Autonomous railway system using artificial intelligence

The autonomous railway system with AI-controlled cylindrical wheels and safety features addresses hunting motion and mechanical switch reliance, enhancing stability and efficiency.

WO2026139923A1PCT designated stage Publication Date: 2026-07-02PARK KYEUNGSIK

Patent Information

Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
PARK KYEUNGSIK
Filing Date
2025-12-25
Publication Date
2026-07-02

Smart Images

  • Figure IB2025063454_02072026_PF_FP_ABST
    Figure IB2025063454_02072026_PF_FP_ABST
Patent Text Reader

Abstract

The present invention relates to an autonomous railway system using artificial intelligence, which is a technology for controlling the rotation of wheels by using an artificial intelligence unit mounted on a railway vehicle (10) driving along a rail (500) so that the railway vehicle (10) drives autonomously. According to the present invention, the autonomous railway system comprises an active driving sensor unit (630), a control unit (610), and an artificial intelligence unit, wherein the artificial intelligence unit includes an active driving artificial intelligence unit (600) and an autonomous driving artificial intelligence unit (700). In the railway vehicle (10), a single-axle active bogie (300) including left and right independently rotating cylindrical wheels (110) is directly connected to a car body and arranged, and safety wheels (210) to prevent derailment in an emergency are equipped. Accordingly, the present invention provides the autonomous railway system in which the railway vehicle (10) can autonomously select a driving route without relying on a mechanical switch machine.
Need to check novelty before this filing date? Find Prior Art

Description

Autonomous railway system using artificial intelligence

[0001] The present invention relates to an autonomous railway system using artificial intelligence. More specifically, the invention relates to an autonomous railway system using artificial intelligence that enables an active railway vehicle to travel along a rail without derailing and to smoothly pass through curved sections without friction and noise, comprising a wheel unit including a cylindrical wheel that rotates and travels independently on the left and right sides and a safety wheel that prevents derailment in an emergency, and a wheel control system that actively controls the rotational speed of the cylindrical wheel by artificial intelligence.

[0002] Furthermore, the present invention relates to an autonomous railway system in which a railway vehicle independently selects a driving direction and passes through a branching section composed of fixed rails without relying on externally physically moving mechanical switches, and belongs to the field of technology that implements autonomous driving by independently determining the direction of travel and steering by controlling the vehicle's driving state and wheel rotation state using artificial intelligence.

[0003] Furthermore, the present invention relates to an autonomous railway system that solves the problems of hunting motion and reduced driving stability that occurred in conventional conical wheels by utilizing the characteristic that the outer diameter of a cylindrical wheel is constant to limit the driving distance control variable to the rotational speed of the wheel and controlling the rotational speed with an artificial intelligence-based control system.

[0004] Conventional railway vehicles adopt a structure in which a pair of left and right wheels are fixed to a single axle using conical wheels to travel along the rails. These fixed-axle conical wheels have an inclined outer surface, which prevents derailment by providing a geometric restoring force that returns the wheel center to the center due to the difference in contact diameter if the wheel center deviates from the rail center. Furthermore, in curved sections, the vehicles are configured to pass through the curved rails without a separate steering mechanism by utilizing the difference in contact diameter between the left and right wheels.

[0005] However, in such conical wheel structures, the center-restoring action is combined with inertia, inevitably causing hunting oscillation in which the center of the wheel repeatedly crosses the center of the rail. This hunting oscillation not only causes lateral swaying of the railway vehicle, thereby degrading ride comfort, but also amplifies vibration loads as the driving speed increases, increasing the risk of derailment and causing problems such as wear on the wheels and rails and track damage.

[0006] Furthermore, because the difference in contact diameter of conical wheels is insufficient to adequately accommodate the difference in travel distance between the left and right rails that occurs in sharp curves, the wheel flange comes into contact with the side of the rail in these sections. This results in increased noise and friction, as well as accelerated wear of the wheels and rails. Although derailment prevention rails are installed on the inner rails to prevent these problems, this approach has limitations, including the generation of severe noise, a complex structure, and high maintenance costs.

[0007] In conventional railway systems, railway vehicles cannot independently select their direction of travel; instead, a manual driving method is employed where trains follow tracks opened by external mechanical switches at branching points. Consequently, complex and expensive mechanical switches are mandatory at these points, and failures or malfunctions in these switches can lead to serious accidents such as derailments. Furthermore, since the operation of these switches and train travel are controlled by a centralized control system, expanding the system is difficult, and flexibility and efficiency are limited.

[0008] Although a structure in which the left and right wheels rotate independently has been proposed to mitigate friction problems occurring when passing through curves, conventional independently rotating wheels primarily utilize conical wheels, causing the effective diameter of the wheel to vary depending on the contact position with the rail. Consequently, it is difficult to accurately predict the relationship between the wheel's rotation speed and the actual travel distance, which limits the ability to precisely control the driving direction and position. Furthermore, some technologies electronically synchronize the left and right wheels to operate them effectively like a fixed axis; as a result, the problem of hunting motion remains unresolved, and the system is unable to move beyond manual driving.

