t[t].
e_i(t) refers to the right next event that (Srt_s, Arr_s, Dpt_s, End_s) train [i] will have as of the (current) time [t].
Srt_s denotes start (usually s = 1) in station [s].
Arr_s and Dpt_s denote arriving at station [s].
End_s denotes terminating at station [s] (where usually s = M) .
x_i(t) denotes the (real time) location of train [i] at (current) time [t].
v_i(t) denotes the (real-time) speed of train [i] at (current) time (t).
ci(t) denotes the congestion of the cabin having the highest congestion among the plurality of cabins constituting the train [i] at the (present) time [t].
c_si(t) denotes the platform congestion of the station at which the train [i] will stop at the (current) time [t].
The location (x_i(t)), velocity (v_i(t)), course (p_i(t)) and event (e_i(t)) shown in FIG. 5 are requisite factors. Congestion, delay, and train interval are factors computable based on the requisite factors. The next activity is the same factor as the event.
FIG. 6 is a diagram for explaining a method of classifying exceptional circumstances by scenario according to at least one embodiment.
The action scenarios against exceptional circumstances that the train control apparatus 110 classifies are as shown in FIG. 6.

① Scenario 1 is the detour route determination scenario for the trailing train when the preceding train is obstructed. For example, scenario 1 means that a decision is to be made for whether or not to change the arrival route of the delayed trailing train where the preceding train is stopping on the main line.
② Scenario 2 is the pass sequence adjustment scenario due to track resource contention. For example, scenario 2 means, where the main track and the auxiliary main track are scheduled to have respective trains made ready for departure until the main track train departs ahead of the auxiliary main track train, that a decision is to be made for whether to allow the auxiliary main track train go ahead of the main track train when delayed.
③ Scenario 3 is the reverse train line adjustment scenario against train reverse delay. For example, scenario 3 refers to a decision situation on whether or not to adjust the reverse train course at the time of the train reverse delay so that the destination train arrives right on the opposite departure line to shorten the reverse time.