[0009] When using cylindrical wheels with a constant outer diameter, the travel distance is determined precisely in proportion to the wheel's rotational speed, enabling accurate control of the travel distance and structurally eliminating the hunting motion that occurred in conventional conical wheels. Since cylindrical wheels lack the geometric center-restoring force provided by conical wheels, new control technology is required to precisely control the wheel's rotational speed to prevent derailment and ensure stable travel along the rail.

[0010] The first objective of the present invention is to provide an autonomous railway system utilizing artificial intelligence that allows a railway vehicle to independently select its direction of travel and pass through branching sections, moving away from the manual driving method in conventional railway systems where the railway vehicle travels along rails physically opened by external mechanical switches. The purpose of this autonomous railway is to reduce dependence on complex and expensive mechanical switches and centralized control systems.

[0011] The second objective of the present invention is to provide an autonomous railway system using artificial intelligence that can fundamentally eliminate the hunting motion inevitably caused by the geometric characteristics of conventional conical wheels, thereby reducing vibration, noise, and wear on wheels and rails during railway vehicle operation, and ensuring driving stability even in high-speed driving environments.

[0012] The third objective of the present invention is to provide an autonomous railway system using artificial intelligence that determines the driving direction itself through accurate control of the driving distance by applying a cylindrical wheel with a constant outer diameter instead of a conical wheel whose effective diameter changes depending on the contact position with the rail, thereby clarifying the relationship between the number of rotations of the wheel and the driving distance, and limiting the driving distance control variable to the rotational speed of the wheel.

[0013] The fourth objective of the present invention is to provide an autonomous railway system using artificial intelligence capable of active driving, which enables a railway vehicle to travel along the rail without derailing and to smoothly pass through curved sections by actively controlling the rotational speed of cylindrical driving wheels that rotate independently on the left and right sides using artificial intelligence.

[0014] The fifth objective of the present invention is to provide an autonomous railway system using artificial intelligence that utilizes an artificial intelligence algorithm to achieve stable control by overcoming the time difference in which the effect occurs when changing the rotational speed to correct the error of the wheel center deviating from the rail center.

[0015] The sixth objective of the present invention is to provide an autonomous railway system using artificial intelligence that ensures the safety of a railway vehicle at any time by preventing derailment through safety wheels that do not come into contact with the rail during normal driving conditions, but support the inner surface of the rail in emergency situations such as emergency braking, external collision, or loss of control.

[0016] To solve the above problem, the autonomous railway system using artificial intelligence according to the present invention comprises a sensor unit mounted on a railway vehicle (10) that travels along a rail (500) with cylindrical wheels that rotate independently left and right to detect the driving state, an artificial intelligence unit that processes driving data input from the sensor unit, and a control unit that controls the rotation of the wheels according to the control of the artificial intelligence unit. The artificial intelligence unit is configured to determine the driving state of the railway vehicle (10) based on the driving data and to autonomously control the rotation of the wheels according to the result of the determination.

[0017] In addition, the artificial intelligence unit comprises a first-stage active driving artificial intelligence unit (600) that controls the rotation of left and right cylindrical wheels so that the railway vehicle (10) travels along the rail without derailing, and a second-stage autonomous driving artificial intelligence unit (700) that controls the railway vehicle (10) to select its own driving direction at the rail branching section (510). Thus, the present invention is configured to perform autonomous driving functions in stages by implementing active driving and autonomous driving hierarchically separated.

[0018] Additionally, the active driving artificial intelligence unit (600) controls the left wheel control unit (611) and the right wheel control unit (612) respectively through the control unit (610) to independently control the rotational speed of the left cylindrical wheel (111) and the right cylindrical wheel (112). To this end, the active driving sensor unit (630) is configured to include one or more of a position information sensor (631), an image information sensor (632), and a load information sensor (633), and provides relative position and driving status information between the rail (500) and the wheel to the artificial intelligence unit (600).

[0019] Additionally, the above-mentioned railway vehicle (10) may be configured to use a battery mounted on the vehicle as a power source so that it can be driven without an external power supply. The battery consists of a battery module (810) placed in the lower part of the vehicle or inside the vehicle body, and is configured to be replaceable by a battery exchange system (800) while the railway vehicle (10) is stopped at a station. Such a battery-based power configuration is applied independently of the driving judgment and control logic of the autonomous driving artificial intelligence unit (700) or the active driving artificial intelligence unit (600), and corresponds to an example of a configuration that provides a choice of power supply method without changing the driving control structure of the railway vehicle.

[0020] Additionally, the wheel is composed of a pair of cylindrical wheels (110) that are separated left and right and can rotate independently, with the outer surface having a cylindrical shape, and the cylindrical wheels (110) are composed of a wheel unit (100) and mounted on a 1-axle active bogie (300). The 1-axle active bogie (300) is configured to rotate horizontally with respect to the driving direction, and the driving direction of the railway vehicle (10) is actively controlled according to the control of the artificial intelligence unit.

[0021] In addition, the above-mentioned 1-axis active bogie (300) is further equipped with a safety wheel (210) having a structure that supports the inner surface of the rail (500) at four points to prevent derailment in abnormal conditions where lateral movement or horizontal rotation occurs while the railway vehicle (10) is traveling along the rail. The safety wheel (210) does not come into contact with the rail during normal travel, but supports the inner surface of the rail in an emergency situation to prevent the derailment of the cylindrical wheel (110).

[0022] In addition, the autonomous railway system according to the present invention can be applied not only to railway vehicles for passenger transport but also to freight trains for transporting cargo. In this case, the freight train is equipped with a two-axle bogie including a pair of axles arranged front and rear to travel along a rail (500), and the two-axle bogie can be implemented as one embodiment in which a plurality of one-axle active bogies (300) are structurally combined. The artificial intelligence control unit is configured to detect and control the driving state of the driving wheels or bogies mounted on the two-axle bogie, and this corresponds to an example of a configuration that extends the driving control principle applied to the one-axle active bogie.

[0023] In addition, the freight train may use a battery mounted on the vehicle as a power source to enable operation without an external power supply, and the battery may be configured to be replaceable.

[0024] This battery-based freight train configuration applies the same principles as the battery exchange systems used in passenger railway vehicles, enabling the operation of autonomous railway systems even in freight transport environments.

[0025] In addition, the artificial intelligence control unit according to the present invention may be configured to control the driving state of each freight train even when multiple freight trains are operated in a platooning mode where they are interconnected and driving, or in a distributed mode where they are driven independently. In this case, platooning or distributed driving refers to a difference in the mode of operation where the same autonomous driving control principle is applied to multiple railway vehicles, and does not mean the additional introduction of a new driving control algorithm or judgment logic.

[0026] In addition, the autonomous driving method according to the present invention is performed by including a driving information input step (S10), a target rotational speed calculation step (S20), an interval data measurement step (S30), a reference value comparison step (S40), a target rotational speed correction step (S50), an individual speed command generation step (S60), and an inverter control step (S70).

[0027] In the above control process, the artificial intelligence preemptively corrects the rotational speed of the cylindrical wheel (110) by reflecting the error caused by the measurement delay and the time difference in power transmission.

[0028] In addition, the autonomous driving method according to the present invention may include a power transmission step in which the rotational force generated by the inverter control step (S70) is transmitted to a cylindrical wheel through a driving motor to drive a railway vehicle.

[0029] In addition, in the target rotational speed correction step (S50), an artificial intelligence algorithm may be applied to preemptively correct errors caused by the time difference between the time of calculating the target rotational speed and the time of actual power transmission. Such preemptive correction is an example of a control supplementation means to maintain a proportional relationship between the rotational speed of the cylindrical wheel and the travel distance.

[0030] According to the present invention, a railway vehicle (10) can move away from a manual driving method in which it travels along a rail physically opened by an external mechanical switch and can pass through a branching section (510) by selecting its own driving direction under the control of an autonomous driving artificial intelligence unit (700). Accordingly, the waiting time of the railway vehicle is eliminated, increasing the operating speed, and reducing the burden of installing and maintaining complex and expensive mechanical switches.

[0031] The present invention applies a cylindrical wheel (110) having a cylindrical shape on its outer surface and independently controls the rotational speed of the left and right wheels using an active driving artificial intelligence unit (600), thereby eliminating the hunting motion that inevitably occurs in conventional conical wheels. As a result, left-right vibration of the railway vehicle is reduced, driving stability is improved, and maintenance costs for the track and vehicle are reduced.

[0032] The cylindrical wheel (110) according to the present invention has no flange, and the 1-axle active bogie always maintains a constant direction of travel with the rail (500), so no slip occurs between the running surface of the cylindrical wheel and the rail in curved sections and branching sections, thereby reducing noise and wear caused by friction, which were problems in the prior art. Accordingly, railway noise is significantly reduced, and the lifespan of the wheel and rail is extended, thereby reducing maintenance costs.

[0033] The present invention improves the reliability of driving control by limiting the wheel driving distance control variable to the wheel rotation speed and controlling it proactively using an artificial intelligence algorithm. In particular, through a control procedure including a target rotation speed correction step (S50) and an inverter control step (S70), precise control is possible to overcome time delays occurring during driving and reduce errors.

[0034] The present invention implements a hierarchical distinction between Level 1 active driving and Level 2 autonomous driving, thereby enabling the application of phased autonomous driving functions by first introducing only Level 1 active driving while maintaining existing railway infrastructure. Accordingly, it is possible to transition to an autonomous railway system while reducing the risk of system introduction.