The train control apparatus 110 checks whether e_i-1(t) = Dpt_s that preceding train event e_i-1(t) arrives (Dpt_s) at station s. The train control apparatus 110 checks whether e_i(t) = Arr_s that subject train event e_i(t) is an arrival event (Arr_s). The train control apparatus 110 determines whether p_i(t) = p_i-1(t) that subject train course p_i(t) and preceding train course p_i-1(t) are the same. The train control apparatus 110 confirms whether x _i-1(t)-x_i-1(t)≥X₁ that the value obtained by subtracting the preceding train location value (x_i-1(t)) from the scheduled preceding train location value (x _i-1(t)) is equal to or greater than the set allowable train delay limit (X₁). The train control apparatus 110 checks whether x_i-1(t)-x_i(t)≤X₂ that the value obtained by subtracting the subject train location value (x_i(t)) from the location value x_i-1(t) of the preceding train is equal to or less than the set allowable train interval limit (X₂). The train control apparatus 110 determines, when the above-described scenario 1 conditions are all satisfied as e_i-1(t) = Dpt_s, e_i(t) = Arr_s, p_i(t) = p_i-1(t), v_i-1(t) = 0, x _i-1(t)-x_i-1(t)≥X₁, and x_i-1(t)-x_i(t)≤X₂, that it is scenario 1: the detour route determination scenario for the trailing train when the preceding train is obstructed.
The train control apparatus 110 checks whether e_i-1(t) = Dpt_s that preceding train event e_i-1(t) arrives (Dpt_s) at station s. The train control apparatus 110 checks whether ei(t) = Dpt_s that subject train event e_i(t) arrives (Dpt_s) at station s. The train control apparatus 110 determines whether p_i(t) ≠ p_i-1(t) that subject train course p_i(t) and preceding train course p_i-1(t) are unequal. The train control apparatus 110 checks whether v_i-1(t) = v_i(t) = 0 that preceding train speed v_{i -1}(t) and subject train speed v_i(t) are 0. The train control apparatus 110 confirms whether x _i-1(t)-x_i-1(t)≥X₁ that the value obtained by subtracting the preceding train location value (x_i-1(t)) from the scheduled preceding train location value (x _i-1(t)) is equal to or greater than the set allowable train delay limit (X₁). The train control apparatus 110 determines, when the above-described scenario 2 conditions are all satisfied as e_i-1(t) = Dpt_s, e_i(t) = Dpt_s, p_i(t) ≠ p_i-1(t), v_i-1(t) = v_i(t) = 0, and x _i-1(t)-x_i-1(t)≥X₁, that it is scenario 2: the pass sequence adjustment scenario due to track resource contention.
The train control apparatus 110 checks whether e_i(t) = End_s that subject train event e_i(t) is the end station arrival event (End_s). The train control apparatus 110 checks whether x _i-1(t)-x_i-1(t)≥X₁ that the value obtained by subtracting the subject train location value (x_i(t)) from the scheduled subject train location value (x _i(t)) is equal to or greater than the set allowable train delay limit (X₁). The train control apparatus 110 determines, when the above-described scenario 3 conditions are both satisfied as e_i(t) = End_s and x _i-1(t)-x_j-1(t)≥X₁), that it is scenario 3: the reverse train line adjustment scenario against train reverse delay.
FIG. 7 is a schematic block diagram of a train control learning server according to at least one embodiment.
The train control learning server 130 according to at least one embodiment includes, but not necessarily limited to, an information transmitting and receiving unit 710, an exceptional circumstance occurrence criterion unit 720, and an exceptional circumstance threshold determining unit 730. The train control learning server 130 may include other components.
The respective components included in the train control learning server 130 may be connected to a communication path connecting software modules or hardware modules in the apparatus and may operate organically with each other. These components communicate by using one or more communication buses or signal lines.
Each component of the train control learning server 130 shown in FIG. 7 represents a unit for processing at least one function or operation, and may be implemented as a software module, a hardware module, or a combination of software and hardware.
The information transmission/reception unit 710 receives subject train information and neighboring train information from a train control apparatus mounted in a plurality of trains when they enter the train depot.
The exceptional circumstance occurrence criterion unit 720 updates the exceptional circumstance occurrence thresholds by comparing the subject train information and the neighboring train information with the scheduled operating information. The exceptional circumstance occurrence criterion unit 720 transmits the exceptional circumstance occurrence thresholds to the subject train.
The exceptional circumstance threshold determining unit 730 updates the exceptional circumstance classification thresholds by learning the exceptional circumstance classification thresholds based on the operation state information with respect to the subject train information and the neighboring train information. The exceptional circumstance threshold determining unit 730 transmits the exceptional circumstance classification thresholds to the subject train.
FIG. 8 is a flowchart for explaining a train situation recognition method according to at least one embodiment.
The dynamic train control module 220 receives neighboring train information (real-time position, speed, course, event, train fault occurrence information, cabin congestion or on-board congestion) from a neighboring train through T2T communication, and receives track state information (track fault occurrence information and platform congestion or ground congestion) from the ATS 140 (S810).
In Step S810, the dynamic train control module 220 extracts, from the neighboring train information, the scheduled subject train location value (x _i-1(t)), subject train location value (x_i(t)), set allowable train delay limit (X₁), preceding train location value (x_i-1(t)), and set allowable train interval limit (X₂).
The dynamic train control module 220 checks whether exceptional circumstances occur based on at least one of neighboring train information and track state information (S820) .
In Step S820, the dynamic train control module 220 confirms that exceptional circumstances occur when it determines x _i-1(t)-x_i-1(t)≥X₁ that the value obtained by subtracting the subject train location value (x_i(t)) from the scheduled subject train location value (x _i(t)) is equal to or greater than the set allowable train delay limit (X₁), or when it determines x_i-1(t)-x_i(t)≤X₂ that the value obtained by subtracting the subject train location value (x_i(t)) from the preceding train location value x_i-1(t) is equal to or less than the set allowable train interval limit (X₂).
The dynamic train control module 220 extracts the train fault occurrence information included in the neighboring train information. The dynamic train control module 220 determines that exceptional circumstances occur based on at least one of the faulty train location, the time of occurrence of the fault, and the type of the fault included in the train fault occurrence information.