[0035] According to the present invention, a safety wheel (210) that does not come into contact with the rail during normal driving supports the inner surface of the rail at four points in abnormal conditions such as earthquakes, emergency braking, or external collisions, thereby preventing derailment caused by horizontal movement or horizontal rotation of the cylindrical wheel (110). This improves the safety of the railway vehicle even in an autonomous driving environment.

[0036] FIG. 1 is a block diagram showing the overall configuration of an autonomous railway system according to the present invention.

[0037] FIG. 2 is a block diagram showing the hierarchical control structure of the first stage active driving and the second stage autonomous driving of the present invention.

[0038] Fig. 3 is a block diagram of the configuration of an artificial intelligence control device for a cylindrical wheel railway vehicle.

[0039] FIG. 4 is a flowchart illustrating an artificial intelligence control method that pre-controls the rotational speed of a cylindrical wheel.

[0040] FIG. 5 is a plan view showing a fixed rail branching section including a disconnection section for autonomous driving of a railway vehicle.

[0041] FIG. 6 is a plan view showing a continuous fixed rail branching section for autonomous driving of a railway vehicle.

[0042] FIG. 7 is a conceptual diagram showing a battery exchange system installed at a platform for a battery-operated autonomous railway vehicle.

[0043] FIG. 8 is a perspective view showing the configuration of a single-axis active bogie.

[0044] FIG. 9 is a perspective view showing the configuration of a single-axle bogie frame.

[0045] FIG. 10 is a perspective view of a wheel set in which two wheel units and a horizontal movement device are combined.

[0046] Fig. 11 is a plan view of a wheel set.

[0047] FIG. 12 is a perspective view showing the configuration of the vertical connection unit and the suspension unit positioned between the axle bogie frame and the wheel unit, with the axle bogie frame and the wheel unit separated.

[0048] FIG. 13 is a perspective view of a wheel unit combining one cylindrical wheel and two safety wheels.

[0049] FIG. 14 is a perspective view showing two 1-axle active bogies arranged side by side to form a 2-axle bogie.

[0050] FIG. 15 is a perspective view of a two-axle bogie for an autonomous freight train configured by connecting a two-axle bogie frame to two one-axle active bogies of FIG. 14.

[0051] For a detailed description, refer to 'Forms for carrying out the invention'.

[0052] The autonomous railway system using artificial intelligence according to the present invention comprises a railway vehicle (10) that travels along a rail (500), a sensor unit mounted on the railway vehicle (10) to detect a driving state, an artificial intelligence unit that processes driving data input from the sensor unit, and a control unit that controls the rotation of a wheel according to the control of the artificial intelligence unit.

[0053] FIG. 1 is a diagram illustrating the overall configuration of an autonomous railway system using artificial intelligence according to the present invention. The autonomous railway system includes a railway vehicle (10) that travels along a rail (500), an active driving artificial intelligence unit (600) and an autonomous driving artificial intelligence unit (700) that are mounted on the railway vehicle (10) and process information related to driving. The railway vehicle (10) travels while being physically guided by the rail (500), and the artificial intelligence units (600, 700) are configured to perform control corresponding to the driving state of the railway vehicle (10).

[0054] The active driving artificial intelligence unit (600) is responsible for driving control to maintain stable driving along the rail, and the autonomous driving artificial intelligence unit (700) is responsible for control to select a driving path in a branching section. Thus, the autonomous driving railway system according to the present invention is composed of a system in which driving control is performed through the relationship between a railway vehicle (10), a rail (500), and a plurality of artificial intelligence units (600, 700).

[0055] FIG. 2 is a diagram illustrating the control layer structure of an autonomous railway system according to the present invention. The control layer is configured to be divided into a first stage active driving and a second stage autonomous driving in order to functionally distinguish the driving control of the railway vehicle (10).

[0056] Stage 1 active driving is performed by an active driving artificial intelligence unit (600) and is a stage of controlling the rotation state of the wheels so that the railway vehicle (10) travels stably along the rail (500). Stage 2 autonomous driving is performed by an autonomous driving artificial intelligence unit (700) and is a stage of performing control to select the driving direction of the railway vehicle (10) in a branching section.

[0057] As illustrated in FIGS. 1 and 2, the autonomous driving artificial intelligence unit (700) may further include an autonomous driving sensor (710) for collecting information related to driving in a branching section. The autonomous driving sensor (710) is a component that provides information to the autonomous driving artificial intelligence unit (700), and provides image information, location information, and load information, and is positioned separately from the active driving sensor unit (630).

[0058] FIG. 3 is a diagram illustrating the specific configuration of a cylindrical wheel artificial intelligence control device in a railway vehicle running on a single axle cylindrical wheel according to the present invention. The control device includes a sensor unit (630) that collects information related to the running of the railway vehicle (10), and a control unit (610) that processes the information collected from the sensor unit (630).

[0059] The control unit (610) is configured to control the rotational state of the cylindrical wheel (110) by controlling the driving of the driving motor (190) based on driving information input from the sensor unit (630). Additionally, the control unit (610) can be linked with the braking unit (620) as needed to perform control related to deceleration or stopping of the railway vehicle (10).

[0060] FIG. 4 is a flowchart illustrating a method for autonomous driving of a railway vehicle using artificial intelligence according to the present invention, and shows a method for controlling the rotation of a cylindrical wheel in the process of the railway vehicle (10) traveling along a rail (500) in steps.

[0061] The autonomous driving method according to the present invention first includes a step of inputting driving information, including rail information, driving schedule, target speed, and target driving path in a branching section, in a driving information input step (S10). Subsequently, in a target rotational speed calculation step (S20), the target rotational speed of the left and right cylindrical wheels required at the current position is calculated based on the driving information. At this time, the rotational speeds of the left and right wheels are set to be the same in straight sections, and the difference in rotational speeds of the left and right wheels is calculated to correspond to the curvature of the rail in curved sections.

[0062] Next, in the gap data measurement step (S30), the gap between the moving cylindrical wheel and the inner side of the rail is measured using an active driving sensor unit. The gap data is used as information indicating the lateral positional state of the railway vehicle (10). Subsequently, in the reference value comparison step (S40), the measured gap data is compared with a preset reference value to determine the driving state of the railway vehicle (10). In this step, potential future deviation trends are preemptively reflected through artificial intelligence based on time-series prediction using past driving history and current driving state.

[0063] If, as a result of comparing reference values, the interval data is outside the allowable range or is predicted to be outside it, the process proceeds to the target rotational speed correction step (S50). In this step, artificial intelligence is used to correct the target rotational speed by reflecting the time delay between the measurement point of the interval data and the actual wheel driving point, measurement error, and interval deviation. On the other hand, if the result of comparing reference values ​​is within the allowable range, the process proceeds to the individual speed command generation step (S60) to generate individual speed commands corresponding to the left and right cylindrical wheels.

[0064] Finally, in the inverter control step (S70), the rotation of the cylindrical wheel is controlled by precisely controlling the frequency of the inverter that drives the driving motor based on individual speed commands. Additionally, the autonomous driving method may further include a power transmission step that transmits the rotational force generated by the inverter control to the cylindrical wheel to drive the railway vehicle (10), and may be configured to preemptively correct the error caused by the time difference between the power transmission step and the target rotational speed correction step (S50).

[0065] In the autonomous railway system according to the present invention, the railway vehicle (10) travels along the rail (500), and the rail (500) has a branching section (510) where the travel path branches off and a crossing section (520) that intersects each other. As shown in FIG. 5, in the fixed rail branching section including the cut section, a nose rail (570) is placed near the crossing section (520), and the structure is configured such that a gap between the rails is formed when the railway vehicle (10) passes through the branching section (510).

[0066] In this type of disconnection structure, the cylindrical wheel (110) of the railway vehicle (10) passes through a disconnection section between rails while driving. At this time, the railway vehicle (10) is configured to pass through a branching section (510) along a set driving path without physical movement of the rail (500) and without impact occurring in the disconnection section because it supports the adjacent rail using the wide width of the cylindrical wheel.

[0067] In the continuous fixed rail branch section illustrated in FIG. 6, the rail (500) is formed continuously including the intersection (520), and is configured so that the cylindrical wheel (110) always remains in contact with the rail while the railway vehicle (10) passes through the branch section (510).

[0068] In such a continuous structure, there are no breaks, so the rail contact state is continuously maintained during operation. Since the railway vehicle, which has a low risk of derailment because it does not travel at high speeds like a tram, can operate autonomously without safety wheels, the railway vehicle (10) in such a low-speed urban tram passes through the branching section (510) according to the path of the set continuous branching section.

[0069] Both the continuous structure and the disconnected structure have in common that they do not include a mechanical switch that physically moves the rail as in the conventional method, and have a structure in which a branch path is formed while the rail (500) is fixed. Therefore, the branch section (510) and the intersection section (520) can be structurally simplified and can be configured as a fixed rail branch structure that reduces the burden of maintenance by not including a mechanical drive unit.

[0070] The selection of a driving path in the branching section illustrated in FIGS. 5 and 6 is performed by an autonomous driving artificial intelligence unit (700), and the autonomous driving artificial intelligence unit (700) is configured to determine a driving path in the branching section (510) based on location information, driving direction information, or operation information of the railway vehicle (10). At this time, the autonomous driving artificial intelligence unit (700) is configured to select one of the pre-set driving paths according to the structural shape of the rail (500) or the arrangement of the branching section, and does not physically move the rail or change the branching structure itself.

[0071] In addition, unlike conventional turnouts, the fixed rail branch section in the present invention does not include a structure that changes the direction of the rail by an externally operated switch or driving device, and is configured so that the rail shape of the branch section (510) and the intersection section (520) is maintained in a fixed state, and may include a right-direction long rail (530), a right-direction short rail (540), a left-direction long rail (550), a left-direction short rail (560), and a nose rail (570) as main components.