The dynamic train control module 220 extracts the passenger information in the train cabin, which is detected when the subject train included in the neighboring train information starts from the train stop. The dynamic train control module 220 calculates the cabin (on-board) congestion based on the passenger information in the train cabin. The dynamic train control module 220 confirms that exceptional circumstances occur when the cabin (on-board) congestion of the train exceeds the predetermined threshold.
The dynamic train control module 220 extracts the track fault occurrence information included in track state information. The dynamic train control module 220 confirms that exceptional circumstances occur, based on the track fault occurrence location, fault occurrence time, and fault type included in the track fault occurrence information.
The dynamic train control module 220 extracts the waiting passenger information on the next scheduled train stop of the subject train, which is included in the track state information. The dynamic train control module 220 calculates the platform congestion or ground congestion based on the waiting passenger information of the platform at the next scheduled stop. The dynamic train control module 220 confirms that exceptional circumstances occur when the platform congestion or ground congestion exceeds a preset threshold.
The dynamic train control module 220 performs, when it confirms that the exceptional circumstance occurs, to classify the exceptional circumstance into a specific exceptional circumstance of the predetermined exceptional circumstances, and recognize the train situation based on the specific exceptional circumstance (S830).
In Step S830, the dynamic train control module 220 checks whether e_i-1(t) = Dpt_s that preceding train event e_i-1(t) arrives (Dpt_s) at station s. The dynamic train control module 220 checks whether e_i(t) = Arr_s that subject train event e_i(t) is an arrival event (Arr_s). The dynamic train control module 220 checks whether p_i(t) = p_i-1(t) that subject train path p_i(t) and preceding train path p_{i -1}(t) are the same. The train condition recognition unit 216 checks whether v_i-1(t) = 0 that preceding train speed v_i-1(t) is 0. The dynamic train control module 220 confirms whether x _i-1(t)-x_i-1(t)≥X₁ that the value obtained by subtracting the preceding train location value (x_i-1(t)) from the scheduled preceding train location value (x _i-1(t)) is equal to or greater than the set allowable train delay limit (X₁). The dynamic train control module 220 checks whether x_i-1(t)-x_i(t)≤X₂ that the value obtained by subtracting the subject train location value (x_i(t)) from the preceding train location value x_i-1(t) is equal to or less than the set allowable train interval limit (X₂). The dynamic train control module 220 determines, when the aforementioned conditions are satisfied, that it is scenario 1: the detour route determination scenario for the trailing train when the preceding train is obstructed.
The dynamic train control module 220 checks whether e_i-1(t) = Dpt_s that preceding train event e_i-1(t) arrives (Dpt_s) at station s. The dynamic train control module 220 checks whether e_i(t) = Dpt_s that subject train event e_i(t) arrives (Dpt_s) at station s. The dynamic train control module 220 determines whether p_i(t) ≠ p_i-1(t) that subject train path p_i(t) and preceding train path p_i-1(t) are unequal. The dynamic train control module 220 checks whether v_i-1(t) = v_i(t) = 0 that preceding train speed v_i-1(t) and subject train speed v_i(t) are 0. The dynamic train control module 220 checks whether x _i-1(t)-x_i-1(t)≥X₁ that the value obtained by subtracting the preceding train location value (x_i-1(t)) from the scheduled preceding train location value (x _i-1(t) is equal to or greater than the set allowable train delay limit (X₁). The dynamic train control module 220 determines, when all of the above conditions are satisfied, that it is scenario 2: the pass sequence adjustment scenario due to track resource contention.
The dynamic train control module 220 checks whether e_i(t) = End_s that the subject train event e_i(t) is the end station arrival event (End_s). The dynamic train control module 220 checks whether x _i-1(t)-x_i-1(t)≥X₁ that the value obtained by subtracting the subject train location value (x_i(t)) from the scheduled subject train location value (x _i(t)) is equal to or greater than the set allowable train delay limit (X₁). The dynamic train control module 220 determines, when all of the above conditions are satisfied, that it is scenario 3: the reverse train line adjustment scenario against train reverse delay.
The dynamic train control module 220 dynamically controls the trains in accordance with the train conditions (S840) .
In Step S840, the dynamic train control module 220 calculates the severity of exceptional circumstances, and predicts, based on the severity, the cascaded delay time on trains sequentially affected by exceptional circumstances among the trains located on the same path (course) as the preceding train's. The dynamic train control module 220 selects an optimal alternative that minimizes the cascaded delay time among the plurality of pre-stored alternatives. The dynamic train control module 220 controls the trains trailing behind the preceding train to travel along the path according to the optimal alternative.
Although Steps S810 to S840 in FIG. 8 are described to be sequentially performed, they are not limited thereto. In other words, various applications are made possible by changing the sequence of steps illustrated by FIG. 8 or by performing two or more of the steps in parallel, and hence the steps in FIG. 9 are not limited to the illustrated chronological sequences.
FIG. 9 is a flowchart for explaining a method of learning an exceptional circumstances determination threshold by a train control server according to at least one embodiment.
The train control learning server 130 receives the receives subject train information and neighboring train information from the train control apparatus 110 mounted in a plurality of trains when they enter the train depot (S910).
The train control learning server 130 updates the exceptional circumstance occurrence thresholds by comparing the subject train information and the neighboring train information with the scheduled operating information. The train control learning server 130 transmits the exceptional circumstance occurrence thresholds to the subject train (S920) .
The train control learning server 130 updates the exceptional circumstance classification thresholds by learning the exceptional circumstance classification thresholds based on the operation state information with respect to the subject train information and the neighboring train information. The train control learning server 130 transmits the exceptional circumstance classification thresholds to the subject train (S930).
Although Steps S910 to S930 in FIG. 9 are described to be sequentially performed, they are not limited thereto. In other words, various applications are made possible by changing the sequence of steps illustrated by FIG. 9 or by performing two or more of the steps in parallel, and hence the steps in FIG. 9 are not limited to the illustrated chronological sequences.
FIG. 10 is a flowchart for explaining a dynamic train control method according to at least one embodiment.