[0072] Accordingly, the autonomous driving artificial intelligence unit (700) is configured to select one of a plurality of branch paths provided by a fixed rail shape (long rail / short rail / nose rail) based on location information, driving direction information, or operation information, thereby enabling passage through the branch section without a mechanical switch that physically moves the rail.

[0073] As illustrated in FIG. 7, the autonomous railway vehicle (10) of the present invention may be configured to receive power through a battery exchange system (800). The battery exchange system (800) is configured to include a battery module (810) mounted on the railway vehicle (10) and a battery exchange unit (820) installed at a platform or designated location to replace the battery module (810).

[0074] The battery exchange system (800) can be applied as an example of a configuration in which a battery module (810) mounted on the vehicle is used as a power source so that the railway vehicle (10) can run without an external power supply. In this case, the battery module (810) can be configured to be replaceable by a battery exchange unit (820) while the railway vehicle (10) is at a station or stopped, and the replacement process is performed independently of the driving control of the railway vehicle (10).

[0075] The battery exchange system (800) is applied as an example of a power supply means for the operation of an autonomous railway vehicle (10) and is not directly connected to the driving control, branching judgment, or active driving control of the railway vehicle (10). That is, the battery exchange system (800) is responsible only for the energy supply of the railway vehicle (10) and does not participate in the control function or driving judgment function of the artificial intelligence unit (600, 700).

[0076] As illustrated in FIG. 8, the 1-axle active bogie (300) according to the present invention is connected to the body of a railway vehicle (10) and configured to independently support the rotation of the left and right wheels, and to enable horizontal rotation with respect to the driving direction. The 1-axle active bogie (300) is configured to include a 1-axle bogie frame (310) that supports two left and right wheel units (100), and a 1-axle bogie rotation center (320) that is coupled to the 1-axle bogie frame (310) to form a rotation center, and the rotation center (320) is supported by a rotation center support (321).

[0077] The 1-axle bogie frame (310) illustrated in FIG. 9 includes a bogie insertion opening (340) for accommodating a wheel unit (100) and an insertion opening support (330) that supports the bogie insertion opening (340). Additionally, a suspension unit support (350) is formed on the 1-axle bogie frame (310) to enable a connection that allows the wheel unit (100) to move freely in a vertical direction and transmit a vertical load through the suspension unit. Through this structure, the 1-axle bogie frame (310) is configured to allow vertical movement of the wheel unit (100) while stably maintaining the rotation of the entire bogie.

[0078] As illustrated in the drawing, the cylindrical wheel (110) is configured to be rotatably coupled to the cylindrical wheel axle (120). The cylindrical wheel axle (120) is supported by a journal box installed on the wheel support unit (130), and the wheel support unit (130) constitutes the wheel unit of the 1-axle active bogie. The 1-axle active bogie is configured to be coupled to the bogie insertion opening (340) of the 1-axle bogie frame (310) through the bogie insertion rod (170) and the vertical connecting wheel (171).

[0079] As illustrated in FIGS. 10 and 11, a single-axis active bogie (300) is configured through a wheel set (200) comprising two wheel units (100). The wheel set (200) is configured to include a horizontal movement unit (270) configured to allow simultaneous horizontal movement of four safety wheels installed on two wheel units, and a horizontal screw connecting rod (260) connecting them. Accordingly, the wheel units (100) can move horizontally to evenly distribute wear on the cylindrical driving surface of the cylindrical wheels that occurs during driving.

[0080] As illustrated in FIG. 12, a vertical connection unit and a suspension unit are arranged between the wheel unit (100) and the axle bogie frame (310). In this structure, the vertical load of the railway vehicle (10) is transmitted to the wheel unit (100) through the suspension unit support (140), the elastic member (160), and the damper (165). The horizontal force generated during driving is transmitted to the vertical connection unit, which is formed by combining a bogie insertion rod (170) installed on the horizontal support unit (150) of the wheel unit (100) and a bogie insertion opening (340) of the axle bogie frame (310), thereby transmitting the horizontal load through this path. Accordingly, the transmission paths of the vertical load and the horizontal load are structurally separated.

[0081] As illustrated in FIG. 13, the wheel unit (100) is equipped with a cylindrical wheel (110) having a cylindrical shape on its outer surface, and the cylindrical wheel (110) is composed of a left cylindrical wheel (111) and a right cylindrical wheel (112) and is arranged to enable independent left and right rotation. Since the outer diameter of such a cylindrical wheel (110) is constant, the driving distance is determined in proportion to the number of rotations of the wheel, making it easy to control the number of rotations by the active driving artificial intelligence unit (600).

[0082] The rotation of the cylindrical wheel (110) is driven by a driving motor (190), and the rotational force of the driving motor (190) is connected to a power transmission unit (180) through a driving motor connection part (181). The power transmission unit (180) is configured to transmit the rotational force of the driving motor (190) to the cylindrical wheel axle (120).