The train control apparatus 110 checks whether or not exceptional circumstances occur in the preceding train based on the train operation state and the track operation state (S1010) .
The train control apparatus 110 receives, from the neighboring train information, the preceding train event e_i-1(t), subject train event e_i(t), arrival (departure) (Dpt_s) at station 's' , arrival event (Arr_s), preceding train course (p_i-1(t)), subject train course (p_i(t)), preceding train speed (v_i-1(t)), subject train speed (v_i(t)), scheduled preceding train location value (x _-1(t)), scheduled subject train location value (x _i(t)), preceding train location value (x_i-1(t)), subject train location value (x_i(t)), and end station arrival event (End_s).
The train control apparatus 110 classifies the exceptional circumstances by using multiple pieces of information extracted from the neighboring train information, into a specific exceptional circumference among scenario 1 that is a detour route determination scenario for the trailing train when the preceding train is obstructed, scenario 2 that is a pass sequence adjustment scenario due to track resource contention, and scenario 3 that is a reverse train line adjustment scenario against train reverse delay (S1020) .
The train control apparatus 110 calculates the severity of each of the scenarios 1, 2 and 3 resulting from classifying the exceptional circumstances (S1030).
The train control apparatus 110 checks whether the severity is critical for the scenarios 1, 2 and 3 resulting from classifying the exceptional circumstances (S1040). In Step S1040, the train control apparatus 110 receives exceptional circumstance duration ti-i, external notification (input) end time τ_i-1, and pre-learned thresholds T_i-1. The train control apparatus 110 determines the severity is critical when exceptional circumstance duration ti-i and external notification (input) end time τ_i-1 (max (t_i-1, τ_i-1)) are equal to or greater than the pre-learned thresholds T_i-1. The train control apparatus 110 determines that the severity is minor when exceptional circumstance duration t_i-1 and external notification (input) end time τ_i-1 (max (t_i-1, τ_i-1)) are less than the pre-learned thresholds T_i-1.
When the checking in Step S1040 determines that the severity is critical for the scenarios 1, 2 and 3 resulting from classifying the exceptional circumstances, the train control apparatus 110 predicts, based on the severity, the cascaded delay time on trains sequentially influenced by the exceptional circumstances among the trains on the same path (course) as the preceding train's (S1050).
In Step S1050, the train control apparatus 110 generates, based on the cascaded delay time, the course hold offer (alternative 1) and the course change offer (alternative 2) for the train trailing the preceding train.
The train control apparatus 110 checks whether or not the course change offer (alternative 2) can be selected as an optimal alternative that minimizes the cascaded delay time out of the course change offer (alternative 1) and the course change offer (alternative 2) for the train trailing the preceding train (S1060).
When the checking in Step S1060 determines that the course change offer (alternative 2) can be selected as the optimal alternative that minimizes the cascaded delay time, the train control apparatus 110 calculates the number of sequentially affected trains (the number of delayed trains) that occur when the train trailing the preceding train is sent on ahead by the auxiliary main track or detour (dynamically re-routed) following the course change offer (alternative 2) (S1070).
The train control apparatus 110 calculates the number of sequentially affected trains (delayed train number) that occur when exceptional circumstances are maintained following the course hold offer (alternative 1).
The train control apparatus 110 determines whether the course change offer (alternative 2) is better from a comparison between the number of sequentially affected trains (the number of delayed trains) that occur when exceptional circumstances are maintained following the course hold offer (alternative 1), and the number of sequentially affected trains (the number of delayed trains) that occur when the train trailing the preceding train is sent on ahead by the auxiliary main track or detour (dynamically re-routed) following the course change offer (alternative 2) (S1080).
In Step S1080, the train control apparatus 110 selects the course change offer (alternative 2) when the number of sequentially affected trains (the number of delayed trains) that occur when the train trailing the preceding train is sent on ahead by the auxiliary main track or detour (dynamically re-routed) falls below the number of sequentially affected trains (the number of delayed trains) that occur when exceptional circumstances are maintained, by a drop threshold or more.
The train control apparatus 110 selects the course change offer (alternative 2) when the delayed arrival time and congestion that occur when the train trailing the preceding train is sent on ahead by the auxiliary main track or detour (dynamically re-routed) fall below the delayed arrival time and congestion that occur when exceptional circumstances are maintained, by predetermined drop thresholds or more.
As a result of checking in Step S1080 confirming that the course change alternative (alternative 2) is better, the train control apparatus 110 employs the course change alternative (alternative 2) (S1090).
Although Steps S1010 to S1090 in FIG. 10 are described to be sequentially performed, they are not limited thereto. In other words, various applications are made possible by changing the sequence of steps illustrated by FIG. 10 or by performing two or more of the steps in parallel, and hence the steps in FIG. 10 are not limited to the illustrated chronological sequences.
As described above, the dynamic train control method according to at least one embodiment illustrated in FIG. 10 can be implemented by a program and recorded in a computer-readable recording medium. The computer-readable recording medium on which the program for implementing the dynamic train control method according to at least one embodiment includes all kinds of recording devices for storing data that can be read by a computer system.
FIG. 11 is a flowchart for explaining an alternatives application method according to at least one embodiment.
The train control apparatus 110 sets the course hold offer (alternative 1) for the train trailing the preceding train to minimize the cascaded delay time among the plurality of pre-stored alternatives (S1110). The train control apparatus 110 sets the course change offer (alternative 2) for the train trailing the preceding train to minimize the cascaded delay time among the plurality of pre-stored alternatives (S1112).
The train control apparatus 110 predicts (calculates) the number of sequentially affected trains (the number of delayed trains) that occur when exceptional circumstances are maintained following the course hold offer (alternative 1) (S1120). The train control apparatus 110 predicts (calculates) the number of sequentially influenced trains that occur when the train trailing the preceding train is sent on ahead by the auxiliary main track or detour (dynamically re-routed) following the course change offer (alternative 2) (S1122).
The train control apparatus 110 compares the number of sequentially affected trains (the number of delayed trains) that occur when exceptional circumstances are maintained following the course hold offer (alternative 1) with the number of sequentially affected trains (the number of delayed trains) that occur when the train trailing the preceding train is sent on ahead by the auxiliary main track or detour (dynamically re-routed) following the course change offer (alternative 2), and based on the comparison result, the train control apparatus 110 checks whether the course change plan (alternative plan 2) is better than course hold offer (alternative 1) (S1130).
When the checking in Step S1130 concludes that the course change plan (alternative plan 2) is better, the train control apparatus 110 communicates with the ATS 140 to check whether or not the train schedule can be changed toward the course following the course change plan (alternative plan 2) (S1140).
When the checking in Step S1140 concludes that the train schedule can be changed toward the course following the course change plan (alternative plan 2), the train control apparatus 110 takes action on the course change offer (alternative 2) (S1150). When the checking in Step S1130 concludes that the course change plan (alternative plan 2) is unfavorable, the train control apparatus 110 takes action on the course change offer (alternative 1) (S1152).
Although Steps S1110 to S1152 in FIG. 11 are described to be sequentially performed, they are not limited thereto. In other words, various applications are made possible by changing the sequence of steps illustrated by FIG. 11 or by performing two or more of the steps in parallel, and hence the steps in FIG. 11 are not limited to the illustrated chronological sequences.
FIG. 12 is a flowchart for explaining the severity determination method according to at least one embodiment.
The train control apparatus 110 extracts predetermined thresholds T_i from the pre-stored database to determine the severity for a scenario [i]. The train control apparatus 110 receives, from the ATS 140, information on exceptional circumstance duration ti and external notification end time τ_i with respect to the exceptional circumstances classified as past events [i] (S1210).
In Step S1210, the train control apparatus 110 also figures out the driving environment/condition (headway, distance between stations, congestion, etc.) at the relevant past time point, and compares the same with the operating environment/condition in the current situation, and then extracts highly relevant thresholds T_i. For learning and default setting of T_i, the train control apparatus 110 obtains, stores, and manages information on exceptional circumstance duration t_i, external notification end time τ_i and their related information with respect to the exceptional circumstances classified as past events [i].
The train control apparatus 110 checks whether (max (t_i, τ_i))≥T_i that maximum exceptional circumstance duration ti and maximum external notification end time (τ_i) max (t_i, τ_i) are equal to or greater than preset thresholds T_i (S1220).
As a result of the checking in Step S1220, when (max (t_i, τ_i))≥T_i that maximum exceptional circumstance duration ti and maximum external notification end time (T_i) (max (t_i, τ_i)) are equal to or greater than the predetermined thresholds T_i, the train control apparatus 110 determines that the severity is critical (S1230).
As a result of the checking in Step S1220, when (max (t_i, τ_i))<T_i that maximum exceptional circumstance duration t_i and maximum external notification end time (τ₁) (max (t_i, T_i)) are less than the predetermined thresholds T_i, the train control apparatus 110 determines that the severity is minor (S1230) .
Although Steps S1210 to S1230 in FIG. 12 are described to be sequentially performed, they are not limited thereto. In other words, various applications are made possible by changing the sequence of steps illustrated by FIG. 12 or by performing two or more of the steps in parallel, and hence the steps in FIG. 12 are not limited to the illustrated chronological sequences.
FIG. 13 is an exemplary view showing cascaded delay time according to at least one embodiment.
The severity represents the event duration/delay time of the train. The cascaded delay time means the delay time and congestion of all trains sequentially influenced by the relevant event.
The train control apparatus 110 uses past learning data to predict the cascaded delay time and determine the severity.
The train control apparatus 110 extracts the minimum threshold for the duration of the event which enables the event to be determined critical when the exceptional circumstances are classified into the scenarios 1, 2 and 3. When the exceptional circumstances are classified into the scenarios 1, 2 and 3, the train control apparatus 110 can apply differentiated thresholds according to the driving environment/condition (headway, distance between stations, congestion, etc.).
The train control apparatus 110 extracts the minimum threshold for the duration of the event which enables the event to be determined critical when the exceptional circumstances are classified into scenario 'i'. When the exceptional circumstances are classified into scenario 'i', the train control apparatus 110 can apply differentiated thresholds according to the driving environment/condition (headway, distance between stations, congestion, etc.).
The train control apparatus 110 uses the information received at the current point of time via train-to-train (T2T) communications to determine the cascaded delay time and the severity.
When the exceptional circumstances are classified into scenarios 1 and 2, the train control apparatus 110 receives the event classification and the corresponding event duration (the first time point to the present time) through T2T communication. When the exceptional circumstances are classified into scenario 3, the train control apparatus 110 receives the event classification and the corresponding event duration (operation delay time) up to now (the first time point to the present time) through the T2T communication.
When the exceptional circumstance is classified into scenario 'i', the train control apparatus 110 receives the event classification and the corresponding event duration up to now (the first time point to the present time) through the T2T communication. The train control apparatus 110 communicates the predicted end time of the event from the outside (ground control system) to the onboard device.
The operation for determining the severity by the train control apparatus 110 is as follows.
When the exceptional circumstances are classified as scenarios 1 and 2, and when the train control apparatus 110 determines at that time that the recognized exceptional circumstance events are critical, it performs additional adjustment to predict their cascaded delay time. The train control apparatus 110 does not perform separate tasks when the exceptional circumstance events are determined to be minor.