[0083] In the wheel unit (100), a safety wheel (210) is disposed separately from the cylindrical wheel (110), and the safety wheel (210) is configured to rotate around the safety wheel axle (220). The safety wheel (210) does not come into contact with the rail (500) during normal operation, but in an abnormal state where lateral movement or horizontal rotation occurs while the railway vehicle (10) is running, it supports the inner surface of the rail at four points to prevent the cylindrical wheel (110) from derailing. Thus, the active driving function of the cylindrical wheel (110) and the derailment prevention function of the safety wheel (210) are structurally separated and implemented through the configuration of the wheel unit (100) and the wheel set (200).

[0084] As illustrated in FIG. 14, to form one embodiment of a two-axle bogie applied to the autonomous railway system of the present invention, two single-axle active bogies (300) may be arranged in parallel along the direction of travel. Here, each single-axle active bogie (300) has a structure in which it is not directly connected to the body of the railway vehicle (10) as illustrated in FIG. 3, but is connected to a separate intermediate bogie and rotates.

[0085] As illustrated in FIG. 15, two 1-axle active bogies (300) arranged in parallel are connected to each other by a 2-axle bogie frame (360) to form a 2-axle bogie. The 2-axle bogie frame (360) is configured to structurally combine the two 1-axle active bogies (300) to distribute and transmit the load of the railway vehicle (10), and is connected to rotate independently along the rail while maintaining a constant relative position of each 1-axle active bogie (300).

[0086] A two-axle bogie frame (360) has a two-axle bogie rotation center (370) formed therein, so that the entire two-axle bogie is connected to the railway vehicle (10) and is configured to rotate around the two-axle bogie rotation center (370). Since the two-axle bogie, which combines two such one-axle active bogies, always follows the rail even if there is a vertical track deviation in a curved section or a section with a change in cant, the railway vehicle (10) can always travel stably through the two-axle bogie.

[0087] The two-axle bogie configuration illustrated in FIGS. 14 and 15 can be described as an embodiment applicable not only to cases where a single-axle active bogie is directly connected to the car body to transport passengers, but also to freight trains transporting cargo. In this case, the freight train may use a battery mounted on the vehicle as a power source to enable operation without an external power supply, and an embodiment is possible in which the battery is configured to be replaceable through the battery exchange system (800) illustrated in FIG. 7.

[0088] Meanwhile, the artificial intelligence control unit may be configured to control the driving state of each freight train even when multiple freight trains are operated in a platooning mode where they are interconnected and driving, or in a distributed mode where they are driven independently. In this case, the driving mode is described as an example of driving control applied to each of the multiple railway vehicles (10), and does not imply a new driving control structure.

[0089] The two-axle bogies illustrated in FIGS. 14 and 15 are merely examples of embodiments that can be configured using the one-axle active bogie (300) of the present invention and are not to be interpreted as limiting the scope of the rights of the present invention. That is, the present invention exemplarily explains that, while maintaining the structural features of the one-axle active bogie (300), a plurality of one-axle active bogies (300) can be combined as needed and applied to a railway vehicle (10).

[0090] The autonomous railway system using artificial intelligence according to the present invention is applied to a railway vehicle (10) traveling along a rail (500) and has a structure in which driving control is performed by an active driving artificial intelligence unit (600) and an autonomous driving artificial intelligence unit (700). Accordingly, the present invention can be applied as is to an existing railway transportation environment in which a railway vehicle travels along a rail, and can be used industrially without separate large-scale track modifications.

[0091] The present invention can be applied to railway infrastructure including a branch section (510) and a cross section (520) while the rail (500) is fixed. Since such a fixed rail branch structure does not require a mechanical switch to physically move the rail, it can be applied to existing or new railway lines to reduce the burden of maintenance of the branch section.

[0092] The present invention is configured based on a structure in which a single-axle active bogie (300) including a cylindrical wheel (110) and a safety wheel (210) is directly connected to the body of a railway vehicle (10), so it can be applied to various railway vehicles having a single-axle or multiple-axle structure. Accordingly, the present invention can be industrially applied not only to single-axle vehicles but also to railway vehicles including a two-axle bogie structure as shown in FIGS. 14 and 15.

[0093] The autonomous railway system according to the present invention is applicable to an example of an operation method including a battery exchange system (800), and the battery exchange system (800) can be used as a power supply means for a railway vehicle (10). As such, the present invention can be applied regardless of the power supply method, and thus can be industrially utilized in railway transportation environments having various operating conditions.

[0094] Based on the above configuration, the present invention is applicable to passenger railways, freight railways, or railway systems requiring autonomous operation in general, and enables industrial utilization to improve the operational efficiency of existing railway transportation systems. Therefore, the present invention has sufficient industrial applicability in the railway industry.