① The train control apparatus 110 determines that the severity of the exceptional circumstance event is critical when the event duration up to now is greater than the learned duration (default setting).
② The train control apparatus 110 determines that the severity of the exceptional circumstance event is critical when the event duration time inclusive of an externally given end time point is greater than the learned duration.
③ The train control apparatus 110 determines whether the severity is critical or minor, when no end time point is given at the current point of time, by assuming the corresponding event duration ends after an integer multiple of the event duration up to now. Here, a preset value of the integer multiple is updated through learning.

The train control apparatus 110 performs an additional adjustment operation to predict the cascaded delay time when the exceptional circumstance is classified as scenario 3 and when it determines at that time that the recognized exceptional circumstance event is critical. The train control apparatus 110 does not perform separate tasks when the exceptional circumstance event is determined to be minor.

① The train control apparatus 110 determines that the severity of the exceptional circumstance event is critical when the train delay up to now is greater than the learned duration (default setting).
② The train control apparatus 110 handles the severity, assuming that the current level of delay is maintained if there are no additional factors.

The train control apparatus 110 performs additional adjustment to predict the cascaded delay time when the exceptional circumstance is classified as scenario 'i' 3 and when it determines at that time that the recognized exceptional circumstance event is critical. The train control apparatus 110 does not perform separate tasks when the exceptional circumstance event is determined to be minor.

① The train control apparatus 110 determines that the severity of the exceptional circumstance event is critical when the event duration up to now is greater than the learned duration (default setting).
② The train control apparatus 110 determines that the severity of the exceptional circumstance event is critical when the event duration inclusive of an externally given end time point is greater than the learned duration.
③ The train control apparatus 110 determines whether the severity is critical or minor, when no end time point is informed at the current point of time, by assuming the corresponding event duration ends after an integer multiple of the event duration up to now. Here, a preset value of the integer multiple is updated through learning. The severity determination operation by the train control apparatus 110 is as shown in FIG. 12.