Claims

In an autonomous railway system using artificial intelligence, A sensor unit mounted on a railway vehicle traveling along a rail with wheels that rotate independently on the left and right sides, for detecting the driving state of the railway vehicle; An artificial intelligence unit that processes driving data input from the above sensor unit; A control unit that controls the rotation of the wheel according to the control of the artificial intelligence unit; is included. An autonomous railway system using artificial intelligence, characterized in that the artificial intelligence unit determines the driving state of the railway vehicle based on the driving data and autonomously controls the rotation of the wheel according to the result of the determination. In paragraph 1 The above artificial intelligence unit is, An active driving AI unit that performs active driving, which is a first-stage AI-controlled driving in which the above-mentioned wheel travels along the rail without derailing; An autonomous driving railway system using artificial intelligence, characterized by comprising: an autonomous driving artificial intelligence unit that performs autonomous driving, which is a two-stage artificial intelligence-controlled driving in which the wheel itself selects a direction and drives in a rail branching section. In paragraph 2, The above autonomous driving AI unit is An autonomous railway system using artificial intelligence characterized by determining the driving direction of the railway vehicle in the branching section based on driving data including information regarding the speed, position, or driving direction of the railway vehicle. In paragraph 2, The above active driving artificial intelligence unit is It includes an active driving sensor unit that transmits information to enable the railway vehicle to drive stably along the rail without derailing, based on driving data including information regarding the speed, position, or driving direction of the railway vehicle. The above active driving sensor unit is, A position information sensor that measures the error between the center of the rail and the center of the wheel; An image information sensor that measures the error between the rail and the wheel using an image between the rail and the wheel; An autonomous railway system using artificial intelligence characterized by comprising one or more of the following: a load information sensor that measures the frictional force between the rail and the wheel to measure abnormal rotation of the wheel. In paragraph 1, The above control unit is A vector control unit that accurately controls the rotational speed of the running wheels of a railway vehicle; including An autonomous railway system using artificial intelligence characterized by driving a permanent magnet synchronous motor that rotates accurately using the above vector control. In paragraph 1, The above wheel is An autonomous railway system using artificial intelligence characterized by including a pair of cylindrical wheels that are separated left and right, can rotate independently, and have a cylindrical outer surface. In paragraph 6, The above cylindrical wheel It is mounted on a single-axis active bogie configured to rotate horizontally with respect to the direction of travel, and An autonomous railway system using artificial intelligence, characterized in that the driving direction of the railway vehicle is configured to be actively steered according to the rotational speed control of the artificial intelligence control unit. In Paragraph 7, In the above-mentioned single-axle active bogie, In the event that an abnormal condition occurs in which the above-mentioned railway vehicle moves sideways or rotates horizontally while traveling along the rail, An autonomous railway system using artificial intelligence, characterized by further being equipped with safety wheels having a structure that supports the inner side of the rail at four points to prevent derailment. In paragraph 8, The above-mentioned railway vehicle is a freight train that transports cargo, The above freight train is equipped with a two-axle bogie consisting of two single-axle active bogies arranged front and rear to run along the rail, and An autonomous railway system using artificial intelligence, characterized in that the above artificial intelligence control unit is configured to control the driving state of the driving wheel or bogie mounted on the above two-axle bogie. In Paragraph 9, The above freight train uses batteries mounted on the vehicle as a power source so that it can run without an external power supply, and An autonomous railway system using artificial intelligence characterized by the above-mentioned battery being configured to be replaceable. In the 10th, The above artificial intelligence unit is An autonomous driving railway system using artificial intelligence, characterized by being configured to control the driving state of the freight trains so that multiple freight trains can drive in a platooning manner in which they are interconnected, or drive independently in a distributed manner. In paragraph 1, The above-mentioned railway vehicle uses a battery module mounted on the vehicle as a power source so that it can run without an external power supply, and An autonomous railway system using artificial intelligence characterized by the above-mentioned battery module being configured to be replaceable. In Paragraph 12, An autonomous railway system using artificial intelligence, characterized in that the above-described battery module is configured to be replaced while the railway vehicle is stopped at a station. In the autonomous driving of railway vehicles, Operation information input step for inputting operation information such as rail information, operation schedule, target speed, and target route in branching sections; A target rotational speed calculation step for calculating the target rotational speed of both cylindrical wheels required at the corresponding location; A gap data measurement step for measuring the gap between a moving wheel and the inner surface of a rail through an active driving sensor unit; A target rotational speed correction step that uses artificial intelligence to correct the rotational speed of the cylindrical wheel by reflecting the time delay from the interval data measurement step to actual wheel driving and the deviation of the measured interval; An autonomous driving method for a railway vehicle using artificial intelligence, comprising: an inverter control step for precisely controlling the frequency of an inverter that rotates the drive unit of the above-mentioned cylindrical wheel. In Paragraph 14, A permanent magnet synchronous motor is used as the above drive unit, and A power transmission step for driving by transmitting power having a rotational speed generated by the permanent magnet synchronous motor according to the frequency of the inverter to the cylindrical wheel; further comprising A method for autonomous driving of a railway vehicle using artificial intelligence, characterized by using an artificial intelligence algorithm to preemptively correct an error caused by the time difference between the power transmission step and the target rotational speed correction step.