The train control apparatus 110 utilizes T_i (thresholds for determining the severity for scenario [i]) for learning. When the exceptional circumstance is classified into scenario 'i', the train control apparatus 110 picks up the corresponding event duration ti from the neighboring train through T2T communication. The train control apparatus 110 selectively delivers predicted end time τ_i from the ground system to the onboard device by way of information transmitted on the same channel.
For learning and default setting of Ti, the train control apparatus 110 obtains, stores, and manages information on exceptional circumstance duration ti and external notification end time τ_i with respect to the exceptional circumstances that are classified as past events [i].
The train control apparatus 110 also figures out the driving environment/condition (headway, distance between stations, congestion, etc.) at the relevant past time point, and compares the same with the operating environment/condition in the current situation, and then provides highly relevant thresholds T_i.
The operation of the train control apparatus 110 for performing a prediction of the cascaded delay time is as follows.
The train control apparatus 110 sets for prediction targets, when the exceptional circumstances are classified into scenarios 1, 2 and 3, such cascaded delay time as the number of delayed trains, delay of train by the hour and minute, and estimated passenger congestion/travel time increase, with respect to a given event duration. The train control apparatus 110 uses, for a prediction method, a trailing-train train delay time estimation method in a first-come-first-serve (no control) concept.
FIG. 14 is an exemplary diagram showing dynamic path setting according to at least one embodiment.
The train control apparatus 110 sets no change in the train operation order and the route as a precondition for alternative 1.
The train control apparatus 110, when given the temporal (reviewing) range for alternative 1 by externally (ground control system) given end time point information, calculates the chain delay between trains based on the current event of the event-relevant train by taking account of, as far as, the time when the external end time point information is given. The train control apparatus 110, when not given the temporal (reviewing) range for alternative 1 by externally given end time point information, calculates the chain delay between trains by taking account of a time point obtained by multiplying the event duration up to now by an integer multiple, as a default.
The train control apparatus 110 dynamically adjusts the sequence/route of the trains for each scenario as a precondition for alternative 2.
The train control apparatus 110 calculates the chain delay between trains by immediately reflecting alternative 2 at the current time point for the temporal (reviewing) range for alternative 2. The train control apparatus 110 sets the temporal range up to the last train estimated due to the chain delay, to be the temporal (reviewing) range for alternative 2.
The train control apparatus 110 estimates the chain delay by sequentially setting the main track train intervals that can be set based on the braking distance calculated by the ATCS based on the current train speed, to a minimum value.
The train control apparatus 110 estimates the chain delay from the sequence of train delay → increase of arriving passengers → increase in boarding time → increase in stopping time → delay in train departure time → repetition of vicious cycle, by predicting the increase of arriving passengers and then calculating the increase in stopping time (station dwell time). The train control apparatus 110 calculates the number of trains, the delay by the hour and minute per train, and the total delay by the hour and minute which are related to the chain delay.

Although exemplary embodiments of the present disclosure have been described for illustrative purposes, those skilled in the art will appreciate that various modifications, additions and substitutions are possible, without departing from the various characteristics of the disclosure. Therefore, exemplary embodiments of the present disclosure have been described for the sake of brevity and clarity. Accordingly, one of ordinary skill would understand the scope of the disclosure is not limited by the explicitly described above embodiments but by the claims and equivalents thereof.

REFERENCE NUMERALS

110:	train control apparatus
120:	relay apparatus
130:	train control learning server 140: ATS
210:	train condition recognition module
212:	communication unit
214:	running condition determining unit
216:	exceptional circumstance determining unit
218:	train condition recognition unit
220:	dynamic train control module
222:	severity determining unit
224:	cascaded delay time prediction unit
226:	alternatives selection unit
228:	train operation unit
710:	information transmission/reception unit
720:	exceptional circumstance classification criterion unit
730:	exceptional circumstance threshold determining unit

Claims

An apparatus for providing a dynamic train control, comprising:
a train condition recognition unit configured to check whether there is an occurrence of an exceptional circumstance of a preceding train based on a train operation state and a track operation state;

a severity determining unit configured to calculate a severity of the exceptional circumstance;

a cascaded delay time prediction unit configured to predict, based on the severity, a cascaded delay time on trains sequentially affected by the exceptional circumstance among trains located on the same path as the preceding train;

an alternatives selection unit configured to select an optimal alternative among a plurality of pre-stored alternatives to minimize the cascaded delay time; and

a train operation unit configured to perform a control so that a train trailing the preceding train travels in the course following the optimal alternative.
The apparatus of claim 1, wherein the severity determining unit is configured
to receive an exceptional circumstance duration (t_i-1), an external notification end time (τ_i-1), and one or more pre-learned thresholds (T_i-1) with respect to the preceding train,
to determine that the severity is critical when the exceptional circumstance duration (t_i-1) and the external notification end time (τ_i-1 (max (t_i-1, τ_i-1)) are equal to or greater than the pre-learned thresholds (τ_i-1), and
to determine that the severity is minor when the exceptional circumstance duration (t_i-1) and the external notification end time (τ_i-1 (max (t_i-1, τ_i-1)) are less than the pre-learned thresholds (T_i-1).
The apparatus of claim 2, wherein the severity determining unit is configured,
to set, when no external notification end time (τ_i-1) is received, an integer multiple of the exceptional circumstance duration (t_i-1) up to now as an estimated notification end time,
to determine that the severity is critical when the estimated notification end time is equal to or greater than the pre-learned threshold (τ_i-1), determine that the severity is minor when the estimated notification end time is less than the pre-learned threshold (T_i-1).
The apparatus of claim 2, wherein the cascaded delay time prediction unit is configured
to predict the cascaded delay time when the severity of the exceptional circumstance is determined to be critical as of a current time point, and
to leave a separate task of predicting the cascaded delay time unperformed when the severity of the exceptional circumstance is determined to be minor.
The apparatus of claim 4, wherein the cascaded delay time prediction unit is configured
to calculate delayed arrival times for all trains sequentially influenced by the exceptional circumstance among the trains located on the same path as the preceding train, and calculate a congestion corresponding to the delayed arrival times, and
to calculate a detour arrival time which occurs when the train trailing the preceding train is sent on ahead by an auxiliary main track among all trains sequentially influenced by the exceptional circumstance, and calculate a congestion corresponding to the detour arrival time.
The apparatus of claim 5, wherein the alternatives selection unit is configured to select, based on the cascaded delay time, one of a course hold offer and a course change offer for the train trailing the preceding train.
The apparatus of claim 6, wherein the alternatives selection unit is configured
to perform a comparison of a first number of sequentially affected trains that occur when the exceptional circumstance is maintained following the course hold offer with a second number of sequentially affected trains that occur when the train trailing the preceding train is sent on ahead by the auxiliary main track following the course change offer, and
to select one of the course hold offer and the course change offer according to the comparison.
The apparatus of claim 7, wherein the alternatives selection unit is configured
to select the course change offer when the second number of sequentially affected trains that occur when the train trailing the preceding train is sent on ahead by the auxiliary main track, falls below the first number of sequentially affected trains that occur when the exceptional circumstance is maintained by a drop threshold or more, and
to select the course hold offer when the second number of sequentially affected trains that occur when the train trailing the preceding train is sent on ahead by the auxiliary main track equals to or falls below the first number of sequentially affected trains that occur when the exceptional circumstance is maintained by less than the drop threshold.
The apparatus of claim 7, wherein the alternatives selection unit is configured
to select the course change offer when a delayed arrival time and a congestion that occur when the train trailing the preceding train is sent on ahead by the auxiliary main track fall below a delayed arrival time and a congestion that occur when the exceptional circumstance is maintained, by drop thresholds or more, and
to select the course hold offer when the delayed arrival time and the congestion that occur when the train trailing the preceding train is sent on ahead by the auxiliary main track are equal to or fall below the delayed arrival time and congestion that occur when exceptional circumstances are maintained, by less than the drop thresholds.
The apparatus of claim 1, wherein the train condition recognition unit is responsive to the exceptional circumstance that occurs, for classifying the exceptional circumstance into a specific exceptional circumstance of predetermined exceptional circumstances.
The apparatus of claim 10, wherein the train condition recognition unit is configured
to classify the exceptional circumstance based on information that is extracted from neighboring train information, into a specific exceptional circumference among a detour route determination scenario for the trailing train when the preceding train is obstructed, a pass sequence adjustment scenario due to a track resource contention, and a reverse train line adjustment scenario against a train reverse delay,
wherein the information that is extracted from neighboring train information comprises:
a preceding train event (e_i-1(t)), a subject train event (e_i(t)), an arrival (Dpt_s) at a station (s), an arrival event (Arr_s), a preceding train course (p_i-1(t)), a subject train course (p_i(t)), a preceding train speed (v_i-1(t)), a subject train speed (v_i(t)), a scheduled preceding train location value (x _i-1(t)), a scheduled subject train location value (x _i(t)), a preceding train location value (x_i-1(t)), a subject train location value (x_i(t)), and an end station arrival event (End_s).
The apparatus of claim 11, wherein severity determining unit is configured to calculate the severity by using pre-learned thresholds (T_i) which are learned differently for each of the detour route determination scenario, the pass sequence adjustment scenario, and the reverse train line adjustment scenario.
The apparatus of claim 7, wherein the alternative selection unit is configured to select one of the course hold offer and the course change offer by using thresholds employed differently for each of the detour route determination scenario, the pass sequence adjustment scenario, and the reverse train line adjustment scenario.
A method of providing a dynamic train control, comprising:
checking whether there is an occurrence of an exceptional circumstance of a preceding train based on a train operation state and a track operation state;

calculating a severity of the exceptional circumstance;

predicting, based on the severity, a cascaded delay time on trains sequentially affected by the exceptional circumstance among trains located on the same path as the preceding train;

selecting an optimal alternative among a plurality of pre-stored alternatives to minimize the cascaded delay time; and

performing a control so that a train trailing the preceding train travels in the course following the optimal alternative.